NORTH AMERICAN TUNNELING 2004

09096-Prelims.qxd

2/18/04

5:25 PM

Page III

PROCEEDINGS OF THE NORTH AMERICAN TUNNELING CONFERENCE 2004, 17–22 APRIL 2004, ATLANTA, GEORGIA, USA

North American Tunneling 2004 Edited by

Levent Ozdemir Colorado School of Mines, Golden, Colorado, USA

A.A. BALKEMA PUBLISHERS LEIDEN / LONDON / NEW YORK / PHILADELPHIA / SINGAPORE

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-Prelims.qxd

2/18/04

5:25 PM

Page IV

Copyright © 2004 Taylor & Francis Group plc, London, UK All rights reserved. No part of this publication or the information contained herein may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, by photocopying, recording or otherwise, without written prior permission from the publisher. Although all care is taken to ensure the integrity and quality of this publication and the information herein, no responsibility is assumed by the publishers nor the author for any damage to property or persons as a result of operation or use of this publication and/or the information contained herein. Published by: A.A. Balkema Publishers, a member of Taylor & Francis Group plc www.balkema.nl and www.tandf.co.uk For the complete set (book  CD ROM), ISBN 90 5809 669 6 CD ROM: ISBN 90 5809 670 X Printed in The Netherlands

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-Prelims.qxd

2/18/04

5:25 PM

Page V

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Table of Contents

Foreword Levent Ozdemir

XI

Organization

XIII

Session 1 – Differing site conditions as applied to design/build contracts Track 1 – Project management Design review boards – current state of practice A. Elioff & W.W. Edgerton

5

Drawing from past experience to improve the management of future underground projects C. Laughton

15

The ECIS story J.W. Critchfield & B. Miya

21

Track 2 – Security of critical infrastructure and key national assets: use of underground space Internal blasting and impacts to tunnels Wern-ping (Nick) Chen

29

Track 3 – Mechanized tunneling Improvements of the capabilities of cutting tools and cutting systems R. Bauer

37

MTBM and small TBM experience with boulders S.W. Hunt & F.M. Mazhar

47

Joint orientations for TBM performance analysis using borehole geophysics to orient rock cores T. Tharpe, B. Crenshaw & J. Raymer

65

Slurry type shielded TBM for the alluvial strata excavation in downtown area W.R. Jee

73

Estimating ground loss from EPB tunneling in alluvial soils for ECIS project, Los Angeles T.R. Seeley

79

Some aspects of grouting technology for Manhattan tunnels M. Ryzhevskiy & P. Barraclough

87

Track 4 – Specialized urban construction Design and construction of an LRT tunnel in San Jose, CA P.J. Doig

V Copyright © 2004 Taylor & Francis Group plc, London, UK

95

09096-Prelims.qxd

2/18/04

5:25 PM

Page VI

Underpinning design and construction – Atlantic Avenue Station complex rehabilitation, New York, USA A. Grigoryan & L.G. Silano

101

Slurry walls accelerate shaft construction in rock in Los Angeles M.P. McKenna, K.K. So, M.A. Krulc & E. Itzig-Heine

109

Performance of Russia Wharf Buildings during tunneling H.S. Lacy, M.D. Boscardin & L.A. Becker

121

Blasting adjacent to high voltage duct banks K.R. Ott, D.A. Anderson & S.E. Haq

129

Subway rehabilitation – secant wall cofferdams and penetration of tunnel liner V. Tirolo & N. Hirsch

135

Overcoming the complex geotechnical challenges of urban construction T.J. Tuozzolo

143

Session 2 – Subsurface investigations and geotechnical report preparation for design/build projects Track 1 – Risk allocation Risk management in tunneling – occupational safety  health plans for drill and blast and tunnel boring machines A. Moergeli

153

Managing underground construction risks in New York N. Munfah, S. Zlatanic & P. Baraclough

163

Risk allocation in tunnel construction contracts W.R. Wildman

171

Getting back on-track: Exchange Place Station Improvements M.F. McNeilly, S.A. Leifer & G.F. Slattery

177

Influence of geologic conditions on excavation methodology E.C. Wang, L.M. Hsia, C.C. Chang & A.N. Shah

185

Track 2 – Owners opinion forum Discussion and panel talk sessions – no written papers

Track 3 – Non-mechanized construction Santiago’s Metro expands C.H. Mercado, G.S. Chamorro & K. Egger

195

Benchmark for the future: the largest SEM soft ground tunnels in the United States for the Beacon Hill Station in Seattle J. Laubbichler, T. Schwind & G. Urschitz

201

Application of the Press-In Method in East Side Access tunnel project J. Liu & V. Nasri

209

Shotcrete for tunnel final linings – design and construction considerations V. Gall, K. Zeidler, N. Munfah & D. Cerulli

215

Robotic shotcrete applications for mining and tunneling M. Rispin, C. Gause & T. Kurth

223

VI Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-Prelims.qxd

2/18/04

5:25 PM

Page VII

Development of the LaserShell method of tunneling C.M. Eddie & C. Neumann

231

Ground support design and analysis: Exchange Place Station Improvements M.R. Funkhouser & M.F. McNeilly

241

Track 4 – Ground modification for underground construction Cantilever frozen ground structure to support 18 m deep excavation D.K. Chang, P.W. Deming, H.S. Lacy & P.A. van Dijk

251

New chemical grouting materials and delivery equipment technologies G.N. Greenfield & A.C. Plaisted

259

Jet grout bottom seal for cut and cover tunnel T.M. Hurley

265

North airfield drainage improvement at Chicago O’Hare International Airport: soil stabilization using jet grouting D.A. Lewis & M.G. Taube

271

Ground freezing and spray concrete lining in the reconstruction of a collapsed tunnel S.J. Munks, P. Chamley & C. Eddie

277

Ground freezing for urban applications P.C. Schmall, D. Maishman, J.M. McCann & D.K. Mueller

285

Session 3 – Design/build contracting practices Track 1 – Predicting and controlling cost and schedule An economic approach to risk management for tunnels B. Altabba, H. Einstein & H. Caspe Top down construction of Ramp L, Value-Engineering Change Proposal for the Massachusetts Turnpike Authority, Contract CO9A4 W.D. Driscoll & G.A. Almeraris Contemporary methods of budget preparation B. Martin & S. Sadek

295

303 313

Geotechnical mapping methods utilized in the Chattahoochee Tunnel Project, Cobb County, Georgia, USA J. Reineke, J. Raymer, M. Feeney & K. Kilby

319

Value engineered design facilitates Grand and Bates Relief Sewer Tunnel Construction, St. Louis, MO J.R. Wheeler & N.E. Thomson

327

Track 2 – Show me the money Discussion and panel talk sessions – no written papers

Track 3 – Investigation, inspection and rehabilitation Monitoring excavations using 3D Laser Scanning and Digital Close-Range Photogrammetry T. Trupp, L. Liu & Y. Hashash

VII Copyright © 2004 Taylor & Francis Group plc, London, UK

337

09096-Prelims.qxd

2/18/04

5:25 PM

Page VIII

Durability and corrosion protection of support systems in soil and rock tunnels M.R. Jafari, V. Nasri & M. Wone Investigation of complex geologic conditions for the Second Avenue Subway tunnel alignment in New York City, New York C.P. Snee, M.A. Ponti & A.N. Shah

345

357

An automated structural monitoring system for the Federal Reserve Bank of Boston T.L. Weinmann & L. Edgers

363

A deep horizontal boring – technical and contractual issues J. Glastonbury, K. Ott, J. Freitas, B. Russell, M. Wooden, W. Meakin & J. Canale

373

Rehabilitation of the Big Walker Mountain Tunnel in Bristol, Virginia D. Kukreja & P. Moran

381

Corrosion evaluation of the Manhattan rocks and corrosion protection of the rock reinforcement system for subway tunnels M. Ryzhevskiy & M. Berman

389

Rehabilitation of the Amtrak Long Island City ventilation structures S.G. Price

395

Using seismic tomography and holography ground imaging to improve site investigations E.J. Kase & T.A. Ross

401

Track 4 – Machine mining – soft ground to hard rock to everything in between Conditions encountered in the construction of the Braintree-Weymouth Tunnel Project, Boston, Massachusetts D.W. Deere, J. Kantola & T. Davidson

411

The Manapouri Tailrace Tunnel No. 2 construction – a very large TBM tunnel in very strong rock D.W. Deere, S. Keis & C. Watts

421

South Austin Regional Waste Water Treatment Plant Interconnect Tunnel Project S. Cheema, K. Koeller, R. Pohren, G. Sherry & R. Webb

433

Tunneling through an operational oilfield and active faults on the ECIS Project, Los Angeles, CA, USA E. Keller & M. Crow

441

Rock tunneling at the Mill Creek project M. Schafer, B. Lukajic, R. Pintabona, M. Kritzer, T. Shively & R. Switalski

449

Construction of the Dougherty Valley Tunnel, San Ramon, California, USA G.S. Nagle & H. Thom

453

City of Los Angeles Northeast Interceptor Sewer Tunnel Z. Varley, R. Patel & J. McDonald

461

Session 4 – Design/build risk Track 1 – SEM/NATM practices/prescriptive specifications NATM and its practice in the US Wern-ping (Nick) Chen & H. Caspe

473

SEM/NATM design and contracting strategies J. Gildner & G.J. Urschitz

477

VIII Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-Prelims.qxd

2/18/04

5:25 PM

Page IX

Engineers, contractors, and soft-ground tunneling equipment W.H. Hansmire & J.E. Monsees

485

Track 2 – Transit oriented development – making the case for going underground Atlanta West Area Combined Sewer Overflow Storage Tunnel and Pumping Station R.C. Divito, W. Klecan & G.D. Barnes

497

Track 3 – Analysis and design Consideration on machine data and load in TBM excavation for tunnel support selection N. Isago, H. Mashimo, W. Akagi & H. Shiroma

507

A durability design for precast concrete segments for tunnel linings G. Bracher & D. Wrixon

515

Design and construction of the Lindbergh Terminal Station, Twin Cities, Minnesota E.E. Leagjeld, B.K. Nelson, C.R. Nelson, D.L. Petersen, R.L. Peterson & B.D. Wagener

521

Design and impact of the Beacon Hill Station exploratory shaft program C. Tattersall, T. Gregor & M.J. Lehnen

529

Comparison of the predicted behavior of the Manhattan TBM launch shaft with the observed data, East Side Access Project, New York V. Nasri, W.S. Lee & J. Rice

537

Drop shafts – selection principals J.F. Zurawski & E. Petrossian

545

Stability evaluation and numerical modeling Exchange Place Station Improvements J.F. Lupo & M.F. McNeilly

553

Track 4 – Conventional underground construction Tunnel and shaft construction for the Pingston Hydro Project B. Downing, Z. Vorvis, G. Rawlings & P. Kemp

561

Shoal creek raw water intake and pump station construction on Lake Lanier D. Ackerman, R. Wiek & R. Gutridge

571

Design and construction of shafts at the San Roque Project M. Funkhouser, R. Humphries, W. Warburton, J. Daly & E. O’Connor

575

Ten years’ experience using roadheaders to bore tunnels for the Bilbao Metro J. Madinaveitia

581

Rio Piedras Project, San Juan, Puerto Rico B. Fulcher, N. Kofoed, P. Madsen & M. Bartlett

589

Devil’s Slide Tunnels Y. Nien Wang, B. Hughes, H. Caspe & M. Amini

605

IX Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-Prelims.qxd

2/18/04

5:25 PM

Page XI

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Foreword

The theme of this North American Tunneling Conference (NAT 2004) is “Underground Construction – The Sensible Solution to Urban Problems”. This title reflects the increasing importance of locating facilities underground for enhanced security and function of urban areas and to build critical infrastructure for sustainable development. This conference includes papers covering a wide range of subjects dealing with nearly all aspects of underground construction, tunneling and effective utilization of underground space. The papers are grouped under four major tracks. Track 1 addresses the management of underground projects and includes presentations on project management, risk allocation and predicting and controlling cost and schedule. Track 2 includes presentations and panel discussions on issues related to security of critical infrastructure and key national assets, owner’s opinion forum, financing of underground projects and transit oriented development making the case for going underground. Track 3 addresses new advances in technology, including sessions on mechanical tunneling, non-mechanized construction, investigation, inspection and rehabilitation and analysis and design of underground structures. Track 4 covers trials, tribulations and triumphs in tunneling industry by presenting significant case histories. The sessions address specialized urban construction, conventional underground construction and machine mining in soft ground, hard rock and mixed-face conditions. I would like to express my appreciation to NAT 2004 organizing committee, track and technical program chairs, panel members and the authors for their contribution to the success of the conference. The continuing support of cooperating organizations, AMITOS, TAC, NUCA, NASTT and UTRC is also acknowledged. Levent Ozdemir Proceedings Editor

XI Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-Prelims.qxd

2/18/04

5:25 PM

Page XIII

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Organization

NAT 2004 EXECUTIVE COMMITTEE Raymond W. Henn, Conference Chair Lyman Henn, Inc., Denver, CO Refik Elibay, Vice Chair Jordan Jones & Goulding, Atlanta, GA Susan Nelson, Executive Director AUA, Minneapolis, MN William W. Edgerton, Chair, Track I Managing Underground Projects Jacobs Associates, San Francisco, CA Brenda M. Bohlke, Chair, Track II Public Policy and Underground Projects PB Consult, Herndon, VA Robert J.F. Goodfellow, Chair, Track III Advances in Technology URS Corp, Gaithersburg, MD Gary Almeraris, Chair, Track IV Case Studies: Trials, Tribulations and Triumphs of Tunneling Slattery/Skanska, Whitestone, NY George Yoggy, Exhibition Chair GCS LLC, Allentown, PA Thomas Clemens, Technical Tour Chair American Commercial, Louisville, KY Carin Mindel, Exhibition Manager AUA, Minneapolis, MN

SESSION CHAIRS Dan Dobbels, Haley & Aldrich Brian Fulcher, Kiewit Construction Company Michael Goode, Telford Consulting Michael Greenberg, NYC Department of Environmental Protection John Kaplin, Gilbane Building Company Gary Irwin, City of Portland Bureau of Engineering Laurene Mahan, PBConsult, Inc. Bill Mariucci, Kiewit Construction Joseph M. McCann, Freeze Wall Chris Mueller, URS Corporation Galen Nagle, URS Corp.

XIII Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-Prelims.qxd

2/18/04

5:25 PM

Page XIV

Levent Ozdemir, Colorado School of Mines Stephen C. Redmond, Frontier Kemper Constructors Tibor Rozgani, Colorado School of Mines Heiner Sander, ILF Vince Tirolo, Slattery/Skanska

AUA BOARD OF DIRECTORS Officers Raymond W. Henn, President Thomas F. Peyton, President-Elect George D. Yoggy, Past President Hugh S. Caspe, Treasurer Susan R. Nelson, Executive Director Directors Gary Almeraris Charles H. Atherton Brenda M. Bohlke Jack Brockway Thomas Clemens Joseph P. Gildner Michael Greenberg Hugh Lacy Robert A. Pond Gregory L. Raines Kirk Samuelson Don Zeni Designated Representatives Charles W. Daugherty Randall J. Essex D. Tom Iseley Levent Ozdemir

XIV Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-01.qxd

2/18/04

8:13 PM

Page 1

Session 1 Differing site conditions as applied to design/build contracts

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-01.qxd

2/18/04

8:13 PM

Page 3

Session 1, Track 1 Project management

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-01.qxd

2/18/04

8:13 PM

Page 5

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Design review boards – current state of practice Amanda Elioff Parsons Brinckerhoff, Los Angeles, California, USA

William W. Edgerton Jacobs Associates, San Francisco, California, USA

ABSTRACT: This paper summarizes the current state of practice on the use of design review boards, or consulting boards, as used primarily used during design of underground construction projects. It discusses the various types of boards in use and reviews the history, and then, using the results of an industry survey of both owners and consultants, it discusses the purpose and typical uses, provides examples of specific outcomes, and reviews the methods of selection and modes of operation. It summarizes the advantages and disadvantages, evaluates the use of the construction manager to provide design review, and provides recommendations for future users based upon the lessons learned to date.

1 INTRODUCTION

•

Owner agencies have used a number of different methods for evaluating or “verifying” the design of underground facilities before advertising for bids. These methods include: Independent Peer Review, Value Engineering, Boards of Consultants, and Technical Review Committees. For the purposes of this paper, we have developed the following definitions:

•

•

•

This paper focuses on the current state of practice of Boards of Consultants, and Technical Review Committees, and is not intended to evaluate the use of either IPPR or VE panels.

Independent Project Peer Review: An independent panel tasked with design review for some outside party such as financing agency, congressional committee, etc. This process typically includes an in-depth review of criteria, analysis, and calculations. Value Engineering: Formal evaluation of design documents that evaluates design and to some extent anticipated construction methods and is focused primarily on cost objectives. This process typically consists of a one-week workshop with participants specially-trained in value-engineering skills, and results in recommendations for design changes to reduce cost while maintaining objectives. Board of Consultants: A separate board or panel under contract to the owner agency to evaluate the design prepared by the design consultant. This review can be done at specific time periods as the design proceeds, and is intended to determine bigpicture design issues and does not typically review detailed analysis or calculations. Can also be referred to as a Technical Advisory Panel (TAP).

2 HISTORY OF DESIGN CONSULTING BOARDS Owners have relied upon individual consultants to supplement the prime designer for some time. (See Terzaghi (1958) which contains an excellent review of his personal experience serving as an individual consultant on design and construction projects.) Review boards have been employed on major complex public works projects dating at least to the early 19th century during the bridge building era. (Petroski, 1996). Similarly, boards have been assembled for large dam projects under construction by public agencies such as the Tennessee Valley Authority, the US Army Corps of Engineers (COE) and the Bureau of Reclamation. In recent years, major large subway projects, for example the Bay Area Rapid Transit (BART)

5 Copyright © 2004 Taylor & Francis Group plc, London, UK

Technical Review Committee: A Board that is formed as a part of the design team, to evaluate the design progress and solutions on a periodic basis. The type of review and evaluation is similar to that performed by a Board of Consultants, with the primary difference being that the Board’s client is the design firm, rather than the owner agency.

09096-01.qxd

2/18/04

8:13 PM

Page 6

have stated that “The purpose of the Board should be to provide an objective, balanced and impartial view of the overall design and construction progress on a project. The Board should not be used as a substitute for normal consulting services …” Hoek & Imrie (1995). This usually requires an “independent” board that reviews the work of others (e.g., designers) at pre-determined intervals. To achieve this purpose, the owner must keep the review as a separate function and not let the consulting board become a part of the design team, thus insuring its “independence” and the ability for the board to testify (if necessary) as to the design adequacy. (2) More recently, boards have been used to improve both the efficiency and accuracy of the work product, acting as a part of the design team. The work product is typically plans, specifications, other contract documents, and (sometimes) cost estimates. To fulfill this purpose the consulting board (or technical review committee) need not be “independent” but can contribute to the design process at any time, even continuously at certain key period. Since the board or panel is an integral member of the team, the members cannot be presumed to provide an “independent” review of the work products.

System, designed in the 1960s, WMATA (1970s) and Los Angeles Metro (1980s-present) have maintained boards in some form to advise on project design. Within the past 30 years there has been an increase in the use of consulting boards. This may be in part due to the increasing complexity and multi-disciplinary nature of large projects. It may also be due to increasing oversight of the use of public funds and the arguably increased level of litigation resulting from the construction of such large projects. To the extent that this litigation is founded upon the theory of inadequate or defective design documents, both owners and designers are motivated to minimize these problems. 3 SURVEY From April to October 2003, an industry survey was conducted which asked questions concerning (1) the history of the use of such boards, primarily in the underground industry, (2) the purpose and typical uses of these boards as they are currently constituted, (3) typical criteria for selection of board members, (4) various modes of operation, and (5) approximate cost. The survey also solicited feedback from the respondents as to the use of the construction manager providing such design review, the perceived advantages and disadvantages of boards, and recommendations for improvement in the future. The results of this survey are incorporated into this paper. The survey instrument itself is available from the authors upon request. Although respondents were promised confidentiality, the raw data itself, absent attribution, is also available for subsequent researchers upon request. The survey was sent to 95 people identified by the authors as either consultants or owners who have experience either employing or participating on boards. We received 48 replies, a 51 percent response rate. The respondents’ experience represents over 500 boards as users of the process (i.e., receiving advice, either as an owner, designer, or construction manager), and over 300 boards as board members (i.e., providing advice). A list of the projects from upon which the respondents have based many of their comments is included as an Appendix. Projects represented by the experience of the survey respondents include tunnels both in soft ground and rock, transit stations, underground powerhouses, wastewater treatment plants, large dams, highway projects, pipelines, microtunnels, and large diameter shafts.

Occasionally a funding agency will require a design review board, and when that is the case, the owner’s purpose is to fulfill the specific agency requirements. Examples of such agencies are the Federal Energy Regulatory Commission (FERC), and the Federal Highway Administration (FHWA). Other agencies provide internal review teams when there are significant specialty design issues or high-risk elements. In the case of the Federal Transit Administration (FTA), project management oversight consultants (PMOC’s or PMO’s) are be established to “help ensure that grantees [of federal funds] constructing major transit projects have the technical capability to carry out the projects’ design and construction according to accepted engineering principles,” GAO (2000). The PMO function was incorporated into FTA “new starts” projects after some quality, cost and construction management issues occurred in the 1970s and early 80s. The oversight function includes review and evaluation of various project processes to ensure: compliance with statutory, administrative, and regulatory requirements. The PMOC and the other members of the design team typically work together in the design phase, but although the typical operation of PMOC is similar to that of consulting boards, because the purpose is to assure compliance with specific funding agency requirements, it does not serve the same function as a design review board, and the owner may not rely upon it to fulfill the same purpose

4 PURPOSE AND TYPICAL USES Two primary reasons are given for creating a design review board: (1) To reassure upper level decision makers that the design solution is adequate. Previous commentators

6 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-01.qxd

2/18/04

8:13 PM

Page 7

Changes in concept: Examples included changes from an exploratory shaft to an exploratory tunnel, and changes in tunnel ventilation systems to implement European design methods in the United States. More effective methods: In one example, use of explosives was deemed too dangerous by the design team and owner, as neither believed it would be safe or would be acceptable to the public. The consulting board was able to convince the design team that blasting could be done safely, if designed and implemented properly, and this resulted in savings of significant time and money. Contract packaging/Contracting methods: Examples cited were recommendations for pre-qualification of bidders, changes in contract pricing methods (e.g., plugged prices), and owner purchased TBM’s to save schedule. In addition to these examples of specific outcomes, consulting boards have been instrumental in the development of and/or “blessing” the use of new or “never been done before” design and contracting approaches: Acceptance of unique or first time designs can be difficult for some owners, and conservative designers to accept. The boards’ recommendations for additional testing to verify design assumptions was cited in a few cases, such as for special seismic designs, high loading assumptions, and gas barriers.

Table 1. Technical issues addressed by board. Technical issue

Percent of respondents

Geotechnical Engineering Design Methods Estimating/Scheduling Constructability Contracting Methods Equipment Selection/Approval Risk Evaluation/Assessment Special Construction Techniques

93 67 54 89 48 39 67 65

for which a design review board is established. In addition, the selection of PMOC consultants is usually much different, and the criteria and skill of the participants varies significantly from that used on consulting boards. Consulting boards and panels are usually formed to provide advice and make recommendations on certain technical issues. The industry survey indicated the percentage of respondents who have used such boards for these technical issues as given in Table 1. In some cases design consulting boards continue to provide consultation during construction, and in such cases they evaluate construction issues such as verification of design intent, basis of design, and contractor performance. There is little evidence that such design boards are used to resolve disputes between the contracting parties, although in some cases they have advised the owner on pre-dispute technical issues. Consulting boards discussed in this paper are not Dispute Resolution Boards. (For further information on DRB’s, see Matyas et al. (1996)).

6 SELECTION OF BOARD MEMBERS Members are selected for most consulting and/or review boards on the basis of recommendations by the design team or others. In fact, 95% of the survey respondents indicated that this is the most common method. In some cases an RFP or letter of interest is sent to the industry, and the board members selected using this method. 20% of the respondents had used this method, but only 5% used this method exclusively; i.e., did not rely upon recommendations of the design team. The background of board members appears to be quite varied. Members of academia (university professors) are used frequently, as are contractors, construction managers, and other designers. The industry survey indicated the percentage of respondents who have selected members with the backgrounds as given in Table 2. The use of individual consultants with one specific technical specialty is the most popular; and the technical specialties are determined based upon the key issues on the individual project. For most underground projects, board member backgrounds include geology, geotechnical engineering, tunnel boring machine (TBM) design, ground support design, and other specialties as required. Also noted was the use of operations and maintenance staff, the program manager,

5 SPECIFIC OUTCOMES Survey respondents gave numerous examples of outcomes resulting from consulting board meetings. These can be categorized as changes in design approach, construction methods, concepts, and contracting packaging and/or methods: Changes in design approach: Numerous respondents reported alignment changes (higher or lower tunnel profile) to improve excavation conditions, reduce settlement, avoid hazardous material, and subsequently save cost. Other recommendations were additional explorations, to collect more geologic or groundwater data. Changes in excavation method: These included changes from excavation by Tunnel Boring Machine (TBM) to use of the Sequential Excavation Method (SEM), use of closed face and/or pressurized face TBMs in lieu of open face shields to control settlement or mitigate hazardous conditions, and recommendations for changes in excavation sequence.

7 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-01.qxd

2/18/04

8:13 PM

Page 8

Most consulting boards are asked to reduce their observations to a written report at the end of the meeting. The only exception to this policy appears to be when the owner’s primary consideration is to limit the cost of the board, and in such cases results are reported verbally to the design team. In 70% of the reported cases, both verbal and written comments are provided. Also, in addition to comments to the design team, 80% of the survey respondents indicated that the board has provided comments to owner-agency upper level staff at a meeting. This comports with one of the underlying purposes that is to provide independent review.

Table 2. Backgrounds of respondents. Background

Percent of respondents

Academia Construction Contractors Construction Managers Designers

77 64 60 81

retired government employees, and upper management representatives from other public agencies. Owners that continue board operation through construction sometimes change the makeup, deleting academia and designers, and adding ex-contractors and construction managers to better evaluate the construction issues. In at least one case, a board has had access to a separate group of specialists, “… individuals that are not involved in the design of the project but available to serve in an ad hoc capacity to the Board on an asneeded basis on specialty issues” (Shamma et al. (2003)).

8 COST Most boards are composed of senior-level people who operate on a consulting basis and are compensated by the hour. (For information concerning contract provisions for senior-level consultants, see Dunnicliff & Parker, 2002). The hourly rates are relatively high compared to those of the design team, but owners who have used consulting boards report that the limited use of the board’s time, in part because of the ability of most experienced consultants to quickly identify the root issues, results in total cost to the project that is quite low, comparatively speaking, for the value added. Reported cost ranges from 0.5% to 1.5% of the total design fee; less than 0.1% of the construction cost. The total cost of the design review board ranges from $30,000 to $300,000, although the total cost is quite variable depending upon the frequency of meetings, length of the design period, and number of project contracts.

7 MODES OF OPERATION Virtually all of the survey respondents have used consulting boards during the planning and design phases of the project. A surprising 70% have continued the use of these boards into the construction phase, although only 15% have used them in the post-construction phase, presumably to defend contract disputes with the contractors. During the planning phase, most consulting boards meet only one or two times, although some respondents indicated 10 to 12 meetings, and some have experienced quarterly meetings. (These responses were received from individuals who were part of large construction programs that included more than one construction contract.) During design, some boards meet once, and some up to 6–8 times; although most respondents indicated from 2–3 meetings. These meetings are typically held at pre-determined milestone times, such as 30%, 60% and 90% design completion. The type of work reviewed at early periods is typically quite different from that reviewed at later stages of completion. More importantly, the ability and willingness of the design team to accept recommendations at the later stages of design is limited. Typical meetings are from one to five days in length. Summary or relevant documents are usually provided to the board for review in advance, and at the beginning of the meeting the designer makes a short presentation setting forth the key issues and the status of work currently under development. In some cases a tour of the work site is provided, especially if local conditions are critical to the design solution.

9 ADVANTAGES AND DISADVANTAGES 9.1

8 Copyright © 2004 Taylor & Francis Group plc, London, UK

Advantages

There are advantages to using consulting boards, some of which have been previously mentioned. Advice from a senior advisory board provides an independent check on the design criteria, which is helpful because “Those involved in the design and construction of a project can often become so involved in the details of the work that they find it difficult to stand back and take an impartial view of alternate approaches” (Hoek & Imrie (1995)). This advice can also provide the owner with the support to make decisions and design changes when warranted. If completed early enough in the process, it can provide a level of credibility and a “stamp of approval” to the design solution, and also provides the owner with confidence in its designer. Survey respondents confirmed this summary and also provided a range of other advantages summarized in categories as follows:

09096-01.qxd

2/18/04

8:13 PM

Page 9

a format to brief the Agency directors, PMO and other invited guests to all or summary portions of the meeting. These representatives may otherwise have little direct contact with the designer for questions and answers about the design or for relating political issues to the design team. When used during construction, the board can provide the owner with “third party” advice on contract disputes or differing site conditions, which is helpful if there is disagreement between the designer and CM. Risk Management: The board’s assessment of project risk is not likely to be as “sugarcoated” or conservative as that provided by the design team, thus providing more value to the owner agency. Conversely, the board may point out over conservatism in design.

Overall review of project: Coming to the project with experienced, “fresh” sets of eyes, the board’s review and concurrence with the design approach and criteria developed bring additional confidence to owners and engineers. In the process of review, they may point out overlooked issues, and recommend new areas to look into or additional study, such as more geotechnical exploration. For programs with multiple design contracts and no program manager, the board can provide a level of consistency with the design criteria and other factors. Bring additional experience, perspective, and trust: Given the collective years of experience, and worldwide exposure, the board members are able to compare the project at hand to past experiences, i.e., lessons learned from many underground projects. These board members may have access to information about other projects well before it is published. In some cases, negative experiences are never published, thus making it difficult to apply these lessons learned to future projects without the input of people who possess the appropriate first hand knowledge and can relate it confidentially in a venue such as a consulting board meeting. The board can also provide input to specialties that are not present in the personnel on the design team. This generally results in better quality contract documents that are more consistent, constructible, and results in better bid prices. Advice from disinterested “outsiders” may be more acceptable to politicians and the public. One survey respondent attributed the following quote to Walter Douglas: “ you hire a consultant (1) because you face a difficult problem you have never faced so you hire someone who has or (2) you need an expert to say something you could say but it is more believable because of his or her reputation. Point number two is particularly applicable as politicians, agency boards of directors, and others in upper management may be more willing to accept “bad news” if the source is a group of renown experts rather than staff or the design consultant. Focus on Key issues: In the process of preparing for the board meetings, designers and owners, must assemble relevant information for preview and presentation to the board. To have an effective meeting, they must develop the key issues they would like the board to address. These periodic meetings assist the design team by providing a “time-out” from the day-to-day crash program of completing the design documents. This allows for a review of “where are we going” that is of benefit to all participants. Often, it is not until faced with an upcoming meeting that the questions/issues are well defined. Increase Communication: The format of a “roundtable” type discussion, over a day to several days promotes better interaction between the personalities involved and facilitates better understanding of all positions and issues. The board meetings also provide

9.2

9 Copyright © 2004 Taylor & Francis Group plc, London, UK

Disadvantages

While disadvantages of consulting boards were cited less often, survey respondents did report several related to time and cost, disagreement between members, overreliance on board opinions, potential for late changes, and problems associated with the composition of the board: Time and Cost: Time and cost to prepare for and conduct, and present conclusions from the meeting may be as much as a week or two. Not only the board members, but also the project’s top management and designers may be tied up for days at these meetings, impacting the design schedule. The logistics of gathering all board members together can also be somewhat time consuming, since such senior people typically have full calendars. The design schedule can be impacted not only by the time for preparation and meetings, but also by the time required to review and revise the design documents should the board recommend changes. This can be a financial burden for smaller projects. Reliance on the Board: The presence of a consulting board can affect the design team’s view of responsibility for design decisions. Some designers (and owners) may be tempted to use such boards as “cover”, thus allowing them to avoid accountability for their design solution. Potential for Late Changes: The nature of the periodic review can make designers feel they must defend their design because they don’t have time to change it and still meet schedule. This works against the principle of collaboration, and can lead to the designer and/or owner disregarding the board’s advice to stay on schedule. Group Dynamics/Board Composition: Several instances were reported where board members disagreed, were uncooperative, or conversely, were too willing to compromise, resulting in “design by committee syndrome.” Boards and panels, in evaluating all of the issues, may not be able to find any middle ground. As a result, reaching consensus can be difficult,

09096-01.qxd

2/18/04

8:13 PM

Page 10

•

and such consensus is important for owners whose primary purpose is to convince upper level management that the design is adequate. On the other hand, there is also the disadvantage of “Design by Committee” which is different than “achieving consensus.” Without an effort to identify the reasons for all recommendations, when presented with a choice between alternate approaches or actions, there is a tendency to do both, thus resulting in an over-conservative design solution. Not all board members are helpful, nor do they all understand their role: Some want to be the designer, some may want to manage the entire process, some merely want to obstruct the designer’s progress for competitive reasons, and some simply are not qualified to be on the board. Opinionated panel members can also be counter-productive leading to conflict and/or delay.

•

• •

10 USE OF THE CONSTRUCTION MANAGER FOR DESIGN REVIEW There have been some suggestions that the owner could use its construction manager (CM) to perform the design review functions that are sometimes done with a consulting board or technical review board. The arguments in favor of this approach include: 10.1

• • • • •

•

Arguments in favor of using the CM

The CM’s construction experience will benefit the design solution. A document review from the point of view of a construction specialist will identify inconsistencies in the documents. Better bids will result if the contractors know that the CM has been involved at an early stage. The CM may be better equipped to consider market conditions and the advantages/disadvantages of different contract size and packaging strategies. It’s a good way to get the CM up to speed on the design intent, and have them buy into the design, thus reducing future disagreements with the designer. Review can be continuous rather than at specific times, allowing more timely and therefore less expensive design modifications than would be possible by waiting for the next scheduled consulting board meeting.

10.2

10.3

Can both be used?

Many users feel that both methods should be used, with the consulting board used early in the process, and the CM used later to provide constructability input and a consistency check of the documents prior to advertising. Many respondents said that the most important element in both methods is the use of knowledgeable personnel. 11 RECOMMENDATIONS After evaluating the comments from industry representatives, we offer the following recommendations for improving the results and the success of consulting boards. These recommendations are summarized by category: Purpose/Use, Member Selection, and Operation.

Arguments against using the CM 11.1

On the other hand, there are several disadvantages of using the CM instead of a separate design consulting board to perform these functions:

•

The future construction phase services contract may bias the CM’s recommendations to what the owner wants to hear. A separate consulting board would provide more independence from the process, i.e., the board has no self-interest in the outcome. Bringing on the CM earlier will increase the owner’s costs, and could encourage a postponement of the CM procurement, thus delaying critical input into the design. In addition, once mobilized, it is difficult to cut back CM’s costs if the project schedule slips due to public resistance or financing difficulties. Separate review boards are usually lower cost, as a result of a board’s “spot” reviews of specific issues, as opposed to the CM’s continuous review as a part of the design process. The CM may be perceived as less technically qualified (i.e., credible) on key issues. Because the design is undefined, it is difficult to identify the key issues and thus select the appropriate CM staff early enough in the design process to make beneficial use of their input. Also, procuring the CM at such an early stage may result in reliance upon staff that is subsequently unavailable when the construction starts. Using the construction manager adds an extra layer of review that cause delay and confusion of responsibility, especially if the CM firm is also a designer.

•

The CM staff tends to be generalist in nature, and the industry expert who can provide the technical expertise to the specific design problem is typically an individual consultant, not part of a CM firm.

10 Copyright © 2004 Taylor & Francis Group plc, London, UK

Purpose/Use

Review panels are most effective when retained and used aggressively during the concept design phase. It is more difficult to change direction when panels are convened after the project design is well advanced. Consider having a two part process, consisting of an early board to address the overall

09096-01.qxd

•

•

•

•

•

8:13 PM

Page 11

•

approach, and a subsequent technical board with some construction experience to address technical and constructability issues. Define the purpose and scope of the board for each specific project. Decide whether it is to be “independent” or an “integral part” of the design team. Confirm the meeting frequency and use these meetings as milestones in the design schedule. Determine what output is required, and write the scope of work for the board to define all of the above. Make the individual consulting contracts compatible with the purpose. For “independent” boards, ensure that the board members have a separate contract with the owner agency, not through the design engineer. For boards that are expected to function as an “integral part” of the design team, contracting through the designer is acceptable and perhaps preferable from a risk and liability perspective. On major projects, use a standing Board of Consultants to achieve consistency across separate projects that are all part of a large program.

11.2

•

2/18/04

•

11.3

•

Member selection

•

Select consultants whose technical expertise match the specifics of the project. In order to achieve a balance of expertise, select members with a diversity of experience (i.e., contractors, other owner reps, peers, and academic folks). Include representatives from the end-user; e.g., maintenance and operations personnel. Include experienced constructability experts. Recently retired construction executives can be effective panel members. Formally interview prospective board members before appointment in order to determine availability, understanding of public projects, predetermined positions and whether the board would have an “open” mind to innovations. Identify potential conflicts of interest, which could include competing design firms, a prospective member with contractor clients who may be potential bidders, or previous representation of third parties who may be opposed to the project. The board should have no stake in the outcome of the project. If so it can lead to some questionable “advice” and conclusions. Appoint consultants who are supportive of each other, and in particular, ask “... if they are willing to work with specific other potential members” (Hoek & Imrie (1995)). If the designer is solely responsible for selection of the board members, there is anecdotal evidence that the same members appear repetitively on many different boards. Whether this practice is beneficial is subject to debate, however given an ongoing consultant-designer relationship, the owners’ interests may not be considered using this approach.

• • • • •

Operation

Plan for meetings sufficiently in advance so that they are well organized and to ensure that the board members have a good understanding of the project. If information is limited, the board may not be able to raise critical issues, and its effectiveness will be limited. “Failing to keep the Board advised of critical decisions or events” and “Meeting only when the project is in trouble and expecting the Board to somehow rectify the problem or protect the parties” are ways to misuse the Board (Shamma et al. (2003)). The owner and/or designer should make a brief presentation at the beginning of the meeting to establish ground rules and bring the board up to date on recently completed studies, investigations, etc. Ask the board to reply to specific needs on the project. Be specific about the type of review or recommendations requested. Allow time before the meeting for the board to review documents, and after the meeting to think about and document their recommendations. Develop the client/consultant relationship so that it is not only technical and professional, but also business-like. Develop a rapport between the owner, designer, and the board members. Do not permit the board to “direct” the design. This can happen with a very aggressive and assertive consultant on the board. Give consultants feedback on what worked and what did not. This allows for continuous improvement in the process. To avoid diminishing the Board’s effectiveness, the Owner should be careful not to tell the Board what the Owner wants to hear (Shamma et al. (2003))

12 CONCLUSION As one respondent aptly wrote, “Each large underground project has common elements and elements that are unique to the individual project. How one uses boards is a function of the project, funding sources, public and political involvement, and how to get the best thinking and advance it with real

11 Copyright © 2004 Taylor & Francis Group plc, London, UK

Consider member personality. “Personality is also critical, since an effective Board consists of individuals unafraid of stating their opinions but who, on the other hand, do not attempt to dominate with dogmatic or irrational behavior” (Hoek & Imrie (1995)). Avoid those whose history indicates a trend toward becoming the “savior of the project.” Pay particular attention to selection of a chairman, to organize and direct the board’s operations so that the board collaborates with the other members of the design team.

09096-01.qxd

2/18/04

8:13 PM

Page 12

News. (December 2002 and subsequent discussion: March 2003, June 2003, and September 2003.) Hoek, E., & Imrie, A.S., 1995. Guidelines to establish project consulting boards. International Water Power and Dam Construction. August 1995. Matyas, R.M., Mathews, A.A., Smith, R.J., & Sperry, P.E. 1996. Construction Dispute Review Board Manual, McGraw-Hill. Petroski, Henry 1996. Engineers of Dreams: Great Bridge Builders and the Spanning of America. Vintage Books. Shamma, J.E., Tempelis, D., & Nakamura, D. 2003. Board of Consultants – A Requirement for Hard Rock Tunneling Projects. Proceedings of the Rapid Excavation Tunneling Conference; Society for Mining, Metallurgy, and Exploration. Terzaghi, K. 1958. Consultants, Clients and Contractors. Journal of the Boston Society of Civil Engineers. January 1958. United States General Accounting Office, 2000. Mass Transit, Project Management Oversite Benefits and Future Funding Requirements, Report to Congressional Requesters.

world aspects.” By incorporating the advice and recommendations of expert consultants through the process of a formally established design review board, the design and construction of underground and heavycivil projects can be completed more effectively and with less project risk. However, establishing an efficient design review board, selecting the right members, and operating it successfully requires the active consideration of the purpose for which the boards’ recommendations are solicited. Owner agencies contemplating the use of a design review board should consider lessons learned from previous projects, and take an active part in the establishment, member selection, and operation of the consulting board. The efficient use of both consulting boards and construction managers can contribute positively to project success.

ACKNOWLEDGEMENTS APPENDIX – REPRESENTATIVE PROJECTS

The authors wish to acknowledge the following participants who have provided a significant amount of information or assistance in the development of this paper: Alistair Biggart Hugh Caspe Pete Douglass Herb Einstein Refik Elibay Joe Guertin Bill Hansmire Ray Henn Roger Ilsley Jon Kaneshiro Gregg Korbin Jim Lammie Jack Lemley Dan Meyer Lew Oriard Harvey Parker Ralph Peck Pete Petrofsky Tom Peyton Ed Plotkin Bill Quick Wolfgang Roth Tim Smirnoff Joe Sperry Fred Estep Kim Chan

Examples of projects, which have used Consulting Boards or Panels – some of these projects are in various stages of completion at this writing. The authors acknowledge that project names may not be accurate and are reported as provided in survey:

Ron Drake Paul Gilbert Joe Gildner Paul Gribbon Richard Harasick Geoff Hughes George Morschauser Priscilla Nelson Joe Pratt Martin Rubin John Shamma Lily Shraibati Ralph Tripani Al Wattson Lee Wimmer Howard Lum Tom Kuessel Judy Cochran Ed McSpedon Richard Proctor Gordon Revey Rube Samuels Gordon Smith Francis Fong John Ramage Birger Schmidt

Subways Tunnels Sound Transit Central Link LRT, Seattle, WA Bay Area Rapid Transit Project, (BART), CA Washington Metropolitan Area Transit Authority Los Angeles Metro Rail, CA Baltimore Metro, MD Eastside Access, New York, NY Buffalo LRT, Buffalo, NY Shepard Line, Toronto, Ontario, Canada San Diego LRT Extension, San Diego, CA Tri-Met Tunnels, Portland, OR Water and Sewer Tunnels City of L.A., Central Outfall Sewer Rehabilitation Chattahoochee Interceptor, Cobb County, GA Mercer Street Tunnel, Seattle, WA Narragansett Bay Comm., CSO, Providence, RI MWRA Inter-Island Tunnel, Boston, MA MetroWest Water Supply, Boston, MA Claremont Tunnel seismic upgrade, Oakland, CA North Dorchester CSO, Boston, MA East Central Interceptor Sewer (ECIS), L.A., CA North East Interceptor Sewer (NEIS), L.A., CA MWD, Inland Feeder System, Los Angeles, CA South Bay Ocean Outfall, San Diego, CA Milwaukee Water Pollution Abatement, WI Upper Diamond Fork Project, Provo, UT

REFERENCES Dunnicliff, J., & Parker, H.W., 2002. The Care and Feeding of Individual Consultants and Their Clients, Geotechnical

12 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-01.qxd

2/18/04

8:13 PM

Page 13

Glenwood Canyon Tunnel, CO Central Artery Tunnel, Boston, MA Doyle Drive, San Francisco, CA Whittier Access Tunnel, AK Malmo City Tunnel, Sweden Wolf Creek Pass, CO Cumberland Gap tunnels, TN and KY Pinglin Highway Tunnels, Taiwan

Stanley Canyon, CO Wasatch County Water Efficiency Project Nancy Creek Tunnel, Atlanta, GA Colombia Slough CSO, Portland BES, OR Brightwater Conveyance Tunnels, Seattle, WA Baumgartner Interceptor Tunnel, St. Louis, MO Detroit River Tunnel, MI SWOOP, San Francisco, CA

Other Underground Uses

Highway Tunnels

Superconducting Super Collider, Dallas, TX Positron Electron Project (PEP), Palo Alto, CA

Interstate H-3, HI Devils Slide Tunnel, CA

13 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-02.qxd

2/18/04

8:13 PM

Page 15

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Drawing from past experience to improve the management of future underground projects Chris Laughton Fermi National Accelerator Laboratory, Illinois, USA

ABSTRACT: The high-energy physics community is currently developing plans to build underground facilities as part of its continuing investigation into the fundamental nature of matter. The tunnels and caverns are being designed to house a new generation of particle accelerators and detectors. For these projects, the cost of constructing the underground facility will constitute a major portion of the total capital cost and project viability can be greatly enhanced by paying careful attention to design and construction practices. A review of recently completed underground physics facilities and related literature has been undertaken to identify some management principles that have proven successful in underground practice. Projects reviewed were constructed in the United States of America and Europe using both Design-Build and more traditional Engineer-Procure-Construct contract formats. Although the physics projects reviewed tend to place relatively strict tolerances on alignment, stability and dryness, their overall requirements are similar to those of other tunnels and it is hoped that some of the principles promoted in this paper will be of value to the owner of any underground project.

vertical shaft and numerous chambers and caverns up to 25 m in span. Sadly, the Superconducting Super Collider (SSC) project, the largest such project so far attempted, was terminated before tunnel construction was complete. This project, perhaps above all others referenced, stands as an excellent example of what can be achieved when good contracting practices tailored to underground construction are adopted. The physics community is now developing a new set of accelerator projects, including the Tera ElectronVolt Superconducting Linear Accelerator, the Next Linear Collider and the Very Large Hadron Collider. The scope of underground construction for these facilities will be larger than any so far undertaken. Rock tunnel housings as currently envisaged will range in length from approximately 50 to 250 km. In addition, a number of new proposals for detector-based underground experimental programs are being developed, notably relative to the study of beta and neutrino particles, at sites in Brazil, France, Japan, Russia and the USA. Effective management of underground design and construction is a critical focus of the planning process as these projects move forward. The goals of this planning are to deliver satisfactory facilities quickly at an affordable price (“better, faster, cheaper”).

1 INTRODUCTION Over the past twenty years the particle physics community has built a number of underground projects worldwide. Underground sites are preferred for many experiments because the groundmass overlying the facility acts to block the passage of particles and/ or radiation that could otherwise have a deleterious impact on the experiments and/or the surrounding environment. Underground accelerator-based projects constructed in this timeframe include the Super Proton Synchrotron, the Large Electron Positron and the Large Hadron Collider located at the European Particle Physics Laboratory, in Switzerland and France; various projects at the Deutches Elektronen-Synchrotron in Germany and the Stanford Linear AcceleratorCollider, Superconducting Super Collider Laboratory and Neutrinos at Main Injector (NuMI) projects in the USA. A number of underground detector sites have also been constructed in this same timeframe, notably including excavations made within existing mine boundaries at the Creighton, Homestake, Kamiokande, and Soudan mines or located adjacent to road tunnels within the Fréjus, Mont Blanc and Gran Sasso alpine massifs. The combined underground scope of these projects totals close to 100 km of tunnel, 10 km of

15 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-02.qxd

2/18/04

8:13 PM

Page 16

2 UNDERGROUND PLANNING

3 TUNNELING IS DIFFERENT

The main design and construction phases of a rock tunnel project are shown in the flowchart in Figure 1. The flowchart is based on that proposed by the International Tunneling Association and discussed by Lowe (1993). This chart outlines the basic steps in tunnel design and construction from alignment through construction. The flowchart omits reference to some key tasks, notably those associated with estimating and scheduling the work. However, the flowchart does provides a framework for the discussion that follows in which ten general principles are proposed to support an effective tunnel design and construction process.

Decisions made at the start of the project will have a great influence on project outcome. As far as a tunnel project is concerned, probably the most critical decisions that need to be addressed at the outset are related to preparing the owner for changes to his normal construction practices. The owner may need some convincing that “normal” business practices may not work so well underground. “First-time” tunnel owners, in particular, may see no particular benefit or need to change established ways of doing business and will need convincing that the changes are worth the effort, notably because

•

Site Investigation & Alignment

• •

Rock Mass Characterization

Of course, the underground project may go smoothly or encounter problems irrespective of whether an owner decides to take such precautions. However, such precautions are warranted in order to be responsive to the particular vagaries of the underground project. It will take more effort in the short-term, but will provide for more effective protection of the project over time. If the owner can be convinced of the value of these changes up front, the rest should be easy!

Excavation Methods & Means & Structural Elements

Detailed Design & Modeling Experience Estimation Bypass

Design & Safety Criteria Review

4 FAMILIARITY WITH LOCAL CONDITIONS An early understanding of the host rock mass conditions is a critical element in the design process. To evaluate a site’s suitability, regional and locationspecific geologic information will need to be gathered. Information should be collected on rock units, structural folds and faults, groundwater and in situ stress regimes. This geological information will need to be assimilated and interpreted at an early stage in design in order to characterize the rock mass along the alignment(s) and provide input for concept constructability and engineering analyses. Early acquisition and interpretation of this data is key in support of the design process. This data will help quickly eliminate showstopper situations and avoid much of the “wheel-spinning” (multiple layouts, designs and drafting work) that can occur during design and can consume a sizeable amount of a highly limited resource. At the earliest stage of design, shown in Figure 1, adequate site investigation data can generally be drawn

Accept/Reject

Risk Assessment & Contract Structure

Tunnel Construction

Field Observation

Stability

Figure 1. Tunnel design process flowchart.

16 Copyright © 2004 Taylor & Francis Group plc, London, UK

normal design and construction partner(s) may not be able to provide the types and breadth of support necessary for underground construction significant resources will need to be expended on site investigation and this work will need to start early the bid documents may need to be changed to address the added elements of risk that tunnel construction brings to contracting.

09096-02.qxd

2/18/04

8:13 PM

Page 17

from field visits and desk studies. In all but the remotest of areas, published matter can be found to support desk studies (e.g. topographic and geologic, land use mapping and related studies). The design team should also seek to supplement the public domain data sets with specific information on construction projects of a similar nature undertaken in the region. As underlined by Trautman and Kulhawy (1983) such information can most readily be tracked down with the help of a local “geo-practitioner” (geologist, engineering geologist, geological engineer). Such individuals will know where the data is and, more importantly, know how to access it. Their familiarity with local formations and involvement on other projects will prove invaluable to the team throughout design and construction. Every design team needs access to such a professional, particularly at the outset of the project when data acquisition and rock mass characterization skills are at a premium.

Rock Conditions Excavation Size Structural Behavior Excavation Shape

Excavation Support Initial Stresses

Figure 2. Factors influencing the structural behavior of a tunnel, after Sutcliffe et al. (1990).

with due regard to the constraints of the construction process results in a more practical design and ultimately provides for a more affordable and lower risk construction product. A more integrated design strategy that involves the contractor can also provide for a more innovative approach to tunneling (Songer and Molenaar, 1996) and help to lower risks associated with unreasonable end-user demands.

5 CONTRACTORS’ DESIGN INPUT By the time a basic rock characterization has been attained for a site, key underground end-user requirements will also need to have been established. These requirements will typically include a definition of the space and environmental needs of the operating systems as installed. In this regard, the physics end-user is likely to focus on issues such as foundation stability, dryness and alignment given that the success of their operations (accelerator and/or detector) will be highly dependent upon these aspects of the opening’s performance. However, before decisions are made and drawings developed defining alignment and crosssectional requirements, the end user should be made aware that some compromises might be needed if the facility is to be built economically. Absolutes in precision, stability and watertightness cannot be met easily in a natural, variable rock material and the needs of the experiment will need to be balanced against the practical constraints that the ground mass imposes. To reach the economic compromises discussed above, the requirements setter(s), the designer(s) and builders should ideally have an opportunity to discuss the factors that will impact tunnel behavior, as shown in Figure 2. Ironically, contractors, who undoubtedly have the best appreciation of the constraints of tunnel construction and are the ones who will ultimately price and build the facility, are often completely excluded from all stages of the design process. A way needs to be found, regardless of the contract format, to solicit the input of the tunnel builder in order to establish an understanding of the process and build-up confidence in the practicality of the design (Atkinson et al., 1997). A tunnel design developed

6 CASE HISTORY BENCHMARKING One basic question that needs to be addressed during design is that of precedent. Have similar tunnels been built before? And if they have, what was the outcome? Such questions usually emanate from the owner or their representative who are interested in understanding exactly what kind of situation they have gotten into! These are reasonable questions for which the owner should expect comprehensive answers. Underground projects with similar rock mass and construction methods and means should be researched and made available for the design team to review. Some papers and reports that have compiled tunnel project data bases include the United States National Committee on Tunneling Technology (USNCTT) (1984), Sinha (1986), Parkes (1988), the Association Française des Travaux en Souterrain (AFTES) (1994), and, Nelson et al. (1994). These databases are recommended as a resource for anyone seeking an objective evaluation of case histories, they describe mining performance and problems encountered over the length of the tunnel. In addition to the compiled data base material listed above, tunnel construction issues are often reported in a number of industry journals and in conference proceedings such as those of the Australian Tunneling Conference, International Tunneling Association, North American Tunneling Conference, Rapid Excavation and Tunneling Conference and Tunneling Symposium.

17 Copyright © 2004 Taylor & Francis Group plc, London, UK

Excavation Method

09096-02.qxd

2/18/04

8:13 PM

Page 18

Whatever the level of risk anticipated on the job it is important to find a mechanism that allows this risk to be objectively expressed and communicated to others. To manage such risks effectively the impacts of risk on cost and schedule are perhaps best expressed under a series of “what if ” scenarios. These scenarios are needed to complement the deterministic cost and schedule performance reporting systems and will serve to remind management that although underground problems are not shown as activities on the schedule the possibility of encountering them is real! Even the most thorough site investigation of the most uniform geologic conditions will not be able to completely define the scope of an underground construction contract. Some surprises from the natural material should always be anticipated along the way and an effort should be made to provide management with a clear expression of risk as an integral part of the normal reporting process.

The owner’s confidence in the viability of the “tunnel plan” will be improved if comparable case history data can be compiled and assimilated. The owner will be even more convinced if visits to similar sites can be organized. Examination of case studies also serves as a reality check on plans. A similar case whose outcomes are inconsistent with current projections may raise useful questions or may point to key parameters that differ between the projects. 7 INTEGRATED ENGINEERING In the title of their 1979 paper, Curtis and Rock frame the problem of working on structural linings underground as follows: “Tunnel Linings – Design?” This title is a simple acknowledgement that ground loading on a tunnel lining is difficult to predict even in the most homogeneous of groundmasses. This uncertainty can result in conservatism and/or complexity in design; for example, the use of thick cast-in-place linings to support an otherwise strong rock mass. The over-design of the final lining is difficult to avoid when loading conditions cannot be predicted with great certainty. Key to minimizing such overdesigns is a consideration of the ground’s ability to contribute to the long-term stability of the opening. To this end there is a need to better integrate the geotechnical engineer’s knowledge in to the structural engineer’s model. Such integration may allow greater opportunity for a discussion of the strengths of the rock mass and ultimately result in the streamlining or even elimination of a “permanent structure.”

9 CONTRACTING STRATEGIES Nowadays, design and build is commonly held to have distinct advantages over more traditional Engineering-Procurement-Construct contracting, but design and build will not always provide the best solution. Under the right circumstances, a design and build approach may save the owner time and money and offer the individual contractor the best opportunity to integrate the design needs of construction with their preferred methods and means. As Cording (1985) notes, “The separation of design and specifications from the contractor’s planning create unnecessary impediments and adds unnecessary costs to the project.” However, there are circumstances where the owner may wish to maintain greater active control of the underground project through its execution, notably where public interest is high and/or architectural features are an important part of the project. As pointedout by Boye and Eskensen (2003) the argument for design and build is weakened as public involvement in the permanent works design (geometry, layout, aesthetics) and complexity of the contract interfaces increases. As the needs for prescriptive language in design and construction is reduced, the case becomes stronger for leaving the contractor greater flexibility in his/her choice of methods and means within the framework of the design and build contract option.

8 RISK MANAGEMENT Risks associated with underground construction are notoriously difficult to describe and quantify and setting realistic expectations for scope, cost and schedule is always a major challenge. Risks underground are strongly influenced by a number of factors, including the diversity/complexity of the geology, the density of the site investigation coverage, the amount and relevance of compiled case history information, the flexibility of the selected mining methods and means and the skill-set of the construction team. Risk analyses should be performed at critical junctures during design and construction to ensure that risks are properly characterized. Risk analysis should be performed to identify the types of risk to which the project is exposed and provide for an estimate of their frequency of occurrence, and the severity of their impact, ultimately in terms of cost and schedule. Management should use such information to decide upon the type and extent of mitigation required for each type of risk event.

10 ORGANIZING FOR SUCCESS All of the issues discussed above, while important, are secondary when compared to the need for assembling and maintaining a good project team to manage the work. Care should be exercised in the selection of

18 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-02.qxd

2/18/04

8:13 PM

Page 19

the framework of discussion. Issues should be framed in such a way that participants are not asked to answer leading questions and attention should be made to ensure that individuals are not placed in positions where conflicts of interest might arise. The review process should encourage frank and open discussion between participants aimed at comprehensively addressing agenda topics and answering specific questions. Review outcomes should include a single attendee-reviewed document that faithfully records the topics discussed, findings and recommendations. Any review recommendations that require follow-up should be addressed and appropriate actions taken.

Department of Energy

M&O Contractor

Texas Commission

AE/CM Contractor(s)

Design Consultants

Technical Systems

Construction Contractors

Figure 3. Management organization for the SSCL, after USNCTT (1989).

12 LESSONS LEARNED Many of the decisions made during the course of a tunnel project are experience-driven. Despite improvements in rock mass modeling and the prediction of mining performance the industry is likely to remain heavily dependent on this “experience factor” for the foreseeable future. Within the industry there is an ongoing need to share and learn from our collective experiences, both good and bad. The industry, cannot afford to let every owner learn from his/her own mistakes. If past successes and failures go unreported opportunities for improved practices will be lost and the same common errors will continue to be repeated. A more concerted effort is needed to methodically analyze and openly discuss the underlying reasons for success and failure of tunnel jobs. Sharing these experiences would allow the tunneling protagonists the opportunity to get smarter more quickly and allow potential owners better insight in to the workings of the underground construction industry.

all team members whether searched and selected from in-house staff or out-sourced. At a minimum, candidate members should be expected to demonstrate a requisite level of individual and corporate competence, and work products should be provided that exemplify the candidate’s ability to fulfill projectspecific roles. Focus should be placed on judging the relevance of past experience (similar requirements, geology, methods and means, etc.). When there is inadequate expertise within the owner’s existing organization, responsibility for the management of the design and/or construction may be delegated, as shown in Figure 3. Here the SSCL Architect/Engineer and Construction Manager (AE/ CM) team was carefully selected following guidelines setout by the US National Committee on Tunneling Technology, Geotechnical Board (USNCTT, 1989). The selected AE/CM (Parsons Brinckerhoff and Morrison Knudsen) provided a dedicated team of experienced professionals to the SSCL project. The project was managed to cost and schedule up until its termination in the early 1990s.

13 CONCLUSIONS Digging a hole underground is not as simple as it sounds. Cost and risk are potentially much higher than they are for equivalent surface-based or cut and cover structures. Tunneling really does present the owner with a different set of construction challenges than he/she may be accustomed to dealing with. At the outset of the tunnel project, focus should be placed on educating the owner to the particular vagaries of the underground contract. As work commences attention should be paid to developing an early appreciation of the site in general and the rock mass in particular. During the design, focus should be placed on properly integrating the end-user and engineering needs of the facility with the construction preferences of the contractor.

11 THE VALUE OF REVIEWS Technical reviews are a common part of most large tunnel projects. They can be regarded as a distraction from the core project objectives, but if properly run they can provide valuable opportunities for improved communication and learning between project members and ultimately result in a better project. Reviews are most likely to be effective if the agenda is established ahead of time and if participants are invited based on their ability to address agenda items. In some instances, an individual may be nominated to play the role of “devils advocate” to encourage and broaden

19 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-02.qxd

2/18/04

8:13 PM

Page 20

Underground Transport, Future Developments in Technology, Economics, and Policy, Boston, MA, US., April, pp. 121–141. Lowe, P.T. (1993), “The Planning and Design of the Prospect to Pipehead Tunnel.” Proceedings, 8th Australian Tunneling Conference, Sydney, Australia, 24–26 August, pp. 21–27. National Research Council (1989), “Contracting Practices for the Underground Construction of the Superconducting Super Collider”, Washington DC, p. 99. Nelson, P.P., Al-Jalil, Y.A. and Laughton, C. (1994), “Tunnel Boring Machine Project Data Bases and Construction Simulation,” Geotechnical Engineering Center Report GR. 94-4 to the National Science Foundation, December. Parkes, D.B. (1988), “The Performance of Tunnel-Boring Machines in Rock,” Construction Industry Research and Information Association, Special Publication No. 62, p. 56. Songer, A.D. and Molenaar, K.R. (1996), “Selecting DesignBuild: Public and Private Sector Owner Attitudes.” Journal of Management in Engineering, November– December, pp. 47–53. Sutcliffe, M.L., Rogers, S.F., Whittaker, R.N. and Roberts, B.H. “Integrated Approach to Geotechnical Assessment of Rock Tunnel Stability and Performance.” Proceedings of Tunnel Construction ’90, London, UK, April 1990, pp. 145–153. Trautman, C.H. and Kulhawy, F.H. “Data sources for Engineering Geologic Studies.” Bulletin of Association of Engineering Geologists, Vol. XX, No. 4, pp. 439–454. US National Committee on Tunneling Technology (1984), “Geotechnical Site Investigations for Underground Projects.” National Research Council, Washington DC, National Academic Press.

For tunneling particular attention should be placed on establishing and updating expectations for costs and schedule performance. Regardless of the contracting strategies and the instruments chosen to mitigate and/or allocate risks, the owner will need to be regularly briefed on issues of project risk as tunnel projects are vulnerable to critical path delays. Reviews can be valuable tools for providing fresh technical and contractual insights to the management team. During construction, the contract will require active management in order to ensure that contract provisions are met and, that ground conditions are evaluated and timely decisions made as necessary. At the end of each tunnel job the process and outcome should be objectively reported so that any lessons learned can serve as a reference and guide for other owners and industry professionals alike. REFERENCES AFTES, Working Group No. 4 on Mechanization of the Excavation Process (1994), “Fiche Signalétiques de Chantiers Mécanisés, Recueil 94.” Atkinson, A., Cavilla, J. and Wells, J. “Securing the Contractor’s Contribution to Buildability in Design.” Project Report 27, CIRIA. London 1997. Boye, C. and Eskesen, S.D. (2003), “Specifying underground Works – the Challenge of Developing the Optimal Requirements.” Proceedings Underground Construction Conference, London September, 2003 pp. 509–520. Cording, E.J. (1985), “Constraints on Tunneling Technology,” Proceedings of the Conference on Tunneling and

20 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-03.qxd

2/18/04

8:14 PM

Page 21

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

The ECIS story J.W. Critchfield Parsons Brinckerhoff, Los Angeles, CA, USA

B. Miya Bureau of Engineering, City of Los Angeles, CA, USA

ABSTRACT: At $240 M, the East-Central Interceptor Sewer (ECIS) is the largest construction contract ever awarded by the Los Angeles Department of Public Works. Four Earth Pressure Balance (EPB) Tunnel Boring Machines, in an urban setting excavated the 18.4 km tunnel. Construction management issues that affected the completion cost and schedule involved construction access, design changes, permeation grouting, existing utilities, community issues, tunneling mishaps, and unforeseen conditions. These challenges were met and overcome by the combined efforts of the City and the Construction Contractor, to complete a vital infrastructure improvement, as mandated by the State of California.

1.2

1 INTRODUCTION 1.1

Ground conditions along the tunnel alignment include a thin surficial layer of fill overlying alluvial and marine sediments. There are three generally recognizable deposits, from the ground surface downward: Recent alluvium – inter-fingered layers of streamdeposited loose to dense silty and sandy soils with gravel, cobbles and boulders, including some local deposits of soft organic soils. Encountered in about 5 to 10% of the tunnel. Lakewood formation – alluvial and shallow marine deposits including layers, lenses and pockets of generally dense silty sands and sandy silts, with gravel, cobbles and boulders. Encountered in about 80% of the tunnel. San Pedro formation – deep marine deposits composed of hard silts and clays with zones of dense sand and gravel. Encountered in about 10 to 15% of the tunnel. Hydrogen sulfide and methane were encountered in these deposits. The tunnel alignment is within or near several oil fields. Contaminated soil and groundwater were encountered at some work sites. Several active and inactive faults are present along the tunnel alignment. The most significant is the Inglewood Fault, located in the vicinity of the Baldwin Hills, near the downstream end of the project. The regional groundwater table is generally 25 to 50 m below tunnel invert. However, water is above the

Project description

The North Outfall Sewer – East Central Interceptor Sewer Project (NOS-ECIS) is designed to divert wastewater from the middle portion of the existing 80 yearold NOS. This will provide increased capacity to handle wastewater flows and allow the NOS to be rehabilitated. The project is the first phase of an urgently needed program required to prevent sewer overflows during storm events. A Cease and Desist Order (CDO) mandated construction, under threat of heavy financial penalties, from the California Regional Water Quality Control Board. An 18.4 km-long tunnel was constructed from East Los Angeles, westward to Culver City, along the alignment shown in Figure 1. The depth to tunnel invert ranges from approximately from 10 to 30 m, with a maximum depth of about 110 m under the Blair Hills. The excavated diameter was 4.7 m. The finished inside diameter is 3.35 m. The alignment was divided into five tunnel drives. Other project elements include 7 shafts, 31 maintenance holes, 11 junction structures for future connections, 2 diversion structures, a 90 m-long siphon and a 250 m-long micro tunnel, and a connection to NORS and other sewer lines. Conduits for fiber optic cables are incorporated into the final tunnel lining.

21 Copyright © 2004 Taylor & Francis Group plc, London, UK

Ground conditions

09096-03.qxd

2/18/04

8:14 PM

Page 22

Figure 1. Project location.

tunnel crown at the east end of the project, beneath the Los Angeles River. Water is also perched within the various soil layers and trapped within fault zones. 1.3

Construction environment

The alignment is largely within public right-of-way, but does cross beneath several private properties and structures. The neighborhoods near the construction sites include residential and commercial areas with numerous schools, churches and other sensitive facilities. Accordingly, site access and work hours are restricted and limitations are imposed on construction noise and vibration. Hundreds of existing structures are present within the zone of influence above the tunnel excavation. The tunnel is directly under several buildings and major utilities, the Interstate 10 and 110 Freeways, the Los Angeles River Channel, the Metropolitan Transportation Authority (MTA) Blue Line, Union Pacific and Amtrak Railroads, and the new Alameda Corridor facilities. 1.4

Figure 2. Completed sewer pipe in tunnel.

tunnel lining consists of segmental pre-cast concrete rings, installed and grouted at the tail of machine. Sections of Precast Concrete Cylinder Pipe (PCCP), lined with polyvinyl chloride (PVC) as corrosion protection, were installed inside the tunnel to complete the sewer, as shown in Figure 2. Cellular concrete was used to fill the annulus between the carrier pipe and

Construction methods

EPB tunnel boring equipment was specified as the primary means to control ground loss during excavation, limit surface settlements, and prevent damaging existing structures. Four Lovat EPB tunnel boring machines (TBM) excavated five sections of tunnel. The initial

22 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-03.qxd

2/18/04

8:14 PM

Page 23

An Owner-Controlled Insurance Program (OCIP) and a Project Labor Agreement (PLA) were included in the construction contract. A local hiring goal was incorporated into the PLA.

the tunnel segments. Fiber optic conduits are embedded in the annular grout. Permeation grouting, with cement or chemicals, was specified to improve ground strength prior to excavation beneath critical structures. No compaction grouting or structural underpinning was expected to be necessary. A geotechnical instrumentation program was used to monitor ground movements around the tunnel and measure effects on existing structures. A preconstruction survey was also conducted to document the condition of existing structures. Conventional mining methods was also used. Staged excavation and support installation was needed to excavate beneath the siphon and to construct starter tunnels and breakouts for junction structures. The connection to the active NORS was specified to be accomplished almost entirely underground, due to work site restrictions at the downstream end of the project. Micro-tunneling technology was used to construct portions of the siphon and for a primary connecting sewer. Large shafts and junction structures were constructed within soldier pile and slurry wall shoring systems. Small diameter shafts and maintenance holes were be installed by drilling. 1.5

2 MANAGEMENT CHALLENGES Issues and events that affected the cost of construction and the schedule for completion are outlined below. 2.1

Property acquisition and construction access issues soon became apparent, once ECIS was underway. The City struggled to obtain dozens of underground easements, often from intransigent property owners. Shaft construction for the Siphon Inlet cut off driveways to four homes. Planned back alley access turned out to be impracticable and City real estate agents scrambled to arrange for construction of a temporary driveway across the private properties. Work areas and traffic control plans on city streets needed to be revised to accommodate larger than anticipated equipment. Arrangements for a construction site at the North Outfall Relief Sewer (NORS) connection in Culver City took over a year longer than expected, due to numerous third-party complications.

Construction contract

Bids were opened in November 2000. A low bid of $240 M was submitted by the Joint Venture of Kenny, Shea, Traylor & Frontier-Kemper. The Engineer had estimated construction costs at $255 M. The Board of Public Works established a construction budget of approximately $280 M, including the bid amount, plus $10 M for insurance and $30 M as a construction contingency. Funding is provided entirely by local sources. The Contract, awarded in January 2001, is the largest construction contract ever for the LA Board of Public Works. Notice-to-proceed was given in February 2001. The original contract duration was 1000 calendar days, giving a contract completion date in midNovember 2003, to meet a completion deadline imposed by the State of California. 1.6

2.2

Design changes

A major design change was necessary on the ECIS project. The original design called for cast-in-place concrete structures at the bottom of each maintenance hole, some of which have junction structures for future sewer connections. The Contractor used 3-D Computer modeling to demonstrate that cast-in-place construction was impracticable and proposed an alternative scheme using pre-fabricated concrete and steel pipe. Eight unneeded maintenance holes were deleted to offset extra costs. The configuration of a slurry wall shaft at the upstream end of the project had to modified to make it constructable on a small work area, immediately adjacent to operating railroads. Additional modifications were made to facilitate connection of the Northeast Interceptor Sewer (NEIS), by another contractor. The NORS connection work site needed extensive site development work. Additional temporary shoring was constructed that allowed recovery of a tunnel boring machine and the installation of carrier pipe, as part of a schedule recovery strategy. Odor control facilities were added to the contract after award. Additional chemical injection and air scrubbers were needed to control odor during construction of sewer connections. Interim carbon filters

Project management

Design work was performed by the City of Los Angeles Bureau of Engineering, with assistance from Parsons Brinckerhoff Quade & Douglas. Construction management duties were carried out by a combined team of Bureau of Engineering and Bureau of Contract Administration personnel, together with consultant staff led by a Joint Venture of Parsons Brinckerhoff Construction Services and Brown & Root Services.

23 Copyright © 2004 Taylor & Francis Group plc, London, UK

Construction access

09096-03.qxd

2/18/04

8:14 PM

Page 24

Figure 4. Maintenance hole/tunnel intersection.

to advertising the contract for construction. A oneblock long section of the large concrete box culvert had to be demolished and reconstructed at higher elevation. This entire effort was performed on a time and materials basis, at a total cost of $4.6 M, exceeding the contract allowance by $1.2 M. Relocation of underground utilities and overhead lines was a ubiquitous problem at maintenance hole sites. At one location, the presence of critical overhead electric and underground services combined with high-speed traffic control issues made it impracticable to use a large drill. The 3 m diameter maintenance hole had to be hand-excavated to a depth of 24 m.

Figure 3. Maintenance hole installation.

were added at three sites to serve until permanent Air Treatment Facilities can be constructed for the completed sewer system.

2.5

Community issues

The largest single extra cost was for permeation grouting work. Several unit price bid items were included in the contract, assuming that both cement and chemical grout would be injected from the surface and from within the tunnel. The contract also assigned responsibility for design of the grout program to a specialty subcontractor. The contractor design was completely different, including only chemical grout and working only from the surface. This resulted in some unused bid items, for cement and underground work, and massive quantity overruns on chemicals and surface work items. The total volume of grout used was 11,000,000 liters, roughly the same as the design. The total cost was $12.6 M, exceeding the bid price by $5.5 M. The engineer estimate had included $14 M for permeation grouting.

Respect for the community is of paramount importance to the Board of Public Works. Restrictions were placed on work hours, noise, vibration levels and traffic, to reduce the burden of construction activity on neighborhoods. Costs for mitigating construction noise, vibrations and disruption exceeded the contract allowance, mainly for the construction of additional noise barriers, and for additional traffic controls to maintain access to properties. Prevailing wages requirements of the Project Labor Agreement were misunderstood by the trucking subcontractor, in preparation of their bid. The City issued a $2.8 M change order to remedy the situation. The narrowly defined PLA local hiring goal of 40% proved impracticable. Using the PLA definition, local hiring peaked at 26%, but averaged 14% overall. The percentage of Los Angeles residents on the project was 30%.

2.4

2.6

2.3

Permeation grouting

Existing utilities

EPB tunneling was mandated to control the excavation, limit ground loss, and minimize surface settlement.

City surveyors discovered that a major LA County storm drain conflicted with the ECIS tunnel, just prior

24 Copyright © 2004 Taylor & Francis Group plc, London, UK

Tunneling mishaps

09096-03.qxd

2/18/04

8:14 PM

Page 25

Extra work was required at the NORS connection because the geometry of the existing structure was different than expected and the flow level within the operating sewer was higher due to operational changes since the design was completed. 3 PROJECT COMPLETION 3.1

The critical path of the schedule was directly effected early in the project. Archaeological work and the delay associated with real estate dealings for the temporary driveway at the siphon inlet added 99 calendar days to the contract duration. During the course of the work, additional delays began to accumulate. Major identifiable delays were associated with obtaining and developing the NORS work site and with repairing the tunnel mishaps. More insidious delays resulted from the cumulative and inter-related effects of design changes, and other extra work activities. In June 2003, the City and Contractor negotiated a “global schedule agreement”. Under terms of the agreement all schedule issues to-date were resolved by development of a revised schedule which will place ECIS in service by June 2004 and complete all contract work by August 2004. The agreement also provided for payment of $2.0 M and for release of a portion of funds retained from progress payments. Subsequently the City approached the State Board to request an extension of the CDO completion date. The State accepted the City’s explanation and justification for the delay based upon unforeseen conditions, and granted an extension consistent with the global agreement.

Figure 5. Chimney caused by ground loss during a tool change.

There was no more important technical objective on the project, even though the application for ECIS tunneling in dry sandy ground was somewhat unusual and controversial. The EPB tunneling machines were generally effective in achieving the goal of controlling ground loss and limiting surface settlements. However, lapses in application of the EPB techniques did result in over-excavation. The process of stopping the TBM to maintain the cutterhead sometimes resulted in unintended ground loss. Chimneys, as shown in Figure 5, formed above the TBM during a tool change stops along the Alameda Corridor, which had just opened to commercial service. An extensive program of Compaction Grouting was performed to repair the ground along a 300-meter section of the alignment. A short section of tunnel lining was deformed and some heaving of the railroad tracks resulted from the compaction grouting operations. 2.7

3.2

Construction cost

The approximate cost to complete the project, as estimated in November 2003, is summarized in Table 1, along with a breakdown of extra costs. At $259.2 M, the expected final completion cost exceeds the Engineer Estimate by 2% and is 8% more than the original bid price. The total cost is considered acceptable, at 96% of the originally assigned budget.

Unforeseen conditions

A “mono” (grinding stone) artifact was recovered during soldier pile drilling at the siphon inlet shaft. The State Historical Preservation Officer required the shaft excavation to proceed in lifts, under supervision of the Project Archeologist. An abandoned oil well casing was found within the work shaft at the upstream end of the project. A specialty contractor was needed to investigate and remove a portion of the old casing. The hand-excavated maintenance hole encountered two additional differing site conditions. An unmarked sewer line conflicted with the excavation and had to be relocated. Unexpected flowing ground conditions required dewatering and grouting work to complete the excavation.

3.3

Lessons learned

ECIS has been one of the most challenging projects ever undertaken by the Department of Public Works. The quality of the completed work is excellent. The City considers the project successful with respect to schedule and budget. Key lessons learned are summarized as follows: Construction access. Acquire easements and work sites prior to beginning construction. Allow

25 Copyright © 2004 Taylor & Francis Group plc, London, UK

Schedule

09096-03.qxd

2/18/04

8:14 PM

Page 26

Table 1. Cost to complete summary. Engineer estimate Original contract Bid Contingency funds (12%) Insurance Assigned budget

$255.0 M $240.3 M $29.7 M $10.0 M $280.0 M

Extra costs Design changes Construction access Permeation grouting Existing utilities Unforeseen conditions Community issues Tunnel mishaps Total extra cost Cost to complete

$2.0 M $1.8 M $5.5 M $1.6 M $2.9 M $3.2 M $1.9 M $18.9 M $259.2 M

Committee is helpful to elevate utility issues for resolution. Community issues. Provide adequate funding allowances to meet commitments to the community. Tunneling. EPB has been shown to be the tunneling method of choice and it will be specified on future projects in Los Angeles. Unforeseen conditions. Cooperation and perseverance between the Owner and Contractor, with help from the various stakeholders can overcome unforeseen conditions.

ACKNOWLEDGEMENTS The authors would like to acknowledge the contributions and support of the following individuals:

•

realistic space for construction equipment and operations. Design changes. Review designs thoroughly for constructability. Permeation grouting. Consider implications of bidding schemes to reduce the potential for unintended consequences. Existing utilities. Accurate as-built information about existing utilities is difficult to find. An Executive

• • •

26 Copyright © 2004 Taylor & Francis Group plc, London, UK

Commissioners Valerie Shaw and Ellen Stein, Los Angeles Board of Public Works. City Engineer Vitaly Troyan and Deputy City Engineer Tim Haug. Inspector of Public Works Stan Sysak and Bureau of Contract Administration Inspectors John Reamer, Chris Smith and George Stofila. Project Director Ted Budd, and Patrick Kenney, of the Kenny Shea Traylor Fontier-Kemper Joint Venture.

09096-04.qxd

2/18/04

8:15 PM

Page 27

Session 1, Track 2 Security of critical infrastructure and key national assets: use of underground space

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-04.qxd

2/18/04

8:15 PM

Page 29

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Internal blasting and impacts to tunnels Wern-ping (Nick) Chen HNTB Corporation, Boston, MA, USA

ABSTRACT: During an external explosion event, the ground would shelter a nearby tunnel and the explosion impact to the tunnel may be minimum; however, if the explosion occurs in the tunnel, the implication from the explosion event may be enormous. Publications and researches for internal blast loading and its impact to tunnels are limited. The intents of this paper are to review general blast loading phenomenon and structure response to this type of loadings; to recommend load factors and dynamic material strength during a blast event for limit state design; to qualitatively estimate the impacts of internal blast loadings to tunnels; and to propose feasible hardening countermeasures and other security alternatives. Literature review was performed to explore current policies and design criteria for blast loadings on tunnels.

tunnels is seldom, it can happen. What are the sources that cause internal blast in a tunnel? It may be from trucks with chemical explosives, may be from the transportation of military vehicles, and may be from terrorist attacks. Next questions are: What is the special condition for a blast in a tunnel that is different from that of surface structures? What is the size of this blasting source? How is the blasting pressure determined? What is the response of tunnel structures from blast attack? What is the material behavior of structures in a blast event? What are exiting policies and criteria for design and prevention of blasting in tunnels? The purpose of this paper is to address these issues.

1 INTRODUCTION Traditionally, tunnels, as others underground structure, are very resilient to dynamic loadings, which include seismic events. The reason of this phenomenon is that once the ground is stabilized by tunnel supports during excavation, the redistribution of stresses in the ground occurs and eventually the ground reaches a new state of equilibrium and is self-supported. After this new state of equilibrium is reached, man-made ground supports are no longer needed in theory, since the inherent strength of the ground is mobilized. During an internal blasting event, with the composite effect from tunnel supports and the ground, the blasting impact to tunnel itself may be small. Local concrete spalling of tunnel lining is likely to occur, but the overall integrity of the tunnel remains. On the other hand, its damages to life and functional systems of the tunnel may be tremendous. This is especially critical for transportation tunnels. Consequences and damages to transportation tunnels from an internal blasting event can include:

• • • •

2 BLAST PHENOMENON A unique feature of blast loading in tunnels is its confinement. Overpressure builds up in tunnels, from a blast event, at different phases. First, the incident pressure reaches tunnel walls and generates reflected overpressure. Because of confinement, this reflected overpressure generates re-reflected overpressure. This process produces a series of blast waves of decaying amplitude. While this is happening, the second loading phase develops as the gaseous products of detonation independently causing a build-up of pressure, the gas pressure. The phenomenon is different from a free-air burst that is remote from any reflecting surface. The free-air blast is categorized as spherical airburst. In applying the spherical airburst to the hemispherical surface

Life safety issues from blasting overpressure, falling debris, fires, smokes, and flooding, Damages to transportation vehicles, Damages to ventilation systems, and Damages to tunnel structures, causing lining spalling and cracks, which may subsequently cause inundation to the tunnel if the tunnel is subaqueous and the inflow is excessive and can’t be stopped.

Tolerance for these consequences is null, but how do we detect and prevent it. Though blasting incident in

29 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-04.qxd

2/18/04

8:15 PM

Page 30

The most widely used blast wave scaling approach is developed independently by Hopkinson (1915) and Cranz (1926), which is commonly referred as the cubic root scaling.

burst, such as an explosive sitting directly on the ground or in a tunnel, an enhancement factor of 1.8 is generally recommended. The hemispherical blast pressure may also be obtained from blast chart based on tests. Figure 1 displays the pressure and time relationship after a blast, where negative pressure occurs after time t0. Figure 2 shows the Shock-reflection phenomenon in a region where , incident angle, is greater than 45° (Norris et al., 1959). 2.1

(d1/d2)  (W1/W2)1/3

(1)

(R1/R2)  (W1/W2)1/3

(2)

W1 and W2 are charge masses of charge diameters d1 and d2, respectively. Ranges at which a given overpressure is produced can thus be calculated from Equation (2), where R1 is the range at which a given overpressure is produced by W1 and R2 is the range at which the same overpressure is generated by W2. It is inferred that the charge weight is inversely proportional to the cubic of the standoff distance, R; therefore, the best way of mitigating a blast event is to increase the standoff distance. Spherical blast pressure, airburst, can be obtained from classic derivation or from blast curves generated by experimental testing such as those generated by the Departments of the Army, the Navy, and the Air Force (1990). Analytic results by Brode (1955) are listed below.

Blast loading

The industry standard to determine the magnitude of an explosive is in terms of its equivalent weight to TNT. Most explosive device used by terrorist attack in the US is a mixture of Ammonium Nitrate and Fuel Oil (ANFO), which is about 80% equivalency to TNT. Conversion factors for other explosives to TNT equivalent weights can be found in many other literatures, such as Conrath’s (1999).

Ps  6.7/(Z3)  1 bar for Ps  10 bar

(3)

Ps  0.975/(Z)  1.455/(Z2)  5.85/(Z3)  0.019 (4) bar for 0.1  Ps  10 bar Z is the scaled distance, given by Z  R/W1/3

Ps is the peak static overpressure. R is the distance from the center of a spherical charge in meters and W is the charge mass expressed in kilograms of TNT. Hemispherical blast pressure can be obtained by multiplying the results from the spherical results by the factor 1.8 or from blast curve generated by the Department of the Army, the Navy, and the Air Force (1990), as shown in Figure 3, where, Pr (psi) is the reflected overpressure, ta (ms) shock arrival time, td positive phase duration, Lw (ft) positive phase wave length U shock front velocity (ft/ms), and is and ir side-on and normally reflected impulse (psi-ms), respectively. Analytical hemispherical blast pressure can be derived based on works by Rankine and Hugoniot (1870) and Liepmann and Roshko (1957) as:

Figure 1. Pressure–time relationship after a blast.

Pr  2Ps [(7po  4Ps)/(7po  4Ps)]

(6)

where, po is the ambient air pressure. Published testing result for blasting pressure in confined space, such as tunnel, is not available. This blast pressure must consider shock wave re-reflection

Figure 2. Shock-reflection in a region where  is greater than 45° (Norris et al.).

30 Copyright © 2004 Taylor & Francis Group plc, London, UK

(5)

09096-04.qxd

2/18/04

8:15 PM

Page 31

•

phenomenon. Baker et al. (1983) has estimated this pressure as: PrT  1.75Pr1

(7)

PrT is the total peak pressure and Pr1 is the first reflected pressure as identified in Figure 3 or Equation (6).

2.2

2.3

•

Dynamic structure strength

The strain rate of a material will increase in a fast rate of loading condition. In such a condition, the mechanical properties of the material behave differently. Concrete and steel strengths are usually higher in a fast rate loading condition than at in a static loading condition. The factor by which the static stress is enhanced to calculate the dynamic stress is called the dynamic increase factor (DIF). Typical values are shown in Tables 1 to 3. Tables 1 to 3. Dynamic Increase Factor (DIF) for Design (Mays and Smith, 1995).

Structural response to blast loading

The positive duration, td, of a blast wave and the natural period of a structure determine the response characteristic of the structure. The structure shall be design in accordance with its response behavior as described below, where  is the natural frequency of the structure.

•

Dynamic response (0.4  td  40) – True dynamic loading only occurs when the plosive phase duration of a blast wave is equivalent to the natural period of a structure, and is seldom occurred in underground structures.

Quasi-static response (40  td ) – When the natural period of the structure is much less than the positive phase duration of a blast wave, the structure will be fully displaced before the decay of the blast load. Such loading is quasi-static or pressure loading condition, such as the gas pressure in a tunnel after blast. Impulse response (0.4  td ) – When the positive phase duration of a blast wave is much less than the natural period of a structure, the blast wave decays significantly before the structure has had time to respond. Most blast events in tunnels have this type of response, since tunnel structures, in combination with the ground, are massive.

Table 1. Concrete Type of stress

fdcu/fcu

Bending Shear Compression

1.25 1.00 1.15

Table 2. Reinforcing bars Type of stress

fdy/fy

fdu/fu

Bending Shear Compression

1.20 1.10 1.10

1.05 1.00 –

Table 3. Structural steel Type of stress

fdy/fy*

fdu/fu

Bending Shear Compression

1.20 1.20 1.10

1.05 1.05 –

* Minimum specified fy for grade 50 steel or less may be enhanced by the average strength increase factor of 1.10.

Figure 3.

where, fy – static yield stress, fdy – dynamic yield stress, fu – static tensile strength, fdu – dynamic tensile strength, fcu – static concrete compressive strength, and

Hemispherical blast curve (TM 5-1300, 1990).

31 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-04.qxd

2/18/04

8:15 PM

Page 32

Table 4. Threat parameters for tunnel design. Very High 2,000 lb

High 500 lb

Medium 100 lb

Low 50 lb

From 12,000 lb Truck

From 5,000 lb Truck

From 4,000 lb Car

From 4,000 lb Car

gas pressure build-up in blasting events; however, it does not protect personnel in the tunnel from injuries, since the injuries are directly associated with the initial overpressure and debris (or fragmentation if a bomb is cased). It is, therefore, concluded that detecting and preventing adverse explosives from entering tunnels is the primary countermeasure for blasting events in tunnels.

fdcu – dynamic concrete compressive strength. The modulus of elasticity for steel and concrete are insensitive to loading rates. Also, since blast loading is an ultimate event, its design load factor shall be set to unity. 2.4

4 POLICIES AND DESIGN CRITERIA An attempt was made to review existing policies and design criteria for tunnels under internal blasting events. Most of these documents are from military facility programs and policies triggered from tragic events between 1990s and earlier 2000s caused by terrorist attacks.

Threat parameters

Blast threat to tunnels is most likely from explosives or bombs in vehicles. Therefore, the maximum possible load from a blast will be a function to the size of the vehicles that carry the explosives. Table 4 is a recommendation for threat parameters for private-sector facilities (Conrath et al., 1999), which is suitable for tunnel design purpose as well. It also includes the likelihood of each threat occurrence and its associated vehicle size.

4.1

The section summarizes results from literature reviews of the latest available policies under blasting or terrorist attack events. These documents include:

•

3 COUNTERMEASURES Besides detecting and preventing blast threats from the outside of tunnels, physical countermeasures for tunnels under blasting events include:

• • • • • •

Structure hardening to improve structure resistance to blasting load, Provide shielding around tunnel, such as tube in tunnel and separating tunnels from direct exposure to blasting, Provide shielding around critical structural elements, including ventilation and fire fighting systems, Provide mechanisms in a tunnel to automatically detect and isolate blasting events and prevent their spreading (a blast proof automatic venting system would be required), Ground strengthen around tunnels by contact grouting and consolidation grouting, Provide external groundwater cut off mechanisms around, above, or in the tunnel by ground improvement, slurry walls, and internal automatic bulkheads.

•

•

Structural hardening is not cost-effective for blasting events in tunnels, since it is difficult and costly to design each element of a tunnel system to be blastproof. Even so, the life and safety issues for users in the tunnels can’t be guaranteed. For example, blast-proof automatic venting system is a countermeasure against

32 Copyright © 2004 Taylor & Francis Group plc, London, UK

Policies

Use of Underground Facilities to Protect Critical Infrastructures, Summary of a Workshop (Little et al., 1998) – It is a summary for workshop conducted to discuss the use of underground facilities for protection of critical infrastructures. This workshop discussed findings of the President’s Commission on Critical Infrastructure Protection (PCCIP) and key issues in going underground, but no policy issue was addressed. A Guide to Updating Highway Emergency Response Plans for Terrorist Incidents (AASHTO, 2002a) – This document addresses the existing state and DOT emergency management plans and practices, the standard view of the terrorist threat since 9/11, and a process guidance as to how state Departments of Transportation (DOTs) can update their emergency response plans. No specific policy is addressed in this document. A Guide to Highway Vulnerability Assessment for Critical Asset Identification and Protection (AASHTO, 2002b) – This guide was developed as a toll for state DOTs to (1) assess the vulnerabilities of their bridges, tunnels, roadways, and inspection and operation facilities, (2) develop countermeasures to deter, detect, and delay the consequences of threats, (3) estimate the capital and operating costs of such countermeasures, and (4) improve security operational planning for better protection against future acts of terrorism. This document addresses mostly surface structures. It does not mention specifics to tunnels or underground structures.

09096-04.qxd

•

2/18/04

8:15 PM

Page 33

Though not design criteria, several design manuals developed by federal agencies (mostly for military use) are helpful for general blast deigns. These manuals would be used to assist the development of specific criteria for blasting designs for civilian facilities and tunnels in the future. These manuals include (their distribution may be limited):

National Need Assessment for Ensuring Transportation Infrastructure Security (AASHTO, 2002c) – This document examines three key security planning program areas: (1) protecting critical mobility assets, (2) enhancing traffic management capabilities, and (3) improving state DOT emergency response capabilities. It estimates the total costs for the proposed initiatives, including capital investment and operations and maintenance expenses during the TEA-21 six-year reauthorization period. Annual cost for tunnel related security program and 54 critical tunnels are identified; however, emphasis of this document is still on bridges and surface facilities.

• •

From above reviews, it is clear that most of these documents provide guidelines and guidance in handling infrastructure security and threat identification and prevention, but not policies. Furthermore, most of these documents address on surface facilities, such as buildings, highways, and bridges. Document that directly addresses tunnel policies does not exist.

•

• 4.2

Design criteria

Civilian blasting design criteria for infrastructures does not exist either. Most design criteria for facilities are developed by US federal agencies for federal properties and most design manuals are derived by the US Department of Defense, Department of State, and General Services Administration for antiterrorism requirements for military, embassy, and federal facilities. The following sections review these criteria and design manuals. They could be used as guides in developing specific tunnel documents in the future. Most criteria that were reviewed are for federal surface facilities. Their applications to civilian infrastructures and tunnels are not direct and must be revised. These criteria include:

• •

• •

• •

GSA Security Criteria (GSA, 1997) – This document has been used for new facility designs and has been the basis of performance standards in retrofit analyses of existing buildings. ISC Security Design Criteria for New Federal Office Buildings and Major Modernization Projects, (ISC, 2001) – This document is fundamentally built from GSA Security Criteria. Its purpose is to adopt GSA criteria to suit all federal agencies. This document was review by NRC in 2003. Major comments by NRC are that though the intent of this document is performance based, its performance-based design process is unclear and explicit statement that mandates the use of the ISC criteria for all projects is missing. The intent of this document is clear, but its execution may be an issue since it is not mandated.

•

5 CONCLUSION Conclusions drawn from this paper are:

•

33 Copyright © 2004 Taylor & Francis Group plc, London, UK

Structures to Resist the Effects of Accidental Explosions (U.S. Departments of the Army, Navy, and Air Force, 1990). It is the mostly used publication by both military and civilian organizations. A Manual for the Prediction of Blast and Fragment Loadings on Structures, DOE/TIC-11268 (U.S. Department of Energy, 1992). This manual provides guidance for facilities subject to accidental explosions and aids in the assessment of the explosion-resistant capabilities of existing buildings. Protective Construction Design Manual, ESL-TR87-57 (Air Force Engineering and Services Center, 1989). This manual provides procedures for the analysis and design of protective structures exposed to the effects of conventional (non-nuclear) weapons. Fundamentals of Protective Design for Conventional Weapons, TM 5-855-1 (U.S. Department of the Army, 1986). This manual provides procedures for the design and analysis of protective structures subjected to the effects of conventional weapons. Design of Structures to Resist Nuclear Weapons Effects, Manual 42 (ASCE, 1985). This manual was prepared for civilian use, and has been widely distributed throughout the world. The Design and Analysis of Hardened Structures to Conventional Weapons Effects (DAHS CWE) (DNA, 1995). This new Joint Services manual, written by a team of more than 200 experts in conventional weapons and protective structures engineering. Security Engineering, TM 5-853 (U.S. Department of the Army, 1993). Terrorist Vehicle Bomb Survivability Manual (Naval Civil Engineering Laboratory, 1988). This manual contains information on vehicle barriers and blast survivability for buildings. Structural Design for Physical Security – State of the Practice Report (ASCE, 1995). This report is intended to be a comprehensive guide for civilian designers and planners who wish to incorporate physical security considerations into their designs or building retrofit efforts.

Blast wave propagation in tunnels is complicated. Blast pressure for design must take into

09096-04.qxd

• • • • • • •

2/18/04

8:15 PM

Page 34

consideration the shock wave re-reflection phenomenon. No blast testing result for tunnels is available. This paper presents a simplified procedure for blasting pressure in tunnels and provides recommended blast loads for tunnel designs. This paper provides countermeasures for tunnels under blast events; however, the best defense for blast events is early detecting and preventing adverse explosives into tunnels. Mandate security policy for civilian infrastructures does not exist in the US. Unified security policy for tunnels does not exist. It varies from state to state and from tunnel to tunnel. Mandate civilian blast design criteria for infrastructures do not exist. Blast design criteria and manual for tunnels does not exist. Blasting events in tunnels can happen. The need to address issues, polices, and design criteria for blasting in tunnels are immediate.

Baker, W.E., Cox, P.A., Westine, P.S., Kulesz, J.J. and Strehlow, R.A. (1983) “Explosion Hazards and Evaluation,” Elsevier. Brode, H.L. (1955) “Numerical Solution of Spherical Blast Waves,” J. App. Phys., Mo. 6, June. Cranz, C. (1926) “Lehrbuch Der Ballistik,” Springer, Berlin. Conrath, E.J., Krauthammer, T., Marchand, K.A. and Mlakar, P.F. (1999) “Structural Design for Physical Security – State of the Practice,” American Society of Civil Engineers. Department of the Army, the Navy, and the Air Force (1999) “Structures to Resist the Effects of Accidental Explosions,” Revision 1 (Department of the Army Technical Manual TM 5-1300, Department of the Navy Publication NAVFAC P-397, Department of the Air Force manual AFM 88-22), November. General Service Administration (1997) “GSA Security Criteria,” October. Hopkinson, B. (1915) British Ordance Board Minutes 13565. Interagency Security Committee (2001) “ISC Security Design Criteria for New Federal Office Buildings and Major Modernization Projects,” May. Little, R.G., Pattak, P.B. and Schroeder, W.A. (1998) “Use of Underground Facilities to Protect Critical Infrastructures, Summary of a Workshop,” National Academy Press. Mays, G.C. and Smith, P.D. (1995) “Blast Effects on Buildings,” Thomas Telford. Norris, C.H., Hansen, R.J., Holley, M.J., Biggs, J.M., Namyet, S. and Minami, J.K. (1959) “Structural Design for Dynamic loads,” McGraw-Hill Company, Inc. National Research Council (2003) “ISC Security Design Criteria for New Federal Office Buildings and Major Modernization Projects – A Review and Commentary,” The National Academies Press. Rankine, W.J.H. Phil. (1870) Trans, Roy, Soc., 160, pp 277–288.

REFERENCES American Association of State Highway and Transportation Office, in cooperation with the Federal Highway Administration (2002a) “A Guide to Updating Highway Emergency Response Plans for Terrorist Incidents,” May. American Association of State Highway and Transportation Office, in cooperation with the Federal Highway Administration (2002b) “A Guide to Highway Vulnerability Assessment for Critical Asset Identification and Protection,” May. American Association of State Highway and Transportation Office (2002c) “National Need Assessment for Ensuring Transportation Infrastructure Security,” October.

34 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-05.qxd

2/18/04

8:16 PM

Page 35

Session 1, Track 3 Mechanized tunneling

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-05.qxd

2/18/04

8:16 PM

Page 37

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Improvements of the capabilities of cutting tools and cutting systems R. Bauer VOEST-Alpine Mining Tunneling & Construction

ABSTRACT: Newest improvements of the capabilities of cutting tools and cutting systems for hard rock conditions within the Icutroc research project. Underground excavations for infrastructural development and extraction of minerals in urban areas is increasingly becoming a necessity. Wherever conditions and circumstances permit mechanical rock excavation methods such as roadheaders, drum miners or tunnel boring machines are used. However in the past, certain situations (e.g highly abrasive material and/or high strength of the material to be cut) precluded the usage of certain mechanical excavation methods such as roadheaders. In such cases drill and blast was the only economical and practical alternative. With the development of Icutroc, an exiting new opportunity for cutting rock that provides numerous practical, logistical, environmental, and safety benefits, was introduced and is on the edge of making a big impact on the construction and mining world in the US. The continuous, mechanical cutting process provides an excellent opportunity for automatization with a high potential for various cost reductions. Furthermore it is very often the only viable solution in urban, congested areas where drill and blast is restricted or prohibited.

industrial initiator. Further partners were two customers (Thyssen Schachtbau and Somincor) as well as three research institutes (Seibersdorf Research Institute, Vienna, Armines CGES, Paris and the mining engineering department of the Montanuniversity Leoben, Austria)

1 PROJECT OBJECTIVE What does Icutroc mean and what implications will it have for the future of the North American tunnel and construction world? Icutroc is a corporate research and development project that was partly funded by the European community. Its original main target was to develop the necessary cutting tools to be able to apply higher cutting forces to economically cut higher rock strengths. When VOEST Alpine Bergtechnik, situated in Zeltweg, Austria took the initiative in 1995 to start a research and development program under the acronym “Icutroc” the goal was to extend the range of economic applications for the existing roadheader lines by moving into territories of harder and more abrasive rock types. The project objective was to achieve the required target with a type of roadheader that does not exceed 130 metric tons of operating weight and to stay within a range of 300 kW installed power on the cutterhead. The reason for these premises were that the maneuverability of the machine shouldn’t be sacrificed neither should the investment cost for this roadheader type exceed an acceptable, economic range. 1.1

1.2

Research partnership

The research program incorporated VOEST Alpine Bergtechnik and Sandvik Rock Tools as the main

37 Copyright © 2004 Taylor & Francis Group plc, London, UK

The Icutroc research approach

The Icutroc research approach was characterized by a combined development work addressing the necessary improvements of the cutting system and the machine system. Additionally Icutroc aimed to significantly improve the mechanical and wear properties of the cutting tools. In detail the whole project included: Development and refining of cutting systems and processes followed by simulations of their real world behavior by using computer-aided modeling and FEM-calculations. Better understanding of the interaction between rock/rock mass behavior and its influence on the cutting process. New concepts and material designs of cutting tools and new production technologies to manufacture these tools. Laboratory testing of these units. Testing and optimization of the newly developed system by civil engineering and mining end users under practical conditions.

09096-05.qxd

2/18/04

8:16 PM

Page 38

Figure 1. Icutroc’s project target for the development of a hard rock roadheader.

techniques in order to harmonize the factors of the cutting process with the characteristics of the cutting machine. Thereby taking into consideration the rock properties, geometric parameters of the cutting unit and the operating characteristics of the cutting system (cutting speed, sump in depth and machine stiffness). A finite element model of the complete machine has been set up for the simulation of the elastic behavior of the system. Concepts to reduce the elasticity of the cutting system were investigated to meet the required stiffness parameters for hard rock cutting. A substantial improvement of the overall system stiffness for the cutting action could be achieved by the boom stabilization system acting on the hydraulic boom cylinders. The boom stabilization system allows for a better adherence to the preset cutting depth as well as an improved compliance with a uniform swivel process. Together with the added benefit of reduced vibrations (the boom stabilization system also reduces the “bouncing” of the boom significantly) the shorter overall path length of picks further improves pick life. Development of a new rock mass rating specifically adapted for the new generation of roadheader technology, which increased the quality and reliability of performance prediction tremendously. In order to gain a revised RMR, two approaches were used first the theoretical net cutting rate based on cuttability of intact rock thereby reflecting the machine characteristics. Second the effective net cutting rate directly measured on site reflecting the actual operating conditions. Outcomes of this investigation were an exceptional correlation between NCReff/NCRtheor and the

Depending on the toughness and abrasivity, economically cutting of rock hardness up to 200 Mpa. Provide all conceptual prerequisites to implement updated control and data logging facilities, as they are required due to project conditions. The economic and environmental significance of this project is emphasized by the fact that the research work was funded by the European union within the Brite-Euram III program for Industrial and Material Technologies, managed by the European CommissionDG XII. 1.3

Development of a new cutting process

Using the newly developed VOEST Alpine cutterhead design software accounting for parameters such as optimized forces, cutting depth, cutting distances, slew and feed speed of the cutter boom, cutter head diameter, geological parameter, etc. the research project was able to design lacing layouts that resulted in the highest possible cutting efficiency. These cutting systems employing low speed cutting and providing greater power at the cutterhead in connection with the development of a new generation of cutting tools, were able to cope with the higher forces to be expected when cutting rock above 130 Mpa.

2 DEVELOPMENT OF AN ADAPTED MACHINE SYSTEM The research project included intensive modeling of the complete system using computer simulation

38 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-05.qxd

2/18/04

8:16 PM

Page 39

Figure 2. Specifically for Icutroc developed cutter heads designed with VOEST Alpine’s proprietary software. Rating of uniaxial compressive strength

Rating of block size Block size [m3]

Rating 20

UCS [MPa]

Rating

>0,6

1–5

15

0,3–0,6

16

5–25

12

0,1–0,3

10

25–50

7

0,06–0,1

8

50–100

4

0,03–0,06

5

100–200 >200

2 1

0,01–0,03 <0,01

3 1

Rating of joint conditions Surface

Aperture

Wall/Fill

Rating of orientation of joint set Rating

Influence on cuttability

Rating

rough

closed

hard, dry

30

very favorable

ⴚ12

slightly rough

<1 mm

hard, dry

20

favorable

ⴚ10

slightly rough

<1 mm

soft, dry

10

fair (and if block size < 0,03m3)

ⴚ5

smooth

1–5 mm

soft, damp

5

unfavorable

ⴚ3

very smooth

>5 mm

soft, damp to wet

0

very unfavorable

0

Figure 3. Revised rock mass rating approach for Icutroc roadheader technology.

RMRrev. No significant difference for different rock types and different roadheader types, and most important the higher effect of parting systems on cutting performance at low cutting speed.

The ratio between net cutting rate and effective net cutting rate was used as a measurement for the increase of performance including the influence of rock mass conditions.

39 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-05.qxd

2/18/04

8:16 PM

Page 40

10 Erzberg Premadio

9

Stillwater Athens

Pozzano

NCReff / NCRtheor

8 NCReff/NCRtheor = 46,537RMRrev-0,9877 R2 = 0,9072

7 6 5 4 3 2 1 0 0

10

20

30 RMR (revised)

40

50

60

Figure 4. Evaluation of rock mass influence on roadheader performance (slow cutting speed).

10 Balsareny Bogdanka Saudi Arabia Blumenthal

9

NCReff / NCRtheor

8 7

Wujek Prosper (AM 105) Prosper (AM 85)

Borynia Erzberg Szieroszowice

6 5

NCReff/NCRtheor = 9,4302RMRrev-0,5614 R2 = 0,8256

4 3 2 1 0 0

10

20

30 RMR (revised)

40

50

60

Figure 5. Evaluation of rock mass influence on roadheader performance (high cutting speed).

9

Trendline for High Cutting Speed (~3m/s)

NCReff/NCRtheor

8 7 6

Trendline for Low Cutting Speed (~1.4m/s)

5 4 3 2 1 0 0

10

20

30 RMRrev

40

Figure 6. Comparison between high (3 m/sec) and low (1.4 m/sec) cutting speed.

40 Copyright © 2004 Taylor & Francis Group plc, London, UK

50

60

09096-05.qxd

2/18/04

8:16 PM

Page 41

considerable less developed rock mass features (higher RMRrev) to the cutting process. It gets evident that low cutting speed provides more time to activate parting systems, whereby the higher available pick forces lead to an increased influence zone ahead of the pick tip. With increasing values for RMRrev the encountered difference in influence decreases.

After evaluation and utilization of the gained data from the ratio NCReff/NCRtheor VOEST Alpine was able to either modify existing correlation (e.g. RQD, RMR) or to establish a new rock mass classification system adapted for the new mechanized cutting systems. Why do we see such a tremendous effect between low cutting speed and increased cutting performance? First slow cutting speed means more available time for an activation and subsequent response to the existing parting system. Second higher pick forces resulting from a redesigned gearbox causing an extended stress zone in front of the pick crushing area. The third reason is improved design parameters for the cutter head (e.g. increased pick spacing assists activation of parting systems). Fourth there is a strong relationship between heat generation and cutting speed leading to rapid weakening and failure of the temperature sensitive tungsten carbide pick whereas low speed cutting and the improved cooling system reduce the heat generation significantly. With the chosen ordinate NCReff : NCRtheor the application of these diagrams for rock-mass-related performance prediction is provided. Of significant practical importance is the finding that a “stiff ” machine system applying low cutting speed (1.4 m/s) adds significant more rock mass contribution at

3 DEVELOPMENT OF A NEW COOLING AND DUST SUPPRESSION SYSTEM Heat generation on the pick tip was found to be the second biggest reason (second only to the results of extensive impact energy) for pick failure during the cutting process. In order to overcome this problem a new pick cooling system ensuring optimal cooling at acceptable water flow rates had been developed (the amount of necessary water could be reduced by approximately 25%). In combination with an intermittent sector controlled spraying system this innovation reduced the specific wear and tear on picks by 50%. 3.1

Cutting groove spraying system (VAB “Wet Head”)

A high-pressure water jet is pointed directly to the zone of the maximum dust creation. This system

Figure 7. The new pick cooling system results in a significant reduced wear and tear compared with any existing system. (Pick wear pattern during the cutting of sandstone with 160 Mpa and a Cerchar abrasivity  2,8).

41 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-05.qxd

2/18/04

8:16 PM

Page 42

Figure 8. The newly developed cutting groove spraying system guarantees an optimal dust reduction and heat suppression.

we are now able to compose the structure of cobalt tungsten, cobalt grains and cobalt perfectly – brick by brick – to a desired composition, grain size distribution and morphology which results in a pick without porosity or any other imperfections. Accompanied by FEM calculations, numerous lab tests and research on wear mechanism and thermophysical properties, the life expectance of the new cutting tools over former grades was more than doubled. Various real world site tests confirmed the results from theory and test rig.

results in a drastically reduced dust emission due to dust suppression directly at the source. Areas where the intermitted cutting trace spraying system is most beneficial are:

• • • •

Optimized respirable dust reduction due to the suppression at the dust source Minimum water consumption Minimal tool consumption due to efficient pick cooling Prevention of possible explosion due to cutting trace cooling to values below eventual ignition temperature.

Tests performed on a roadheader in underground conditions confirmed the effectiveness of the cutting groove spraying systems. The total dust concentration measured one meter behind the cutter head was found to be reduced by more then 50% with the spraying system activated during the cutting process. The total dust concentration at about 4 m behind the face was reduced from 143,6 mg/m3 at the reference system to 14,6 mg/m3 with the new, low speed cutting. 3.2

4 SUMMARY OF THE ICUTROC PROJECT AND ITS PRACTICAL IMPLICATIONS The most important outcome of Icutroc was the realization of a much more efficient and effective cutting process. Figure 10 shows the energy requirements of an Icutroc machine per unit (m3) of rock cut in comparison to an optimized existing system. It is clearly evident that Icutroc shows a significant reduction of the required specific energy using the new, low speed cutting process. The combination of higher pick forces at a low rotation speed and the stabilized and controlled guiding of the cutter arm resulted in a more economic

Development of new cutting tools

A complete new generation of cemented carbide grades – the so called “S-grades” – have been developed and patented by Sandvik. The concept means that

42 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-05.qxd

2/18/04

8:16 PM

Page 43

Figure 9. Newly developed cutting tools.

A combination of these outcomes was integrated into the latest generation of roadheaders within the product development project “Icutroc”.

handling of much harder material and a significant expansion of the expected cutting performance. Furthermore the low cutting speed reduced the tool consumption dramatically and led thereby to a more economic and efficient overall performance. Lower cutting speed resulted in significant lower temperatures at the pick tip, which reduced the weakening of the heat sensitive tungsten carbide pick tip, and increased pick lifetime accordingly. Another positive aspect of the lower cutting speed was the decreased impact impulse that does not exceed the mechanical strength of the pick tip anymore.

5 EXAMPLE OF AN ICUTROC APPLICATION IN THE REAL WORLD 5.1

Two VOEST Alpine ATM 105-IC roadheader extent the current line 2 from Henry Bourassa station to

43 Copyright © 2004 Taylor & Francis Group plc, London, UK

Operating results at the Metro Montreal

09096-05.qxd

2/18/04

8:16 PM

14 kWh/m3

Page 44

Approx. tool consumption (DP 8080)

6 kWh/m3

16 1 pick/m3

0.25 pick/m3

Speed energy kWh/m3

14 1

12

0.9

10

0.8

8

0.7 0.6

6

0.5

4

0.4

2

0.3 0.2

0 Reference system

New system

0.1 0 Reference system

Figure 10. Energy requirements per unit of cut rock.

Figure 12. Comparison of tool consumption between Icutroc and a traditional reference system, pick consumption was reduced to one fourth of the reference system (Material of about 160 Mpa and a cerchar abrasivity of 2.5).

Approx. cutting rate (incl. sump in and shearing) 28 m3/h

New system

35–50 m3/h

50

The material to be cut is predominantly limestone with an unconfined compressive strength between 46 Mpa and 140 Mpa. The average unconfined compressive strength over the whole range of material to be cut is 120 Mpa. The thickness of the fossiliferous and crystalline limestone as well as the shale insertions has varied within a wide range. Shale layers ranging from 0.1 inches to about 12 inches can practical be disregarded because of the scarcity of their occurrence. Additional the project faces massive limestone with very low to no bedding planes. Occasionally there are diabase dykes, with a diameter of up to 2 m and a hardness of 300 Mpa that need to be excavated as well. According to VOEST Alpine’s developed system a RMR of 30 was calculated which would enhance the net cutting rate calculated from the unconfined compressive strength test by a factor of 1,6 leading to an average cutting performance of 35 m3 and an average pick consumption of 0,1 pick/m3. Reality during one year of excavation proved even these very ambitious performance predictions a great underestimation. The red line indicates that the actual net cutting rate was ranging between 30 and 65 m3 per net cutting hour. That results in an average of 43,8 m3 per cutting hour, which is about 10–15% higher than predicted. The green line represents the pick consumption per m3 excavated rock, it can clearly be seen how

45 Cutting rate m3/h

40 35 30 25 20 15 10 5 0 Reference system

New system

Figure 11. Comparison of cutting rates between Icutroc and a reference system without the new technology (Material of about 160 Mpa and a cerchar abrasivity of 2.5).

Laval. The extension is 5.2 km and will be partially excavated by open cast. The last part of the metro extension will be the most challenging part of this metro excavation. This part involves the underground excavation of approximately 600 m of double lane tunnel running under a River, with an intersection to the single line tunnel. At the intersection area a maximum open span of over 16 m with a minimum 9 m rock coverage to the riverbed at a 34 m water head will require a sequential tunnel advance.

44 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-05.qxd

2/18/04

Sample #

8:16 PM

Page 45

BTS [MPa]

UCS [MPa]

Rock Type

CAI

from

to

mean

from

to

mean

from

to

mean

2001/086/01

Shale; C02 (TF-33-01, 88'3''– 90')

45,43

71,01

63,18

3,35

6,75

4,92

0,36

0,94

0,65

2001/086/02

Limestone – shaly, fossiliferous; C01 (TF-26-01, 82'2''– 84'3'')

63,55 134,78

102,26

4,81

10,89

6,81

0,55

1,15

0,82

2001/086/03

Limestone – slightly shaly, crystalline; C04 (TF-29-01, 96'8''– 98'6'')

88,09 117,05

105,87

5,11

9,81

7,50

0,29

0,84

0,62

2002/003/01

Limestone – fossiliferous, fine grained; C01 (TF-50, 33'5''– 34'3'')

93,12 117,10

102,14

4,58

9,28

7,37

0,44

0,75

0,63

2002/003/02

Limestone – crystalline, shaly; C01 (PF-2, 63'2''– 65')

58,08

92,84

78,25

4,12

5,49

4,98

0,59

0,82

0,68

2002/069/01

Diabase – dyke; C05 (TF-67, 19.25 –20.85m)

272,83 346,05

300,61

10,75

13,16

12,06

1,44

1,73

1,55

2002/069/02

Limestone – crystalline; C05 (TF-67, 57.0 – 58.2m)

63,77

82,62

70,20

3,31

7,91

5,92

0,68

1,00

0,82

0,45 0,40 0,35 0,30 0,25 0,20 0,15 0,10 0,05

Date

Daily NCR

Currently Average NCR

Daily SPC

Figure 14. Typical average cutting performance during one year at the job site.

45 Copyright © 2004 Taylor & Francis Group plc, London, UK

Currently Average SPC

30.04.03

27.04.03

24.04.03

21.04.03

18.04.03

15.04.03

12.04.03

09.04.03

06.04.03

03.04.03

31.03.03

28.03.03

25.03.03

22.03.03

19.03.03

16.03.03

13.03.03

10.03.03

07.03.03

04.03.03

01.03.03

26.02.03

23.02.03

20.02.03

17.02.03

14.02.03

11.02.03

08.02.03

05.02.03

02.02.03

30.01.03

27.01.03

24.01.03

21.01.03

18.01.03

0,00

Specific Pick Consumption [picks/solid m3]

95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 15.01.03

Net Cutting Rate [solid m3/h]

Figure 13. Summary of the rock test results at Montreal.

09096-05.qxd

2/18/04

8:16 PM

Page 46

the occurrence of diabase influenced the pick consumption. On April 21st one dyke reached up to 2 m on one part of the face, causing a pick consumption of 0,6 picks/m3 during that day. However, in general the pick consumption arrived at an average of 0,064 picks per m3, which is about 30% below the predicted values. In terms of actual daily productivity the machine was averaging 434 m3 per 2 times 10-hour shifts, which translates to nearly 10 m of daily face advance. The best daily advance was close to 16 m. The biggest advantages of the mechanical excavation in Montreal are the positive environmental impact

and the compliance with a nearly perfect profile by a largely avoidance of any over cut. While drill and blast had ongoing vibration problems exceeding vibration limits with hundreds of complaints filed by the community the roadheader operator had no complains at all. It was also recognised that the mechanical excavation does not damage the immediate rock integrity, it results in ground stability improvements and does not open vertical joint systems for ground water inflow. Because of the delicate excavation under the river, close to an existing station in operation, the project owner actually demanded the use of a roadheader for any further excavation work.

46 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-06.qxd

2/18/04

8:17 PM

Page 47

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

MTBM and small TBM experience with boulders S.W. Hunt & F.M. Mazhar MWH

ABSTRACT: Boulders present a significant and potentially costly challenge when mining small (less than 3 m) diameter tunnels. This paper provides an overview of methods generally used to cope with boulders and summarizes experiences with boulders during forty tunneling cases. The results indicate that microtunneling with drag cutters and no face access significantly increases the risk of a stuck machine. The addition of roller cutters or tunneling with face access reduces the risk of getting stuck. Use of multiple features for handling boulders result in the least risk of getting stuck. When specifying or selecting tunnel boring machine features, the cost and schedule consequences of getting stuck should always be considered.

The results of a thorough desk study should provide the framework for the next step in boulder characterization: site investigation.

1 BOULDER CHALLENGES 1.1

Overview

Boulder occurrence is an extremely important consideration for design and construction of pipelines with microtunnel boring machines (MTBMs) and small diameter tunnel boring machines (TBMs). When large boulders or abundant quantities are anticipated, a rugged MTBM or TBM features such as face access, roller cutters, and a durable cutting wheel are keys to reducing the risk of getting stuck or experiencing expensive problems. 1.2

1.2.2 Site investigation Site investigation for tunnel projects with a boulder risk should be phased and focused. Unless substantial previous site investigation data is available, the site investigation should not be a single phase of routinely spaced borings. Instead, it should be phased with each phase designed to reduce uncertainties determined from the previous phase, starting with the desk study results. The most appropriate methods of site investigation depend on the geology and uncertainties. Hunt & Angulo (1999) discuss subsurface exploration methods for identification of boulders and cite most of the available pertinent references at that time. Since 1999, several additional papers involving site investigation for boulders have been published. Frank & Daniels (2000) discuss use of ground penetrating radar for boulder identification. Frank & Chapman (2001) describe how geologic setting, site reconnaissance, conventional borings, roto-sonic borings and large diameter (0.9–1.2 m) auger borings were used to characterize cobble and boulder conditions for the Big Walnut Augmentation/Rickerbacker Interceptor tunnel project in Columbus, Ohio. An important conclusion reached or implied by both Hunt & Angulo (1999) and Frank & Chapman (2001) is that a site investigation relying on one method, e.g. conventional borings, is unlikely to be sufficient unless there is significant previous local tunneling

Boulder characterization

Boulder characterization is essential to proactive management of tunneling risk in bouldery ground. A concise discussion of the main characteristics of boulder occurrence is given in Hunt & Angulo (1999). Boulder characterization generally requires both a thorough desk study and a focused site investigation program. 1.2.1 Desk study Boulder characterization should start with a thorough desk study that:

• • •

Determines the geologic setting and character of units and formations that may be encountered. Obtains available previous pertinent site investigation data in the project area. Finds and assesses available local tunneling case histories for cobble, boulder and abrasive ground conditions and impacts.

47 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-06.qxd

2/18/04

8:17 PM

Page 48

made of many types and materials, but can be generally grouped into two categories – roller cutters and drag cutters (Cigna & Ozdemir, 2000). Roller cutters are generally used for full-face hard rock, mixed-face and cobbly-bouldery ground and include:

experience in similar geologic conditions. Furthermore, a single phase of site investigation is also unlikely to be as successful or cost-effective as a phased investigation where during later phases increasingly more specific subsurface investigation and sampling methods are utilized. 1.3

• • • •

Boulder baselining

Boulder baselining is more than boulder characterization. The data and conditions found during the desk study and site investigation need to be converted into anticipated boulder frequencies, sizes and matrix conditions for the planned tunneling methods. Hunt & Angulo (1999) attempted to use probabilistic methods to relate boulder indications from borings to boulder encounters in tunnels, but did not have success with the method. Instead, they developed a semiempirical method that correlates boulder conditions encountered in borings and tunnels for similar local geologic units. Hunt (2002) described several case histories where the method was used since 1999 and concluded that “Boulder quantities can be reasonably estimated if boulder occurrence records from similar geologic units are correlated with indications of boulders from properly logged borings.” Other methods of quantifying boulder occurrence are discussed in Hunt & Angulo (1999). More recent papers have not presented specific methods for quantifying and baselining boulders, but have provided significant useful data on boulder conditions encountered in tunnels. DiPonio et al. (2003) discuss the occurrence of 2169 boulders within approximately 12.8 km of 2.1 and 2.3 m ID tunnels in the Detroit, Michigan area. Cronin & Coluccio (2003) discuss the occurrence of 34,300 boulders within approximately 2.5 km of 3.7 m ID tunnels in Portland, Oregon. Both papers provide useful data on frequency and size distribution of the boulders. 1.4

Single disk cutters Single and multiple mini-disc cutters Strawberry (button) roller cutters Multiple row carbide insert cutters

Examples are shown in Figure 1. Drag cutters are generally more efficient at mining fine-grained soil without cobbles, boulders and hard rock layers. Drag cutters include:

• • • • •

Chisel teeth Block scrapers Plow teeth Pick (bullet) teeth Blade cutters Examples are shown in Figure 2.

single disk cutter

double disk cutter

triple disk cutter with carbide inserts

carbide button cone cutter

two row carbide insert cutter

five row carbide insert cutter

Figure 1. Examples of roller cutters.

TBM features

TBM features substantially influence the potential impact of boulder occurrence. The features of most relevance include cutter types, face access, cutting wheel opening size, mucking system and cutting wheel torque and thrust. Other relevant features include: rock crusher type, cutting head armor (abrasion resistance), and ability to retract the cutting head 10–20 cm or more if to reestablish cutting wheel rotation. 1.4.1 Cutters Cutter type and configurations are extremely important TBM features that affect advance rate productivity, ability to excavate and fracture boulders, and extent of cutter and cutting wheel wear. Cutters are

chisel teeth on cutter arm

blade cutter

plow tooth

pick or bullet teeth

block scrapers

Figure 2. Examples of drag cutters.

48 Copyright © 2004 Taylor & Francis Group plc, London, UK

chisel tooth cutter

09096-06.qxd

2/18/04

8:17 PM

Page 49

hold boulders in place during fracturing. In soft or loose ground, boulders may become dislodged and roll around on the cutting wheel causing cuter damage and high face abrasion until fractured small enough for ingestion. If the ground is soft or loose enough, partially fractured boulders may pushed aside rather than ingested. The matrix shear strength issue has been discussed by a number of authors including: Navin et al. (1995), Becker (1995), Dowden, & Robinson (2001), and Goss (2002). Goss completed a PhD thesis on this subject and summarizes the issues and his results in the referenced 2002 paper. Goss concluded that the two most important parameters are the actual in-situ shear strength of the soil matrix and the shear or unconfined compressive strength of the boulders. Based on finite element modeling and case history evaluations, Goss suggested that: “…when the rockto-soil shear strength ratio is greater than 600:1, boulders cannot be broken by disc cutters.” Goss provided data on 18 case studies that generally support this conclusion. The case histories that are summarized subsequently in this paper suggest that disc and multiple row carbide insert cutters are essential to minimization of the risk of getting stuck on microtunneling and small TBM drives where there is no face access or access is problematic due to ground and groundwater conditions. This data will show that matrix shear strength was generally not the most important factor causing stuck TBMs. Although lower productivity and higher cutter wear were generally experienced when boulders were encountered in soils with low matrix shear strength, rugged TBMs with roller cutters were generally able to advance through bouldery ground even if the matrix soil was soft or loose.

A combination of roller and drag cutters are often used to maximize mining flexibility and TBM performance. Drag cutters are generally preferred for tunneling in mostly fine-grained soils because of higher cutting efficiency, however, drag cutters are vulnerable to rapid wear and breakage when mining through abrasive cobbly and bouldery ground. Furthermore, drag cutters are generally not capable of fracturing large boulders with unconfined compressive strengths over approximately 100 Mpa. In order to help minimize drag cutter wear Dowden and Robinson (2001) recommended that disc cutters be included whenever larger numbers of hard boulders are anticipated. A decision to include roller cutters for mining in cobbly or bouldery ground should not only depend on TBM productivity and cutter wear, but also on the risk and consequences of getting stuck. The risk of getting stuck depends on boulder size, cobble and boulder concentrations and distribution (nested vs. scattered), rock strength, TBM diameter, cutting wheel opening size, rock crusher capability, face access and method of face pressurization (slurry shield, earth pressure balance by pressure relieving gates, earth pressure balance by screw auger). Nishitake (1987) discusses roller cutter types and configurations and presents the results of tests on cutting effectiveness. Specific recommendations for earth pressure balance machine features in bouldery ground are given. Friant and Ozdemir (1994) discuss the development of mini-disc cutters for microtunneling and small diameter TBM tunneling and compare their effectiveness to other drag and roller cutters. A study of cutter efficiency at the Colorado School of Mines (Ozdemir, 1995) resulted in a conclusion by Ozdemir that: “In summary, the mini-disc cutter offers many significant advantages over any other type of cutting tolls currently used on microtunneling machines. These advantages include very high cutting efficiency: high penetration rates; low machine thrust, torque and power requirements; excavation capability in any type of soil and hard rock:* low initial cost: low replacement costs; ease of replacement and the elimination of the need for a cutter shop; true-rolling feature (meaning reduced torque and power requirements compared to button or multi-kerf cutters); greater lifetime and drive lengths compared to carbide cutters; and significantly reduced fines, meaning less slurry cleanup requirements.”

2 CASE HISTORIES 2.1

1.4.2 Disc cutter effectiveness in soil Cutter configuration and choice of both roller and drag cutters will significantly impact success at mining through bouldery ground. One concern with use of disc and multiple row carbide insert cutters is the soil matrix shear strength or density and its ability to

49 Copyright © 2004 Taylor & Francis Group plc, London, UK

Data

A search of tunneling case history articles, papers and project files resulted in 40 cases from 36 projects for this study. Extracted data for these 40 cases are listed in Table 1 (sheets 1a-1h). The primary references for the cited data are listed in the bottom row. These references may have additional information of interest. Some of the references lack all the desired pertinent data. In many cases boulder quantities and sizes were not thoroughly reported, often because documentation of boulder quantities and sizes is not practical (Hunt, 2002). Where boulder quantities and sizes were estimated by the Authors based on an interpretation of reported information, this data is preceded with an asterisk (*).

09096-06.qxd

2/18/04

8:17 PM

Page 50

Table 1a. Case histories 1a–4. Case no.

1a

1b

2

3

4

Location

Clearview Snohomish River Undercrossing, 1st Attempt – Snohomish, WA SJP Alluvial sand, gravel and cobbles

Clearview Snohomish River Undercrossing, 2nd Attempt – Snohomish, WA SJP Alluvial sand, gravel and cobbles

Tolt Pipeline No. 2, Snoqualmie River – Seattle, WA SJP Sand, gravel, and cobbles

Tolt Pipeline No. 2, Bear Creek – Seattle, WA

Swamp Creek – Snohomish County, WA

SJP Silty clay till, sand and gravel outwash

RCP Glaciolacustrin silty clay to beach sand

Loose to dense

Loose to dense

Loose to dense

Hard to dense

21.3

21.3

24.4

5.5

Hard to med. dense 7.6

22.9–33.5

22.9–33.5

25.9

10.7

10.1

339.9 339.9

339.9 339.9

652.3 213.4

113.4 113.7

315.5 169.2

177.4 1575

339.9 1575

213.4 2286

113.4 1905

108.2 1219

*8

*17

*400

50

15

1 *0.30%

*2 *0.10%

* 80 *2.50%

10 0.90%

4 0.20%

1016

*508

914

508

1219

65% Igneous and metamorphic

32% Igneous and metamorphic

27% Igneous and metamorphic

100% Igneous and metamorphic

STBM Isecki Unclemole None Block scraper drag cutters

STBM Lovat MTS

40% Igneous and metamorphic 234.4 STBM Soltau RVS 800 STS None Triple disc roller cutters and chisel teeth Cutters and cone crusher

STBM Soltau RVS 600 None Five row carbide insert roller and strawberry cone cutters Cutters and cone crusher

STBM Isecki TCC 1000 Yes Block scraper drag cutters

Jacked pipe type Primary soil type (with varying quantities and sizes of boulders) Soil matrix Consistency/density Maximum groundwater head, m Approx. tunnel depth, m Tunnel length, m Selected drive length, m Length achieved, m Excavated diameter (ED), mm No. of boulders in drive No. of large boulders Est’d % boulders (by Vol) Max. boulder Size mm Boulder size/ED, % Rock types Rock Qu, MPa TBM type TBM make Face access TBM cutters

Fracture method

Cone crusher

TBM advance Impact Cutter, cutterhead Damage

STBM stuck and abandoned Bits broken and worn

Remedial Measures

Sank new shaft, re-mined w/new STBM

Risk of getting stuck Consequence of stuck References

Very high $3 million Staheli, 2003

None Multi-row carbide insert roller cutters and chisel teeth Cutters and cone crusher None Considerable cutterhead surface wear NA

Drag teeth badly worn, disc cutters ok NA

Minor

Low $3 million Staheli, 2003

Low $1 million Molvik et al., 2000

Low $1 million Molvik et al., 2000; Beieler et al, 2003

* Estimated from best available data – reliability uncertain.

50 Copyright © 2004 Taylor & Francis Group plc, London, UK

Cone crusher EPBM w PRG stuck at 355 ft Major-one cutter arm torn off Recovery tunnel mined to finish drive, remove EPBM Very high $400,000 Genzlinger, 1995

09096-06.qxd

2/18/04

8:17 PM

Page 51

Table 1b. Case histories 5–7. Case no.

5

6

7a

7b

7c

Location

Mercer St. Railroad Crossing TunnelsSeattle, WA RCP Glaciolacustrine silty clay

NEHLA undershore crossing, Keahole Point, Hawaii SJP Coarse sand and gravel with cobbles

Folsom east IIB Test A Sacramento, California RCP Alluvial sand, gravel and cobbles

Folsom east IIB Test B Sacramento, California RCP Alluvial sand, gravel and cobbles

Folsom east IIB Sewer Sacramento, California RCP (2 pass) Alluvial sand, gravel and cobbles

Hard

Med. dense

Loose to dense

Loose to dense

Loose to dense

12.2

13.7

0.0

0.0

0.0

15.2

13.7

10.7

10.7

24.4

213.4 106.7

353.6 176.8

304.8 152.4

304.8 152.4

906.8 431.6

106.7 3048

176.8 1746

61.0 1778

26.2 1778

431.6 3048

2

*10

*32

*72

*1900

1 *0.03%

*2 *0.10%

*7 *4.00%

*5 *3.00%

*300 *2.00%

*762

*610

889

610

889

25% Igneous and metamorphic

35% Basalt

50% Hard igneous

34% Hard igneous

29% Hard igneous

138–270 STBM Soltau RVS 800 AS None Multi-row carbide insert roller cutters Button bits and cone crusher

186–268 OFRW Akkerman

186–269 STBM Akkerman

Yes Chisel teeth

None Chisel and bullet teeth

Passed through head

Cone crusher

TBM advance Impact

STBM jammed/ stuck 6 times

STBM stuck, replaced by OFRW TBM

Cutter, cutterhead Damage

Minor

Remedial measures

Cutting wheel freed by retracting head 10 cm High $100,000 Smith, 1995: Miller, 1995a

OFRW stuck, replaced by another OFRW TBM Severe wear of drag cutters, cutterhead TBM removed in rescue shaft, drive ended High $100,000 Staheli et al., 1999

186–276 OFRW Lovat 121 PJ/RL -8600 Yes Chisel, block scraper and carbide bullet teeth Jackhammer for occasional v. large boulders Periodic cutter replacement

Jacked pipe type Primary soil type (with varying quantities and sizes of boulders) Soil matrix consistency/density Maximum groundwater Head, m Approx. tunnel depth, m Tunnel length, m Selected drive Length, m Length achieved, m Excavated diameter (ED), mm No. of boulders in drive No. of large boulders Est’d % boulders (by Vol) Max. boulder size mm Boulder size/ED, % Rock types Rock Qu, MPa TBM type TBM make

EPBM-PRG Lovat M-102

Face access TBM cutters

None Chisel teeth

Fracture method

Hyd. Splitting, most passed through head,

Risk of getting stuck Consequence of stuck References

Very low $500,000 Genzlinger, 1995

51 Copyright © 2004 Taylor & Francis Group plc, London, UK

Severe wear of drag cutters, cutterhead STBM removed in rescue shaft, drive ended Very high $100,000 Staheli et al., 2000

Severe wear of cutters, cutterhead

Very low $500,000 Castro et al., 2001

09096-06.qxd

2/18/04

8:17 PM

Page 52

Table 1c. Case histories 8–12. Case no.

8

9

10

11

12

Location

Carmichael Water Crossing of American River – Carmichael, CA RCP Alluvial sand and gravel with cobbles

CCWA Santa Ynez River Crossing, Mission Veijo, CA

Peralta Blvd Sanitary Sewer, Fremont, CA PCP Mixed clay, gravel and cobbles

North Mission Valley Interceptor, San Diego, CA

SJP Claystone, silty sand and cobbles

Dense

Hard to dense

Pacific Coast Highway Sewer, Phase II, Santa Monica, CA RCPP Sand and gravel with cobbles and some clay seams Soft to stiff

Stiff

Dense to hard

15.2

11.6

3.0

27.4

16.5–23.5

6.7

4.9–6.1

6.7

219.5 219.5

365.8 248.4

1,036.3 365.8

914.4 128.0

1,402.1 85.3

219.5 1524

248.4 1549

365.8 1930

128.0 635

21.3 780

*10

*5

*200

*16

*100

*1

*1

0

0

*20

*0.10%

*0.10%

*0.50%

*1.00%

*1.00%

*457

1549

381

*457

457

30% Hard igneous 186–269 STBM Soltau RVS 600 A-S None Block scraper and carbide bullet teeth

100% Igneous 241.3 STBM Soltau RVS 600 A5 None Multi-row carbide insert roller cutters and button cones Cutters and cone crusher

20% Igneous

72% Hard igneous

STBM Herrenknecht AVN 1500T Yes Chisel teeth and disc cutters

STBM Akkerman

59% Igneous 69 STBM Isecki TCC600

None Chisel and bullet teeth

None Block scraper drag cutters

Cutters and cone crusher

Cone crusher

Cone crusher STBM stuck 3 drives

Jacked pipe type Primary soil type (with varying quantities and sizes of boulders) Soil matrix consistency/density Maximum groundwater head, m Approx. tunnel Depth, m Tunnel length, m Selected drive Length, m Length achieved, m Excavated diameter (ED), mm No. of boulders in drive No. of large boulders Est’d % boulders (by Vol) Max. boulder Size mm Boulder size/ED, % Rock types Rock Qu, MPa TBM type TBM make Face access TBM cutters

Fracture method

Cone crusher

TBM advance impact

VCP Cemented silty sand, silty clay and cobbles

3.7

Cutter, cutterhead damage Remedial measures

Moderate

Minor

Minor

Steering and grade difficulties from boulders Moderate

NA

NA

NA

NA

Risk of getting stuck Consequence of stuck References

High

Low

Low

High

2 recovery shafts, 1 recovery trench Very high

}

Over $500,000

$100,001

$100,000

$100,000

Castro et al., 2001

Miller, 1996

Rush, 2000; Rush 2002

Miller, 1997a

Miller, 1997b

52 Copyright © 2004 Taylor & Francis Group plc, London, UK

Moderate

09096-06.qxd

2/18/04

8:17 PM

Page 53

Table 1d. Case histories 13–16. Case no.

13

14

15a

15b

16

Location

Deicing Fluid Line, Denver Int. Airport – Denver, CO SJP Hard silty clay and layered soft shale

CSO Separation RNC 5318 – Omaha, NE

Lincoln Way Drive 1 -Ames. IA

Lincoln Way Drive 2 -Ames. IA

MN-320 Sewer Minneapolis, Minnesota

RCP Alluvial sand and gravel

RCP Silty clay till w sandy gravel outwash

RCP Silty clay till w sandy gravel outwash

FMCP Silty clay till, gravelly sand outwash

Hard

Loose-medium dense 7.6

Very stiff to hard 9.1

Very stiff to hard 9.1

Very stiff 4.6

9.8

12.2

3.7–12.2

3.7–12.2

10.4–12.2

316.4 241.7

231.6 231.6

929.0 111.6

929.0 132.3

431.3 253.0

241.7 1092

231.6 2159

103.9 787

132.3 787

140.8 1524

*10

*10

*10

17

16

0 *0.05%

*5 *1.00%

*2 *0.70%

3 *1.00%

6 2.00%

356

762

610

610

1372

33% Igneous

35% Igneous

77% Granite, other igneous

77% Granite, other igneous

90% Gabbro, granite

Rock Qu, MPa TBM type TBM make

OFRW Akkerman

OFRW Akkerman 720

Yes Chisel teeth

Yes Chisel teeth

Fracture method

Pneumatic hand tools 10 hour delay

Blasting

Cone crusher

Stuck on large boulder at 10 ft advance

Stuck at 341 ft

Major cutter wear and broken teeth Recovery trench, added disc cutters Very high $100,000 Schumacher & Ellis, 1997; Najafi & Varma, 1996

STBM Herrenknechht AVN600 None Double-row carbide insert and disk cutters and plow teeth Cutters and cone crusher STBM w rock head successful – stuck at launch using head with drag cutters Major cutter wear and broken teeth Pulled back TBM, changed to head with disc cutters Low $100,000 Schumacher & Ellis, 1997; Najafi & Varma, 1996

STBM Akkerman

Face access TBM cutters

STBM Herrenknechht AVN600 None Chisel teeth

Jacked pipe type Primary soil type (with varying quantities and sizes of boulders) Soil Matrix consistency/density Maximum groundwater head, m Approx. tunnel depth, m Tunnel length, m Selected drive length, m Length achieved, m Excavated diameter (ED), mm No. of boulders in drive No. of large boulders Est’d % boulders (by Vol) Max. boulder size mm Boulder size/ED, % Rock types

TBM advance impact

Cutter, cutterhead damage

Minor

Minor

Remedial measures

Man access to face to split, dislodge boulders Low $5,000 Coss, 1993

Man access to face to blast boulders Moderate $10,000 King et al., 1997

Risk of getting stuck Consequence of stuck References

53 Copyright © 2004 Taylor & Francis Group plc, London, UK

None Chisel and bullet teeth

Cone crusher STBM stuck twice, 2 rescue shafts Half of teeth gone, rest very worn Cuttingwheel replaced once, TBM replaced Very high $3 million Hunt, 2003

09096-06.qxd

2/18/04

8:17 PM

Page 54

Table 1e. Case histories 17–21. Case no.

17

18

19

20

21

Location

Menomonee River Water Pipeline – Marrinette, Wisconsin SJP Silty clay till w outwash pockets

Oklahoma Ave. Relief Sewer – Milwaukee, WI

Ramsey Ave. Relief Sewer – Milwaukee, WI

Miller 37th & State Area MIS – Milwaukee

S. Penn. Ave. Relief Sewer – Oak Creek, WI

RCP Silty clay till w pockets of outwash sand Very stiff

12.2

7.6

12.2

RCP Silty clay till and cobbly sand and gravel outwash Very stiff – hard to very dense 3.0

RCP Silty clay till w thick layers of outwash

Very stiff

RCP Silty clay till w outwash pockets & boulder clay Very stiff

10.4–15.2

10.7–14.9

6.1–19.8

15.2–21.3

6.1–12.1

268.2 268.2

874.8 874.8

941.2 201.5

390.4 405.4

2,497.8 179.8

268.2 1422

874.8 1537

195.1 997

405.4 2642

179.8 1397

7

151

7

346

160

2

71

1

60

21

0.10%

0.40%

0.40%

1.60%

2.50%

762

762

610

1067

914

54% Hard igneous

50% Igneous erratics and dolomite

40% Igneous erratics and dolomite

STBM Soltau RVS 600 None Multi-row carbide insert roller cutters and chisel teeth Disc cutters and cone crusher

OFRW Decker Yes Chisel teeth

61% Gabbro and dolomite 206.8 STBM Isecki Unclemole None Block scraper drag cutters

65% Igneous erratics and dolomite 206.8 OFRW Decker Yes Chisel teeth

Cone crusher

Blasted boulders 60 times Hand mined from cut wheel twice to replace cutters

Jacked pipe type Primary soil type (with varying quantities and sizes of boulders) Soil matrix consistency/density Maximum groundwater Head, m Approx. tunnel depth, m Tunnel length, m Selected drive length, m Length achieved, m Excavated diameter (ED), mm No. of boulders in drive No. of large boulders Est’d % boulders (by Vol) Max. boulder size mm Boulder size/ED, % Rock types Rock Qu, MPa TBM type TBM make Face access TBM cutters

Fracture method

Blasting

OFRW Modifed Decker Yes Chisel teeth

Very stiff to hard or dense 13.7

Severe – wear and broken drag teeth NA

Very low

Moderate – worn, broken drag teeth Recovery tunnel mined to finish drive and remove STBM High

Blasted boulders 177 times Slower production and delays for blasting boulders Moderate – worn, broken drag teeth NA

Low

Low

$1 million

$100,000

$600,000

$100,000

$10,000

Vadnais, 2002

Hunt, 2002

Hunt, 1999

Hunt, 2002

Hunt, 1999

TBM advance Impact

Stuck at 640 ft

Cutter, cutterhead Damage

Minor cutter wear

Minor cutter wear

Remedial measures

NA

NA

Risk of getting stuck Consequence of stuck Reference nos.

Low

54 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-06.qxd

2/18/04

8:17 PM

Page 55

Table 1f. Case histories 22–26. Case no.

22

23

24

25

26

Location

Oak Creek Southwest Relief Sewer – Oak Creek, WI RCP Silty clay till with cobble and boulder lag zone Very stiff to hard or dense 15.2

Elgin Interceptor – Elgin, IL

Northeast Interceptor Sewer – Libertyville, IL

Weller Creek Sewer Project – Arlington Hts, IL

Evanston CSO Phase IV – Evanston, IL

RCP Silty clay till and outwash sand and gravel Very stiff to hard or dense 12.2

FMCP Silty clay till w sandy gravel outwash Very stiff to hard 10.7

RCP Silty clay to clayey silt till

RCP Silty clay to clayey silt till

Very stiff to hard 12.2

Very stiff to hard 9.1

16.8

6.1–14.3

10.7–14.0

15.2

11.6–12.2

1,524.0 1,524.0

1,313.1 405.4

929.0 152.4

3,116.0 393.2

335.3 182.9

1,524.0 1397

405.4 1981

152.4 1397

393.2 3353

182.9 2159

156

45

*10

*20

*5

80

20

*2

*5

*2

1.60%

1.60%

*0.18%

*0.04%

*0.20%

1219

1016

*914

*762

*1067

87% Igneous erratics and dolomite

51% Igneous erratics and dolomite

65% Igneous erratics and dolomite

23% Igneous erratics and dolomite

49% Igneous erratics and dolomite

Rock Qu, MPa TBM type TBM make

OFRW Decker

OFRW Lovat

OFRW Lovat

Face access TBM cutters

Yes Chisel teeth

Yes Chisel teeth

OFRW Akkerman 720C Yes Chisel teeth

Fracture method

Blated boulders 80 times

Blasting

STBM Herrenknechht AVN 1200 None Chisel teeth and block scraper drag cutters Cutters and cone crusher

TBM advance impact

Slower production and delays for blasting boulders Moderate – worn, broken drag teeth NA

Stopped 24 times to blast large boulders

Cone crusher

Moderate – worn, broken drag teeth NA

Low

Moderate

Moderate – worn, broken drag teeth Relief shafts bored, large boulders below MTBM High

$50,000

$100,000

Hunt, 2002

Hunt, 1999

Jacked pipe type Primary soil type (with varying quantities and sizes of boulders) Soil matrix consistency/density Maximum groundwater head, m Approx. tunnel depth, m Tunnel length, m Selected drive length, m Length achieved, m Excavated diameter (ED), mm No. of boulders in drive No. of large boulders Est’d % boulders (by Vol) Max. boulder size mm Boulder size/ED, % Rock types

Cutter, cutterhead damage Remedial measures

Risk of getting stuck Consequence of stuck Reference nos.

55 Copyright © 2004 Taylor & Francis Group plc, London, UK

Yes Chisel teeth

Most past through, blasted large boulders

Hydraulic split large boulders

Minor cutter wear

Minor cutter wear

NA

NA

Low

Low

$25,000

$100,000

$100,000

Rickert et al., 1999, Westcon.net

Hunt, 1999

Miller, 1995b

09096-06.qxd

2/18/04

8:17 PM

Page 56

Table 1g. Case histories 27–31. Case no.

27

28

29

30

31

Location

Marathon Oil Storm Sewer – Indianapolis, IN

Downriver DRSTS Contract No. 4 – Wayne Co. MI

PSE&G Power Tunnel, Jersey City, NJ

Chelsea River Crossing – Boston, MA

Jacked pipe type Primary soil type (with varying quantities and sizes of boulders) Soil matrix consistency/density Maximum groundwater head, m Approx. tunnel depth, m Tunnel length, m Selected drive length, m Length achieved, m Excavated diameter (ED), mm No. of boulders in drive No. of large boulders Est’d % boulders (by Vol) Max. boulder size mm Boulder size/ED, % Rock types

FMCP Hard silty clay till w sandy gravel outwash

RCP Silty clay till and hardpan till with cobbles

SJP Clayey silt and sand and rubble fill

SJP Silty clay till and alluvial silty sand

New Neponset Valley Force Main – Canton, MA RCPP Sand, silty sand, some gravel, cobbles, peat

Hard to very hard or very dense 3.7

Very stiff to very hard 15.2

Stiff

Dense

9.1

Hard to very dense 24.4

6.7–7.6

18.3

10.7–11.3

7.3–10.4

8.2

684.3 239.9

4,572.0 4,572.0

283.5 94.5

299.9 299.9

430.1 430.1

239.9 1758

4,572.0 2642

94.5 965

299.9 1961

430.1 2108

8

995

*5

32

5

3

283

*1

6

0

0.17%

0.20%

*0.50%

*0.15%

0.01%

610

1219

*508

914

457

35% Granite, other igneous

46% Igneous erratics and dolomite

53% Igneous, Metamorphic

47% Granite, other igneous

22% Granite, other igneous

OFRW Akkerman

OFRW Lovat M-120

STBM Soltau RVS 600

EPBM -PRG Lovat M77

Face access TBM cutters

Yes Chisel and bullet teeth

Yes Chisel and bullet teeth

STBM Soltau RVS 250A/S No Chisel teeth

Yes Chisel and bullet teeth

Fracture method

Hydraulic split large boulders

Blasted boulders 10 times – most past through

Disc cutters and cone crusher

Yes Multi-row carbide insert cutters, bullet and chisel teeth Disc cutters and cone crusher

Minor NA

Moderate – worn, broken drag teeth NA

Minor cutter wear NA

Minor cutter wear NA

Low

Low

Minor cutter wear Relief shaft needed to remove Hand timber piles High

Low

Low

$100,000

$100,000

$100,000

$1 million

$100,000

Garrett, 1992

DiPonio, et al., 2003 Miller, 1994

Tarkoy, 2001

Boscardin, 1997

Rock Qu, MPa TBM type TBM make

TBM advance impact Cutter, cutterhead Damage Remedial measures

Risk of getting stuck Consequence of stuck Reference nos.

56 Copyright © 2004 Taylor & Francis Group plc, London, UK

4.6

All boulders passed through

09096-06.qxd

2/18/04

8:17 PM

Page 57

Table 1h. Case histories 32–36. Case no.

32

33

34

35

36

Location

Kelvin Valley Wastewater, Scotland UK

Sudden Valley Sewer, Lancashire UK

East Dock Relief Sewer, Dundee, Scotland UK

Carronvale Sewer, Scotland UK

Jacked pipe type Primary soil type (with varying quantities and sizes of boulders) Soil matrix consistency/ density Maximum groundwater head, m Approx. tunnel depth, m Tunnel length, m Selected drive length, m Length achieved, m Excavated diameter (ED), mm No. of boulders in drive No. of large boulders Est’d % boulders (by Vol) Max. boulder size mm Boulder size/ED, % Rock types

SJP Silty sand w cobbles, schist and sandstone

FCMP Silty sand and gravel w cobbles and clayey zones

GRP Silty clay, sand and gravel over sandstone, basalt

RCP Fill, clay, silt and fine sand

Neva River Undercrossing – St. Petersburg, Russia RCP Silty fine sand and gravel, some clayey zones

Hard

Dense

Stiff to dense

6.0

8.0

Soft to stiff and loose to medium dense 4.0

20.1

9.1–12.2

10.0

10.0

5.0

25.0

438.3 178.9

570.0 570.0

930.0 301.0

132.0 132.0

774.0 774.0

178.9 1956

570.0 1460

264.0 2200

132.0 900

774.0 2540

*5

*100

*20

*9

*25

0

0

*5

1

*5

*0.02%

*0.50%

*0.50%

*0.50%

*0.03%

406

330

762

610

508

21% Schist, other

23% Igneous

35%

68%

20% Granite, other igneous

STBM Wirth-Soltau BH1920/1600 Yes Triple disc cutters and chisel teeth

EPBM Markham OKMS

STBM Herrenknecht AVN 1600 Yes Disc cutters and chisel teeth

EPBM Howden-716

STBM Herrenknechht AVN 2000D Yes Double disc cutters, chisel and plow teeth Disc cutters and cone crusher

Rock Qu, MPa TBM type TBM make Face access TBM cutters

Fracture method

Yes Disc cutters, bullet and chisel teeth

Disc cutters and crusher arms

Disc cutters

Cutter, cutterhead damage Remedial measures

Minor cutter wear NA

Minor cutter wear NA

Risk of getting stuck Consequence of stuck Reference nos.

Low

Yes Disc cutters and chisel teeth

Minor cutter wear NA

NA

Low

Disc cutters and cone crusher Stuck at 866 m in nested cobble and boulder backfill Minor cutter wear Relief shaft sunk to remove nested cobbles and boulders Moderate

Moderate

Medium

$500,000

$100,000

$100,000

$500,000

$1 million

Gehlen & Huhn, 2002

Jones, 1990

Fleet & Owen, 2002

Clarke, 1990

Wallis, 2002

TBM advance impact

57 Copyright © 2004 Taylor & Francis Group plc, London, UK

Plenum rock crusher

Loose to dense or stiff

Minor

09096-06.qxd

2/18/04

8:17 PM

Page 58

The following abbreviations were used in the table for jacked pipe type: SJP RCP RCPP VCP FMCP PCP GRP

all cases 30% stuck = 12/40

Steel jacking pipe Reinforced concrete pipe Reinforced concrete pressure pipe Vitrified clay pipe Fiberglass polymer composite pipe Polymer concrete pipe Glass resin pipe

drag cutters 37% stuck = 10/27 roller or roller and drag cutters 15% stuck = 2/13 drag cutters

The following abbreviations were used in Table 1 for tunnel boring machine (TBM) type: STBM OFRW EPBM EPBM-PRG

67% stuck = 8/12

Slurry tunnel boring machine Open face rotary wheel machine Earth pressure balance machine w screw Earth pressure balance machine with pressure relieving gate (PRG)

20% stuck = 2/10

roller or roller and drag cutters

drag cutters 13% stuck = 2/15 roller or roller and drag cutters

2.2

Analysis of case history data

0% stuck = 0/3

Table 1 summarizes data from 40 microtunneling and small diameter tunneling cases where boulders were encountered. The following paragraphs provide an analysis of the data.

Figure 3.

OFRWs, EBPMs 11% stuck = 2/18

Distribution of stuck TBMs.

pressure relieving gates (PRGs) and face access. None of the TBMS for these cases became stuck. Of the 27 cases involving tunneling with only drag cutters, 13 cases involved open-face rotary wheel machines (OFRW) with face access. Two of the OFRW machines or 13 percent became stuck (Cases 7a and 14). In the Case 7a, the drag cutters and cutting wheel of the OFRW were so badly damaged by cobbles and boulders upon becoming stuck that the TBM was removed from the project (Staheli et al., 1999). In Case 14, a very large boulder was encountered. The stuck OFRW and boulder were removed with a recovery trench rather than by splitting or blasting the boulder. Subsequently, roller disc cutters were added to the cutting head for the remaining tunnel drives. In the other 11 OFRW cases, face access allowed any large boulders that would not pass through the cutting wheel and mucking system to be blasted or split ahead of the cutting wheel. In general, the ground at these headings had sufficient stand-up time for face access without air pressure. In some cases, grouting or localized dewatering was required for face access. While blasting or splitting boulders through face access generally prevented the TBMs from getting stuck, but it did not prevent cutter damage. In some cases, No. 20 for example, chisel cutters were so badly worn and broken by cobbles and boulders during each of two 340–350 m long drives that hand-mining in front of the cutting wheel was required to replace chisel cutters before pipe jacking could proceed. Fortunately, increased side friction from set-up during cutter repairs was not sufficient to cause stuck drives.

2.2.1 Stuck STBMs For the case histories evaluated in this paper, the tunnel boring machines were stuck (unable to advance without intervention) a total of 12 times in 40 cases (30 per cent overall stuck rate). Figure 3 shows a graphical summary of the results for stuck TBMs. Slurry shield tunnel borings machines (STBMs) were stuck during 10 of the 40 cases (a 25 percent stuck rate) and during 10 of the 22 cases where STBMs were used (45 percent stuck rate). Of the 10 stuck cases, 8 involved STBMs with drag cutters only. The other two cases were STBMs that had some roller cutters. The STBM for one of the stuck cases had only roller cutters (Case 6). This STBM was jammed by cobbles and boulders six times, however, cutting wheel retraction capability of about 10 cm allowed the head to be freed each time resuming roller cutter fracturing and allowing the drives to be finished without other intervention (Miller, 1995a). The other stuck STBM had combined drag and roller cutters (Case 34). This STBM encountered a fill with nested cobbles and boulders that required a relief shaft to remove the obstructions. Twenty-seven of the 40 cases involved tunneling with drag cutters only. Of these 27 drag cutter only cases, 12 cases involved tunneling with STBMs. The STBMs with only drag cutters became stuck during 8 of these 12 cases – a 67 percent stuck rate. Of the 27 cases involving tunneling with only drag cutters, 2 cases were mined with EPBMs having

58 Copyright © 2004 Taylor & Francis Group plc, London, UK

STBMs 45% stuck = 10/22

09096-06.qxd

2/18/04

8:17 PM

Page 59

Thirteen of the 40 cases in Table 1 involved tunneling with roller cutters only or a combination of roller and drag cutters. Ten of these 13 cases involved STBMs with combined roller and drag cutters or roller cutters only. Two of these machines or 20 percent became stuck (Cases 6 and 34). In Case 6, which was previously discussed, the STBM became jammed six times and was freed each time by retracting the head about 10 cm to re-establish cutting wheel rotation. The cutters and rock crushers then successfully fractured the obstructing boulders and no other intervention was required. In Case 34, the STBM encountered a nested cobble and boulder fill zone that required a relief shaft to remove the obstructions. Three of the 13 cases with combined roller and drag cutters involved EPBMs (with screw augers). None of these machines became stuck. In summary, STBMs with only drag cutters became stuck 67 percent of the cases (8 of 12), while STBMs with combined drag and roller cutters or only roller cutters became stuck for 20 percent of the cases (2 of 10). TBMs with face access (OFRW machines and EPBMs with pressure relieving gates) became stuck 13 percent of the cases (2 of 15). EPBMs with combined roller and drag bits did not get stuck.

large boulder cases 43% stuck = 12/28 drag cutters 50% stuck = 10/20 roller or roller and drag cutters 15% stuck =2/8 drag cutters 80% stuck = 8/10

29% stuck = 2/7

STBMs 59% stuck = 10/17

drag cutters 20% stuck = 2/10 roller or roller and drag cutters

OFRWs, EBPMs 18% stuck = 2/11

0% stuck = 0/1 Figure 4. Distribution of stuck TBMs with maximum boulder size 33% of excavated diameter.

2.2.2 Boulder size factor Maximum boulder sizes encountered as reported in the references or as estimated based on information provided are listed in Table 1. Assuming that cobbles and boulders smaller than 33 percent of the excavated diameter can generally pass through the cutting wheel and be crushed or pass through a conveyor mucking system without further fracturing, then boulders larger than 33 percent would require fracturing, pushing aside or removal by relief shaft or tunnel intervention in order for a TBM to proceed. For some TBMs, particularly EPBMs with a screw auger, the digestible boulder size is smaller than 33 percent of the excavated diameter and may be as low as 5 to 10 percent of the excavated diameter. A total of 28 of 40 or 70 percent of the cases in Table 1 encountered boulders larger than 33 percent of the excavated diameter. In 43 percent (12 of 28) of the large boulder cases, the TBM became stuck (Figure 4). Of these 28 large boulder cases, 10 involved STBMs with drag cutters and 80 percent (8 of 10) of them were stuck. Ten of the 28 large boulder cases involved OFRWs with drag cutters and 20 percent (2 of 10) of them were stuck. Seven of the 28 large boulder cases involved STBMs with only roller or roller and drag cutters and 29 percent (2 of 7) of them were stuck. One of the 28 large boulder cases involved EPBMs and it was not stuck. A comparison of the results presented in Figure 4 to those in Figure 3 shows that the frequency of stuck TBMs is greater when the maximum boulder size

encountered is greater than approximately one third of the excavated diameter. 2.2.3 Boulder volume (frequency) Table 1 includes estimated boulder frequencies and volumes as a percentage of the excavated tunnel volume. The listed total boulder and large boulder (greater than 50 cm in size) quantities were based on available project data or estimated based on reported information on boulder encounters. The most uncertain data is preceded by an asterisk (*). Boulder volumes were estimated using an Excel spreadsheet that based on a normal distribution of common sizes using an estimated mean size, limited maximum size and assumed standard deviation. Boulder volumes were computed as 0.7 D3 where D is the average diameter for sub-rounded boulders (Hunt & Angulo, 1999). This method worked well for several cases where better boulder size data was available (e.g. Cases 18–23), but should only be considered a crude estimate of boulder volume. A total of 23 of 40 or 58 percent of the cases in Table 1 encountered estimated boulder volumes that were equal to or more than 0.4 percent of the excavated tunnel volume. The 0.4 percent division was arbitrarily based on experience indicating that boulder impacts were generally more significant when this relative volume was reached or exceeded. In 30 percent

59 Copyright © 2004 Taylor & Francis Group plc, London, UK

roller or roller and drag cutters

09096-06.qxd

2/18/04

8:17 PM

Page 60

feature that allowed the cutting wheel to be retracted about 10 cm to allow cutting wheel and roller cutter action to resume. Several methods of removing boulder obstructions are discussed in Hunt, 2002. Staheli and Hermanson, 1997 in a paper entitled “What to Do When your Head Gets Stuck” discuss some additional alternatives for dealing with boulder obstructions. In most cases of stuck TBMs without face access, the remedial alternatives are limited and often result in a recovery shaft or tunnel, particularly if severe cutter damage has occurred from the boulder encounters prior to getting stuck. If face access is available, more remedial options exist for boulder obstruction removal. The least risky method, if the ground and groundwater conditions allow it, would be open-face mining with an OFRW machine or other type of shield or hand-mining operation. Combined drag and roller cutters would help minimize cutter wear and damage. Boulder blasting or hydraulic splitting can generally be achieved in a cost-effective manner if ready face access exists in ground with suitable standup time. Having face access makes use of drag cutters less risky because the larger boulders can be accessed and fractured before excessive, damaging grinding occurs. While this method generally prevents the TBM from becoming stuck, it does not prevent problems. The cobbly and bouldery ground in Case 7a so severely damaged the drag cutters and cutting arm that the OFRW machine was removed and mining with it was discontinued. In Case 20, the cobbly and bouldery ground caused such severe wear to chisel drag cutters that a cutter repair chamber had to mined in front of the TBM on each of both 200 m long drives even though 60 boulder obstructions were blasted. If an EPBM is utilized due to poor standup time or variable ground conditions, the most flexibility for boulder obstruction removal would result if semipressurized mode mining with a pressure relieving gate is allowed. This method was successful for Case 31 where a small number of boulders were encountered. Similar results were experienced in Milwaukee on the Northshore 9 Collector System tunnel (Goss, 2002; Budd & Cooney, 1991). However, if cobble and boulder quantities are high, flood doors and pressure relieving gates may be destroyed or severely damaged (Castro et al., 2001; Cronin & Coluccio, 2003). The data in Table 1 indicate that for 8 of 12 stuck TBM cases, a recovery shaft or tunnel was mined to remove the TBM. In 6 of these 8 cases, the stuck TBM was removed and replaced by a different TBM or cutting wheel or tunneling was discontinued. For the other two cases, the TBM resumed mining after cutter and cutting wheel repairs. Boulder obstruction removal shafts were completed for 2 of the 12 stuck TBM cases (Cases 24 and

cases w boulder vol. 0.4% exc. vol. 30% stuck = 7/23 drag cutters 35% stuck = 6/17 roller or roller and drag cutters 17% stuck =1/6 drag cutters 67% stuck = 6/9

25% stuck = 1/4

roller or roller and drag cutters

STBMs 54% stuck = 7/13

drag cutters 25% stuck = 2/8 roller or roller and drag cutters

OFRWs, EBPMs 20% stuck = 2/10

0% stuck = 0/2 Figure 5. Distribution of stuck TBMs when estimated boulder volume 0.4 percent of excavated volume.

(7 of 23) of the larger boulder volume cases, the TBM became stuck (Figure 5). Of these 23 larger boulder volume cases, 9 involved STBMs with drag cutters and 67 percent (6 of 9) of them were stuck. Two of the 23 larger boulder volume cases involved OFRWs with drag cutters and 24 percent (2 of 8) of them were stuck. Four of the 23 larger boulder volume cases involved STBMs with only roller or roller and drag cutters and 25 percent (1 of 4) of them were stuck. Two of the 23 larger boulder cases involved EPBMs with drag and roller cutters and neither of them became stuck. The results for stuck TBMs related to larger boulder volume (or more total boulders) are similar to those for maximum boulder size. However, of significance is that 5 of 17 or 29 percent of the stuck TBMs occurred when estimated boulder volumes (frequencies) were small. Four of 5 or 80% of these cases involved STBMs only with drag cutters. 2.3

Remedial measures if TBM gets stuck

If a TBM becomes stuck because of cobble and boulder obstructions, a relief or recovery shaft or tunnel may be required to remove the obstruction and repair or remove the TBM, unless the TBM has face access. Case 6 (Miller, 1995a) describe how a STBM with roller cutters, but without face access was able to unjam itself and avoid being stuck by use of a TBM

60 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-06.qxd

2/18/04

8:17 PM

Page 61

stuck on this project also exceeded US 3 million and resulted in a completion delay of nearly one year. For Case 19, the original TBM (a STBM with drag cutters) was removed within a hand-mined relief tunnel. Problems with groundwater control and flowing ground required significant compensation and permeation grouting resulting in over three months delay and US $600,000 in contractor expenses. After the boulder obstruction was removed, the contractor found that site friction setup had caused the pipe string to also become stuck.

34). Tunneling was then completed without TBM removal. In one case (No. 14), an open-face TBM was stuck on a very large unanticipated boulder that resulted in significant delays before permitting and preparation for blasting could be achieved. 2.4

Cost and schedule consequences of getting stuck

Specifications and planning should be focused on minimizing the chances of a stuck TBM, particularly if the consequences are severe. The consequences of getting stuck should be evaluated on every project. The delay and cost consequences of getting stuck can be very high when one or more of the following constraints would prevent or complicate access for boulder or TBM removal by a recovery shaft or tunnel:

• • • • • •

3 CONCLUSIONS 3.1

Buildings, railroads, restricted access roadways, utilities or other facilities. Environmental impact restrictions below water courses, wetlands or other protected areas. Presences of contaminated ground or groundwater, particularly if hazardous waste is involved. Tunnels deeper than about 15 m. Large potential damages for delay of completion. Boulder conditions are too severe to continue with TBM requiring another TBM to be mobilized and launched.

This study of microtunnel and small diameter TBM encounters with boulders clearly shows that microtunnel boring machines equipped with drag cutters and no face access (Figure 6) have a high risk of becoming stuck when boulders are encountered. The risk of becoming stuck increases as the maximum boulder size or frequency increases. The risk of a STBM or EPBM becoming stuck decreases if roller cutters replace or are combined with drag cutters (Figure 7). Microtunnel boring machines that were equipped with roller or a combination of roller and drag cutters became stuck approximately one third to one fourth less often.

A row listing approximate or roughly estimated cost consequences from getting stuck or potentially getting stuck is provided in Table 1. Many if not most of the estimated cost consequences are probably higher than listed. More accurate information on cost consequences of becoming stuck was available for Cases 1, 4, 16 and 19. For Case 1a, the original TBM (a STBM with drag cutters) and over 150 m of jacked pipe had to be abandoned and replaced with a new launch shaft and STBM with roller and drag cutters (Case 1b). The additional cost was over US $3 million and approximately one year of delay. For Case 4, a STBM with drag cutters became stuck below a multi-lane interstate highway. A recovery tunnel was excavated to remove the badly damaged TBM and finish the drive. The additional costs were reported over US $400,000. For Case 16, the cutters and cutting wheel of the original TBM (a STBM with drag cutters) was totally replaced after becoming stuck the first time. After becoming stuck the second time, the TBM was removed and sent back to the manufacturer. It was replaced with an OFRW machine having face access for boulder blasting, but also which required expensive dewatering to provide suitable standup time and minimize settlement damage. The cost of becoming

Figure 6. STBM with drag cutters and no face access.

61 Copyright © 2004 Taylor & Francis Group plc, London, UK

The risk of getting stuck on boulders is highest with a TBM equipped with only drag cutters and no face access

09096-06.qxd

2/18/04

8:17 PM

Page 62

3.3

Soil matrix shear strength is important, but boulders in weak soil may not result in a stuck TBM

The ratio of rock shear strength to soil matrix shear strength is an important factor, but having a ratio greater than 600:1 may not result in a stuck TBM, particularly if the cutting wheel is well armored to resist abrasion and some strawberry cone or multi-row carbide insert cutters are included on the cutting wheel. The STBMs and EPBMS used for Case Nos. 1b, 2, 10, 35 and 36 had combined roller and drag cutters and were able to handle boulders without becoming stuck even though soft or loose soil zones were reported that probably resulted in shear strength ratios less than 600:1. Even if the TBM does not get stuck, which is important, the risk of cutter damage and much slower progress than desired may result when boulders are encountered in a soft or loose matrix than if the boulders are embedded in stiffer or denser ground that is capable of holding boulders for effective fracturing. Goss, 2002 listed four case studies where disc cutters failed to fracture boulders as desired. If a more rugged TBM with roller cutters is selected for handling boulders in weak ground, the potential costs of higher cutter and cutting wheel wear, slower productivity and higher risk of getting stuck should be carefully compared to the costs and risks of other tunneling alternatives such as open-face tunneling with face stability provided by compressed air, grouting or dewatering. Figure 7. Rugged STBM with drag cutters and roller cutters.

3.2

3.4

Multiple or redundant boulder handling capabilities should result in the least risk of getting stuck

The potential consequences of getting stuck should be carefully considered when specifying or selecting the TBM. In many cases the cost of mobilizing a more rugged TBM with more boulder handling features is well worth the reduction in risk of getting stuck that results. In some situations the cost of getting stuck exceeds several million dollars.

Bouldery ground results in a risk of getting stuck that can be significantly reduced by careful selection of TBM components, but remains a risk that cannot generally be eliminated. The least risk of becoming stuck results if multiple, redundant TBM features are provided so that more than one method can be used if the primary method of handling boulders does not work sufficiently. For example, a STBM with drag cutters and no face access allows few if any alternatives other than a recovery shaft or tunnel if it becomes stuck. If roller or roller and drag cutters are used, the STBM has capability of fracturing boulders larger than those ingestible and being further fractured by a cone crusher. If capability for cutting wheel retraction is added, jammed cobbles and boulders are more likely to be handled without outside intervention. If an option exists for face access or back-loading cutters for manual boulder fracturing when needed or cutter changing, the risk of getting stuck would be even less.

REFERENCES Abramson, L., Cochran, J., Handewith, H. & MacBriar. 2002. Predicted and actual risks in construction of the Mercer Street Tunnel. In Ozdemer, L. (ed). Proceedings Of The North American Tunneling 2002: 211–218. Rotterdam: Balkema. Becker, C. 1995. The Choice Between EPB- and Slurry Shields: Selection Criteria by Practical Examples. In Williamson, G.E. & Fowring I.M. (eds). Proceedings, 1995 Rapid Excavation and Tunneling Conference. Chapter 31, 479–492. Littleton CO: SME. Beieler, R., Gonzales, D. & Molvik, D. 2003. City of Seattle – Tolt Pipeline No. 2 Bear Creek and Snoqualimie River

62 Copyright © 2004 Taylor & Francis Group plc, London, UK

The consequences of getting stuck should be carefully considered

09096-06.qxd

2/18/04

8:17 PM

Page 63

Microtunnels. Proceedings of North American No-Dig 2003, NASTT, Las Vegas, April, Paper A-2-03. Bennett, D. & Wallin, M.S. 2003. American River Crossings: Then and Now. Proceedings of North American No-Dig 2003, NASTT, Las Vegas, April, Paper D-2-02. Boscardin, M., Wooten, R.L. & Taylor, J.M. 1997. Pipe Jacking to Avoid Contaminated Groundwater Conditions. In Proceedings of Trenchless Pipeline Projects – Practical Applications:135–141. Boston, MA. June 8–11, 1997. ASCE Pipeline Division. Budd, T.H. & Cooney, A.M. 1991. Milwaukee’s North Shore 9 Collector System – A Case History. In Wightman, W.D. & McCarry, D.C. (eds). Proceedings, 1991 Rapid Excavation and Tunneling Conference: 349–378. Society for Mining Metallurgy and Exploration, Littleton, CO. Castro, R., Webb, R. & Nonnweiler, J. 2001. Tunneling Through Cobbles in Sacramento, California. In Hansmire, W.H. & Gowring, I.M. (eds). Proceedings 2001 Rapid Excavation and Tunneling Conference: 907–918. Littleton, Colorado: SME. Cigla, M. & Ozdemir, L. 2000. Computer Modeling For Improved Production of Mechanical Excavators. In Proceedings of Society for Mining, Metallurgy and Exploration (SME) Annual Meeting, Salt Lake City, UT, February 2000. Clarke, I. 1990. Carronvale Sewer Project. No-Dig International. April 1990: 17–19. Coss, T.R. 1993. Pascal Tunnels Under New Denver Airport. Trenchless Technology, May/June. 1993:41–42. Cronin, H.E. & Coluccio, J.J. 2003. The True Cost of Boulders in a Soft Ground Tunnel. 2003. In Robinson, R.A. & Marquardt, J.M. (eds), Proceedings 2003 Rapid Excavation and Tunneling Conference. Littleton, Colorado: SME: 535–539. DiPonio, D.D., Manning, F.B. & Alberts, J.B. 2003. An Encounter with Bolulders During Soft Ground Tunneling in Wayne County, Michigan: A Case History. In Robinson, R.A. & Marquardt, J.M. (eds), Proceedings 2003 Rapid Excavation and Tunneling Conference. Littleton, Colorado: SME: 522–534. Dowden, P.B. & Robinson, R.A. Coping with Boulders in Soft Ground Tunneling. 2001. In Hansmire, W.H. & Gowring, I.M. (eds). Proceedings 2001 Rapid Excavation and Tunneling Conference: 961–977. Littleton, Colorado: SME. Ellis, M. 2003. Northeast Interceptor Sewer – Libertyville, Illinois. Westcon.net. Fleet, J. & Owen, D. 2002. Difficult Drives in Dundee. World Tunneling, 2002:216–218. Frank, G. & Daniels, J. 2000. The Use of Borehole Ground Penetrating Radar in Determining the Risk Associated With Boulder Occurrence. In Ozdemir, L. (ed). Proceedings Of The North American Tunneling 2000: 427–436. Rotterdam: Balkema. Frank, G. & Chapman, D. 2002. “Geotechnical Investigations for Tunneling in Glacial Soils,” In Hansmire, W.H. & Gowring, I.M. (eds). Proceedings 2001 Rapid Excavation and Tunneling Conference: 309–32, Littleton, Colorado: SME. Friant, J.E. & Ozdemir, L. 1994. Development of the High Thrust Mine-Disc Cutter for Microtunneling Applications. No-Dig Engineering. June 1994: 12–15.

Garret, R. 1992. Refined Solutions at Indianapolis. North American Tunneling Supplement to World Tunneling. May 1992: N27-N30. Gehlen, H. & Hunn, C. 2002. Microtunneling in the Scottish Highlands. Trenchless Technology International. Aug 2002: I-16 – I-17. Genzlinger, D.D. 1995. Teamwork Overcomes Tunneling Difficulties. Trenchless Technology. Feb. 1995: 34–36. Goss, CM. 2002. “Predicting Boulder Cutting in Soft Ground Tunneling,” In Ozdemer, L. (ed), Proceedings Of The North American Tunneling 2002: p37–46. Rotterdam: Balkema. Hunt, S.W. 1996. Evaluation of Represented and Encountered Subsurface Conditions to Determine Merit of Differing Site Conditions Claims – Elgin Interceptor. STS Consultants Report to Michels Pipeline Construction Company. November 20, 1996. Hunt, S.W. & Angulo, M. 1999. Identifying and Baselining Boulders for Underground Construction. In Fernandez, G. & Bauer (eds), Geo-Engineering for Underground Facilities: 255–270. Reston, Virginia: ASCE. Hunt, S.W., Bate, T.R. & Persaud, R.J. 2001. Design Issues For Construction of a Rerouted MIS Through Bouldery, Gasoline Contaminated Ground, In Proceedings of the 2001 – A Collection Systems Odyssey Conference. Session 6. Alexandria, VA: Water Environment Federation, Inc. Hunt, S.W. 2002. Compensation for Boulder Obstructions. In Ozdemir, L. (ed), Proceedings Of The North American Tunneling 2002: 23–36. Rotterdam: Balkema. Hunt, S.W. 2003. MWH Files: MN-320 Project, Minneapolis, Minnesota. Jones, M. 1990. Sudden Valley Sewer Project. No-Dig International. April 1990: 22–24. King, J., Najafi M. & Varma, V. 1997, Pipe Jacking Operation Completed in Flowing Ground. Trenchless Technology, Sept. 1997: 88–89. Mazhar, F. 1995. Flood Control and Combined Sewer Overflows. Harza Engineering Company Project Profile. 1p. Miller, P. 1994a, Microtunneling Delivers Transmission Crossings. Trenchless Technology, Aug. 1994: 36–37. Miller, P. 1994b. Pipe Jacking Delivers Nearly Two-Mile CSO Sewer. Trenchless Technology, Dec. 1994: 26–27. Miller, P. 1995a. Nova Battles Waves for Seawater Recovery Tunnels. Trenchless Technology, Oct. 1995: 26–28. Miller, P. 1995b. L.J. Keefe Conquers Squeezing Clays for Relief Sewer. Trenchless Technology, Dec. 1995. Miller, P.J. 1996. West Coast Microtunneling Finds Niche. Trenchless Technology, Mar. 1996: 40–43. Miller, P. 1997a. MT Project Uses Pipe Array. Trenchless Technology, Feb. 1997: 28–29. Miller, P. 1997b. San Diego Project Marks Technologies and Teamwork. Trenchless Technology, Oct. 1997: 22–24. Molvik, D., Breeds, C.D., Gonzales, D. & Fulton, O. 2003. Tolt Pipeline Under-Crossing of the Snoqualmie River. In Robinson, R.A. & Marquardt, J.M. (eds), Proceedings 2003 Rapid Excavation and Tunneling Conference. Littleton, Colorado: SME: 396–403. Najafi, M. & Varma, V. 1996, Two Firsts for Iowa – Microtunneling and RCPP. Trenchless Technology, Dec. 1996:36–37. Navin, S.J., Kaneshiro, J.Y., Stout, L.J. & Korbin, G.E. 1995. The South Bay Tunnel Outfall Project, San Diego,

63 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-06.qxd

2/18/04

8:17 PM

Page 64

California. In Williamson, G.E. & Gowring, I.M. (eds). Proceedings 1995 Rapid Excavation and Tunneling Conference: 629–644. Littleton, Colorado: SME. Nishitake, S. 1987. Earth Pressure Balanced Shield Machine to Cope with Boulders. In Jacobs, J.M. & Hendricks R.S.. (eds). Proceedings, 1987 Rapid Excavation and Tunneling Conference. Chapter 35,. 552–572. Littleton CO: SME. Ozdemir, L. 1995. Comparison of Cutting Efficiencies of Single-Disc, Multi-Disc an Carbide Cutters for Microtunneling Applications. No-Dig Engineering. March 1995. 18–23. Rickert, W.R. & Galantha, M.A. 1999. Northeast Interceptor Meets Future While Respecting the Environment. Public Works. April 1999: 50–53. Rush, J.W. 2000. West Coast Microtunneling. Trenchless Technology. Apr. 2000: 28–31. Rush, J.W. 2002. Microtunneling Key to California Earthquake Repair Project. TBM Tunnel Business Magazine. Aug. 2002: 22–23. Rush, J.W. 2002. Northwest Boring Completes WorldClass Microtunnel. Trenchless Technology. Oct. 2002: 28–31.

Schumacher, M. & Ellis, M. 1997. Conquering Glacial Till in Ames, Iowa, Proceedings of No-Dig ’97: 455–461. NASTT, Seattle, April 1997. Session 4B-3. Smith, M. 1995. NEHLA Undershore Crossing. North American Tunneling. June 1995: N16-N20. Staheli, K., Bennett, D., Maggi, M.A., Watson, M.B. &. Corwin, B.J. 1999. Folsom East 2 Construction Proving Project: Field Evaluation of Alternative Methods in Cobbles and Boulders. In Fernandez, G. & Bauer (eds). Geo-Engineering for Underground Facilities: 720–730. Reston, Virginia: ASCE. Staheli, K. & Duyvestyn, G. 2003. Snohomish River Crossing: Bring on the Boulders, Success on the Second Attempt. Proceedings of North American No-Dig 2003, NASTT, Las Vegas, April, Paper B-4-03. Tarkoy, P.J. 2001, Challenges & Successes in MicroTunneling on the Chelsea River Crossing. Proceedings of 5th International Microtunneling Symposium – BAUMA 2001. 16p. Vadnais, P. 2002. Personal Discussion with Steve Hunt on Marinette Water Main River Crossing Project. Wallis, S. 2002. Remotely controlled passage under the Neva. World Tunneling. Feb. 2002: 25–27.

64 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-07.qxd

2/18/04

8:18 PM

Page 65

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Joint orientations for TBM performance analysis using borehole geophysics to orient rock cores T. Tharpe, B. Crenshaw, and J. Raymer Jordan, Jones & Goulding, Inc., Norcross, Georgia

ABSTRACT: A borehole televiewer was used to orient rock core for joint analysis as part of the Atlanta CSO Tunnel Geotechnical Investigation. The Atlanta CSO tunnels are about 30 feet in diameter and will be bored through medium grade gneiss and schist. The televiewer is a wireline geophysical tool that uses sonic waves to map the traces of individual joints around the inside of the borehole wall. From these traces, the televiewer software calculates the strike, dip, and aperture of each joint. The televiewer was used in 51 core holes averaging 300 feet in depth, and was much less expensive, faster, and more accurate than using oriented core. The televiewer provided joint characteristics, an acoustic velocity log to indicate areas of weathered or blocky ground where core recovery is typically poor, and a graphical picture of the borehole that can be used to orient the core for more detailed analysis. Each joint in the televiewer data was correlated to the core and classified in terms of RMR and Q parameters. Machine breaks and core damage were easily recognized because they occurred only in the core, but were not read by the televiewer in the borehole wall. A stereoplot of the joint data was made for each borehole and for the project as a whole. These stereoplots were used for three purposes: (1) kinematic wedge analysis and support design; (2) classifying the ground into different baseline types; and (3) estimating the potential benefits of the fractures on TBM performance using the Norwegian Fracturing Factor criteria.

During this process, all joints were described in terms of type, fit, roughness and planarity, and alteration. Orientation and geotechnical property data gathered during the rock core fracture analysis was used for numerous aspects of the Atlanta West Area CSO Project. Orientation data was used to generate stereonets for the Geotechnical Data Report (GDR) as well as stereonets used in various analyses conducted for the purposes of design and baselining. Geotechnical property data was used to aid in prediction of ground conditions along the tunnel alignment.

1 INTRODUCTION A digital acoustic televiewer (DATV) was utilized as a part of the City of Atlanta Combined Sewer Overflow (CSO) Tunnel geotechnical investigation. The geotechnical investigation for the CSO Project included geological mapping, 77 core borings, geotechnical analysis of the core, laboratory testing of the rock properties, and DATV logging. Televiewer logging was conducted in 51 of the 77 completed core borings. The DATV is a geophysical tool that provides highresolution data that can be used to determine dip direction and dip angle of planar features intersecting a borehole (Keys, 1990). Both travel time and amplitude of the acoustic signal are recorded and displayed in real time. The data is analyzed, and orientation of downhole features is recorded. The orientation of the joints delineated from the televiewer data was transferred to corresponding joints in the rock core. The rock core was then oriented to North based on identified joint azimuths, and a North line was scribed on the core. Fractures not oriented from televiewer data (low angle fractures, healed fractures, microfractures, etc) were identified and oriented from the geophysically derived northern orientation.

2 PROJECT DESCRIPTION Like many older cities of its size, Atlanta’s sewer system consists of areas where stormwater and sanitary sewer flows (wastewater), are collected in the same pipe, known as a combined sewer system; and areas where they are collected in separate pipes, known as a separated system. Approximately 85 percent of Atlanta’s system is separated, mainly in residential areas that have developed within the last 75 years. The remaining 15 percent of the system, the older section that serves the core of the City, consists of a

65 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-07.qxd

2/18/04

8:18 PM

Page 66

commonly weathered away, leaving zones of broken, water-bearing rock that are more easily eroded to form topographic valleys and draws. Gouge and shattered rock are present in some of the lineaments. Also found in the Atlanta region are large, subhorizontal stress relief fractures (Cressler et al., 1983). These fractures may be visualized as a low arch in cross section with the largest opening occurring at the midpoint. Fractures are typically open and are very transmissive. For this reason, they may produce a large flow of water when encountered in wells or tunnels. A flushflow of approximately 2,000 gpm was encountered in the Atlanta Three Rivers Tunnel when a stress relief fracture was intercepted. Geophysical information was useful in recognizing and characterizing these features along the West Area CSO alignment because these fractures are not observable at the surface.

combined sewer system. It was constructed in the late 1800s through 1920 to carry both stormwater and wastewater to a treatment facility, where treatment would occur before the water was discharged into the Chattahoochee River. During dry weather, all flows are conveyed to the treatment facility. When rainfall occurs, flows are still conveyed to the treatment facility. However, during larger rainfall periods, sewer capacity is sometimes exceeded, resulting in portions of the flow being diverted from the wastewater treatment facility to a combined sewer treatment facility that provides a lesser degree of treatment (screening and disinfection). This condition is known as a Combined Sewer Overflow (CSO). The CSO Storage and Treatment System plan involves capturing, storing and conveying CSOs. The overflows are stored in large underground tunnels in bedrock. The captured CSO volume is conveyed to a separate treatment system for removal of suspended solids and other pollutants, undergoing disinfection before discharge to receiving waters. The plan has two components: the construction of the new West Area CSO Storage Tunnel and the East Area CSO Storage Tank. The West Area CSO Project consists of two large diameter tunnels, one smaller diameter connecting tunnel, three intakes, one pump station, one overflow shaft and tunnel, and four additional construction shafts. The two large diameter tunnels are the North Avenue Tunnel and the Clear Creek Tunnel. Both of the large diameter tunnels are about 27 feet in excavated diameter and about 24 feet in finished diameter. Both tunnels will be excavated at an average depth of 200 feet below land surface. The North Avenue Tunnel is approximately 23,333 feet long, and the Clear Creek Tunnel is approximately 20,783 feet long.

4 GEOPHYSICAL INSTRUMENTATION An acoustic televiewer provides a digital record of the location, character, and orientation of any features in the casing or borehole wall that alter the reflectivity of the acoustic signal. These include diameter and shape of the borehole, drilling or lithology induced rugosity, differences in rock hardness, and structural features such as bedding, fractures, and solution openings. The acoustic televiewer provides a magnetically oriented, 360-degree, image of the acoustic reflectivity of the borehole wall (Keys, 1990). Because the collected data is spatially oriented, it can be used to calculate the dip azimuth and dip direction of planar featured that intersect the borehole. 4.1

The digital acoustic televiewer tool used for this study utilizes an acoustic transducer that operates at a frequency of .5-MHz. The acoustic transducer, which functions as a transmitter and receiver, rotates at 12 revolutions/second and digitizes 256 data samples per revolution, which allows for the collection of highresolution data. The high frequency induced by the acoustic transducer is reflected from the borehole wall and is received by the instrument. An internal flux-gate magnetometer is triggered each time the acoustic transducer rotates past magnetic north. The signal from the magnetometer is transmitted to the recording equipment at land surface (Keys, 1990).

3 BEDROCK DESCRIPTION The West Area CSO Tunnel is located in the Piedmont region of the southeastern United States. The ground along the West Area CSO Tunnel generally consists of medium-grade metamorphic rocks that have been intruded by granitic rocks in some places. The bedrock in the project area has undergone intense deformation, weathering, erosion, and some regional uplifting. The bedrock is overlain by approximately 15 to 120 feet of soil and partially weathered rock. Lineaments are a common feature in parts of the Piedmont, including the area of the Atlanta West Area CSO Tunnel. Lineaments in the Piedmont are typically long, narrow topographic valleys and draws. In many cases, these lineaments represent surface expressions of subsurface features, such as fracture zones. These fracture zones may become cemented with minerals at depth. At shallower depths, these cements are

4.2

Field methods

Upon completion of the coring process, the resulting borehole is logged with geophysical instruments. Before DATV logging is conducted, it is necessary to verify that the tool’s internal magnetometer is triggering

66 Copyright © 2004 Taylor & Francis Group plc, London, UK

Instrument function

09096-07.qxd

2/18/04

8:18 PM

Page 67

be achieved if the rock coring rig vacated the drilling site, followed by the immediate geophysical logging of the core hole, and then followed by the grouting of the borehole by a second crew dedicated only to borehole grouting. This procedure may not be possible for all projects, and stand-by charges, from drilling contractors, may be incurred during the logging of the borehole. When initiating a logging program, site access issues must be considered. For the West Area CSO project, a two-wheel drive vehicle was utilized for logging services, and the project was conducted in an urban environment, which allowed for easy access to nearly all of the drilling locations. However, some locations could not be accessed due to the terrain and vehicle limitations. Data collected in geologic settings where the geology is very homogeneous, and where regional dip is very shallow, is difficult to use for rock core orientation. In homogenous, shallow-regional-dip geologic settings, there may be very few distinctive planar features that intersect a given borehole. In order for the rock core orientation process to be successful, a number of distinctive features, which alter the acoustic reflectivity of the borehole wall, must intersect the borehole. In sections of featureless core, the orientation process is not possible unless strong geologic layering is present. If the geologic features can be delineated through data analysis, then orientation can be accomplished in very competent rock that is not fractured. In the geologic setting in which the City of Atlanta CSO project took place, gneissic banding in the bedrock is often very prevalent. In instances where competent bedrock, devoid of fracturing, was encountered, gneissic banding was used to orient sections of rock core. During the course of this project, it was determined that it is difficult to use shallow dipping planar features to orient rock core. When delineating planar features using the televiewer data, the margin of error for strike and dip determination increases with features that have a dip of less than 5°. Although shallow dipping features can be delineated with the DATV, it is problematic to transfer this data to the rock core and should be avoided. Additionally, when examining both the DATV data and the rock core, steeply dipping planar features are more easily and quickly identified than are shallow dipping features. Thus, the rock core orientation process is simplified by first using distinct, steeply dipping fractures (dip 20°) to establish the north axis on the core, followed by the manual measurement of shallow dipping fractures that exist in the core. The collection and interpretation of DATV data can be complicated by a number of factors. Of these factors, incorrect gain settings and insufficient tool centralization within the borehole severely impact the quality and usefulness of the collected data. The resolution of the collected data is determined by the gain setting. If the gain setting is too low then the acoustic

Figure 1. Schematic diagram for logging setup used during West Area CSO Project investigation.

on magnetic north, and an electronic calibration file is downloaded to the tool before logging commences. After it has been verified that the tool is functioning properly, the televiewer is advanced down the borehole with an electric drawworks utilizing a fourconductor wire line (Figure 1). The wire line is secured to the geophysical instrument by a watertight locking cable head. Once the tool has reached the terminal depth of the borehole, the logging procedure begins. The borehole is logged from the bottom of the core hole up to land surface at approximately 5 feet/minute. While logging, the data is recorded and displayed in real-time using a laptop computer. 5 CONSIDERATIONS FOR DATV LOGGING In part, the geologic setting in which a project or study area lies will determine the usefulness of DATV data for the purposes of orienting rock core. A geologic setting that is conducive to DATV logging is needed. Excellent televiewer data can be collected in a competent, fractured bedrock environment such as that of the Piedmont of Georgia. Televiewer data collected in the Piedmont of Georgia often shows distinctive, open, weathered and steeply dipping fractures that are easily identifiable in the core. The logging speed for the DATV tool is approximately 5 feet/minute. Depending on the investigation depth for a given geotechnical study, it could take two to three hours to complete a round of DATV logging. Time was a consideration for the West Area CSO project, thus, it was determined that a substantial time savings could

67 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-07.qxd

2/18/04

8:18 PM

Page 68

formation to remain mostly intact. Thus, the televiewer can be used to determine fracture orientation in highly fractured and contorted areas where orientation data cannot be determined from rock core samples. Also, drilling induced mechanical breaks and post drilling core damage are easy to identify because these features are not present in the televiewer data. Thus, data quality from the analyses of the rock core is improved because mechanical and post drilling core breaks are more easily discarded from the data set.

signal will be too weak to accurately detect features within the borehole. Conversely, if the gain setting is too high, the borehole will be flooded with acoustic signal, and data resolution will be greatly decreased. Poor tool centralization will also decrease data resolution. The acoustic signal must be introduced perpendicular to the borehole wall. If the televiewer is not centered within the borehole, travel time of the signal is altered because the acoustic transducer will either be closer to, or farther away from, the borehole wall. The DATV is a complex geophysical tool. A highly experienced equipment operator, or geophysical contractor, is needed to perform the logging services. Geophysical equipment is composed of complex electronic components. If a problem with the equipment should arise, a great deal of time could be spent trying to diagnose and repair the source of the malfunction. Additionally, it is possible for an inexperienced operator to collect data, which at first glance seems reasonable, but may prove to be useless once the data analysis process is initiated. An experienced geophysicist is also needed not only for data integrity, but for data interpretation as well. In order to use orientation data, derived from geophysics, to orient rock core, it is necessary to have a skilled geophysicist complete a thorough interpretation of the collected data. The geophysicist must be accustomed to analyzing geophysical data collected in complex geologic settings. It is also helpful if the geophysicist has knowledge pertaining to tunneling or underground construction and is able to recognize and delineate subsurface planar features that are important in tunnel construction.

7 CORE ORIENTATION USING GEOPHYSICAL DATA To collect data on healed fractures, to describe joint characteristics, and to verify geophysical data, rock core was oriented using features from geophysical logs. The methodology used to orient rock core is outlined in the following sections. 7.1

Where a is arc length (in the same units as r), r is the radius of the rock core, and is the azimuth of the down dip expression of the feature in degrees. For the Atlanta West Area CSO Tunnel, HQ core was used with a diameter  61.1 mm (r  30.55 mm). The calculated arc length (a) is the distance from the down dip expression of the feature to North measured counter clockwise around the core when the core is viewed from the top (Figure 2). A core trough was used to orient individual core runs (core runs were 10 feet for the project) and the North orientation was marked for the length of the run with a magic marker.

6 POSITIVE ASPECTS OF DATV USE During the drilling process, rock cores were removed from the inner collection tube, placed in wooden core boxes, and then labeled. The boxes were then transported to a warehouse facility where the core could be analyzed. When transferring the core to the boxes, and during transport of the core, the samples may be disturbed. During transport, core samples obtained in heavily contorted zones had a tendency to crumble and take on the appearance of gravel. These contorted zones are often characterized by fracture sets with a dip of 20°. Joint orientations within these zones were often highly varied. Utilizing DATV, it is possible to determine fracture orientation in highly contorted zones, which are often weathered, or in zones where geologic unloading or exfoliation has occurred. Additionally, the orientation of planar features within these zones can be determined because the televiewer is viewing the contorted zones in situ. The drilling process somewhat disturbs the investigated area, however, lithostatic and hydrostatic pressure enables the

7.2

Rock core orientation

The orientation process began by identifying a distinct feature within the cored interval to be oriented (the bottom 160 feet of the borehole). A distinct feature is defined as a joint that can be identified definitively within the core based on televiewer data. Distinct joints include, but are not limited to, the following:

•

68 Copyright © 2004 Taylor & Francis Group plc, London, UK

Method overview

To orient the core, features recorded on the televiewer geophysical log were identified on the rock core. The down dip azimuth from the televiewer log was then used to orient the down dip expression of features identified on the rock core. The orientation of the down dip expression of the feature was used to orient the core to North. The core was oriented to North using a diameter tape measure and the following conversion:

A single joint within a zone of otherwise featureless rock core

09096-07.qxd

• • • •

2/18/04

8:18 PM

Page 69

Each identified feature was assigned an identification number. The rock core was marked with the identification number and televiewer data. Recorded televiewer data included the dip, dip azimuth, up dip depth, and down dip depth. After identification and verification of initial orientations, the process was carried out again on adjacent runs. For every run, an attempt was made to identify and orient at least one feature. North orientations between identified features and adjacent oriented runs were then compared. If no feature could be identified in a run, the North orientation was carried on from an adjacent oriented run by aligning the ends of runs.

A joint that is in close proximity to a unique feature, such as a series of partially penetrating joints, a crushed or soil zone, vugs, etc. Cross cutting joints A joint which terminates into another joint A series of closely spaced, similarly oriented joints.

After identification of a distinct feature, the down dip expression of the feature was identified and the dip azimuth was used to mark North on the rock core, as described previously. Care was taken to align rock core pieces based on the fit of fractures and breaks. When first beginning to orient a borehole, a series of closely spaced joints was the preferred distinct feature. Identification of a number of closely spaced joints allowed several features to be oriented and north to be compared between them. After aligning the core in the core trough, a line was marked for the length of the run along the 0° (north) azimuth (Figure 3).

8 APPLICATION OF ORIENTED DATA Oriented core data was used in various analyses for the Atlanta West Area CSO Project. For design purposes, orientation data was used to conduct wedge analyses for design of support methods to be used in tunnel segments. Joint descriptions collected during joint orientation were used to calculated Q and RMR values and used to predict ground conditions along the tunnel alignment. Additionally, core orientation data was used to make stereonets for all oriented boreholes along the tunnel alignment, showing adjacent tunnel azimuth, for the Geotechnical Data Report (GDR). Orientation data was also used in baseline data interpretations. An analysis to determine the fracturing factor (ks-tot) was done based on the work of Bruland (1998). A detailed description of the use of oriented core data in the determination of ks-tot values is presented in the following sections.

Figure 2. Three-dimensional view of rock core demonstrating how a dip azimuth from geophysical data is transferred to the rock core and North is established. An example calculation is provided.

8.1

Fracture factor (ks-tot) is an index value used to estimate the benefit that fractures provide to Tunnel Boring Machine (TBM) performance. The value is determined based on fracture spacing (St class) and the orientation of fractures to the tunnel axis (). The methodology for calculating fracture factor was developed by NTNU and is described by Bruland (1998). The method is empirically based and was developed using post-excavation tunnel mapping. The method has been used successfully for calculating fracture factor values from rock for this and other tunnels in the Atlanta area (Dollinger, 2002). All variables needed to calculate fracture factors for the Atlanta West Area CSO Tunnel were determined during the core orientation and analysis process. These variables included fracture orientation and the location of the fracture along the borehole. The Atlanta West Area CSO Tunnel will consist of two intersecting tunnel alignments, the North Avenue Alignment and the Clear Creek Alignment. These two alignments will be bid separately, therefore, separate fracture factor

Figure 3. Picture showing fracture that is being used to orient rock core. The down dip expression has been identified (indicated by arrow on core pointing to left), the fracture has been numbered and geophysical data recorded, and the appropriate rotation has been calculated to orient the core to North. A geologist marks the North orientation along the entire core run along the core trough.

69 Copyright © 2004 Taylor & Francis Group plc, London, UK

Fracture factor calculation

09096-07.qxd

2/18/04

8:18 PM

Page 70

analyses were conducted for each alignment. The following sections will refer specifically to the North Avenue Alignment; the analysis of the Clear Creek Alignment was carried out using the same methodology. 8.2

Determination of fracture sets

Fractures sets were determined using a stereonet created using oriented rock core data along the North Avenue Alignment. These boreholes were located to sample random, non-biased subsurface conditions along the alignment. This method insured fractures represented typical conditions along the tunnel alignment. The stereonet included all fractures that were open or healed with weak cementing minerals, such as soft zeolites. All fractures were weighted by the secant of the dip to the core to account for the orientation bias of vertical boreholes. A stereonet of fracture orientations obtained through the analysis of geophysical data was compared to the stereonet created from oriented core analysis (Figure 4). These stereonets compare favorably when analyzing fracture sets. Variation in fracture set orientations is generally low, within approximately 10–15° for dip azimuths and dips. Fracture set intensity varies due to the criteria used to determine which fractures would be represented. Geophysical data (Figure 4a) represents only open fractures and fracture Set 1, foliation, is more strongly represented. Core data (Figure (4b) considers both open and weakly healed fractures, it includes more high angle fractures which are generally filled, and therefore fracture Sets 2 and 3 are more strongly represented. A visual analysis of the stereonet was used to group fractures into major fracture sets. Two of the major fracture sets identified along the North Avenue Alignment were further divided into fracture sub-sets (Figure 4b). This was necessary because of the highly undulatory nature of the fracture sets along the alignment. Boundary limits in terms of dip azimuth and dip were determined for each set and an analysis was done to separate fractures into Sets 1, 2, and 3. All fractures which fell outside of these boundary conditions were considered to be random, not affecting the ks-tot value, and were not considered further. Sets 1 and 2 were further divided into sub-sets based on boundary dip azimuth and dip conditions. Each fracture within a major fracture set was included in a fracture sub-set. Sub-division of the major fracture sets was necessary to account for the effects of undulations in fracture factor calculations. 8.3

Figure 4. Stereonets of fractures along North Avenue Alignment. a) Open fractures identified from interpretation of geophysical logs. b) Open and weakly healed fractures identified from core orientation. Major joint set boundaries are represented by bold dashed lines and are denoted by numbers. Sub-set boundaries ate denoted by dotted lines and denoted by lower case letters following number designation of major fracture set.

calculating the spacing along the rock core between two adjacent fractures of a set. This distance was then multiplied by the cosine of the average dip of the major fracture set to derive the length of an orthogonal line between two parallel planes. The length of this line is the effective spacing between two fractures.

Calculation of fracture spacing (St classes)

Once fracture sets were established, the spacing between all fractures of a major fracture set within a borehole was calculated. This was done by first

70 Copyright © 2004 Taylor & Francis Group plc, London, UK

Fracture Class (St) O O-I II II III IV

8:18 PM

Page 71

NTNU (cm)

CSO Tunnel spacing (cm)

massive 160 80 40 20 10 5

greater than 200 200 to 100 100 to 50 50 to 25 25 to 12.5 12.5 to 6.3 6.3 or less

IV 4.0

Figure 5. Fracture class classifications used the West Area CSO Tunnel. Adapted from Bruland (1998).

3.0 III-IV

2.0 III II-III 1.0

0.0

The effective fracture spacings were divided into spacing groups corresponding with NTNU St classes (Figure 5). All effective spacing values in each St group were summed and divided by the total effective distance of core evaluated to calculate a relative percentage of rock mass that contains each St class.

8.4

II I 0 0

10

20

30 40 50 60 alpha angle ()

70

80 90

Figure 6. Diagram for estimating fracturing factor (ks) from alpha angle () and fracture spacing (St class). Adapted from Bruland (1998).

The ks values for each major fracture set were used to calculate the ks-tot value for each tunnel segment using the following equation:

Calculation of fracture factor (ks-tot)

Both alignments of the Atlanta West Area CSO Tunnel contain numerous curves, and therefore, various segments of the alignment have different azimuths. Since fracturing factor values are dependent on the relationship between fracture orientations and the tunnel axis, alignments were divided into straight line segments between points of intersections of curves or between a point of intersection and a shaft. The North Avenue Alignment contained ten straight line segments. The ks-tot values were calculated for each straight line segment. To calculate ks-tot, a fracture set fracture factor (ks) value must be determined for each major fracture set. First, an alpha angle was calculated using average fracture orientation data for each fracture sub-set (or major fracture set if no sub-sets were designated) and each tunnel segment azimuth. These alpha angles were used to obtain a raw fracture factor (ks-raw) value for each NTNU St class (Figure 6). The ks-raw values did not take into account the relative abundance of fracture spacing (St classes) within the rock mass or the relative abundance of fracture sub-sets within major fracture sets. The ks-raw values were first adjusted to reflect the relative abundance of the St classes in the rock mass and then summed to obtain a ks-sub value for each tunnel segment. The ks-sub values were adjusted to reflect the relative abundance of the fracture sub-set in the major fracture set and summed to obtain the ks value of the fracture set in each tunnel segment. This step was not necessary for major fracture sets with no fracture sub-sets.

as described by Bruland (1998), where n is the number of fracture sets. The ks-tot values for each tunnel segment were provided in the baseline documents for use by the contractor and were used by the engineer for making TBM performance assessments for use by the owner. 9 SUMMARY Geophysical data was successfully used to orient rock core for the Atlanta West Area CSO Project. A strong correlation between the location of joint groupings was observed for stereonets produced from orientation data derived from geophysical methods and orientations of joints measured from the core. This indicates that features were correctly identified between the geophysical logs and the core, because if the incorrect joints were identified, then the joint groupings on the stereonet derived from core data would have been rotated differently then those from geophysical analysis. The successful implementation of geophysics as an orienting tool allowed for the acquisition of a large amount of orientation data at a reasonable cost to the project. Additionally, joint orientations obtained from geophysics supplemented the core orientations in zones that were too heavily weathered or fractured to measure manually. These orientations would not have been obtainable using traditional orientation methods. Orientations within

71 Copyright © 2004 Taylor & Francis Group plc, London, UK

Fissure Class

2/18/04

Fracturing Factor (ks)

09096-07.qxd

09096-07.qxd

2/18/04

8:18 PM

Page 72

orient core, but also served as a secondary verification of the subsurface conditions that exist along the tunnel alignment in addition to the rock core. This benefit is not associated with any of the traditional orientation techniques. The cost effectiveness of this technique over tradition orientation methods as well as the benefits associated with this technique merit its consideration for many tunneling projects.

these zones are of great benefit to understanding the ground conditions in and around them. These zones are particularly important during tunnel investigations because they represent non-typical conditions that have the greatest potential to negatively impact the project. The large amount of data collected from oriented core allowed numerous, thorough analyses to be conducted for the purpose of tunnel design and contract baselining. Stability analyses were conducted for many regions of the tunnel based on fracture characteristics within each region. From these analyses, regions of the tunnel where wedge failures may occur were identified. Additionally, stability analyses and computer modeling were used to design and test tunnel support. Ground conditions were characterized based on joint characteristics associated with each joint through the calculation of Q and RMR values. Lastly, oriented joint data was used to calculate fracture factors for each segment of the West Area CSO Tunnel. These values quantify the benefit of fractures to TBM performance. The use of geophysical televiewer data proved to be a viable, cost-efficient core orientation option for the CSO Tunnel. Not only did it provide a means to

REFERENCES Bruland, Amund, 1998, Hard rock tunnel boring: Advance rate and cutter wear: Trondheim: Norges teknisknaturvitenska-pelige universitet. Cressler, CW, Thurmond, CJ, and Hester, WG, 1983, Ground water in the greater Atlanta region, Georgia: Information Circular 63, Georgia Geologic Survey. Dollinger, Gerald L, Raymer, and John H, 2002, Rock mass conditions as baseline values for TBM performance evaluation: in North American Tunneling 2002, Ozdemir (ed), Swets & Zeitlinger, Lisse. Keys, W, Scott, 1990, Techniques of Water-Resources Investigations, U.S. Geological Survey, Book 2, Chapter E-2.

72 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-09.qxd

2/18/04

11:51 AM

Page 73

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Slurry type shielded TBM for the alluvial strata excavation in downtown area Wang Ruel Jee Colorado School of Mines, Golden, CO

ABSTRACT: The soil and rock types relevant of construction lot 9 of Seoul subway line number nine of Yeoi Island area show a distinct difference between the eastern and western part of the alignment especially in the eastern part. The tunnel is expected to be located entirely in alluvial deposits, predominantly in sandy gravels with low strength of alluvial soils with presence of groundwater. For the western part, the geological conditions are considerable more complex, varying between hard rocks, residual soils and alluvium, with the higher proportions in weathered rocks. The alluvial deposits probably dominated by sandy gravels including some boulders are expected to be favorable for the application of a slurry shielded TBM with the higher underground water levels. For a more detailed assessment, reliable grain size distribution curves of the alluvial deposits along the tunnel alignment are investigated to select the optimal selection of separation plant and suitable advance rate of the slurry type shielded TBM. Along the whole alignment closed mode tunneling with active face support will be required, due to the low strength of the alluvial soil and the presence of groundwater. There are two basic options for the machine type to be applied: A Slurry type shielded TBM or Earth pressure balanced shielded TBM. For specific details concerning the technical machine operation and the safety precautions of the tunnel construction on this alluvial deposits are described in this paper.

1.1 Slurry type shielded TBM

1 INTRODUCTION OF THE SHIELDED TBM FOR SHALLOW TUNNEL CONSTRUCTION

Tunneling with slurry type shielded TBMs has been proven to be a safe excavation method causing low settlement in all kinds of loose ground with higher ground water condition. This method has another advantage more easy to handle the boulder problems during the tunnel excavation. The tunnel face is supported by a

Fundamentally, the main aim of geotechnical investigations is to identify the ground conditions for a proposed underground structure. Based on the results of the geotechnical investigation, designers can make decision on the selection of optimal construction methods, and also get the suitable idea of the underground structures as well as to prepare the geomechanical parameters for the numerical analysis of the safety precautions. Since the construction plan of Seoul subway line number 9–9 was established in the shallow alluvial deposits where the subway tunnel’s permanent stability should be guaranteed during the construction periods and also the operation periods continuously, and so it was hard to select the suitable tunneling methods in such poor ground conditions. It was considered that for the entire given alignment, closed-mode tunneling with active face support shall be required, due to the low strength of the alluvial soils and the presence of high level groundwater. There are two machine types to be used in such a poor ground conditions, a slurry type shielded TBM (Fig.1), and an earth pressure balance shielded TBM (EPB), which are discussed below.

Figure 1. Typical layout of the slurry type shielded TBM.

73 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-09.qxd

2/18/04

11:51 AM

Page 74

Hydroshield

Slurry Shield

Figure 3. Separation plant of the Slurry type shielded TBM (Busan, Korea). EPB Shield

Coarse gravel Medium gravel

1 10-1 10-2 10-3

slurry, generally, a suspension consisting of water and bentonite or clay. Through this filter cake, the pressurized suspension in the excavation chamber balances the earth and water pressure. The excavated material mixes with the suspension fluid and it will be pumped to a surface separation plant. Recycled suspension fluid will be fed again into the front working chamber of the shielded TBM. In comparison to other systems, economical usage of this system is mainly determined by the separation efficiency, speed, and suspension requirement, and the permeability of tunneling ground which are directly connected with the advance rate of this tunneling machine and how could they reduce the noise through the separation works in downtown area is the critical problem in urban construction. Slurry shield operates most efficiently in non-cohesive, watersaturated soil where the particle size distribution ranges between coarse silt/fine sand and coarse gravel. However, for operation in fine cohesive soils, it is difficult to perform the special technical installation of the separation plant with complex and expensive operation with higher energy consumption (Fig. 3). The optional installation of hydraulic rock crusher in front of the suction inlet will break boulders (Maximum 500 mm diameter) down to size corresponding to the diameter of the slurry discharge pipes (Diameter: 100–150 mm) (Fig. 4).

Fine gravel

10-4

Coarse sand

10-5 10-6

Fine sand Sandy, silty clay Silt

10-7 10-8 10-9 10-10

Clay

10-11 -12

10

Figure 4. Optimal particle sizes for the better implementation of Slurry & EPB shield.

utilize the material as excavated by the cutting wheel serves as a support medium. The support pressure is mainly influenced by the following two processes (Fig. 2).

• •

The forward thrust of the TBM determines, with cutter head rotation, the volume of excavated material, which is forced into the working chamber by advance of the TBM. The rotation of the screw conveyor determines the removal rate of excavated material from the working chamber.

The excavated material in the screw conveyor is forced to form a “Plug” which acts as a seal against the pressure from the working chamber, and allows the excavated material to be discharged at atmospheric pressure onto a conventional belt conveyor. Transport of the excavated material through the tunnel can be by belt conveyor, track-bound vehicles, dump trucks or, solid handling pumps to the surface.

1.2 Earth pressure balance shielded TBM Apart from the high separation costs and environmental hazards involved, the confined space in a big city like Tokyo, Seoul, EPB machines were developed as a substitute system against the slurry type TBM. Operation principle of EPB is very simple to

74 Copyright © 2004 Taylor & Francis Group plc, London, UK

10

Permeability Factor K [m/s]

Cobbles

Figure 2. Schematic figures showing the principles of the shielded TBM.

09096-09.qxd

2/18/04

11:51 AM

Page 75

Table 1. Comparison of two different shield machine types.

Face support

Earth pressure support

Liquid support

•

Dull regulation and control characteristic of the support pressure, influenced by many parameters Increased risk in soils with high permeability; face support in coarse grain size (sandy gravel with cobbles up to  250 mm, k  1*102 to 1*104 m/s) requires a lot of experience in EPB tunneling Increased blow-out risk in case of low overburden (For example, km 15  500: 1.0 m to pipeline, km 15  520: 2.5 m to surface, km 15  900: 0.7 m to foundations of Seoul Subway Line No.9) High expenditure for soil conditioning especially in gravel

•

•

•

•

• •

Exact and quickly responsive support pressure regulation in unstable face support situations (e.g. due to high permeability) Increased safety against settlements in permeable soils (sands, gravel) and at low overburden due to the liquid face support – lowered blow-out risk Increased safety in case of low distance bridge foundations

Ram loads

•

Increased ram loads due to noncompressible nature of the support medium, possible damage to the lining

•

Moderate ram loads due to cushion effect of slurry (and compressed air)

Material discharge

•

Difficult pressure suppression in highly permeable soils (k  1 102 to 1 104 m/s)

•

Pressure reduction over pipe length

Wear

•

Ground contact of tools and chamber structures on all sides, high wear in particular on screw conveyor and cutter head Uncertainties concerning long-lasting functionality of disks, increased number of interrupts for tool inspection/ replacement expected

• •

Reduced wear on cutter head, tools and chamber structures Wear in tubes and pumps exists but can be controlled

• Time consumptive procedure (clearing the working chamber from soil) • Difficulties in generating a membrane if

•

High stability due to betonies filter cake and air pressure

•

•

For Hydro-/Mixshield: submerged wall gate valve allows work under atmospheric conditions, full face support still maintained

•

Chamber access (face support with pressurized air)

tunnel face is unstable, possible difficulties in restart

Service and repair of screw conveyor/ in suction area

Entire evacuation of working chamber necessary (in particular for works on front spiral), high expenditure for service under atmospheric conditions

To assist in forming the “Plug” the excavated material may require conditioning. The conditioning agents frequently include bentonite, long chain polymers, and foam. Usually the cutterhead of EPB machines has a more closed design as compared to slurry type machines, and voids in the front plate allow for passage of excavated material into the working chamber. As for the boulder handling, the EPB machines cannot be fitted with a rock crusher because that the excavated material has to be mixed within the working chamber to produce a homogeneous support medium, therefore, boulders have to be broken exclusively by roller discs. In order to replace the cutter tools, the working chamber must be emptied from soils and filled with compressed air in the same manner as a slurry type machines. EPB machines are

most in cohesive clayey-silty and silty-sandy soils, which allow the formation of an appropriate earth mixture in the shield chamber with any water conditions. 1.3 Hydroshield TBM The principle of hydroshield is a modified slurry type shield system. The most important design elements of this machine are the separation of the excavation chamber by the submerged wall and the support pressure at the fluid supported tunnel face is regulated. A submerged wall with a gate at the invert of the TBM separates the front compartment from a second chamber, where a compressed air cushion maintains and controls the fluid pressure. Therefore, face support

75 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-09.qxd

2/18/04

11:51 AM

Page 76

settlement during the TBM operation. Actually, the advantage of given face stabilization a TBM based on the hydroshield principle is favored over a slurry type shielded TBM. Therefore, the designer recommended a hydroshield (Slurry) TBM with stone crusher and a cutterhead, allowing to change of cutting tools. In general, the design of TBM is the responsibility of the machine manufacturer according to the given geological conditions. The cutter head of the TBM must be equipped with suitable tools for the whole range of expected soil types. For this project, the capability to cut hard rock must be provided for tunneling in the western section of the project area. Further, in the alluvium tunneling, provisions should be included for breaking up and removal of boulders embedded in the ground matrix. It is the contractors responsibility to identify the ground conditions and has to meet all requirements of the technical specifications. The control of ground movements shall be maintained to minimize ground surface settlement by the loss of ground and potential damage to all adjacent structures and services. The machine characteristics shall comply with the following technical data.

and fluid circulation are decoupled, and so it is possible to compensate for eventual volume imbalance quicker and more accuracy than an ordinary slurry type TBM. The gate in the submerged wall can be closed with a hydraulic sliding door, then, it is available to empty the second chamber for maintenance work on the stone crusher or for clearance of the suction inlet, while the excavation face is still supported by fluid in the front compartment. 1.4 Mixshield TBM The mixshield is a proprietary name used by Herrenknecht in Germany for a shield TBM to be adapted to variable ground conditions. The mixshield provides the possibility to operate the TBM alternatively in EPB modes or in slurry mode. Table 1 shows the two different shield machine types regard to practical machine operation and tunneling safety. 2 OPTIMAL SHIELD MACHINE SELECTION FOR THE ALLUVIAL ISLAND The ground conditions relevant of Yeoi-island section show a distinct geological difference between the eastern and western part of the alignment. In the eastern part between km 15  860 and km 17  020, the subway tunnel is expected to be located entirely in alluvial deposits, predominantly in sandy gravels. For the western part between km 14  956 and km 15  860, the geological conditions are considerably more complex, varying between hard rocks, weathered rocks, residual soils, and alluvium, with the higher proportions in weathered rocks. The alluvial deposits probably dominated by non-cohesive sandy gravels under water saturation are expected to be favorable for the application of a slurry shield. For a more detailed assessment reliable grain size distribution curves of the alluvial deposits are necessary to deem the selection of machine types. Weathered rock and residual soil contain a higher percentage of clayey material and are expected to be more suitable for EPB shield excavation. However, through the excavation of core stones, boulders, and hard rocks machine will produce chips instead of an earth mixture. In addition, the use of plastifying additives must be considered carefully with regard to environmental protection regulations for free muck disposal. The present decision is to employ only one machine for the entire TBM tunneling, because of the consumptive construction time and the total tunnel length, only 4 km. Through these tunneling conditions, slurry shielded type TBM is highly recommended to choose a machine with liquid face support to be able to tunnel through the alluvial layers safely and with minimum

3 TREATMENTS OF EXISTING BOULDERS DURING TUNNELING The presence of cobble, boulder, and corestones can present significant difficulties during the shield tunnel operation in alluvium deposits. Design the shielded TBM should be based on the information presented in geotechnical investigation report. The TBM should have the capability to handle boulder and cobbles. Normally, modern TBM should be provided with a robust cutterhead with double bladed cutter disks, which can break boulders and remove the debris. The cutter head and the mucking system should design to operate in the presence of this type of materials (Fig. 5). During the tunnel excavation by shielded TBM, it is essential to monitor and record continuously all operation parameters, especially such parameters that may affect ground movements and settlements. As far as bigger boulders, cobbles, and corestones are concerned, they may be broken by twin disk cutters, but sometimes these boulders will be the main reason of delay in the construction schedule. A slurry type machine shall be equipped with a stone crusher optionally, which can crush boulders up to a size of 500 mm in diameter. Practically, some of the boulders are not crushed by the cutterhead because the soft soil matrix allows the boulder to move, in such a case, the TBM machine must be stopped and prepared for intervention inside the excavation chamber. If the corestones are located at the front of cutterhead in a

76 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-09.qxd

2/18/04

11:51 AM

Page 77

to prevent water inflow. The working miner can enter into the chamber through the airlock to work at the face front of the cutterhead. The boulder can be broken up with hydraulic breakers or drilled with hand held drilling machines, and then crushed by blasting or expanding mortar. After the crushing the boulders, the TBM can resume the normal boring operation. 4 GEOPHYSICAL PROBE SYSTEM FOR THE BETTER TUNNEL OPERATION No matter how they investigate the ground conditions at the design stage, it is hard to avoid the unforeseen geological conditions during construction work because of variations in ground conditions, the variation of ground materials will be much easier to predict with a forward looking geophysical technique. A possible method is seismic prospecting in soft ground that is currently used in EPB and slurry shield machine like TSP by Amberg consultant Switzerland. The Sonic soft ground probing system by Herrenknecht uses acoustic reflection to record contrasts in physical properties within soil formations. A special coded acoustic signal travels into the ground after being emitted by a sonic transmitter, installed on one of the cutting wheel arms. These methods allow detecting and localizing irregular bodies and obstacles within a range of about 50 m ahead of the tunnel face in maximum. However, accuracy and classification results of these surveys are still far to reach for the practical purposes at the tunneling sites. For the improve of this kind of blind shielded TBM, boulder detection technology shall be developed for the better machine operation to reduce the risk of adaptation of Shield tunneling technology.

Figure 5. Boulders should be handled properly during tunnel excavation. Table 2. Optimal technical specification of the slurry type shielded TBM Shield type

Slurry/Hydroshield

Design static pressure Cutter head type

3 bar Option 1: cutter head with spokes and rim Option 2: closed cutter head with minimum passages of 200 mm Drag picks and twin roller disk cutters Left and right 3500–4000 Nm 10000 hours 10000–12000 kN 5000 kN 2000 mm

Tools Direction of rotation Torque Main bearing lifetime Maximum thrust force Emergency thrust force Push ram stroke (for 1500 mm segments) Slurry pump capacity Tail shield seals Tail shield seal grease lines Man lock type Man lock operating pressure Tail shield articulation Tail shield grout injection lines Tail shield grout injection sensor Max. advance rate Possible ring assembly time

1200 m3/h 3 rows of brushes plus steel plates outside 8 No., 4 No. per chamber

5 RESULTS This subway tunnel project shows an example for a railway tunnel construction in soft alluvial deposits with higher ground water levels in a congested urban area. Although the prevailing ground conditions do not indicate corestones and boulders along this planned alignment, designer shall choose the tunnel excavation method by considering the poor ground conditions to increase the safety during the construction period. It is proven that the ground water could be controlled by slurry type shielded TBM. In this project area, most of the overburden depth ranges from 10 m to 20 m with low earth pressures. By adopting the proper construction sequence, geophysical probe system is proposed for detecting the boulders ahead of the tunnel working face and to prepare the suitable boulder handling to keep the given construction time schedule. Finally, this slurry type shielded TBM is suggested as an optimal tunnel excavation method in

Double chamber 3 bar Yes 4 No. with cleaning facilities 4 No. 60 mm/min 20 min

shallow depth, it is possible to crush them by the direct drilling work from the surface, and so on. To gain access to the excavation chamber, it will first be emptied, keeping the required pressure by means of compressed air in order to support the tunnel face and

77 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-09.qxd

2/18/04

11:51 AM

Page 78

picks, Ph.D Thesis, The University of New South Wales, Sydney, Australia, May, 1992. Maidle, B., Herrenknecht. M. & Anheuser. L., Mechanised Shield Tunnelling, 1995. Belling, W. & Eisenbach, R., Schwierigkeiten und Stillstande beim Shildvortrieb mit Flussigkeitgesestutzter Ortbrust und Uberwindung der Storfaktoren Durch Einsatz Eines Neuen, Veranderten Schildes, Forschung und Praxis 33, 1990. Fong, M.L., Bednarz, S.L, Boyce, G.M. & Irwin, G.L. History and explortation redefine Portland’s West Side CSO Tunnel alignment. North American Tunneling 2002, Seattle Washington USA.

such a soft, poor alluvial deposits, including the possibility of boulders with water saturated conditions. This will also help overcome the noise and vibration problems at nearby buildings with low settlement by the strong slurry pressures in downtown area of Seoul. REFERENCES Muirwood, A.M. The circular tunnel in elastic ground. Geotechnique 25, No.1, pp 115–127, 1975. Jee, W.R. An assessment of the cutting ability and dust generation of Polycrystalline diamond compact insert

78 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-10.qxd

2/18/04

11:51 AM

Page 79

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Estimating ground loss from EPB tunneling in alluvial soils for ECIS project, Los Angeles Theo Robert Seeley City of Los Angeles, Department of Public Works, Bureau of Engineering, Los Angeles, CA, USA

ABSTRACT: Urban environments require a higher degree of ground loss control than other areas. To accomplish this the City of Los Angeles required pressurized face support for tunneling in alluvial soils above the water table on the East Central Interceptor Sewer tunnel. To monitor the ground loss the City’s Geotechnical Engineering Division along with the Contract Administration and Soils Laboratory gathered data from the tunneling operation to estimate the ground loss. This paper will present the various methods used to gather data as the operation progressed, and methods used to analyze the data to determine if the contractor was in compliance with the specifications. Specifically to determine expected bulk volumes of muck for the various soil types encountered. This was then used to determine where ground treatment would be attempted. Ground treatment after tunneling was performed at various recommended locations. The success of the ground treatment provided feedback on how accurate the methods of analyses were. This case study will present the various methods of gathering the data, the methods of analyzing the data, and the accuracy of the analyses made from the data.

mode. The specifications allowed only 19 mm of settlement at the ground surface. The tunnel alignment was then broken into four units. The ECIS alignment with its four units is presented below in Figure 1.

1 INTRODUCTION Historical background

The East Central Interceptor Sewer (ECIS) consists of approximately 18.46 km (11 miles) of new sewer to add additional capacity to the City of Los Angeles existing North Outfall Sewer (NOS). The NOS was constructed in the 1920s and is now flowing at full capacity. Although the NOS is strictly a sanitary sewer at times of heavy rains it has overflowed by raising maintenance hole covers in the low lying sections of the city. Various routes and construction methods were considered for nearly a decade prior to construction of ECIS. The primary concerns were to build a new trunk line across the Los Angeles basin that would gravity flow with the least amount of disruption to the citizens. As late as 1997 a cut-and-cover option was considered. This was dismissed as too disruptive to the citizens. Past tunnel projects including the new metro rail system have also been disruptive due to excessive settlement. The primary goal of the ECIS tunnel would have to be the control of settlement. The City of Los Angeles enlisted the help of a Technical Advisory Panel to develop a plan to minimize the ground settlement. Their recommendation to accomplish this was to require that the tunneling be done using new tunnel boring machines (TBM) that would be operated in earth pressure balance (EPB)

1.2

The west to east alignment of ECIS crosses the Los Angeles Basin from the tunnel’s lowest point on the

Hollywood

From San Fernando Valley

101

5

it

Un

1

Unit 2

Unit 3

Downtown L.A.

t4

Blair Hills

Uni

110

NORS

Culver City LAX

Baldwin Hills

International Airport

To Hyperion Treatment Plant

0

Figure 1. Project location.

79 Copyright © 2004 Taylor & Francis Group plc, London, UK

Geology and alignment

2

4

6

Los Angeles River

1.1

8

710

10 km

09096-10.qxd

2/18/04

11:52 AM

Page 80

Figure 2. Geologic cross-section.

To the east of the hills, the alignment traverses the nearly flat topography of the Ballona Gap and the flood plain of the Los Angeles River established prior to 1825. Uplift along the Newport-Inglewood Fault zone resulted in ponding and the formation of swamps and buried peat deposits in the low lands just east of the Baldwin Hills. Interbedded peat and silty clay deposits have been observed to a depth of 12 m in the area (City of Los Angeles, 1995).

west side of the Baldwin Hills to its highest point on the east bank of the Los Angeles River. Unit 1 passes under the Baldwin Hills and then turns north into a low-lying flood plain. Unit 2 crosses the low-lying flood plain running upstream along the former historic Los Angeles River alignment from west to east. Unit 3 continues east under a slightly steeper portion of the flood plain and finally turns north for the last 540 m. Unit 4 begins north and quickly turns east again for about 3 km and then turns north along the west side of the Los Angeles River channel. Finally the last 400 m turns east passing under the concrete lined section of the Los Angeles River to its terminus at the Mission and Jesse shaft. The surface feature of the Los Angeles Basin covers an area 75 by 30 km wide. It is primarily a lowland coastal plain that slopes gradually southward and westward toward the Pacific Ocean. The plain is interrupted by a series of small hills created by the uplift of the Newport-Inglewood fault. The fault trends northsouth across the west side of the coastal plane. The uplift of the Newport-Inglewood Fault created the Baldwin Hills on the west side of the fault. This fault zone is also responsible for creating the structural traps for the oil fields found along the NewportInglewood Fault zone (Yerkes et al., 1965).

1.3

1.4

Unit 2

Unit 2 tunneling encountered a wider variety of geologic conditions that varied from buried peat deposits to very dense sand and gravel deposits. The middle 1.6 km

80 Copyright © 2004 Taylor & Francis Group plc, London, UK

Unit 1

Unit 1 geologic conditions encountered by the tunnel machine were favorable in terms of controlling ground loss. For all but the northern most 300 m, the tunneling encountered the very stiff to hard clays and silts of the San Pedro Formation that form the Baldwin Hills. The 300 m north of the San Pedro Formation consisted of very dense silty sand and very hard clayey silt.

09096-10.qxd

2/18/04

11:52 AM

Page 81

Unit 3 East, it was refurbished and reinserted into the Grand Avenue Shaft to mine Unit 3 West.

of this 4.8 km reach was in a mixed face condition. The upper half of the face was in younger alluvium that included the peat deposits, while the lower half was in the very dense granular deposits of the Lakewood Formation. The shallowest ground cover condition was approximately 5 m and coincided with the mixed face condition. Also, the groundwater table (GWT) fluctuated in this area over time. The GWT at Maintenance Hole 7, near the middle of Unit 2, varied from 6.5 m to 8.5 m below the ground surface. The tunnel at that location is 7 m below the ground surface. 1.5

1.8

At the request of the contractor, the TBM design was modified to move the auger screw up to allow for a larger main bearing on the cutter head. The contractor also had the TBMs modified to accept a muck ring and pressure relief gates. The Geotechnical Baseline Report (GBR) discussed a similar modification to an EPB machine used for the North Outfall Replacement Sewer. The final conclusion of the GBR on that issue was that the pressure relief gate does not maintain pressure in the chamber and therefore there is no earth pressure support. The first TBM started east out of the Unit 3 Grand Avenue Shaft Site with the muck ring modification. After about 100 m of mining the contractor chose to remove the muck ring primarily due to ground loss problems. Muck rings were not used again on the ECIS project.

Unit 3

Unit 3 geology was considered to be relatively favorable with the tunnel alignment completely in the dense granular soil deposits of the Lakewood Formation and above the water table. The cover ranged from 10 m at the west end to 25 m at the east terminus. The ground surface sloped uniformly up stream at a rate of 0.26% steeper than the sewer. 1.6

Unit 4

Unit 4 the tunnel alignment was also completely in the dense granular soils of the Lakewood Formation. Near its eastern terminus it is below the groundwater table which is locally higher in the vicinity of the Los Angeles River. Near the middle of Unit 4 the ground cover decreases to 10 m of Lakewood Formation overlain by 7 m of younger alluvium and fill soils. 1.7

Modifications to TBM

2 CONSTRUCTION MONITORING 2.1

Construction overview

The contract was awarded January 5, 2001, to Kenny, Shea, Traylor, Frontier-Kemper J.V. The construction started February 14, 2001. First tunneling occurred out of the Grand Avenue shaft site on December 14, 2001 with the last hole through on September 26, 2003 at the west end of Unit 4. The project was mined with four identical Lovat TBMs with a cut diameter of 4.714 m. The TBMs met the City’s requirement for new site customized Slurry or EPB-TBMs capable of working above and below the GWT. The tunnel segments have an inside diameter of 4.150 m and are 200 mm thick. The straight segments were 1.524 m long and curved segments for a 150 m radius were 1.375 m long. Tunneling began out of the east side of the Grand Avenue Shaft Site excavating Unit 3 East. The second TBM started out of the Siphon Shaft at the west end of Unit 2 mining eastward. The third TBM started mining Unit 4 from the east end proceeding west and south toward Unit 3. The fourth TBM started mining Unit 1 heading west and south from the outlet side of the Siphon Shaft. After the first TBM completed

2.2

Analyzing data

The data obtained by these five sources was then analyzed and sent to the Construction Management (CM) Team. When the analyzed data indicated that excessive ground loss was occurring, the CM Team notified the contractor so they could modify their mining operation, as they deemed necessary.

81 Copyright © 2004 Taylor & Francis Group plc, London, UK

Sources of information

The primary goal of the ECIS project was to build a new sewer with the minimum amount of disruption to the public. Therefore, construction monitoring was a high priority. This was done in five different ways. First, the ground surface along the alignment was monitored continuously by full time survey crews. Second, the contract required that the contractor install 268 Multiple Point Borehole Extensometers MPBX for the City to monitor settlement at depth. Third, the City’s inspection division monitored the mining operation continuously and estimated the muck volume for each shove. Fourth, the City’s soil laboratory sent technicians to obtain samples of the muck as it was being mined by the four TBMs. And fifth, the contract required that the TBMs be equipped with automatic recording system for data collection. The data was transmitted to the surface for both the City’s and the contractor’s use.

09096-10.qxd

2/18/04

11:52 AM

Page 82

tunnel elevation. However, this did not always give reasonable values. The typical spacing of boring information was on the order of one boring every 50 m to 100 m. This is because driven samples in very dense granular soil tend to disturb the sample and give a lower dry density. Also, in the space between borings the mined soil would typically change consistency several times. However, the boring logs did show that soil in the Lakewood Formation was consistently very dense. To resolve this problem GED requested the compaction tests to determine the laboratory maximum dry density. Then, based upon the fact that the mined soil in Units 3, 4, and the east end of Unit 2 are very dense, a relatively high percentage of the laboratory Maximum Dry Density was found to give reasonable results. The percentages used were 92% initially and later we found that 95% relative compaction gave more reasonable results.

3 SOIL SAMPLING 3.1

Goals of soil sampling

The goals of soil sampling were to provide: 1. Samples of the condition muck to test the quality of conditioning. 2. Samples for determine the wet and dry density of the conditioned muck. 3. Samples for estimating the in-place density of the mined soil. 4. Database of mined soils for future comparison and create a permanent record of the materials mined, since EPB mining precludes any regular mapping or sampling from the face. 3.2

Sampling methods used

The primary test used to determine the muck conditioning was the slump cone test. This utilized the same cone test procedure used for concrete testing. For determining the wet density, samples of the muck were poured into 150 mm diameter 300 mm high cylinders with a minimum amount of compaction effort. To simulate the compactive effort of muck falling off the end of the conveyor belt into the muck car, the technician filled the cylinder in three lifts. The side of the cylinder was tapped lightly after each lift. Excess soil from the final lift was then carefully trimmed and the cylinders were then capped. Full capped cylinders were then weighed in a field lab. Given the known volume of the cylinder, the weight was then converted to determined wet density. A representative sample was then taken from the cylinder to determine the moisture content and calculate the dry density of the muck in the tunnel prior to transporting to the shaft. This was important because the inspectors estimated, to the nearest cubic meter, the volume of mined muck at the heading. When the muck was saturated granular soil, it was observed that the muck would consolidate during transportation. Therefore, the technicians would only add a small amount of compactive effort to simulate muck falling from the conveyor belt to the muck car. The muck for a typical shove would fill three to four muck cars. Muck cars were designed to hold 10 m3 when filled to the brim. Samples were taken from the second or third car. This was to avoid sampling muck that was on the conveyor belt or in the plenum when the machine was not mining. 3.3

3.4

During the initial stages of mining, the City of Los Angeles, Division of Standards, under the direction of GED, started taking samples of the muck as it was being loaded into the muck cars from the conveyor belt. This sampling operation was moved to the surface as the tunnel length increased. The majority of the sampling was done at the surface as the muck was being removed from the shafts. Near the end of mining Unit 4, which was the longest drive, GED requested sampling the muck at the heading and then take a second sample from the same muck car as it was being unloaded at the ground surface. The Unit 4 shaft site is located at Station 18460. The results of this study are presented on Figures 3 and 4 where moisture content and bulking factor are compared. The bulking factor is a ratio of the in-place density to the muck density. Therefore, two tests are required for each data point, and six of the seven sets of points are within 1.5%. An attempt was also made to compare the slump of the muck at the heading and at the shaft; however, all samples in this study exhibited zero slump. On one occasion earlier in the project the slump was taken in the tunnel of Unit 2 at Station 5092.0 and again at the surface. The elapsed time between samples was one half-hour. The slump at the heading was 12 mm and at the shaft 3 mm. The one sample at Station 13340 shows that a higher moisture content and lower bulking factor was measured at the heading of the tunnel. This sample was classified as GP-GM. Its counter part that was sampled at the shaft from the same muck car was classified as SM silty sand. This variability in soil type appears to be the reason for the difference in test results. The other sample pairs did not have this variability.

Determing in place density

Near the beginning of the project, boring log data was used for the dry density of the in-place soil at the

82 Copyright © 2004 Taylor & Francis Group plc, London, UK

Checking soil sampling procedures

09096-10.qxd

2/18/04

11:52 AM

Page 83

a ground loss was to manually drill up from the crown of the tunnel approximately 3 m and fill voids above the tunnel. This method had a short window of time. Within a matter of hours the mining operation would build enough rings to pass the location where the miners could stand on top of the trailing gear and hand drill up. Also, the grout was pumped under pressure using the pump that was near the front of the trailing gear. Therefore, a quick qualitative test was needed to decide if the muck counts indicated lost ground or just bulked up soil. Observations early on were that the clean SP sands had a higher probability of creating chimney-type cavities. They were also found to have a lower bulking factor. The calculation for the bulking factor required drying the soil back to optimum moisture and running a laboratory maximum dry density test. This typically took a half a day or more. Therefore, it was decided to experiment with the Sand Equivalent test to help determine if there was a problem or not. The California Test 217, Method of Test for Sand Equivalent was developed as a field test primarily for cut and cover pipe-laying operations. Its primary use in construction is to determine if proposed pipebedding material is acceptable. What makes pipebedding material acceptable is its ability to flow into the cavity under a pipe laid in a trench when a small amount of water is added. Ironically, this ability to flow into open spaces is what made the SP sands so difficult to mine without loosing ground. When the screw conveyor was taking soil out of the plenum faster than the advance of the TBM, then the clean SP sands would drain into the TBM like sand in an hourglass. The sands of the Lakewood formation were typically very dense, and dry with virtually no cementation. The vibration of the TBM was enough to loosen the dry sand and cause it to flow. The only thing preventing ground loss was the EPB pressure. Our observation was that if the SE was over 30% and the EPB pressure was low or fluctuating, then ground loss was likely to occur. A comparison of the bulking factors and SE has been attempted to find a correlation. The results were somewhat erratic, but in general, a SE over 30% indicated the soil would have a low bulking factor and was more likely to run if not supported. Soils with a SE below 30% have a longer stand up time and higher bulking factor.

Figure 3. Moisture content comparison.

Figure 4. Bulking factor comparison.

4 LABORATORY TESTS 4.1

Development of tests

Initially, the laboratory testing consisted of the moisture content (ASTM D2216-98) and density of the muck along with grain-size analysis. After about 100 m of mining, the laboratory maximum dry density (ASTM D1557-91) was added. Conventional concrete slump tests (ASTM C143/C143M-00) as described in EFNARC (2001) were also run on the muck as it was being excavated. These were performed both at the heading and at the shaft site. Atterberg limit tests were run on the cohesive soils. The primary problem with most of these tests was the time it takes to run them. By the end of the project we had added visual classification of dry, pasty, or wet.

5 ANALYSES FROM TEST RESULTS 5.1

4.2

Sand equivalent tests

The use of the sand equivalent test came out of a need for fast turnaround on the laboratory tests. The problem was that the time to react to a lost ground condition is very short. The primary method used to correct

83 Copyright © 2004 Taylor & Francis Group plc, London, UK

Using sand equivalent tests

The primary use of the SE test was to help the inspector in the field decide if ground loss was occurring or not. For example, if the muck volume per shove stayed the same but the SE increased from 20 to 45, then it was very likely that the bulking factor had decreased.

09096-10.qxd

2/18/04

11:52 AM

Page 84

5.3

Therefore, the only reason for the muck volume to stay the same is the TBM was taking in more ground than it was before. At that point, the inspector would require more ground loss check grouting and ask the operator to modify their mining operation. When the laboratory had completed the tests necessary for calculating the bulking factor, that data was relied upon instead of the SE test. 5.2

The sampling and laboratory test results from the mixed face condition encountered in Unit 2 were not satisfactory. The primary problem is that mixing two soils of different gradation creates a new soil that does not resemble the two parent soils and the laboratory analyses are useless. 6 OVER EXCAVATION CHARTS

Bulking factor calculations

The over excavation charts were developed to take into account all of the pertinent lab and field data to determine the amount of soil mined per shove. As the laboratory data started coming in and a method of estimating the bulking factor was developed, it became clear that analyses could be made to calculate the ground loss. This was done on a spreadsheet and then graphed on a bar chart. The spreadsheet takes into account the inspector’s estimated volume of muck, in-place density, density of the muck, theoretical grout volume and actual annulus grout volume. Using these factors against the theoretical cut volume it was possible to develop the bar charts that indicated the amount of ground loss per shove. The results of these bar charts had to be moderated against the fact that there was one sample for every 9 to12 rings in the Lakewood Formation. There was some variability in the muck density. This typically showed up as abrupt changes in the bar charts. Using a moving average solved this. The bar charts presented on the next page used a moving average of three muck density values. The in-place density was taken from averaging laboratory maximum dry densities over a range where the mined soil strata was similar. Examples of over excavation charts are included as Figures 6 and 7. These were chosen to demonstrate that the volume and location of cavities could be estimated. Figure 6 shows a drop in the over excavation volume between 11706 and 11715. This area was pressure grouted from the spring line of the tunnel up to several meters above the tunnel for a Maintenance Hole. The Maintenance Hole was eliminated from the project after the grouting was completed. Between Station 11790 and 11800 there was an attempt to fill cavities simply by drilling a 200 mm diameter hole and filling it with grout. This drill and fill repair method was not satisfactory to the City. However, GED observed that 11 of the 14 borings drilled at that location either hit grout from a previous boring drilled and filled the year before, or found cavities. The previous boring took 9 m3 of slurry to fill. This drill and fill method was not successful because the borings caved about half way down to the tunnel crown. The total amount of grout pumped in the 14 holes was 19 m3. So the total fill to date is 28 m3 and more compaction grout will be required. From the chart the last

The bulking factor analysis that was developed by GED for the ECIS project and first reported by Crow and Holzhauser (2003) is simply a ratio of the dry density of the muck as it is being loaded into the muck cars from the conveyor belt to its in-place dry density prior to being mined. Figure 5 shows how this was derived. The analysis is simple; the main problem, as discussed above, is obtaining the correct data to put in the equation. The dry density obtained from typical soil sampling methods is lower than the in-place density. This is due to sampling disturbance. Then, there is the problem of simulating the compactive effort of muck falling from the conveyor belt into the muck car. The laboratory results from this project indicated that the cohesive soils are the difficult soils to estimate the compactive effort for muck falling into a muck car. However, the cohesive soils have a long standup time and are usually not a problem for ground loss. The granular soils, that are a problem for ground loss, gave more consistent results. This along with the fact that the bulking factor for granular soils is typically only 10% to 20%. This smaller range in bulking factor means that the analysis will usually be close to the true bulking factor. Analysis of the expected excavated soil volume by use of the Bulking factor BEPB In situ

Excavated Air

Air Water

VA,e

VA,i VW,i ,GW,i

Water

VW,e, GW,e

Solids

VS,e, GS,e

Ve

Vi Solids

VS,i ,GS,i

Note: V : Volume [m3] Index i: in situ G : Weight [kN] Index e: excavated γd : Dry unit weight [kN/m3]

Bulking Factor due to EPB-tunneling: Dry unit weight of soil in situ: Dry unit weight of excavated soil: WithVG,i = VG,e use (2) and (3) in (1): Note:

Excavated soil volume includes bulking due to boring process and addition of conditioning agents

BEPB =

Ve - Vi Vi

(1)

γd,i

=

GS,i Vi

(2)

=

GS,e Ve

(3)

γd,e

BEPB

Limitations on calculating the bulking factor

γd,i - 1 [-] = γ d,e

(4)

92% of the max lab dry unit weight was taken as the dry unit weight of soil in s

Figure 5. Bulking factor analysis.

84 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-10.qxd

2/18/04

11:52 AM

Page 85

7 MONTHLY TUNNEL SUMMARY CHARTS About one third of the way through the tunneling a monthly set of charts was developed and delivered to the construction management team. These charts were produced in 100-m sets. Figure 8 as shown on the previous page, is an example of these charts. The purpose of these Tunnel Summary graphs is to summarize the quality of the tunneling. They were summaries of the extensometer data, ground survey data, inspector’s non-compliance and their related job orders, and the estimated ground loss from the over-excavation charts. The charts were color coded to indicate where there was no major ground loss

FACE STATION

Figure 7. Over excavation chart 118 to 119.

Figure 8. Monthly tunnel summary charts.

85

118+98.93

118+92.81

118+86.69

118+80.57

118+74.45

118+68.33

118+62.21

118+56.09

118+49.97

118+44.46

118+37.73

118+31.61

118+25.49

118+19.37

118+13.25

118+07.13

FACE STATION

Figure 6. Over excavation chart 117 to 118.

Copyright © 2004 Taylor & Francis Group plc, London, UK

ECIS UNIT 3E STATION 118+00 TO 119+00 20.0 17.0 14.0 11.0 8.0 5.0 2.0 -1.0

118+01.01

OVER EXCAVATION [m3]

117+91.83

117+85.90

117+79.78

117+73.66

117+67.54

117+61.42

117+55.30

117+49.15

117+43.03

117+36.91

117+30.79

117+24.67

117+18.55

117+12.43

117+06.31

ECIS UNIT 3E STATION 117+00 TO 118+00 17.0 14.0 11.0 8.0 5.0 2.0 -1.0

117+00.19

OVER EXCAVATION [m3]

7 shoves took in about 38 m3. In this same area a reference hole was drilled 6 m north of the tunnel centerline. It did not cave or encounter cavities. The reference hole only took the theoretical volume of grout to fill. Compaction grouting was performed in 6 holes drilled along the centerline between Stations 11856 to 11863.5. The total for all 6 grout holes was 12 m3. However, prior to compaction grouting exploratory borings were drilled at 11859 and 11861 and they each took 5 m3 of grout. So a total of 22 m3 were pumped or poured into an area five shoves long for an average of 44 m3 per shove. In this area the ground loss was calculated to be about 10 m3 per shove.

09096-10.qxd

2/18/04

11:52 AM

Page 86

However, in Units 3 and 4 the dense granular soil of the Lakewood Formation proved to be more difficult to mine. 5. Preparing Tunnel Summary charts on a monthly basis provided a good reference to keep track of the tunneling, and the repair of the few areas where ground loss occurred. 6. Sampling the muck at the shaft site gave reasonable results.

problems, apparent problems, problems that were severe enough to result in a non-compliance and areas where the contractor had completely filled the cavities created by the ground loss. These were colored green, orange, red, and yellow, respectively. The charts also showed the amount of cover relative to the tunnel size. Graphics were also developed to show where various types of grout were used to fill cavities and there estimated elevations and volumes. Below the tunnel cross-section the survey data and the MPBX results were plotted. Where the surface settlement exceeded the allowable 19 mm or the bottom anchor of the MPBX exceeded 32 mm, they were plotted in red. The primary problem with the MPBX and survey data is the spacing between points. This is where the over-excavation charts were used to define the limits of the problem areas. These areas have since become known as the Red Zones. The City of Los Angeles is presently negotiating with the contractor to have the cavities filled. It should be noted that the majority of these charts were colored green meaning that the majority of the tunneling was satisfactory.

ACKNOWLEDGMENTS The Author wishes thank the other members of the Geotechnical Engineering Division who helped develop the charts and graphs used on the ECIS Project, and the members of the Division of Standards Soil Laboratory that obtained and tested the soil. Without this data the development of the methods to estimate ground loss would not have been possible.

REFERENCES 8 CONCLUSIONS

Crow, M.R. & Holzhauser, J. 2003. Performance of Four EPB-TBMs Above and Below the Groundwater Table On the ECIS Project Los Angeles, CA, USA. Proceedings of Rapid Excavation and Tunneling Conference, Society for Mining, Metallurgy, and Exploration, Littleton, Colorado, USA p. 905–931 EFNARC. 2001. Specification and Guidelines for the specialist products for Soft Ground Tunneling. ENFARC, Aldershot, UK Yerkes, R.F., McCulloh, T.H., Schoelhamer, J.E. & Vedder, J.G. 1965. Geology of the Los Angeles Basin, California – An Introduction, U.S. Geological Survey Professional Paper 420-A

1. The volume of ground loss and its limits can be estimated. 2. The bulking factor calculations give reasonable results except in split face condition. 3. Sand Equivalent test is useful in tunneling as a quick test to check on changes in the material that is being mined. 4. During the design of ECIS, Unit 2 was believed to be the area where ground loss would be a problem due to the shallow cover and mixed face condition.

86 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-11.qxd

2/18/04

11:53 AM

Page 87

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Some aspects of grouting technology for Manhattan tunnels M. Ryzhevskiy STV Incorporated, New York, USA

P. Barraclough Parsons Brinckerhoff Quade & Douglas, Inc., New York, USA

ABSTRACT: Conventional methods of tunneling in hard but jointed and heavily sheared rock formations, particularly those containing groundwater, may require additional pre-excavation rock mass reinforcement or improvement methods. These methods can include pre- and post-excavation grouting for both purposes of rock stabilization and controlling groundwater inflow into the excavation. Published geological data for the Manhattan area indicates that rock mass quality is generally good to very good, although it is locally complex, containing heavily faulted and sheared zones of high permeability, which are expected to be unstable. Driving hard rock TBM’s through these zones can pose significant problems and pre- and post-grouting methods should be considered in order to minimize the risk of rock mass failure and water inundation. The methodology and procedures for pre- and post-grouting during the excavation of TBM driven tunnels have been developed and are described within this article.

phase defines the regional structure of Manhattan with the axial plain striking N35°E and generally dipping toward the south-southwest at approximately 10° to 15°. Published information identifies the presence of four major joint sets within the Manhattan Schist. However further to recent and extensive ground investigation and interpretation the structure appears to be more complex than historical data suggests, revealing significant localized faulting, shearing, alteration and folding. The existing shear zones have been identified and categorized as major and minor. Major shear zones occur on a scale of approximately 10 ft to 100 ft (3 m to 30 m) and are characterized by the original rock being sheared, brecciated, and rehealed in a mylonite matrix. The fractures are often coated with secondary minerals. The boundary between the brecciated and the undamaged rock mass is distinctive, with this zone including clusters of open infilled joints and secondary mineralization. Minor shear zones occur on a onefoot (0.3 m) scale with their associated clusters of infilled, stained, mineralized joints and slickensides being apparent on a 10-foot (3 m) scale. Manhattan Island is bound by the East River to the east and the Hudson River to the west and is slightly above sea level. Groundwater levels measured in borehole standpipes range from 15 feet to 60 feet (5 m to 20 m) below street level. The quality of the

1 INTRODUCTION Manhattan Island is the heart of New York City and is one of the most urbanized cities, housing some of the most expensive real estate in the US and the world. Manhattan imposes serious construction considerations owing to its high density of buildings including historic residential districts and hi-rise commercial and residential properties, often with deep basements. In addition to the buildings is a highly developed infrastructure with many existing tunnels and other underground structures. In Manhattan it is therefore very important that during construction, systems are provided to mitigate risk and impact caused by tunneling. 2 GEOLOGY & HYDROGEOLOGY Manhattan is underlain by Proterozoic and Ordovician metamorphic rocks, locally known as the Manhattan Prong. These metamorphic rocks are characterized by three lithologies comprising schist, gneiss and marble, although the greater part of Manhattan is dominated by the more erosion resistant schist and gneiss. The rocks of the Manhattan Prong have been subjected to multiple tectonic episodes including folding, faulting and intrusion, resulting in an intensely folded and locally sheared rock mass. The prominent fold

87 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-11.qxd

2/18/04

11:53 AM

Page 88

of ground conditions. The principle of this method is to replace the water or air contained in the ground (pores, voids, cracks, joints) with a liquid material, which sets to a solid during short interval of time. There is an abundance of choices for the liquid but most commonly used are cement mixes, sodium silicate or organic resins. In tunneling there are two major purposes for grouting: to create a barrier against groundwater flow and to increase shear strength of the rock mass in order to maintain stability of the excavation, reduce settlement and ground movement. Shallow tunnels are often grouted from the surface, so that grouting and excavation procedures can go on simultaneously. Grout patterns would normally be rows of boreholes forming squares, with the spacing between boreholes determined by local conditions. When conditions preclude working from the surface, and when tunnels are very deep, grouting can be done from the tunnel face. The thickness of grouted area surrounding the tunnel varies and depends on the ground conditions and the purpose of the grouting (ground strength and/or permeability).

rock mass controls the permeability of the Manhattan Schist and generally the permeability of the undisturbed and unweathered rock is very low. In faulted and sheared zones the permeability is considerably higher as the network of fractures behaves as conduits for the groundwater. The permeability of the discontinuities is influenced by several factors including roughness, tightness and presence of joint infill. The coefficient of permeability has been derived from in-situ packer tests and typically varies between 105 to 107 cm/sec. However in shear zones the permeability ranged between 104 to 106 cm/sec and it is anticipated that widely spaced, open, steeply dipping fractures may transmit groundwater at greater rates than indicated by packer tests. This is especially true during excavation of new underground openings where flowing groundwater entering the excavation can wash out the clay infill from slickensided joints and cataclasite within shear zones, which can result in the rock mass permeability being in the order of 102 cm/sec. 3 CONSTRUCTION CONSIDERATIONS Construction experience in New York City and the findings of geotechnical programs indicate that the rock mass is generally good quality, stable and generates moderately low amounts of water. Only shear zones are expected to be unstable and to be sources of significant groundwater flow. Typically excavation of any underground structure in similar conditions with drill and blast or TBM methods will require initial support, such as rock bolts/dowels, shotcrete and occasionally steel ribs or lattice girders. The initial rock support systems are designed to prevent failure of blocks and loosened rock mass from the crown and sidewalls of the tunnels. These support systems in the hard rock formations usually take the form of fully grouted dowels and resin anchored rock bolts in combination with steel reinforced shotcrete. Often fiber-reinforced shotcrete is used as opposed to steel wire mesh. Tunneling in unstable rock formations where stand-up time of the excavation is limited due to densely distributed discontinuities, shear zones, and water saturated zones, will require additional special pre-excavation reinforcement/improvement methods to increase the rock mass quality, thus avoiding rock instability and controlling groundwater into the tunnel. In cases where the rock mass conditions dictate the need for pre-excavation rock stabilization, pre-excavation grouting techniques can be utilized.

5 PRE-EXCAVATION GROUTING Specific procedures for pre-excavation grouting from inside the tunnel space can be developed for any tunnel configuration, shape, diameter and excavation method. The philosophy of the pre-excavation grouting is to the increase rock mass properties or to seal a limited area ahead of the face and around the tunnel, using a grout of suitable strength, low permeability to water and of high durability. In general the pre-excavation grouting method involves filling all (or as much as possible) fissures, cracks and voids for a distance of a least 50–80 feet (10–25 m) ahead the tunnel face and 180° above the springline or 360° all around the tunnel. After finishing one round of pre-excavation grouting, 70–80% of the grouted length can be excavated. Subsequent to excavation of the improved rock mass, probing of the rock ahead of the TBM face will ascertain whether additional pre-excavation grouting is necessary. This cycle may be repeated as long as required, depending on geological conditions. To achieve a “dry” tunnel, postexcavation grouting may be required later in addition. Access to the tunnel face with the drill and injection equipment is very important for effective preexcavation grouting. For tunnels excavated with drill and blast method, access to the tunnel face is not often a problem. TBM’s are rarely designed to facilitate injection drilling close to the face, due to the lack of access. However, modern TBM’s are capable of performing pre-excavation grouting over or through their shield and sometimes through the cutterhead, although the latter is a very sophisticated method.

4 GROUTING METHODS Grouting is an established and common technique in modern tunneling. Grouting can be used in most types

88 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-11.qxd

2/18/04

11:53 AM

Page 89

Figure 1. Fault zone prediction ahead of the TBM.

Figure 2. 1st round of pre-excavation grouting.

When pre-excavation grouting is to be adopted for a project, it is important to identify the area and/or volume of ground that needs to be treated and where it is with respect to the tunnel face (Fig. 1). For this purpose probe holes must initially be drilled ahead of the cutterhead to establish the nature of the rock mass and groundwater conditions (location, flow rates and pressure). Tunneling in intensely fractured and crushed rocks, where groundwater inflow is minimal will only require an improvement of the rock mass properties for maintaining stability. This is accomplished by pre-excavation grouting over the TBM shield up to 180° above springline. The angle at which the rig can drill the holes into the tunnel wall will govern how far ahead of the TBM the grout can be injected. The usual set-up for this drill equipment has to allow for drilling holes at a minimum of 5–10° inclined to the tunnel axis. If the holes are drilled at angles greater than this, then the drill holes will be too far away from the tunnel perimeter, greatly reducing the efficiency of the grouting. If the inflow of the groundwater from the probe holes exceeds a predetermined threshold, it will be required to drill grout holes 360° all around the TBM to achieve control of the groundwater. Thus the anticipated flow of groundwater to be encountered plays a major role in the selection of equipment and its configuration. In the cases where the amount of the groundwater can be tolerated during the construction period and where the rock mass will only require increasing its stability to prevent the rock falls behind the shield of the TBM, pre-excavation grouting would be specified above of the TBM cutterhead only (Figs 2 & 3). In these circumstances the following action should be executed: probing ahead of the TBM, pre-excavation grouting of the potentially unstable area ahead and above of the TBM shield and post-excavation grouting of the invert section to control groundwater inflow in to the tunnel.

Figure 3. Subsequent round of pre-excavation grouting.

The procedure will start with drilling a probe hole near the anticipated unstable rock formation or fault zone ahead of the tunnel face, which should be drilled up to 100 feet (30 m) in length at the 12 o’clock position. If this probe hole indicates any potential problem, such as the presence of weak rock (higher drilling rate), loss of flush water or high groundwater inflow through the drilled probe hole, then two more probe holes should also be drilled at the 9 and 3 o’clock positions to verify the findings of the first probe hole. If the additional probe holes indicate poorer rock mass quality ahead, then an appraisal will be required to judge the extent of pre-excavation grouting. Different grouting trigger levels should be adopted for specific projects. The following pre-excavation grouting trigger levels may be used for assessment whilst probing:

•

89 Copyright © 2004 Taylor & Francis Group plc, London, UK

Sudden loss of drilling water; over 50% (an abrupt change in the amount of water returning to the surface or face usually signifies that the drill has reached a highly permeable horizon)

09096-11.qxd

•

•

2/18/04

11:53 AM

Page 90

Sustained (over half an hour) groundwater inflow through any probe hole drilled ahead of the tunnel face, over the 25 gallons/minute or fi gallons/ minute/feet (100 liters/minute or 3 liters/minute/ meter) Total groundwater inflow from all drilled holes during an half an hour period exceeded 50 gallons/ minute (200 liters/minute).

If pre-excavation grouting is necessary, based on trigger levels, the grout holes should be drilled above the tunnel springline as shown on the Fig. 2, Section A. The grout holes should be drilled at least 50 feet (15 m) ahead of the TBM cutterhead through the TBM shield at an inclination between 5°–10° to the tunnel axis. The distance between the grout holes is to be determined according to local rock properties and its grout penetration value. Previously gained experience in similar rock conditions suggests that the maximum distance between the grout holes of 5 feet (1.5 m) should be used. Based on former experience in Manhattan rocks, cementitious grouts are recommended as primary injection materials. Cement can enter fissures as small as 0.3 mm and therefore has a comparable penetration/infiltration capacity as silicate or acrylic resins, but makes a much more durable, more economical and more environmentally friendly solution. The main components of cementitious grout are water and generally Type I or II Portland cement and can be altered by using other cement types, such as Type III (high early strength), Type IV (low heat of reaction), or Type V (resistance to chemical attack). By varying the water to cement ratio it is simple to change the grout’s bleeding rate, subsequently altering its plasticity and ultimate strength. Mixing the main components with additives, such as bentonite, sodium silicate, dispersants, retarders, and accelerators, will also change the grouts properties. Microfine cement is an alternative version of cement that overcomes the difficulty of using Portland cement grouts in low permeability ground.

Figure 4. Post-excavation grouting.

face, once the TBM has passed (Fig. 4). Post excavation grouting alone is usually wasteful as it is costly and often unsuccessful, but in combination with pre-excavation grouting it is very effective. Using cementitious pre- and post-excavation grouting improves the water tightness of the rock mass, achieving lower water flow rates into tunnels typically between 0.1 to 2.5 gallons (0.5 to 10 liters) per minute per 300 feet (100 m) of the tunnel. In cases where the seepage rates into tunnels are required to be more stringent, governed by the sensitivity of the environment and construction, chemical grouting (organic resin) can be utilized as an alternative to cement. 7 CONCLUSION Tunneling under Manhattan generally does not pose significant stability problems as the majority of the rock mass is good quality, therefore the methods detailed within this article are not always required. Published information identifies the existence of several locally complex and significant fault and shear zones in which the rock mass can be expected to be unstable and to yield high rates of groundwater. When such ground conditions are anticipated these techniques described herein can be considered. Among existing pre- and post-excavation rock mass improvement/reinforcement methods available, grouting is very effective in combination with TBM driven tunnels.

6 POST-EXCAVATION GROUTING Most tunneling projects cannot tolerate large volumes of groundwater during construction of the final structure, therefore groundwater ingress is required to be limited, which can be done by using post-excavation grouting in addition to pre-excavation grouting. Clearly, pre-excavation grouting all around the tunnel is more effective than post-excavation grouting alone, but it will significantly slow TBM advance rates. For this reason, when ground conditions permit, it is favored to perform pre-excavation grouting of the rock ahead and above the springline and then post-excavation grouting of the invert area, from behind the tunnel

REFERENCES Merguerian, C. 2002. Brittle Faults of the Queens Tunnel Complex, NYC Water Tunnel #3. In G.N. Hanson, Ninth Annual Conference on Geology of Long Island and

90 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-11.qxd

2/18/04

11:53 AM

Page 91

metropolitan New York. 20 April 2002, Sate University of New York at Stony Brook, NY, Long Island Geologists Program with Abstracts, 116 p. Snee, C., Sarkar, S., Benslimane, A., Stewart, C., and Osborne, C. 2003. Rock Mass Characterization for the Manhattan East Side Access Project. In P. Cullgan et al (ed.), Soil Rock America 2003 (The 12th Panamerican Conference for Soil Mechanics & Geotechnical

Engineering and the 39th US Rock Mechanics Symp.), June 22–25, 2003, Vol. 1, pp. 129–136. Ryzhevskiy, M. 1987. The main principles of the new technologies for construction tunnels in unstable rock formations. Energetic Construction. Moscow, N7. Ryzhevskiy, M. 1988. The advance experience of the chemical ground improvement by jet grouting. VPTI Transstroj. Moscow.

91 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-11.qxd

2/18/04

11:53 AM

Page 87

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Some aspects of grouting technology for Manhattan tunnels M. Ryzhevskiy STV Incorporated, New York, USA

P. Barraclough Parsons Brinckerhoff Quade & Douglas, Inc., New York, USA

ABSTRACT: Conventional methods of tunneling in hard but jointed and heavily sheared rock formations, particularly those containing groundwater, may require additional pre-excavation rock mass reinforcement or improvement methods. These methods can include pre- and post-excavation grouting for both purposes of rock stabilization and controlling groundwater inflow into the excavation. Published geological data for the Manhattan area indicates that rock mass quality is generally good to very good, although it is locally complex, containing heavily faulted and sheared zones of high permeability, which are expected to be unstable. Driving hard rock TBM’s through these zones can pose significant problems and pre- and post-grouting methods should be considered in order to minimize the risk of rock mass failure and water inundation. The methodology and procedures for pre- and post-grouting during the excavation of TBM driven tunnels have been developed and are described within this article.

phase defines the regional structure of Manhattan with the axial plain striking N35°E and generally dipping toward the south-southwest at approximately 10° to 15°. Published information identifies the presence of four major joint sets within the Manhattan Schist. However further to recent and extensive ground investigation and interpretation the structure appears to be more complex than historical data suggests, revealing significant localized faulting, shearing, alteration and folding. The existing shear zones have been identified and categorized as major and minor. Major shear zones occur on a scale of approximately 10 ft to 100 ft (3 m to 30 m) and are characterized by the original rock being sheared, brecciated, and rehealed in a mylonite matrix. The fractures are often coated with secondary minerals. The boundary between the brecciated and the undamaged rock mass is distinctive, with this zone including clusters of open infilled joints and secondary mineralization. Minor shear zones occur on a onefoot (0.3 m) scale with their associated clusters of infilled, stained, mineralized joints and slickensides being apparent on a 10-foot (3 m) scale. Manhattan Island is bound by the East River to the east and the Hudson River to the west and is slightly above sea level. Groundwater levels measured in borehole standpipes range from 15 feet to 60 feet (5 m to 20 m) below street level. The quality of the

1 INTRODUCTION Manhattan Island is the heart of New York City and is one of the most urbanized cities, housing some of the most expensive real estate in the US and the world. Manhattan imposes serious construction considerations owing to its high density of buildings including historic residential districts and hi-rise commercial and residential properties, often with deep basements. In addition to the buildings is a highly developed infrastructure with many existing tunnels and other underground structures. In Manhattan it is therefore very important that during construction, systems are provided to mitigate risk and impact caused by tunneling. 2 GEOLOGY & HYDROGEOLOGY Manhattan is underlain by Proterozoic and Ordovician metamorphic rocks, locally known as the Manhattan Prong. These metamorphic rocks are characterized by three lithologies comprising schist, gneiss and marble, although the greater part of Manhattan is dominated by the more erosion resistant schist and gneiss. The rocks of the Manhattan Prong have been subjected to multiple tectonic episodes including folding, faulting and intrusion, resulting in an intensely folded and locally sheared rock mass. The prominent fold

87 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-11.qxd

2/18/04

11:53 AM

Page 88

of ground conditions. The principle of this method is to replace the water or air contained in the ground (pores, voids, cracks, joints) with a liquid material, which sets to a solid during short interval of time. There is an abundance of choices for the liquid but most commonly used are cement mixes, sodium silicate or organic resins. In tunneling there are two major purposes for grouting: to create a barrier against groundwater flow and to increase shear strength of the rock mass in order to maintain stability of the excavation, reduce settlement and ground movement. Shallow tunnels are often grouted from the surface, so that grouting and excavation procedures can go on simultaneously. Grout patterns would normally be rows of boreholes forming squares, with the spacing between boreholes determined by local conditions. When conditions preclude working from the surface, and when tunnels are very deep, grouting can be done from the tunnel face. The thickness of grouted area surrounding the tunnel varies and depends on the ground conditions and the purpose of the grouting (ground strength and/or permeability).

rock mass controls the permeability of the Manhattan Schist and generally the permeability of the undisturbed and unweathered rock is very low. In faulted and sheared zones the permeability is considerably higher as the network of fractures behaves as conduits for the groundwater. The permeability of the discontinuities is influenced by several factors including roughness, tightness and presence of joint infill. The coefficient of permeability has been derived from in-situ packer tests and typically varies between 105 to 107 cm/sec. However in shear zones the permeability ranged between 104 to 106 cm/sec and it is anticipated that widely spaced, open, steeply dipping fractures may transmit groundwater at greater rates than indicated by packer tests. This is especially true during excavation of new underground openings where flowing groundwater entering the excavation can wash out the clay infill from slickensided joints and cataclasite within shear zones, which can result in the rock mass permeability being in the order of 102 cm/sec. 3 CONSTRUCTION CONSIDERATIONS Construction experience in New York City and the findings of geotechnical programs indicate that the rock mass is generally good quality, stable and generates moderately low amounts of water. Only shear zones are expected to be unstable and to be sources of significant groundwater flow. Typically excavation of any underground structure in similar conditions with drill and blast or TBM methods will require initial support, such as rock bolts/dowels, shotcrete and occasionally steel ribs or lattice girders. The initial rock support systems are designed to prevent failure of blocks and loosened rock mass from the crown and sidewalls of the tunnels. These support systems in the hard rock formations usually take the form of fully grouted dowels and resin anchored rock bolts in combination with steel reinforced shotcrete. Often fiber-reinforced shotcrete is used as opposed to steel wire mesh. Tunneling in unstable rock formations where stand-up time of the excavation is limited due to densely distributed discontinuities, shear zones, and water saturated zones, will require additional special pre-excavation reinforcement/improvement methods to increase the rock mass quality, thus avoiding rock instability and controlling groundwater into the tunnel. In cases where the rock mass conditions dictate the need for pre-excavation rock stabilization, pre-excavation grouting techniques can be utilized.

5 PRE-EXCAVATION GROUTING Specific procedures for pre-excavation grouting from inside the tunnel space can be developed for any tunnel configuration, shape, diameter and excavation method. The philosophy of the pre-excavation grouting is to the increase rock mass properties or to seal a limited area ahead of the face and around the tunnel, using a grout of suitable strength, low permeability to water and of high durability. In general the pre-excavation grouting method involves filling all (or as much as possible) fissures, cracks and voids for a distance of a least 50–80 feet (10–25 m) ahead the tunnel face and 180° above the springline or 360° all around the tunnel. After finishing one round of pre-excavation grouting, 70–80% of the grouted length can be excavated. Subsequent to excavation of the improved rock mass, probing of the rock ahead of the TBM face will ascertain whether additional pre-excavation grouting is necessary. This cycle may be repeated as long as required, depending on geological conditions. To achieve a “dry” tunnel, postexcavation grouting may be required later in addition. Access to the tunnel face with the drill and injection equipment is very important for effective preexcavation grouting. For tunnels excavated with drill and blast method, access to the tunnel face is not often a problem. TBM’s are rarely designed to facilitate injection drilling close to the face, due to the lack of access. However, modern TBM’s are capable of performing pre-excavation grouting over or through their shield and sometimes through the cutterhead, although the latter is a very sophisticated method.

4 GROUTING METHODS Grouting is an established and common technique in modern tunneling. Grouting can be used in most types

88 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-11.qxd

2/18/04

11:53 AM

Page 89

Figure 1. Fault zone prediction ahead of the TBM.

Figure 2. 1st round of pre-excavation grouting.

When pre-excavation grouting is to be adopted for a project, it is important to identify the area and/or volume of ground that needs to be treated and where it is with respect to the tunnel face (Fig. 1). For this purpose probe holes must initially be drilled ahead of the cutterhead to establish the nature of the rock mass and groundwater conditions (location, flow rates and pressure). Tunneling in intensely fractured and crushed rocks, where groundwater inflow is minimal will only require an improvement of the rock mass properties for maintaining stability. This is accomplished by pre-excavation grouting over the TBM shield up to 180° above springline. The angle at which the rig can drill the holes into the tunnel wall will govern how far ahead of the TBM the grout can be injected. The usual set-up for this drill equipment has to allow for drilling holes at a minimum of 5–10° inclined to the tunnel axis. If the holes are drilled at angles greater than this, then the drill holes will be too far away from the tunnel perimeter, greatly reducing the efficiency of the grouting. If the inflow of the groundwater from the probe holes exceeds a predetermined threshold, it will be required to drill grout holes 360° all around the TBM to achieve control of the groundwater. Thus the anticipated flow of groundwater to be encountered plays a major role in the selection of equipment and its configuration. In the cases where the amount of the groundwater can be tolerated during the construction period and where the rock mass will only require increasing its stability to prevent the rock falls behind the shield of the TBM, pre-excavation grouting would be specified above of the TBM cutterhead only (Figs 2 & 3). In these circumstances the following action should be executed: probing ahead of the TBM, pre-excavation grouting of the potentially unstable area ahead and above of the TBM shield and post-excavation grouting of the invert section to control groundwater inflow in to the tunnel.

Figure 3. Subsequent round of pre-excavation grouting.

The procedure will start with drilling a probe hole near the anticipated unstable rock formation or fault zone ahead of the tunnel face, which should be drilled up to 100 feet (30 m) in length at the 12 o’clock position. If this probe hole indicates any potential problem, such as the presence of weak rock (higher drilling rate), loss of flush water or high groundwater inflow through the drilled probe hole, then two more probe holes should also be drilled at the 9 and 3 o’clock positions to verify the findings of the first probe hole. If the additional probe holes indicate poorer rock mass quality ahead, then an appraisal will be required to judge the extent of pre-excavation grouting. Different grouting trigger levels should be adopted for specific projects. The following pre-excavation grouting trigger levels may be used for assessment whilst probing:

•

89 Copyright © 2004 Taylor & Francis Group plc, London, UK

Sudden loss of drilling water; over 50% (an abrupt change in the amount of water returning to the surface or face usually signifies that the drill has reached a highly permeable horizon)

09096-11.qxd

•

•

2/18/04

11:53 AM

Page 90

Sustained (over half an hour) groundwater inflow through any probe hole drilled ahead of the tunnel face, over the 25 gallons/minute or fi gallons/ minute/feet (100 liters/minute or 3 liters/minute/ meter) Total groundwater inflow from all drilled holes during an half an hour period exceeded 50 gallons/ minute (200 liters/minute).

If pre-excavation grouting is necessary, based on trigger levels, the grout holes should be drilled above the tunnel springline as shown on the Fig. 2, Section A. The grout holes should be drilled at least 50 feet (15 m) ahead of the TBM cutterhead through the TBM shield at an inclination between 5°–10° to the tunnel axis. The distance between the grout holes is to be determined according to local rock properties and its grout penetration value. Previously gained experience in similar rock conditions suggests that the maximum distance between the grout holes of 5 feet (1.5 m) should be used. Based on former experience in Manhattan rocks, cementitious grouts are recommended as primary injection materials. Cement can enter fissures as small as 0.3 mm and therefore has a comparable penetration/infiltration capacity as silicate or acrylic resins, but makes a much more durable, more economical and more environmentally friendly solution. The main components of cementitious grout are water and generally Type I or II Portland cement and can be altered by using other cement types, such as Type III (high early strength), Type IV (low heat of reaction), or Type V (resistance to chemical attack). By varying the water to cement ratio it is simple to change the grout’s bleeding rate, subsequently altering its plasticity and ultimate strength. Mixing the main components with additives, such as bentonite, sodium silicate, dispersants, retarders, and accelerators, will also change the grouts properties. Microfine cement is an alternative version of cement that overcomes the difficulty of using Portland cement grouts in low permeability ground.

Figure 4. Post-excavation grouting.

face, once the TBM has passed (Fig. 4). Post excavation grouting alone is usually wasteful as it is costly and often unsuccessful, but in combination with pre-excavation grouting it is very effective. Using cementitious pre- and post-excavation grouting improves the water tightness of the rock mass, achieving lower water flow rates into tunnels typically between 0.1 to 2.5 gallons (0.5 to 10 liters) per minute per 300 feet (100 m) of the tunnel. In cases where the seepage rates into tunnels are required to be more stringent, governed by the sensitivity of the environment and construction, chemical grouting (organic resin) can be utilized as an alternative to cement. 7 CONCLUSION Tunneling under Manhattan generally does not pose significant stability problems as the majority of the rock mass is good quality, therefore the methods detailed within this article are not always required. Published information identifies the existence of several locally complex and significant fault and shear zones in which the rock mass can be expected to be unstable and to yield high rates of groundwater. When such ground conditions are anticipated these techniques described herein can be considered. Among existing pre- and post-excavation rock mass improvement/reinforcement methods available, grouting is very effective in combination with TBM driven tunnels.

6 POST-EXCAVATION GROUTING Most tunneling projects cannot tolerate large volumes of groundwater during construction of the final structure, therefore groundwater ingress is required to be limited, which can be done by using post-excavation grouting in addition to pre-excavation grouting. Clearly, pre-excavation grouting all around the tunnel is more effective than post-excavation grouting alone, but it will significantly slow TBM advance rates. For this reason, when ground conditions permit, it is favored to perform pre-excavation grouting of the rock ahead and above the springline and then post-excavation grouting of the invert area, from behind the tunnel

REFERENCES Merguerian, C. 2002. Brittle Faults of the Queens Tunnel Complex, NYC Water Tunnel #3. In G.N. Hanson, Ninth Annual Conference on Geology of Long Island and

90 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-11.qxd

2/18/04

11:53 AM

Page 91

metropolitan New York. 20 April 2002, Sate University of New York at Stony Brook, NY, Long Island Geologists Program with Abstracts, 116 p. Snee, C., Sarkar, S., Benslimane, A., Stewart, C., and Osborne, C. 2003. Rock Mass Characterization for the Manhattan East Side Access Project. In P. Cullgan et al (ed.), Soil Rock America 2003 (The 12th Panamerican Conference for Soil Mechanics & Geotechnical

Engineering and the 39th US Rock Mechanics Symp.), June 22–25, 2003, Vol. 1, pp. 129–136. Ryzhevskiy, M. 1987. The main principles of the new technologies for construction tunnels in unstable rock formations. Energetic Construction. Moscow, N7. Ryzhevskiy, M. 1988. The advance experience of the chemical ground improvement by jet grouting. VPTI Transstroj. Moscow.

91 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-12.qxd

2/18/04

11:54 AM

Page 93

Session 1, Track 4 Specialized urban construction

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-12.qxd

2/18/04

11:54 AM

Page 95

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Design and construction of an LRT tunnel in San Jose, CA P.J. Doig Hatch Mott MacDonald, Pleasanton, California

ABSTRACT: The Santa Clara Valley Transportation Authority (VTA) is constructing significant extensions to its Light Rail Transit (LRT) system in the metropolitan San Jose area. These include the Vasona line that will extend from downtown San Jose to the suburb of Campbell. At busy Diridon station, trains will travel for 480 m in a subterranean section, comprising tightly-curved tunnel and straight approach ramps. The tunnel passes under streets, a bus terminal, rail tracks and platforms, and is very close to the historic station building. The project also includes an extension to the existing pedestrian tunnel, that will connect the station to a new LRT station. The tunnels were built by cut-and-cover methods. The contract for the tunnel project commenced in February, 2001 and was completed in June, 2003. This paper covers the execution of the project, from the latter stages of design to completion of the LRT tunnel structure.

the light rail system. This may yet have an effect on the scale and operation of the Vasona line. A second funding initiative, passed in 2000, extended the sales tax for a further 30 years beyond 2006. This was intended to fund a number of projects, but primarily the extension of BART from Fremont to San Jose. VTA has been investigating ways of utilizing this projected revenue to support its current expenditures, so as to limit cutbacks in service. However, should the economy not rebound, VTA will be faced with difficult choices for its future that may well have a substantial impact on the scope of transportation services offered in the San Jose area.

1 INTRODUCTION San Jose is situated at the southern tip of the San Francisco Bay in northern California. San Jose is the major city in Santa Clara county which has a population in excess of 1.7 million. The San Jose area is more loosely known as Silicon Valley, being the home of hitech icons such as Apple, Intel and Cisco Systems. Per capita income and house prices in the area are amongst the highest in the US. Many people commute from distant communities and traffic congestion is chronic. Public transportation has a high level of public support, which is reflected in numerous funding initiatives. The Vasona LRT extension arose out of an initiative approved by voters in 1996. Known as the Measure B Transportation Improvement Program, the initiative authorized a half-cent sales tax in Santa Clara county, expiring in 2006. It was anticipated that the tax would generate around $1.6 billion in revenue that would be used to fund a specific package of countywide transportation improvement projects. Management of the sales tax is by the Santa Clara County Board of Supervisors. The Santa Clara Valley Transportation Authority (VTA) is the implementing agency for the 1996 Measure B Transportation Improvement Program. VTA is an independent special district responsible primarily for bus and light rail transit (LRT) service. The recent downturn in the economy, particularly the in hi-tech sector, has had a significant effect on sales tax revenue. VTA has been obliged to scale back its operations and curtail plans for future development of

2 VASONA LINE OVERVIEW The LRT system is currently around 50 km in length, with two lines arranged in a rough T-shaped configuration. New lines under construction will add a further 20 km. Phase I of the new Vasona line will be 8 km long, running from downtown San Jose to Campbell in the southwest. The line will have nine stations and is expected to carry 8000 to 9000 riders per day. The LRT system is largely at grade with tracks located in the central divide of city streets. Trains are powered from an overhead contact system. The initial segment of the Vasona line from the existing Guadalupe line to just east of the San Jose Diridon main line station will be over city streets. At the station itself, the line will be in tunnel. Between San Jose Diridon Station and

95 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-12.qxd

2/18/04

11:54 AM

Page 96

Vasona Junction, the line will follow the former Union Pacific Railroad (UPRR) right-of-way, now owned by VTA. The capital cost of Phase 1 of the Vasona line is budgeted at $321 million, of which around $53 million is expected to come from Federal funding. Construction of the Vasona line commenced in early 2001 and is anticipated to be complete in late 2004. Revenue operation is expected to commence in Spring, 2005. 3 SAN JOSE DIRIDON TUNNELS SEGMENT The San Jose Diridon Tunnels project, designated C345, involved the construction of some 490 m of trackway below grade. This section of the line comprises 120 m of approach ramp at the western end, and 95 m of approach ramp at the eastern end. In between, there is 275 m of tightly-curved, double box tunnel. The tunnel is necessary to negotiate the railyard and other facilities at the historic Diridon Station. The station is owned by the Peninsular Corridor Joint Powers Board (JPB) which also operates the Caltrain commuter service. The station serves three local commuter services, Amtrak long-haul passenger service, and UPRR freight trains. The station is also the maintenance facility for much of the rolling stock. The station yard has twelve tracks, five of which share three platforms. One of those tracks is also used by freight traffic. The remaining tracks are for maintenance and storage. The tunnel project involved an intricate series of staging operations in the railyard to allow tracks to be taken out of service so that construction could proceed. Groups of tracks were isolated and removed for a number of months at a time. This allowed a trench to be excavated from surface to accommodate the LRT tunnel. A second, smaller trench was excavated for the extension of the existing pedestrian tunnel. In three locations, the track was supported on a temporary bridge and the trench excavated beneath it.

Figure 1. Construction within the rail yard.

to concerns from JPB about potential effects on the station. The main station building, and other elements including the platform canopies, are designated historic structures and must be preserved. The design therefore proceeded on the basis of a cut and cover tunnel, skirting the station building. Discussions with JPB led to the development of a staging plan for removal and/or bridging of the tracks during construction. Originally, it was thought that most of the tracks would need to be kept in service and therefore have to be placed on bridges. However, JPB reviewed their operational requirements and decided they could function with four through tracks. Discontinuous tracks were to be kept open at the south end of the yard so that they could be used for storage. Figure 1 illustrates the proximity of the west channel construction to the rail yard. VTA’s designer analyzed the storage requirement and this was factored into the staging plan. The plan allowed for the work to be completed in three distinct stages. This was to be accomplished with the installation of three shooflies, one permanent crossover, and a temporary bridge under the freight track. Signaling work associated with the track modifications was not included in the final design. JPB reserved the signaling design to themselves and this was only made available after the construction contract had been let. The contract design also included the temporary shoring system for the open-cut excavation. The choice of a temporary shoring system is normally left to the contractor. However, JPB’s requirement that they approve the system mitigated against leaving this item open in the bid documents. It was known that JPB had a preference for soil mixing due to its limited impact on the operation of adjacent rail lines, and success with the method elsewhere on the system. Also, JPB does not allow tie-backs on their property because of a perception that the system might adversely affect subsoil conditions. Driven piles were specifically prohibited

4 DESIGN The designer for the C345 project was the General Design Consultant (GDC). This comprises Parsons Brinkerhoff Quade and Douglas, in association with MK Centennial and Korve Engineering. Design of the tunnel structures under subcontract to the GDC, was undertaken by San Francisco-based, Biggs, Cardosa Associates, with shoring design by EQE of Oakland. Detailed design was completed in September, 2000. As part of the review process, the design submissions were routinely reviewed by the rail companies. In the early stages, consideration was given to traversing the railyard in conventional, bored tunnel. However, this was rejected due mainly to the cost, and

96 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-12.qxd

2/18/04

11:54 AM

Page 97

because of noise restrictions. Therefore, soil-mix walls with internal bracing became the specified support system. The system was to comprise of a low-strength, reinforced soil-mix wall installed within the ground on either side of the trench prior to excavation commencing. The wall was made up of contiguous bored columns, with every second column reinforced with a steel beam. The generic soil profile for the LRT tunnel indicated fine-grained sediments, coarse-grained sand, and fine gravel to a depth of 17 m. Stiff clay extended below that depth. The groundwater table was at a depth of 4 m, although the design assumed it was 2 m higher. In order to allow for construction in the dry, the soil-mix wall was designed to terminate in the clay. Before excavation could proceed below the water table, the walls had to be closed on all sides. This meant that temporary bulkheads had to be installed across the alignment in some places to give a closed cell. The soil-mix columns were designed to be 75 cm in diameter, positioned at 60 cm centers. This gave a nominal 60 cm thick wall. The reinforcing beams were mostly W610 82 section, and were required to extend around 7 m below the bottom of the excavation. Ground surface was at approximately 27 m, and base of excavation at 20 m. Maximum design pressure on the walls was 115 kPa. The design generally included for two levels of struts, one at surface and one at mid-depth. Top level walings were specified as W460 113, and lower level were W460 177. Struts were 46 cm diameter pipes with 0.65 cm wall thickness. Top level struts had to be preloaded to 300 kN and lower level struts to 600 kN. Given the sensitive nature of the station terminal building and other structures in the immediate locale, it was decided to mandate a settlement monitoring program. This required installation of inclinometers outside the shoring wall, and establishment of elevation points on the adjacent buildings. Groundwater monitoring wells were also to be installed. These all had to be checked on a regular basis, with a requirement that action to be taken in the event of excessive movement.

Figure 2. Drilling soil-mix columns.

manager, South Bay Transit Associates (SBTA). SBTA is a joint venture of Hatch Mott MacDonald and URS, formerly O’Brien Kreitzburg. Vali Cooper & Associates and Booz Allen also provided staff to the project as subconsultants to SBTA. 6 CONSTRUCTION 6.1

5 PROCUREMENT The contract documents were issued for bid in September, 2000. Bids were opened in November, 2000. Three bids were submitted, ranging from $23 to $26 million. The low bid was submitted by Condon-Johnson & Associates (CJA) of Oakland, California. The contract was awarded in February, 2001 with overall completion required in February, 2003. Construction management of the project was by VTA utilizing its own staff and staff drawn from the program

97 Copyright © 2004 Taylor & Francis Group plc, London, UK

Soil mixing

Preliminary work on the project was sufficiently complete to allow soil mixing to commence in April, 2001. CJA used the Geo-Jet® process developed by Verne L. Schellhorne of Aerial Industrial, Inc. CJA is the exclusive licensee for use of Geo-Jet® in the states of California, Oregon, Washington, Utah and Nevada. Figure 2 shows the process in operation. The Geo-Jet® process is similar to other soil-mixing techniques, with some unique features. The process generally involves cutting a column with a single auger soil-cement processor. Cement slurry is ejected at high velocity through nozzles in the auger to create a high shear mix of soil cuttings and cement slurry. The process of forming the in-situ soil-cement columns is monitored by a computer. The specified strength of the finished soil-mix was 1.4 MPa at 28 days. A test program was instituted at the beginning of the project utilizing production columns. This confirmed a ratio of around 0.2:1, dry Portland cement to in-situ soil. For most of the C345 project, the soil-mixing rig was a Link Belt 218 crawler crane with 32 m of leads mounted to the boom. Reinforcing I-beams were inserted with an ABI pile-driver, on a Sennebogen carrier. Later in the project, an ABI rig was also used for soil-mixing. Grout was produced at a mobile plant. This was generally located within 50 m of the soilmixing rig, although it could be much further away as conditions dictated. The plant had storage for up to 160 000 kg of cement.

09096-12.qxd

2/18/04

11:54 AM

Page 98

mitigating delays caused by the factors mentioned earlier.

The soil-mix operation started with the excavation of a starter trench, around 2 m in depth. This was to capture the material that is pushed out of the top of the column by the mixing operation. A length of wall would be identified and the unreinforced intermediate columns installed first, at 120 cm centers. The following day, the reinforced columns were installed between the intermediate columns. As well as providing walls for the open-cut excavation, soil-mix columns were installed for bulkheads. These could be reinforced where they had a structural purpose, or unreinforced where they were for water control. In addition, at one location where the structural bulkhead had to be supported with diagonal bracing, soil-mix columns were installed en masse to act as a buttress. The buttress columns were designed such that a higher strength of 2.1 MPa was required for the soil-cement. Generally, soil-mixing proceeded without major incident, and progress was satisfactory. Production averaged 256 m of columns per 8 hour shift, with the best shift reaching 665 m of columns. The soil-mixing was not continuous due the project having to be done in stages. Access to some stages was delayed due to issues with third parties, notably the railroad companies, utility owners, and private property owners and businesses. Also, there were frequent encounters with obstructions, including hazardous materials, utilities, rail, and the apparent remnants of a grouting program. These various factors combined to delay the contract for around four months. The work was also delayed by difficulties the contractor experienced with alignment of the columns. In the initial stage of the construction, it was found that some sections of soil-mix wall had impinged on the structure, with the occasional column toeing well into the excavation. Nowhere was this sufficient to require complete removal of the column beam. However, some beams had to be cut back and other measures had to be taken, which included some major rework. The contractor addressed this problem in later stages by moving the wall out another 7.5 cm from the structure. This proved to be successful in achieving the required clearances. 6.2

6.3

Once the soil-mix walls had been installed in a particular stage, the excavation was commenced. Sufficient depth of trench was excavated to accommodate the top level of bracing. The top end of the channels and the pedestrian tunnel had only one level of bracing. The deeper areas had two levels. Excavation was generally carried out with a dozer in the cut, pushing to an excavator located on the surface. Once the soilmix walls had been exposed, they were trimmed to the face of the beams. Thereafter, a 7.5 cm-thick layer of shotcrete was applied. The shotcrete was intended primarily to provide a smooth backing for the waterproofing membrane that would be installed later. However, in one location, it had a structural purpose, serving as a key against uplift of the base slab. After the shotcrete had been applied, bracing was installed. This comprised longitudinal walings spanning a number of beams, normally around four or five, on both sides of the excavation. Two transverse struts were then placed between the walings. The walings were H-beams and the struts large pipe sections. All steelwork was pre-fabricated off-site. The struts were then preloaded to a given load, packed and welded in place. Excavation was then continued down to the second bracing level and the procedure repeated. Thereafter, excavation was completed to sub-grade. Upon completion of the excavation, a 10 cm-thick mudslab was placed on the subgrade. A waterproofing membrane was then installed over the top of the slab and up the walls. A second protective mudslab was then poured over the waterproofing so that invert slab construction could begin. The contract documents specified Preprufe® 300 for blind-side application, and Bituthene® 3000 for exposed-side application. Both materials are of sandwich construction and include a layer of HDPE. Joints have to be taped and penetrations repaired. Due to what he perceived as difficulty installing and maintaining the product intact, the contractor offered Paraseal® as an alternative. Paraseal® comprises a butyl membrane with bentonite prills impregnated on the interior surface. The sheets are nailed to the wall, and require only lap joints. The material is intended to be self-healing. At joints and holes, any water entering from outside is expected to encounter bentonite which expands and closes the hole or joint. VTA accepted the alternative and Paraseal® was installed. Experience with the Paraseal was not entirely satisfactory. Whether this related to deficiencies in the system, problems with installation, or construction effects was not entirely resolved. Wherever the material was

Track staging

The general intent of the staging plan contained in the contract documents was generally followed. However, there were significant modifications to the detail. The contractor elected to install two additional temporary bridges, which reduced the requirement for shooflies and improved access. VTA, with the concurrence of JPB, made stage 3 available prior to the completion of stages 1 and 2. Other constraints on work procedures and sequencing were relaxed. These measures had the effect of opening a second front and

98 Copyright © 2004 Taylor & Francis Group plc, London, UK

Tunnel excavation

09096-12.qxd

2/18/04

11:54 AM

Page 99

Figure 3. Interior of tunnel.

Figure 4. West channel with pedestrian tunnel portal at right.

exposed for any period, it was susceptible to degradation by rain or drainage water. This was particularly evident at joints between stages. Also there were problems with properly terminating the material at grade, on the exterior of the channels. The result was that the finished structure leaked. A chemical grout injection program was conducted during the spring of 2003 and this appeared to substantially reduce infiltration. However, this has been during the dry season and it remains to be seen how effective the system is when the rains commence. The contractor has to warrant the material for ten years and so VTA may direct more remedial measures if leakage is a concern. Figure 3 illustrates the generally dry condition of much of the tunnel as at October, 2003.

along with conduits for other systems such as communications and CCTV. Additionally, there was a dry fire line provided with external connections to hydrants and internal stand-pipes. All conduits and pipes were cast into the concrete so that there was no potential for service interruption in the event of a fire or derailment in the tunnel. The tunnel construction contract did not include installation of the LRT track. This was done under a follow-on contract. The tunnel construction included forming shallow troughs in the invert, complete with threaded rebar inserts. The troughs were used later as the foundation for low plinths poured by the track contractor. Rails were affixed directly to the plinths with purpose-built hardware. The tunnel was constructed with vertical and horizontal curves, including a spiral section. This made the track installation process more difficult and some rework was necessary. Track installation was completed in November, 2003.

6.4

Tunnel construction

The concrete construction was relatively routine. Pour lengths were limited to 18 m. Mechanical rebar couplers were used at construction joints between stages. The exterior walls were formed with a series of full height panels, mounted on wheels. They were moved by hand along the tunnel invert, as the shoring made it difficult to pick and land the panels with a crane. The walls in the channels incorporated architectural features and these required particular care when matching form panels. The roof slab was formed with a proprietary shoring system involving interlocking towers. Concrete was delivered into the forms from a pump on surface. Invert and roof slabs, and cantilever walls in the channels, were required to reach specified strength before shoring could be removed. All exposed surfaces received a Class I finish, and the channel walls were treated with anti-graffiti paint. The tunnel included combined system ductbanks at the base of each exterior wall which doubled as emergency walkways. Lighting was provided in the tunnel,

7 CONCLUSION The San Jose Diridon Tunnels project was completed in June, 2003. The project had to overcome numerous challenges presented by underground obstructions and third party issues. The time for completion was extended by four months as a result. However, the original date for access to the tunnel for the track contractor was met, ensuring that the overall project remained on schedule. Despite these issues, the final cost of the contract was only 8% over the original price, and wellwithin VTA’s budget. As Figure 4 shows, only installation of the overhead contact system remains for the tunnel section to be ready for testing. This will begin early in 2004. It is anticipated that test trains will run on the Vasona line in the Fall of 2004, with revenue service scheduled to begin in April, 2005.

99 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-13.qxd

2/18/04

11:54 AM

Page 101

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Underpinning design and construction – Atlantic Avenue Station complex rehabilitation, New York, USA A. Grigoryan & L.G. Silano Parsons Brinckerhoff Quade & Douglas, Inc., New York, New York, USA

ABSTRACT: The Atlantic Avenue Station Complex in Brooklyn comprises three New York City Transit (NYCT) subway stations and the Long Island Rail Road (LIRR) Flatbush Avenue commuter rail terminal. The Atlantic Avenue Station on the IRT subway line is undergoing a major structural rehabilitation. This paper addresses underpinning methods and procedures, as well as the construction of new passageway sections and a new lower-level concourse under operating tracks without interruptions to transit or station operations and minimal impacts to street traffic.

1 INTRODUCTION The Atlantic Avenue Station on the Eastern Parkway Interborough Rapid Transit (IRT) Line in Brooklyn

opened in 1908 and is located under Flatbush Avenue near the intersection of Atlantic Avenue (Figure 1). In addition to providing Eastern Parkway Line NYCT service (Nos. 2, 3, 4 & 5 trains), the station provides

Figure 1. Project location plan.

101 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-13.qxd

2/18/04

11:54 AM

Page 102

Figure 2. Existing structure plan.

connections to the Pacific Street Station (B, M, N & R trains) on the Fourth Avenue Brooklyn–Manhattan Transit (BMT) Line, the Atlantic Avenue Station (D & Q trains), and the LIRR Flatbush Avenue Terminal. The station services over 65,000 passengers daily. A connecting passageway runs under the tracks of this station, skewed across and near the center of the IRT platforms. This passageway, the vital link of this complex with its multiple connections to other subway and commuter rail lines, is being modified and replaced by three new station elements: A new west passageway section, a reconfigured east passageway section, and a new lower-level concourse connecting east and west passageway sections. Meanwhile, the station structure had deteriorated to a state of disrepair as a result of intensive use and irregular or neglected maintenance. Water infiltration had damaged many structural components and the facility was not adequate to handle present and projected future increased passenger flow demands. Coordinating major project tasks involves working closely with the capital construction and operational departments of two major transit agencies (NYCT

and LIRR), with the city’s Department of Transportation for street traffic issues, with several other city agencies for utility relocations, and with numerous subconsultants. 2 EXISTING STRUCTURE The existing subway station utilizes typical NYCT framing, with columns spaced at 4.6 meters (Figure 2). Most of the existing steel framing members are built-up sections, comprised of web plates, angles, channels and cover plates, connected by 22-millimeterdiameter rivets. The majority of existing columns are supported on individual spread footings, measuring 1.5 meters 2.7 meters and extending 0.6 meters below the IRT invert slab (Figure 3). Columns are located along six column lines with roof beams running along each column line, located approximately 2.1 meters below street level. The depth of roof beams within the construction limits is 0.76 meters. The existing reinforced concrete roof structure is supported by roof beams or by roof beams and exterior station walls.

102 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-13.qxd

2/18/04

11:54 AM

Page 103

Figure 3. Existing structure section.

The invert slab of the existing passageway is a nominally reinforced concrete slab on grade. The structural thickness of the invert slab is approximately 0.2 meters. Four subway tracks, generally consisting of wood half-ties and rail assemblies, are positioned directly on the 0.3-meter-thick track slab on grade. The existing columns, platforms, and tracks are supported on the roof of the existing passageway by transfer grillage beams and girders, spanning the passageway at a skew. The existing structure is sealed against water infiltration using a waterproofing membrane. The IRT Atlantic Avenue Station has three passenger platforms: Two side platforms service northbound and southbound local tracks, while a center island platform serves two express tracks. Two stairs lead from the passageway to each of the platforms. In addition, there is an old control house located adjacent to the exterior wall of the southbound local platform. This structure, which has historic value, was temporarily relocated to a place near the construction site. As part of this project, the control house was repaired, rehabilitated, and returned to its original location. A subconsultant along with a specialty contractor handled this task.

3 PROPOSED MODIFICATIONS The existing connecting passageway is reconfigured and widened in accordance with the Atlantic Avenue Master Plan Study, and is remediated by three new station elements (Figure 4):

•

A new west passageway section

• •

The connecting passageway is reconfigured to allow the station to comply with Americans with Disabilities Act (ADA) accessibility provisions. The west section of the connecting passageway is reconstructed to eliminate stairs by installing an ADAaccessible ramp as well as a new ADA elevator. This elevator provides service to the local southbound platform. Meanwhile, the east section of the connecting passageway is reconfigured to provide ADA elevator access to the northbound local platform along with access to the Atlantic Avenue Brighton Line station. A new lower-level concourse is also built to provide ADA elevator access to the express island platform. A primary function of the concourse is to provide ADA elevator access to the express platform. This new passageway/concourse configuration provides the opportunity to expand the section of passageway bounded by stairs to the southbound local and island platforms, thereby alleviating congestion and circulation problems at the existing connecting passageway and providing more stairs to those platforms. The new concourse, spanning between the local southbound platform and express island platform, has a total of five stairways: Two to the local southbound platform and three to the express island platform. In order to increase the level of service for the section of the passageway east of the new concourse, the passageway is widened near a currently overcrowded platform stair. The area is also reconfigured to install an elevator to the northbound platform.

103 Copyright © 2004 Taylor & Francis Group plc, London, UK

A reconfigured east passageway section A new lower-level concourse connecting the west and east passageway sections

09096-13.qxd

2/18/04

11:54 AM

Page 104

Figure 4. Modified structure plan – platform level.

Figure 5. Modified structure section.

4 DESIGN CONCEPT In order to accommodate extensive passageway modifications and new concourse construction, our firm developed several design concepts. Steel was selected

as the material that would provide both the required strength and flexibility for the necessary construction staging, while reinforced concrete was selected for designing the invert slab, walls, and stairs (Figure 5). Due to architectural considerations as well as passenger

104 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-13.qxd

2/18/04

11:54 AM

Page 105

Figure 6. Modified structure longitudinal section.

flow considerations, the plan arrangement of columns in the modified passageway and the new concourse did not follow the grid of existing columns on the platform level above. This presented major structural challenges – both from a design strength and construction staging perspective. Typical existing column loads range from 135 to 180 metric tons. Transfer girders were designed to support existing columns at the platform and track levels (Figure 6). Due to modifications and new construction, the lower-level columns carry axial loads ranging from 155 to 725 metric tons. Some of the columns are also subjected to bending moments due to unbalanced loading from framing girders and beams. Base plates for the columns measure in plan view up to 0.76 by 0.84 meters, with thicknesses reaching 95 millimeters. Most of the steel framing uses ASTM A36/A36M Grade 250 steel (Fy  250 Mpa), with some of the plate girders composed of ASTM A572/A572 Grade 345 steel (Fy  345 Mpa). The sidewalls of the new concourse, modified passageway, and stairs were designed conservatively as cantilever retaining structures, constructed integrally with the invert slab. One of the main reasons for the conservative design involved a constructibility issue: It would have been more expensive to provide temporary supports during all staging procedures until the walls were completed and platforms reconstructed to provide lateral support at top of walls. The invert slab, meanwhile, is designed as a mat foundation. In comparison to the existing framing, the new framing represents a major structural change. The

original columns extend only from the underside of the platform or track bed up to the roof girders. Loads are applied to these columns only on top, at roof level. The new columns, however, extend from the invert slab of the lower-level through the track and platform levels up to the roof level, as seen in Figure 5. Loads are applied to the new columns from various directions at two or three levels. At track level, floor beams are framed into new columns or into transfer girders. Transfer girders carry track and platform loads and transfer them to lower-level columns. Our firm designed special seats to accommodate the connection of existing columns to transfer girders, shown in Figure 6. These seats provide flexibility during construction and fabrication, with most information verified in the field prior to fabrication. This design solution allowed the contractor to fabricate girders and columns to specified dimensions, before gaining access to the bottom of existing columns to measure their exact elevation. Due to IRT northbound local platform modifications, the demarcation line between two stations – the Atlantic Avenue IRT Station and the Flatbush Avenue LIRR Terminal – was moved east from its original location. This presented another structural challenge as the existing platforms of these two stations had different elevations and slopes. Within the construction limits, the difference ranged from approximately 0.15 to 0.6 meters. The existing grade separation between the stations also had a full height chain link fence extending from the platform level up to the underside of the roof girders.

105 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-13.qxd

2/18/04

11:54 AM

Page 106

Figure 7. Typical cross section at Track No. 6.

No columns could be located in the passageway below the new demarcation line to support the platform framing. As a result, there were only two column lines on either side of the demarcation line available to support the platform framing at this location: Column line “A” of the IRT northbound local platform and column line “K” of the LIRR Track No. 6 platform. Transfer girders run along these column lines, as do variable length platform support beams, spanning between column lines A and K. This challenge was resolved by introducing a custom-designed kinked welded plate girder. The plate girders, spaced at 1.5-meter centers along the northbound platform, support platform glass paver panels which, in turn, limit girder flange width to 0.15 meters. Welded plate girders are supported by longitudinal platform girders on column line A at one end and by LIRR Track No. 6 transfer girders at the other end (Figure 7). 5 CONSTRUCTION METHODS The public interest, safety, and convenience were emphasized at every stage of project design and

construction. Since the clear width of the existing passageway was only 4.6 meters, staging procedures were established to maintain the existing opening at all times throughout each phase of construction. Furthermore, criteria were established to maintain access to the stairways providing access from the passageway to the three subway platforms. As there were two existing stairs at each platform, temporary and/or permanent stairs needed to be constructed before the existing stairs could be closed, removed, or modified during construction. Considering the complexity of construction, and numerous operational limiting factors, the client requested the development of two feasible construction methods as part of the design package. Our firm refers to these methods as:

• •

The major difference between the two methods is that the drift method offers minimum interference with existing train and passenger operations, as well as minimum visual impacts during construction.

106 Copyright © 2004 Taylor & Francis Group plc, London, UK

Suggested Construction Method “A” (Drift Method) Suggested Construction Method “B” (Pile Support Method)

09096-13.qxd

2/18/04

11:54 AM

Page 107

Figure 8. Construction method “A”.

Figure 9. Construction method “B”.

Occurring primarily in locations not visible to the public, the drift method makes limited use of heavy equipment, with construction performed in various stages in tight, confined spaces below existing platforms and tracks (Figure 8). The pile support method, on the other hand, offers more flexibility regarding equipment usage. However,

it has a greater visual impact on passengers, as many temporary roof support piles and beams are visible during construction. Although most construction progresses behind shielding, the general public is fully aware of the ongoing work (Figure 9). Both methods required a construction shaft to provide access for the excavation, removal of spoil, and construction of the

107 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-13.qxd

2/18/04

11:54 AM

Page 108

lower concourse. The suggested work sequence was organized in three construction phases following a general preparatory phase. This preparatory phase consisted of work necessary to prepare the existing structure for actual excavation and underpinning efforts required for construction of the lower concourse. During this phase, the conditions of structures and utilities were inspected and verified; the control house was relocated to a storage location; utilities were relocated or protected; and shaft excavation was completed. The three general construction phases for both suggested construction methods A and B included:

• • •

Phase I: The area north of the existing passageway, comprising excavation, underpinning, demolition, and construction of temporary and final support framing. Phase II: The existing passageway area and the southern portion of the lower concourse, consisting of excavation, underpinning, demolition, and construction of temporary and final support framing Phase III: All finish work within the lower concourse

Major construction phase limitations and techniques included:

• • •

• •

Maintaining the 4.6-meter width of the existing passageway beneath the IRT station for any temporary routing of passengers during construction phasing. Maintaining two stairways from the passageway up to each platform. The capacities of these replacement stairways needed to equal the capacity of existing stairways. Limiting major construction requiring contractor access to transit trackways, to one trackway and the adjacent platform. This work required general orders (GO’s) for train diversions that were only available on weeknights or weekends. Field welding to existing steel members was generally not allowed. Constructing temporary platforms and stairways, barricades, and overhead shields to protect passengers and workers during demolition, construction, and phasing.

The contractor elected to use Method B, modified somewhat to reflect his method of underpinning and procedures, as well as to respond to actual field conditions and ever-evolving operational limitations. The contractor also modified the pile types, pile layout, and track support details, as shown in Figure 6.

The underpinning of the existing structure – roof with full live load of street traffic above, fully operational platforms, and four subway tracks with uninterrupted service – was one of the most challenging and complicated issues encountered during the project. Of course, public safety was also a major issue. For several months, the entire subway structure needed to be underpinned, and major structural elements, such as columns, foundations, stairs, platforms and tracks, needed to be rebuilt. The client requested that we check and scrutinize every detail of the contractor’s proposed changes and modifications, necessitating the submittal of sets of calculations with every new detail. Considering it our highest priority to be continuously on alert for any logistical or structural conflicts, or for any possible flaws in calculations, details or procedures, we scrutinized all proposed modifications and procedures to assure public safety and structural integrity. As a result of our efforts, some of the most critical shop drawings were repeatedly returned to the contractor with numerous questions, comments, and clarifications before finally being approved and accepted for construction. This effort was aided by an extensive research effort involving client archives performed prior to the beginning of design in 1998. The acquired information was essential in addressing the many necessary limitations imposed on the design, underpinning methods and procedures, and construction techniques and staging. This is a prime example of the successful cooperative efforts typically exhibited during this complex project among the client, contractor, and consultant.

ACKNOWLEDGMENTS Clients New York City Transit (NYCT), New York Long Island Rail Road (LIRR), New York Structural Designer Parsons Brinckerhoff Quade & Douglas, Inc., New York Project Architect di Domenico & Partners, LLP Contractor Schiavone Construction Company, New York Santop Construction Company, New York

108 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-14.qxd

2/18/04

11:55 AM

Page 109

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Slurry walls accelerate shaft construction in rock in Los Angeles M.P. McKenna & K.K. So Jacobs Associates, Los Angeles, CA, USA

M.A. Krulc Traylor Brothers, Los Angeles, CA, USA

E. Itzig-Heine Ed Heine Construction Services, Leesburg, VA, USA

ABSTRACT: This paper details the challenges associated with the design and construction of a “figureeight” shaped, or dual cell shaft through soft ground and sedimentary rock for the Northeast Interceptor Sewer project in Los Angeles. The Humboldt Street Shaft was excavated as two cells, using a combination of both reinforced and un-reinforced concrete diaphragm walls. The Contractor chose to construct the diaphragm panels to full depth, using a Hydrofraise rather than sinking them only to the top of rock, thus eliminating the need for conventional rock support with shotcrete and ribs or dowels. This method of construction is unusual for sedimentary rock. The two cells varied in diameter and excavated depth, as each served a different purpose. The 21-m-diameter cell was excavated to a depth of 41 m to support tunneling operations and to allow construction of a junction drop structure and maintenance hole. The 12.5-m-diameter cell was only excavated to a depth of 19 m, allowing the construction of a stub-out connection to a future sewer. Other notable aspects of shaft construction included the use of rock anchors through the partition wall below the shallow cell and the use of weep holes through the shaft walls below the top of rock.

1 INTRODUCTION 1.1

Project description

The City of Los Angeles, Department of Public Works, Bureau of Engineering is presently undertaking two major construction projects to provide relief and redundancy for the aging North Outfall Sewer (NOS). These two projects are the Northeast Interceptor Sewer (NEIS) and the North Outfall Sewer – East Central Interceptor Sewer (NOS-ECIS). At the time this paper was written, the joint venture formed by Kenny, Shea, Traylor, and Frontier-Kemper (KSTFK) had completed tunneling for NOS-ECIS project. Meanwhile, a separate joint venture formed by Traylor, Shea, Frontier, and Kenny (TSFK) is currently mining the NEIS tunnel. An overall vicinity plan for both projects is shown in Figure 1. The NEIS project involves the construction of an 8.5-km-long, 2.4-m-inside-diameter (ID) sewer pipeline in a 4.0-m-diameter excavated tunnel, three drop structures, and seven special maintenance holes. The project must be completed by November 30,

2004 in order to comply with a Cease and Desist Order (CDO) deadline imposed by the Regional Water Quality Control Board. When complete, NEIS will convey flows from existing sewers and the future Eagle Rock Interceptor Sewer southward to NOS-ECIS. 1.2

NEIS will extend from a junction structure with NOS-ECIS at the intersection of Mission Road and Jesse Street, northward to the intersection of Division Street and San Fernando Road, in the Glassell Park area of Los Angeles, on an alignment roughly parallel to the east bank of the Los Angeles River. A project alignment map is provided in Figure 2. 1.3

Site-specific description

This paper focuses on the Humboldt Street work shaft, one of three work shafts being constructed for the NEIS project. The Humboldt Shaft is located on

109 Copyright © 2004 Taylor & Francis Group plc, London, UK

Alignment

09096-14.qxd

2/18/04

11:55 AM

Page 110

Figure 1. Vicinity map.

Figure 2. Alignment map.

the site of a former warehouse structure, near the intersection of Humboldt Street and San Fernando Road (see Figure 3). The shaft is 41 m deep on one side and 19 m on the other. The design team planned the shaft to serve three functions, as:

•

•

The design specified that after the tunnels are mined and the carrier pipe is installed in each reach, a combined drop-and-junction structure with associated

drive shaft for the middle-reach earth pressure balance tunnel boring machine (EPBM), tunneling towards the Richmond Shaft;

•

110 Copyright © 2004 Taylor & Francis Group plc, London, UK

receiving shaft for the upper-reach rock tunnel boring machine (TBM) tunneling from the Division Street Shaft; work shaft for the stub-out connection tunnel for future tie-in to NOS.

09096-14.qxd

2/18/04

11:55 AM

Page 111

Figure 3. Site plan.

Figure 4. Geologic profile.

maintenance holes would be constructed in the deep shaft.

(GED) in the project’s Geotechnical Data Report (GDR):

1.4

•

Geology

Three major geologic units are present at the Humboldt Shaft (see Figure 4). The following is a description of each unit, as described by the Los Angeles Bureau of Engineering’s Geotechnical Engineering Division

•

111 Copyright © 2004 Taylor & Francis Group plc, London, UK

Artificial Fill – Variable in soil type along the alignment, ranging from clayey silt to angular gravel and sand. Recent Alluvium, (Qal) – Fluvial and alluvial deposits (channel deposits, point bar deposits, and flood plain deposits) that have been deposited within

09096-14.qxd

•

•

2/18/04

11:55 AM

Page 112

the past 10,000 years (Holocene age). It consists predominantly of cohesionless silty sands, poorly graded to well-graded sands with gravel, and sands with silt and gravel. Old Alluvium, (Qoal) – These deposits are generally considered to have formed between 10,000 and 700,000 years ago (upper Pleistocene age). Brown fine gravel with fine to coarse sand, containing scattered sand with gravel layers and scattered organic fragments in a clay/silt matrix. Puente Formation, Unit 2 (Tp2) – The lower unit of the Puente Formation, an interbedded siltstone, claystone, and sandstone of Miocene age. The Puente Formation is divided into Tp1 and the Tp2 for the NEIS project. The major difference is that the beds of the Tp2 are thicker and notably stronger than the thinner beds of the overlying Tp1.

For the Humboldt Shaft the GBR indicated that up to 1 m of artificial fill could be expected, underlain by 8 to 9 m of medium dense to very dense recent alluvium, a thin layer (0 to 1 m thick) of older alluvium, then Tp2 to the bottom of the excavation. The groundwater table lies at a depth of about 9 m. None of the project borings around the Humboldt Shaft encountered gas, liquid oil, or tar within the Tp2 or alluvial soils. However, natural hydrocarbons were found in several locations along the alignment. Oil and

in this part of the Los Angeles Basin originates in the petroliferous Tp2 and propagates up along the bedding planes through seams of sand and silty sand. Therefore, Cal/OSHA classified this shaft as “potentially gassy” during shaft excavation. 2 CONTRACT REQUIREMENTS 2.1

The Contract Documents prohibit the Contractor from dewatering outside the limits of the Humboldt Shaft excavation. The reason for this restriction is to prevent migration of potential groundwater contamination and to minimize disruption to the natural groundwater flow. Therefore, the design team selected reinforced concrete diaphragm walls (slurry walls) to support the excavation through the alluvium and to prevent lowering of the groundwater table outside of the excavation. The conceptual design of the Humboldt Shaft shown in the Contract Documents is roughly circular in shape, with an adjoining shallow rectangular cell to the south. The conceptual design included 12 wall panels to approximate a ring and seven to enclose the shallow shaft for the NOS diversion structure (see Figure 5), tied together at the surface with a reinforced concrete cap beam. The design team determined the

Figure 5. Plan view of conceptual shaft.

112 Copyright © 2004 Taylor & Francis Group plc, London, UK

Slurry walls

09096-14.qxd

2/18/04

11:55 AM

Page 113

minimum shaft diameter of a circular shaft at this site to be 21 m in the alluvium and 19 m in rock. These dimensions were chosen to accommodate the permanent structures to be constructed within the shaft, as well as to minimize the amount of rock excavated. The design of the circular cell assumes the walls act as a compression ring, carrying the load by thrust in the panels in the ring’s plane, with no internal bracing required. The design of the rectangular shaft included internal steel bracing and additional reinforcing steel in the wall panels to resist bending stresses. Since slurry walls are very rigid and generally do not allow significant ground movement, the lateral earth pressure loading criteria in the contract documents are a triangular distribution based on averaging the active and at-rest earth pressure coefficients (Ka and Ko respectively). In addition to the triangular distribution, the design criteria included an apparent earth pressure envelope based on the same average K value, which was to be used only for the internally braced, rectangular cell. 2.2 Rock reinforcement The geotechnical exploration program indicated that the Puente Formation is a very weak to moderately strong rock, with most unconfined compression test values falling below 5 MPa. The designers felt the rock strength was adequate to resist the compressive stresses due to hydrostatic and horizontal rock pressures in the rock mass around the circular shaft

opening. However, to ensure a ring of intact rock would carry this load in compression, where joint sets and inclined bedding planes are present, additional rock support analyses were performed. These analyses assumed joint orientations and joint strengths developed from data contained in the GDR and GBR, as well as shaft geometry and locations of contacts between rock and soil. The designers calculated an apparent uniform rock loading, based on the force required to resist the movement of a wedge of intact rock sliding along the most prominent joint sets and/or bedding planes. The pressure diagrams in the Contract Documents included a uniform rock pressure of 67 kPa as a minimum requirement for the Contractor to design the rock support. The Contract Documents also required the Contractor to install strip drains with weep holes between the rock surface and the shotcrete to drain water-bearing joints around potentially unstable wedges that intersect the shaft walls. For this reason, the rock loading minimum design criteria did not include hydrostatic pressure for the design of rock reinforcement. The conceptual design for rock reinforcement shown in Figure 6 included two alternatives for ground support rock:

• •

Figure 6 Section view of conceptual shaft.

113 Copyright © 2004 Taylor & Francis Group plc, London, UK

W8 ribs at 1.8 m vertical spacing, with 150 mm of steel-fiber-reinforced shotcrete. Rock bolts at about 1.8 m 1.8 m spacing, 8.5 m long with 150 mm of fiber-reinforced shotcrete.

09096-14.qxd

2/18/04

11:55 AM

Page 114

These alternatives served as a basis for initial support, with provisions in the contract indicating that modification to these designs may be required, depending on conditions observed in the field. 2.3

Geotechnical instrumentation

The contract required the Contractor to install the following three sets of geotechnical instruments around and within the Humboldt Shaft:

• • •

three inclinometers (shown as ▲ on Figure 5); four piezometers (shown as ■ on Figure 5); twelve horizontal multiple-point borehole extensometers.

The inclinometers and piezometers are considered typical, minimum instrumentation for a shaft of this size and depth. The horizontal multiple-point borehole extensometers are specified for the portion of the shaft in rock. Their primary purpose is to measure lateral ground movement and warn of potentially large block movements. If the maximum lateral movement of 25 mm were exceeded, additional rock anchors or steel ribs would be installed to arrest ground movements and maintain stability of the rock mass.

3 CONTRACTOR’S REVISED DESIGN 3.1

Shaft geometry

The conceptual design consisted of two shafts adjacent to one another. The small, shallow shaft consisted

of slurry walls with waler and strut supports. It was intended that the shallow shaft would carry lateral loads by flexure, which necessitated walers and struts for support. The large, deep shaft was comprised of two different support types. In the alluvium and fill, the deep shaft would carry lateral loads by hoop compression. In the Puente Formation, rock anchors and fiber-reinforced shotcrete would carry the loads directly, or ring beams could be used in hoop compression. Using the variety of support systems as described above would have added time and complexity to the Contractor’s operations. Therefore, the Contractor elected to use a “figure-eight” or dual-cell slurry wall shaft (shown in Figures 7 and 8), similar to the concept used for the Richmond Shaft. The small cell is approximately 19 m deep, and the deep cell is approximately 39 m deep. Ed (Itzig) Heine P.E., and Steve Blumenbaum of Alpha Corporation designed the dual-cell shaft for this joint venture. 3.2

The radial walls were designed by the hoop stress method and were considered to be unreinforced compression members. Only the circular band of concrete inscribed within the limits of the slurry wall panels was considered effective in compression. The Hydrofraise (French for hydro-mill) construction method chosen by the Contractor resulted in average chord lengths of 2.4 m. The contractor-proposed chord lengths were much shorter and therefore more efficient in hoop compression than anticipated in the conceptual design.

Figure 7. Plan view of contractor’s revised geometry.

114 Copyright © 2004 Taylor & Francis Group plc, London, UK

Radial walls

09096-14.qxd

2/18/04

11:55 AM

Page 115

The short chord lengths afforded the opportunity omit a cap beam. When long chord lengths are used, a reinforced concrete cap beam is often used at the top of the shaft to provide continuity and resistance to deformation. But with the close approximation of a circle provided by the shorter panels, a cap beam is not required. Steel reinforcement was not needed for the purpose of resisting lateral loads. However, contingency reinforcing was installed, in case the panels were not installed within the specified tolerances. In that event, the panels could span vertically to remedial ring beams or walers. In the Puente Formation, drain holes were provided in the slurry walls to relieve any groundwater pressures. This was mostly precautionary, given the relative impermeability of the formation. 3.3

Center wall

The center wall was designed for different load conditions depending on depth. Where it is a common wall between the two cells, the center wall was designed for compressive horizontal loads coming from the radial loads in the two cells. It was also designed to accommodate a 1.5 m differential soil loading between elevations in each cell. Reinforcement in this area was designed to limit buckling. At elevations below the bottom of the shallow shaft, the center wall behaves differently. It is subjected to lateral earth loads as well as compressive loads coming from the radial wall of the deep cell. In this area, the wall spans vertically, and reinforcement is used for flexural strength. This

load condition controls the design of the center wall reinforcing. The wall spans between 9 m long rock anchors, which are installed on 2 m 4.5 m centers. Reinforcing cages were only terminated at the top of rock in the circular portion of the shaft. At all elevations, the ends of the radial walls were poured integrally with the end panels of the center wall, and reinforcement was provided across the center/radial wall joint. In this way, shear transfer across the joint is ensured. 3.4

Groundwater considerations

In the fill and alluvium, the slurry walls were designed for the combination of earth and hydrostatic pressures. In the Puente Formation, the slurry walls were designed only to support wedges of rock. In rock, the slurry walls confine the rock mass and the ground to support the horizontal rock and hydrostatic pressures present deeper in the rock mass. Hydrostatic pressures behind the slurry wall and in water-bearing joints around potentially unstable wedges intersecting shaft walls are relieved through the use of weep holes drilled horizontally through the slurry walls. Weep holes were not drilled through the walls above the rock to prevent dewatering of the overlying alluvium. The slurry in the alluvium provides a water barrier that the designers did not want to compromise. 3.5

Summary of advantages

There are several advantages to using the dual-cell, full-depth, slurry wall shaft instead of the conceptual design. First, the construction methods were simplified and shortened by using one excavation method. Second, Hydrofraise construction allowed the designer to eliminate the cap beam. The shorter chord lengths also minimized the need for panel reinforcement. Third, replacing the rectangular small shaft with a circular one minimized the need for reinforcement and eliminated the need for walers and struts in the small shaft. Fourth, no setback was required to change from slurry wall to rock anchor support. This not only reduced the footprint of the large shaft, but it also eliminated the need for a cast-in-place ring beam at the transition from slurry wall to rock support.

4 CONSTRUCTION MEANS AND METHODS 4.1

Figure 8. Section view of contractor’s shaft.

The TSFK Joint Venture selected Soletanche Inc. as their slurry wall subcontractor and chose the Hydrofraise

115 Copyright © 2004 Taylor & Francis Group plc, London, UK

Advantages of hydrofraise method for slurry wall construction

09096-14.qxd

2/18/04

11:55 AM

Page 116

excavation method for several reasons, which are described below: 1. Schedule Advantages – Time restrictions placed on the Contractor by the CDO required that shaft construction be expedited. 2. Achieving Tight Tolerances at Depth – The realtime data supplied to the operator from the fraise allows precise alignment of each panel which assures tight vertical joints at depth. 3. Versatility of the Hydrofraise – The Hydrofraise is able to excavate through all types of materials with minimal modification to the cutting tools. 4. Minimal Impact on Environment – The Hydrofraise method imposes minimal impact on the surrounding environment, which was a necessity in the densely populated vicinity of the shaft site. The operating principle of the Hydrofraise is similar to that of a slurry shield TBM, in which the excavated opening is supported by a pressurized suspension that balances the earth and water pressure of the excavation. In most cases, this suspension is a bentonite and water slurry. The slurry acts not only as a support fluid, but also as a transport medium. The ground excavated by the cutting tool is mixed with the support fluid slurry near the excavation face, where it can then be pumped to the surface. A separation plant, usually on the surface, then separates the support fluid from the ground, and the fluid is again pumped to the excavation face. Fresh bentonite can be added as slurry properties dictate. A typical equipment spread for the Hydrofraise method is comprised of a modified crane, a fraise cutting tool, a bentonite slurry mixing and storage facility, a separation/de-sanding plant, and one or two support cranes. The specific layout of the Hydrofraise used in this project can be seen in Figure 9. 4.1.1 Schedule advantages The Hydrofraise can shorten the construction schedule because of its ability to continuously excavate. The tool is lowered under its own weight into a pre-built concrete guide wall, with the cutting wheels turning. It continues excavating until it reaches the desired depth. In conventional clamshell excavation, the continual raising, lowering, and dumping cycles are time consuming. However, as with any sophisticated piece of equipment, the Hydrofraise is susceptible to mechanical and electrical downtime, whereas the clamshell can be kept running with a minimum of specialized maintenance, tools, and equipment. However, on the NEIS project, the Hydrofraise experienced minimum downtime and was therefore able to keep the project on schedule. Slurry wall panel construction at the Humboldt Shaft lasted 49 days, with the crew working two 10-hour shifts. The approximate area of the slurry wall panels is 3,381 m2, in elevation.

Figure 9. “The Hydrofraise evolution II” by Soletanche, Inc.

4.1.2 Achieving tight tolerances at depth One of the major reasons the Contractor chose the Hydrofraise method was due to its precise excavation control. The design assumptions of shaft geometry are dependent upon the construction tolerances that the equipment can achieve. The cutting tool is equipped with inclinometers and tilt meters that are linked to a computer screen in the operator’s cab. The operator is able to read the information provided by the instrumentation in real-time and make corrections as needed. Several features are available to the operator for steering purposes. First, each of the two rotating cutting wheels can be run at variable speeds to correct for left and right misalignment. Second, the entire cutting head can tilt up to 1.5° in the vertical plane of excavation to correct for front and back misalignment. Such precise control over tool guidance enabled the Hydrofraise to excavate panels on the NEIS project within a tolerance of 0.3%, which equates to a variance of only 125 mm from the designed vertical alignment, over a depth of 42 m. 4.1.3 Versatility of the Hydrofraise The Hydrofraise cutting tool can quickly and easily adapt to changes in ground conditions with modifications to the cutters. Both of the cutting wheels can be removed and replaced in a single shift, which allows the Hydrofraise to perform in almost any ground

116 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-14.qxd

2/18/04

11:55 AM

Page 117

Figure 10. Photo of broken teeth.

Figure 11. Excavation in rock.

condition. As an example, at the Richmond Shaft site, which was the first of three slurry walls constructed for NEIS, the Hydrofraise was equipped with selfcleaning cutting paddles to deal with the soft claystone and mudstone present in the area. However, when harder sandstone was encountered at the Humboldt site, the slurry wall Sub-Contractor quickly replaced the paddles with carbide tipped picks. The carbide picks performed well; the only problem was chipping of the carbide tips in the hardest layers of the Tp2 as shown in Figure 10. It was the versatility of the Hydrofraise that enabled the Contractor to further accelerate the schedule by extending the slurry walls through the Puente Formation and eliminating rock bolting and shotcreting from the shaft excavation activities.

concrete of the secondary panels to bond to, thus producing a strong and relatively watertight joint. The concrete was poured using dual tremie pipes in the primary and follow-up panels and a single tremie pipe in the secondary panels.

4.2

Panel construction sequence

The circular shape of the shaft was approximated with short chords because the Hydrofraise is limited to excavating rectangular shaped sections. The chord length for the Humboldt slurry wall ranged from 1.804 m to 2.448 m. The wall was constructed in 43 “bites,” with each bite being one pass of the cutting tool. The wall was also constructed in 19 “panels,” which were either: a primary panel comprised of three bites, a secondary panel comprised of one bite, or a follow-up panel of five bites. A secondary panel separated each primary panel and the follow-up panels were used to create the joint between the two cells. Each primary and follow-up panel was excavated and concreted first, before the secondary panels were excavated. Tight joints between the primary and secondary panels were constructed by spacing the primary panels so that the Hydrofraise cut into the previously poured concrete of the primary panels, while it was excavating the secondary panels. The cutting of the primary panels produced a rough surface for the

4.3

Because the shaft was originally classified as “gassy” by Cal/OSHA, the Contractor chose to drill a test hole prior to excavation of the shaft for the purpose of drawing gas samples. It was hoped that the Cal/OSHA would reclassify the shaft based on favorable results from these samples, and thus allow the use of conventional equipment for the shaft excavation. The shaft was indeed reclassified to “potentially gassy” with special conditions, based on the gas samples, and the Contractor was allowed to proceed with conventional equipment. The first 5.2 m of the shaft, which consisted mostly of artificially backfilled sand and alluvial sand and clay, was excavated from the surface by a Caterpillar 325 excavator. For the next 4 m of excavation, the CAT 325 excavator was placed in the shaft where it then loaded two 3.8 m3 circular muck skips, which were hoisted on a single line by a 125-ton-capacity American 9260 crane, as shown in Figure 11. The crane was previously factory modified for deep tunnel operations. This 4 m of excavation consisted mostly of alluvial sand and clay. At an approximate 7.6 m depth, the soil became sticky and produced a strong hydrocarbon odor. A chemical analysis of the excavated material revealed that the soil contained a high concentration of natural petroleum hydrocarbon, which is not uncommon in the Los Angeles Basin. The soil was classified as “Contaminated” and was removed and dealt with by the Contractor’s environmental subcontractor.

117 Copyright © 2004 Taylor & Francis Group plc, London, UK

Shaft excavation

09096-14.qxd

2/18/04

11:56 AM

Page 118

The Tp2 Formation was encountered at a depth of 9.1 m, at which point the CAT 325 was no longer able to freely excavate the material with a bucket. Consequently, a 53 kN hoe-ram was attached to the CAT 325 to break the hard layers of the Tp2,, while a CAT 312 excavator was put into service to load the broken material into muck skips. The production rate of each piece of equipment was quite evenly matched so that the two excavators could follow each other around the shaft, one breaking material and one loading material. The efficiency of the operation led to a production rate of about 550 m3 per shift, which equated to approximately 1.2 m (in depth) per shift in the large cell. The CAT 312 excavator was utilized in the small cell, since the shaft was not large enough to accommodate the CAT 325. With aggressive bucket teeth, the CAT 312 was able to excavate nearly all the material down to a depth of 26 m unassisted. Where it was not able to dig, a smaller hoe-ram attached to a Case 580 loader assisted the operation by breaking the harder material. With these production rates, the shaft was sunk in 39 working days, utilizing two 8-hour shifts per day. 4.4

Figure 12. Drilling weep holes in the slurry wall.

Rock anchor installation

The rock anchor scheme in the straight center wall of the shaft consisted of four rows of anchors with five anchors per row and a vertical and horizontal spacing of 4 m and 2 m, respectively. The design specified 9 m long, 35-mm-diameter, 1,030 MPa Dywidag Threadbar anchors, which were to be installed in a 70 mm hole and encapsulated in cementitious grout. The Contractor chose to use a Gardner-Denver PR 123 rock drill mounted to an ATD 3800 Air Trac drill carrier (as seen in Figures 12 and 13) for the drilling in the relatively soft rock of the Tp2 Formation, 10-m-long, 100 mm diameter holes could be drilled in a matter of minutes. Shorter holes drilled as weep holes yielded some water immediately after drilling and periodically during a relatively dry rainy season in 2002. Contract Specifications stated that shaft excavation was not allowed more than 1 m below a row of anchors until each bolt was pull tested. In order to gain high early strength and a quick turnaround on the pull test, a prepackaged non-shrink rock anchor grout formulated by Euclid was initially selected for the cementitious encapsulation. It was to be batched and pumped by a Hany IC 310 colloidal mixer. A prepackaged product was selected with the hope that it would reduce batching times and improve quality assurance of the grout, since the Contractor could not afford to halt shaft sinking production in order to reinstall a failed anchor. However, after a number of unsuccessful attempts to mix and pump the prepackaged product, it became apparent that the Euclid material was not compatible with the Hany equipment at the water-to-cement

Figure 13. Drilling for rock anchor installation.

ratio the Contractor needed to achieve. It was decided that a switch would be made to a traditional cementand-water grout mix. Master Builders MEYCO Fix Flowcable was added to the mix to reduce water requirements, increase pumpability, and provide nonshrink properties. After this change was made, the Hany mixer and pump performed flawlessly. Each anchor was pull tested approximately 20 hrs after installation, with all anchors passing, except the last one.

118 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-14.qxd

2/18/04

11:56 AM

Page 119

execution of the NEIS contract. Jacobs Associates provided design services for the Department of Public Works, Bureau of Engineering and prepared the minimum design criteria used by the Contractor to design the Humboldt Shaft. Parsons Brinckerhoff Construction Services/Brown & Root Services (J.V.) and Jacobs Associates continue to work with Baron Miya and Rajni Patel of the Bureau of Engineering as an integrated team managing the construction of the NEIS project. Chris Smith and Richard Calvo of the Bureau of Contracts Administration supervised inspection of the work and ensured that the slurry walls were constructed in accordance with the demanding project specifications. Figure 14. Weep holes drilled through the slurry wall.

REFERENCES

The anchor failure is believed to have been caused by a transient water flow which washed grout out of the hole. Another anchor was installed immediately and passed pull testing. 5 CLOSING REMARKS Excavation of the Humboldt Shaft ended on March 28, 2003. The shaft excavation was never on the project’s critical path. This can be attributed to the successfully planned and implemented slurry wall operation devised by the TSFK joint venture and their subcontractor, Soletanche Inc. The joint venture kept the project on schedule through shaft construction, despite the rigorous demands dictated by the CDO.

McKenna, M. P., Traylor, D. A., Tarralle, B. and Itzig-Heine, E. 2003. Design and Construction of a Deep, Dual-Cell, Slurry Wall Shaft in Soft Ground, Proceedings of the Rapid Excavation and Tunneling Conference, 368–382. City of Los Angeles Department of Public Works, 2001. Geotechnical Baseline Report: Northeast Interceptor Sewer (NEIS. City of Los Angeles Department of Public Works, Bureau of Engineering. City of Los Angeles Department of Public Works, 2001. Geotechnical Data Report: Northeast Interceptor Sewer (NEIS. City of Los Angeles Department of Public Works, Bureau of Engineering.

ACKNOWLEDGEMENTS The authors would like to recognize some of the key firms and individuals responsible for the design and

119 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-15.qxd

2/18/04

11:56 AM

Page 121

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Performance of Russia Wharf Buildings during tunneling Hugh S. Lacy Mueser Rutledge Consulting Engineers, New York, NY, USA

Marco D. Boscardin Boscardin Consulting Engineers, Inc., Amherst, MA, USA

Leslie A. Becker Massachusetts Bay Transportation Authority, Boston, MA, USA

ABSTRACT: Twin transit tunnels were mined through the timber pile foundation system that supports two historic buildings while maintaining the serviceability and occupancy on all seven floors. Ground stabilization using soil freezing was employed to permit tunneling via NATM methods. This paper discusses and compares the anticipated and actual performance of the buildings.

1 INTRODUCTION MBTA’s Silver Line Phase II in Boston (formerly known as the South Boston Piers Transitway Project) includes construction of a transit tunnel below two, 100-year old buildings. The tunnel under the building is a binocular-shaped structure about 8.5 m high by 13 m wide, constructed using NATM techniques. The tunnel passes directly under both buildings with about 4.6 to 7.6 m feet of cover. Lacy et al. (2000) describes how the buildings were supported while tunnel construction occurred below. This paper will focus on the response of one of these buildings, the Graphic Arts Building, the construction activities and how adverse impacts on the building were mitigated. At the time this paper was prepared, the tunnels were nearly 80% complete. 1.1

The buildings

The three Russia Wharf buildings are shown in Figure 1. The buildings are seven-story, historic structures with steel frames and brick and granite masonry facades, circa 1897. The northwest corner of the third building, the Tufts Building, is located immediately adjacent to the tunnel. The buildings have single-story basements used for parking, and the brick exterior bearing walls and steel interior columns are supported on large granite block pile caps and timber piles. The floor system in each building consists of relatively flat arch masonry and concrete barrel vaults spanning

Figure 1. Russia Wharf Buildings, facing east.

between steel beams that frame into the columns. Typical column spacings are 4 m to 4.6 m . The building use is light commercial and tenants include a ship gallery, architectural firms, development companies, a copy/ printing shop, and a restaurant. The tunnels extend diagonally below the Russia Building (foreground), an atrium between the buildings

121 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-15.qxd

2/18/04

11:57 AM

Page 122

RUSSIA WHARF

EXISTING SEAWALL (LOCATION AND DEPTH UNKNOWN)

ELEVATION, FEET (PROJECT)

EDGE OF WOODEN WHART

Figure 2. Site plan.

M.H.W. 110 EL.104.91 M.L.W. FILL 100 EL.95.10 ORGANIC SOIL 90 Fort Point channel 80 70 Tunnel Alignment 60 CLAY 50 GLACIAL TILL 40 30 20 10 0 BEDROCK -10 -20 93+00 94+00 95+00 96+00 97+00 98+00 99+00 100+00 STATION IN FEET

Figure 4. Soil profile along tunnel alignment.

Figure 3. Section through Graphic Arts Building.

and then below the Graphic Arts Building (middle building). The prime concerns relative to the buildings during tunnel construction include safety of the tenants and users of the buildings, protecting the historic fabric of the structures, and maintaining a facility that continues to serve the functions of the tenants without interruption. 1.2

Transit tunnels

The tunnel alignment is shown on Figure 2. The tunnel profile extends deeper to the east as it approaches Fort Point Channel. Figure 3 is a section through the tunnels at the Graphic Arts Building. 2 DESIGN 2.1

Subsurface conditions

The project site is located at the edge of the Boston’s Shawmut peninsula and adjacent to the Fort Point Channel. The buildings are constructed on an area that is filled land created by several episodes of filling of the working waterfront and mudflats during the 19th century. The soil profile along the tunnel alignment is shown in Figure 4 and in general consists of fill over organic silts and clays, over a silty marine clay, over a

dense to very dense silty, sandy glacial till. The fill ranges from 1.5 m (5 ft) to 4.6 m (15 ft) thick generally located above the crown of the tunnel. The fill includes granular and silty soils (primarily glacial till origins) excavated from the uplands to the west and miscellaneous debris ranging from timbers, boulders, cut stone, bricks etc. The organic silts and clays are discontinuous and range from 0 m to more than 3 m (10 f t) thick and are of marine origin (formerly in the mudflats and channel bottom area). The tunnel crown runs along and generally below the organic/marine clay interface. The marine clay is Boston Blue Clay and ranges from 3 m (10 ft) to more than 9 m (30 ft) thick along the tunnel alignment under the buildings. The upper 3 m of the clay is a stiff crust and the remainder of the clay is of medium stiff consistency. For most of the alignment, the tunnel face is primarily in the clay. Below the clay is a very dense silty, sandy glacial till. The invert of the tunnel is in the glacial till at the western end of the tunnel. Water levels are nominally at the level of the adjacent Fort Point Channel. 2.2

The tunnels were planned in a “binocular” shape to minimize total tunnel width yet permit staged construction with reduced impact on the buildings. Figure 3 illustrates the relative position of the two tunnels with the top bench of the tunnel to the right mined first using lattice girders and shotcrete for the initial lining. The secondary lining consisted of castin-place invert and vertical dividing wall and fiber reinforced shotcrete for the remaining curved sections. 2.3

Support of buildings

The Russia Building was underpinned prior to tunneling using high capacity mini-piles installed from within

122 Copyright © 2004 Taylor & Francis Group plc, London, UK

Tunnel construction method

09096-15.qxd

2/18/04

11:57 AM

Page 123

Figure 7. Tunneling procedure.

the load from a row of severed timber piles to arch around the narrow strip being mined at the tunnel face as illustrated in Figure 7.

Figure 5. Russia Wharf Building underpinning plan.

2.4

Figure 6. Temporary column support at Graphic Arts Building.

the building basement. The mini-piles were installed along the edges of and in the center wall between the two tunnels. A system of distribution beams transferred the column and wall loads to the mini-piles as shown in Figure 5. Columns near the tunnels were re-leveled using temporary cribbing and jacks. The Graphic Arts Building was temporarily underpinned and supported on a pad of artificially frozen ground prior to tunneling. As the tunnels were being mined, the timber piles were cut off in the tunnel face and the pile stub in the tunnel roof was embedded in the lining, transferring support of the column above to the tunnel. Temporary cribbing and jacks at the column base (Figure 6) were used to minimize building movement. The frozen ground provided pile cap support causing

123 Copyright © 2004 Taylor & Francis Group plc, London, UK

Groundwater control

The water level in the fill, organic soils and clay is nominally at the level of the adjacent Fort Point Channel (FPC) and approximately 4.6 m to 7.6 m above the crown of the tunnel. The measured groundwater levels exhibit a tidal influence of 0.3 m to 1.0 m depending on distance from the channel. During test pitting in the basements of the buildings, it was noted that gaps were present under the basement slab and that water rushed into the test pits during high tide. In addition, highly conductive voids related to the site filling in the 19th century were expected to be present around timbers and cribbing. Due to the shallow nature of the tunnel crown, the weak and loose condition of the fill and organic soils at or immediately above the tunnel crown and the close proximity of a large source of water (the FPC), positive means of ground water control/cutoff were needed to tunnel safely. This was provided by constructing frozen ground cutoff walls on either side of the tunnel alignment and mass ground freezing over top of the tunnel. In addition, jet grouting was used to form closure between the frozen ground mass and frozen cutoff walls and the existing slurry walls at the west end of the tunnel. During the freezing, the contractor experienced difficulty in reaching target temperatures in the Atrium cutoff wall suggesting the possible presence of voids in the area. This condition was addressed by injecting, at low pressures, limited mobility cement/bentonite grout through a line of cased grout holes to fill larger voids. This was followed by permeation grouting using a thin microfine cement grout to fill the smaller, remaining voids. Ground water control methods were

09096-15.qxd

2/18/04

11:57 AM

Page 124

effective so only limited sumping of water from inside the tunnel was performed. 2.5

Ground stabilization

Ground stabilization consisted of artificial ground freezing above the tunnel spring line below the buildings and grouted spilling across the tunnel roof combined with local dewatering where the tunnel was advanced below the atrium and non-building areas. Tunneling was advanced in 760 mm increments. Ground freezing causes an increase in shear strength through a phase change of soil moisture to ice. Colder frozen ground results in higher frozen strength. Frozen soil temperature requirements ranged from 10°C to 15°C, and were determined based on analysis of stresses in the ground around the mined opening. Freezing of clays, silts and to a lesser extent silty sands causes soil expansion due to the expansion of water during phase change. This had the potential to cause both lateral and vertical ground movement particularly where a large area is being stabilized. This is in contrast to a single line of freeze pipes in a circular pattern for stabilizing the sides of a deep shaft where much smaller ground movement is generally observed. Initial horizontal ground movement occurs during the period where the ground is freezing around freeze pipes and spreading to merge with that of adjacent freeze pipes. The initiation of ground freezing was planned in two stages to minimize heaving of the pile caps. The first stage was to circulate brine in the pipes closest to the pile caps to permit lateral movement resulting in ground heave between the pile caps where the basement floor had been removed. The second stage turned on the flow of chilled brine to freeze pipes further from the pile caps in areas where ground heave would have less impact on the building foundations. Following initial completion of the frozen mat, temperatures were gradually lowered to meet the strength requirements (Figure 8). During this period, additional ground heave occurred due, in part, to increased growth of the frozen zone downward and horizontally at the edges of the frozen pad. Prior to the start of tunneling below the Graphic Arts Building, contours of ground temperature above the tunnel alignment (Figure 9) were evaluated to determine if soil strengths met or exceeded the design requirements. 2.6

Mitigation of movements

The tunnel design incorporated temporary and permanent underpinning, as well as, ground-freezing to support the buildings and control their movements during and after tunnel construction which included cutting out and removing the timber piles and transferring the loads to the tunnel lining or the permanent underpinning. To evaluate the combined impacts

Figure 8. Freeze pipes in place.

Figure 9. Ground temperature contours.

including the tunneling, ground displacements due to the tunneling were estimated using an ABAQUS finite element model (Dr. G. Sauer Corporation, 1999). The effects of the ground freezing on soil properties (Mueser Rutledge Consulting Engineers, 1998) were included in the modeling. The tunneling-related ground displacements were then considered in combination with the open cut-related ground displacements to estimate building distortions and potential for damage. The conclusions of the combined impacts evaluation were that the tunnel construction could maintain the building response in the range typically associated with very slight to slight cosmetic damage. Due to the building foundation loads and the very shallow nature of the tunnel, the tunnel construction included temporary underpinning (Figure 6) to provide additional support to building foundations during the tunneling. The underpinning system also included provisions for adjusting column elevations in the freezing zones to mitigate the effects of potential ground movements due to ground freezing. Based upon input and requirements from the Building Owner, the Historic Commission, the Project Conservator, and the MBTA, the Design Team established threshold and limiting values for heave/settlement and angular distortion of the buildings. Threshold heave/settlements and angular distortions

124 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-15.qxd

2/18/04

11:57 AM

Page 125

3 CONSTRUCTION

10

3.1

Temperature, deg C

5

0

-5

-10

-15

-20

Jul-03

Aug-03

Jun-03

Apr-03

May-03

Mar-03

Jan-03

Feb-03

Dec-02

Oct-02

Nov-02

Sep-02

Jul-02

Aug-02

Jun-02

May-02

-25

Date

03/14/04

12/15/03

09/16/03

06/18/03

03/20/03

12/20/02

09/21/02

Column Lowered 8-21-02 Column Lowered 11-22-02

06/23/02

03/25/02

12/25/01

09/26/01

06/28/01

03/30/01

12/30/00

10/01/00

0.03 0.02 0.01 0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 -0.08 -0.09 -0.1

Column Lowered 7-3-02

Figure 10. Ground temperature history at CPC-06 (Graphic Arts Building).

Change in Elevation (ft)

Building movement and building column adjustments prior to tunneling

Date

Figure 11. Ground movement with time columns C-9 and D-9.

set in the design documents were 6 mm and 1/700, respectively. Where the frozen zone extended close to sensitive structures such as the building heating plant or where the frozen soil groundwater cutoff wall extended close to areas that could be impacted by ground freezing expansion, ground warming pipes were used to prevent expansion. The Contractor also used jet grouting in place of the frozen groundwater cutoff wall to avoid a buildup of lateral load on slurry walls. Following a period of sustained ground freezing to bring temperatures down to required levels and experiencing a high rate of column heave in the area of ground freezing, a program of cycling the freezing system was employed to reduce the rate of heave. This maintenance mode typically involved turning off part of the freezing system for two weeks followed by reactivation for one week as shown on Figure 10 for one of the nearly 50 temperature monitors. This caused a slowly rising average temperature trend which was maintained below required levels. The impact on column heave was dramatic as shown on Figure 11.

During construction, the elevation of each column in the structures was monitored at least weekly, and in local areas more frequently during periods of active tunneling within 15 m of the tunnel face. These data were transmitted to the engineer immediately, and the contractor and engineer each reviewed the data and compared it to settlement and angular distortion limits agreed upon with the Building Owner. Based on the data and observations of the building response, decisions were made regarding when and how much to adjust column elevations to keep the angular distortion within the agreed upon limits. Prior to the start of tunneling in October 2002, the underpinning had been installed and the freezing operations for the cutoff wall and mass freeze over the tunnel had achieved target temperatures. At this time, the measured ground displacements in response those construction activities ranged from 0 mm to a heave of approximately 56 mm. Figure 12 illustrates contours of total column heave without column adjustment. An example of column adjustment is shown on Figure 11. During the pre-tunneling construction period, adjustments to the building column elevations were made periodically to maintain building angular distortions at 1/400 or less after consultation and agreement with the building owner. This equates to a differential vertical displacement of about 10 mm or less between adjacent columns. Note: typical measured vertical column displacements during underpinning activities were in the range of about 5 mm to 10 mm. In some areas, the ability to adjust the columns was limited and the angular distortion limits were further relaxed, but in no case were angular distortions less than 1/200 permitted to develop. Most column pairs sustained angular distortions of 1/700 or less. Columns at the edges of the freezing zone generally required the most frequent adjustment. Column adjustments were generally performed during periods of low occupancy of the building (e.g. 5 a.m.) and were monitored by representatives of the MBTA, the Engineer, and the Building Owner. During this period, column elevation adjustments were necessary, relatively frequently, sometimes a couple of times a month. The building response was consistent with this level of angular distortion with cracks observed in the 1 mm or less range. 3.2

Building movements that developed during the time period that tunneling occurred have two

125 Copyright © 2004 Taylor & Francis Group plc, London, UK

Building movement and column adjustments during and following tunneling

09096-15.qxd

2/18/04

11:57 AM

Page 126

components: displacements (settlements) related directly to the tunnel excavation and displacements (heave or settlement) related to the maintenance freezing of the ground above the tunnel. After mining of the outbound side of the tunnel was complete (approx. March 2003), cumulative ground displacements from the start of MBTA construction ranged from about 6 mm of settlement to 46 mm of heave. Measured displacements directly related to tunnel excavation were settlements generally in the range of 5 mm to 8 mm. Although adjustments were made to column elevations during the tunnel construction, the adjustments were made in response to the ground freezinginduced heave and not the tunnel settlements. Again column displacements and angular distortions were monitored and adjustments were made when the data indicated angular distortions approaching agreed upon limits. Again, most column pairs sustained angular distortions of 1/700 or less. However, one column pair at the edge of the freezing zone sustained angular distortions approaching 1/200 and underpinning and adjustment of a column that had not previously been underpinned was performed to correct the condition. Building response was consistent with the distortions. Observed cracking was generally in the 1 mm or less range, even at the newly underpinned column, with only a couple of instances where preexisting cracks opened more. One particular case was in a stairwell at the edge of the frozen zone where the configuration of the stairwell and entry doors on each floor served to concentrate movements and open an existing crack by 5 mm to 6 mm. During this phase of the work, column adjustments were performed at a frequency of less than 1 per 2 to 3 months. After the mining of the inbound tunnel through the Graphic Arts Building was completed (approx. October 2003), cumulative ground displacements from the start of MBTA construction ranged from about 15 mm of settlement to 41 mm of heave. Measured displacements directly related to tunnel excavation again were settlements generally in the range of 5 mm to 8 mm. During this phase of the work, the ground freeze energy cycling was tuned sufficiently so that no column adjustments were performed. Cycling of ground freezing started in early 2003 (Figure 10) caused settlement of column pile caps (Figure 11) as illustrated in a comparison of contours in Figures 12 and 13. 3.3

Tunnel deformation during and following tunneling

Following completion of the outbound tunnel initial lining interior monitoring of control points mounted on the lining indicated both north and south tunnel walls moving slightly to the south at a rate of 0.1 to 0.7 mm per day with no significant tendency of convergence or expanding of the sidewalls. The southward

Figures 12 and 13. Contours of pile cap heave from 9/02 and 10/03, respectively, (in feet).

126 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-15.qxd

2/18/04

11:57 AM

Page 127

movement appears to be the result of the more predominant mass of frozen ground to the north that continued to expand. Total movement was not significant.

(5) Successfully limiting building distortion required extensive monitoring results to adjust building columns in a timely fashion. This required a high level of collaboration between all parties.

4 CONCLUSIONS ACKNOWLEDGEMENTS This highly unusual and challenging method of constructing transit tunnels below historical buildings resulted from the need to maintain the fully occupied buildings in operation during construction. Successful completion of construction of this project permits the following conclusions: (1) Ground freezing can successfully be employed to stabilize ground below pile caps permitting mining of tunnels, severing of piles and resupport of the building on the tunnel lining. (2) Ground movement during freezing including the creation of forces necessary to lift 7-story buildings are a function of ground conditions, groundwater control during freezing, confinement, frozen ground temperature, and details of the freezing operation as implemented and maintained by the contractor and can not easily be predicted. Measured values can exceed estimated values. During formation of the mass freeze relatively modest ground heave occurred. During subsequent lowering of frozen ground temperatures, the rate of column pile cap heave increased markedly, due in part to secondary freeze affects. (3) Use of cycling of the freezing units to maintain the required temperatures without causing expansion of the frozen ground area was very effective in limiting additional building distortion. The measured building pile cap settlement during and following tunneling appears to be related primarily to cycling of the freeze system rather than tunnel deformation. (4) The method of releveling columns at their base used for this project successfully limited building distortion to acceptable values. The building sustained relatively little distress. The most noticeable cracks were where the frozen ground cut-off walls passed beneath exterior walls where there was no provision to relevel these massive walls.

The authors wish to acknowledge several firms and individuals for the information that form the basis for this paper, including: The Massachusetts Bay Transportation Authority – Project Owner (Mr. D. Ryan, Mr. E. Karpinski, Jr., and Mr. T. Bretto); DMJM+ HARRIS – Prime Design Consultant; Dr. G. Sauer Corporation – Tunnel Design Consultant; Modern Continental Construction Company – General Contractor (Mr. T. Hennings and Mr. R. Cotes); Layne Christensen Company – Ground Freezing Subcontractor (Mr. J. Sopko); Mueser Rutledge Consulting Engineers – Ground Freezing and Underpinning Consultant (Mr. F. Arland, Mr. T. Popoff, Dr. D. Chang, Mr. P. Deming); and GEI Consultants, Inc. – Geotechnical Consultant (Ms. K. Wood).

REFERENCES Dr. G. Sauer Corp., 1999. Memoranda to GEI Consultants, Inc. regarding ABAQUS FEM analyses for Russia Building and Graphic Arts Building, MBTA Contract E02CN15, South Boston Piers, Russia Wharf and Fort Point Channel Tunnel. Boston, MA. GEI Consultants, Inc., 1999. Estimate of combined impacts of CA/T (C17A1 & C17A3) and MBTA Transitway construction on Russia, Graphic Arts, and Tufts buildings, MBTA Contract E02CN15, South Boston Piers, Russia Wharf and Fort Point Channel Tunnel. Boston, MA. GZA Geoenvironmental, Inc., 1998. C17A3/17A1 combined impacts. Memorandum prepared for FST/HNTB, Joint Venture, Central Artery (I-93)/Tunnel (I-90) Tunnel Project. Boston, MA. Mueser Rutledge Consulting Engineers, 1998. Geotechnical analysis of tunnel stability using artificially frozen soil, South Boston Piers Transitway, Section CC03A. Russia Wharf report prepared for Frederic R. Harris.

127 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-16.qxd

2/18/04

11:58 AM

Page 129

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Blasting adjacent to high voltage duct banks K.R. Ott, D.A. Anderson & S.E. Haq Parsons Brinckerhoff Quade & Douglas, Inc., New York, USA

ABSTRACT: Development of appropriate specifications and criteria related to blasting is a necessity for a project to proceed effectively. A vibration limit of 5 cm/sec is often applied to close in blasting, where it is often not appropriate. Rock excavation was required for construction of an elevator pit in an underground station that will be part of New Jersey Transit’s Hudson-Bergen Light Rail Transit System. The new underground station is being constructed within an existing railroad tunnel, the Weehawken Tunnel, which was originally constructed in 1881 as a dual-track freight line. New Jersey Transit purchased the tunnel to incorporate it into the HudsonBergen Light Rail Transit System, which will link the line along the west shore of the Hudson River, through the Palisades Ridge, to areas north and west of the Hudson River in New Jersey. The elevator pit at the stationshaft intersection was excavated as close as 0.70 m from existing in-service 230 kV oilostatic (pressurized oil conduit) electric duct banks using controlled blasting techniques. The two duct banks run through the tunnel along the invert-sidewall corners of the tunnel. Blasting operations were monitored with dynamic foil strain gages, high-frequency geophones, and standard blasting seismographs. Our recommended criterion of 50 cm/sec as a safe maximum for vibration levels and a strain limit of 600 microstrains as a conservative value for protecting the structures from damage were successful. Peak particle velocities of up to 25 cm/sec and strains up to 150 microstrains did not cause damage.

1 INTRODUCTION Development of appropriate specifications and criteria related to blasting is a necessity for a project to proceed effectively. A vibration limit of 5 cm/sec is often applied in close-in blasting, where it is often not appropriate. It can be hard to convince a layperson that a vibration criterion for threshold damage to residential structures doesn’t apply to a duct bank, a pylon, a dam, or a tunnel wall. Damage to credibility can occur with no damage to structures. The best approach is to make sure the specifications are written clearly in the first place, and communicated to concerned parties. Sometimes, measurements other than standard seismographic data are needed to provide assurance that proper protective measures are being taken. We will discuss development of specifications for a project in which an existing structure was in close proximity to proposed blasting, and how the job was completed. Rock excavation was required for construction of an elevator pit in an underground station that will be part of New Jersey Transit’s Hudson-Bergen Light Rail Transit System. The new underground station is being constructed within an existing railroad tunnel, the Weehawken Tunnel, which was originally constructed

in 1881 as a dual-track freight line. New Jersey Transit purchased the tunnel to incorporate it into the Hudson-Bergen Light Rail Transit System, which will link the line along the west shore of the Hudson River, through the Palisades Ridge, to areas north and west of the Hudson River in New Jersey. The major tunnel construction components include the following: 1. Demolition and reconstruction of two portal structures, 2. Construction of enlarged and retrofitted running tunnels with dual track ways from the portals to an underground station, and 3. Construction of an underground station, which includes a large, single shaft for rider access from the ground surface and ventilation and other utilities. The elevator pit at the station-shaft intersection was to be excavated as close as 0.70 m from existing in-service 230 kV oilostatic (pressurized oil conduit) electric duct banks using controlled blasting techniques. The two Public Service Electric & Gas Company (PSE&G) duct banks run through the tunnel along the invert-sidewall corners of the tunnel (Figure 1). Each existing 230 kV line consists of a 230 kV cable installed in a 22 cm pressurized oil pipe. The pressurized steel pipe and two 12.7 cm PVC conduits are encased in

129 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-16.qxd

2/18/04

11:58 AM

Page 130

Figure 1. Weehawken Tunnel cross section with PSE&G duct banks.

Jurassic igneous sill that extends as a ridge for more than 64 km along the west bank of the Hudson River in a northeasterly direction from Jersey City through Hoboken and Weehawken to Haverstraw, New York. Near vertical jointing patterns are characteristic of the Palisades cliff-face and generally extend throughout the body of the sill. The bedrock of the Palisades is a gray to dark gray fine-to medium-grained diabase, and is characterized as strong to very strong, close to moderately fractured, unweathered to slightly weathered. Unconfined compressive strength tests on rock samples varied from 165 to 248 MPa.

3 PREPARATION

Figure 2. Elevation view of the elevator pit and duct bank locations along the tunnel.

unreinforced concrete to form a duct bank in each invert corner. The duct banks abut directly against the rock tunnel sidewalls and are typically cast on top of ballast material on the tunnel invert, although in some places the duct banks rest directly on rock. Construction plans called for both duct banks to remain in the tunnel, but to be relocated from their present location in the invert corners to a pair of new trenches to be cut into the tunnel invert, below future light rail tracks. Prior to relocating the existing duct banks, the elevator pit was excavated (Figure 2). The tunnel construction work is being performed for New Jersey Transit (NJ TRANSIT) by a joint venture of Frontier-Kemper Constructors, Inc., Shea Construction, and Beton und Monierbrau (FKSB). Parsons, Brinckerhoff, Quade and Douglas, Inc. (PBQD) of New York is the design engineer for NJ TRANSIT.

Construction specifications outlined blasting limitations, blast design requirements, and blast monitoring requirements for blasting adjacent to the PSE&G duct banks. During the design phase, a maximum strain criterion was developed to reduce the potential for damage to the duct bank from blast vibrations. A strain limit was used as a means to avoid overstressing the steel conduit of the oilostatic line, which in turn maintains continuous service. Although vibration criteria have been developed for pressurized pipelines (Westine et al., 1978; Dowding, 1996), these are typically buried in soil, and the strain calculations are related to the soil-pipe interaction. We were looking for criteria for a concrete beam, which is the enclosure that stiffens and supports the oilostatic line. Although an exact analog was not available, we knew that strain limits had been used on other close in blasting projects; such as the Folsom Dam project (Revey and Scott, 1999) and a Lock and Dam project in Minneapolis (Tart et al., 1980). Based upon the spall limits in the Tart et al. study, a limit of 600 microstrains was stated in the contract specifications as a conservative value for protecting the structures from damage. Obtaining strain measurements are time-consuming and expensive. Therefore, in addition to the strain measurements, peak particle velocity (PPV) measurements were required to be recorded at the same location, to develop a relationship between strain and PPV. This relationship would be used to assess predicted strain levels based on PPV measurements during future production blasting. However, during the planning of the blasting program, FKSB requested that a vibration criterion be established prior to any blasting for ease of determining allowable charge weights. We used the plane-strain conversion formula (Dowding, 1996): (1)

2 GEOLOGY The rock type along the tunnel consists mainly of the Palisades Diabase. The Palisades Diabase is an early

. where  strain (in microstrains), u  PPV, and c  seismic velocity of the transmitting medium

130 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-16.qxd

2/18/04

11:58 AM

Page 131

(concrete) (Note that the PPV and seismic velocity need to be in the same units.) If we assume the seismic compressional wave velocity of concrete is about 3000 m/sec, a strain criterion of 600 microstrains yields a peak particle velocity of about 175 cm/sec. Meetings with PSE&G and the design and construction team were held prior to start-up of blasting. PSE&G stated concerns about blast induced damage to the duct banks due to the high PPV criterion. They wanted a more conservative value, 5 cm/sec. The lower criterion is in the New Jersey Administration Code and is applicable and appropriate for protection of residential structures and other above ground buildings. With the lower criterion, blasting would not have been feasible. After explaining the source of the 5 cm/sec criterion to PSE&G, and listening to their concerns, we recommended an intermediate criterion of 50 cm/sec as a safe maximum for vibration levels in the duct bank. (Due to advances in blast monitoring equipment since the original contract specification was prepared in 1998, use of a PPV to monitor the duct banks was possible.) Charge weights could then be established, using the Oriard high-confinement equation (ISEE, 1998): (2) where Ds  scaled distance (D/W0.5) Blast design requirements included the following: 1. Smoothwall blasting techniques were specified to minimize overbreak and damage to the final rock surface, 2. Loading density was restricted for perimeter (smoothwall) holes; 0 to 0.6 kg/m, and buffer holes were allowed up to 1.0 kg/m, 3. Burden to spacing ratio of 1.3 to 1.5, and 4. Line drilling along the perimeter of the elevator pit (hole spacing equal to two to four times the hole diameter). We judged that vibration was in fact not the most likely factor that could adversely affect the duct banks. Mass movement due to possible back-break from the confined elevator shaft would potentially have a greater effect on movement of the duct bank. Therefore, we required the line drilling, smoothwall holes, and a blast sequence that provided adequate relief along the walls closest to the duct bank. The contractor developed and submitted a blast plan, which was reviewed by PBQD and PBQD’s blasting consultant, Mr. Gordon Revey. After several reviews and coordination meetings, a plan was accepted and blasting commenced.

1

Figure 3. Plan view of blast monitoring instruments.

The final aspect of the contract requirements included an instrumentation program to monitor the duct banks during blasting. Seismographs were used to monitor PPV and strain gage arrays were used to monitor the strain in the duct bank. The layout of the instruments is presented in Figure 3. An array of 6 seismographs was used for vibration monitoring. Instantel BlastMate II and BlastMate III seismographs were used for monitoring, with the BlastMate III equipment set up opposite of the center of the blast (one on each duct bank). The BlastMate III seismographs were equipped with high frequency geophones, which could record vibration levels up to 250 cm/sec, with a frequency range of 28 Hz to 1 kHz. Strain gage arrays consisted of two sets of two fiber optic strain gages mounted on top of the duct banks. Each set of strain gages were aligned perpendicular and parallel with the tunnel alignment. 4 BLASTING To reassure PSE&G that blasting near the duct banks was not going to impact service, two small test blasts were detonated and vibrations were monitored. These two shots were placed in the tunnel invert near the center of the tunnel about 3.3 m from the duct banks. The first test shot consisted of a 1 m deep hole with a 0.5 kg charge (Austin Powder Co. Ex Gel 40%) in the hole and 0.6 m of 6 mm stone stemming. The shot was matted and initiated non-electrically. The second shot was a 0.9 kg charge in a 1 m deep hole, with 0.45 m of stemming. Both shots were monitored with the seismographs and strain gages mounted on the duct banks. The results were favorable and are discussed in the section below. Following the small test blasts, the next shots were for excavation of the elevator pit. Figure 4 shows the planned sequence of shots for the elevator pit blasting. Shots 1, 2, and 3 were performed as planned; however, shots 4 and 5 were charged and shot simultaneously.

131 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-16.qxd

2/18/04

11:58 AM

Page 132

Figure 5 presents the loading of the holes for the initial elevator pit shot. This shot included 15 loaded holes and a circular pattern of 6 burn holes to provide relief for the shot. The holes were approximately 2.4 m deep of which 0.6 m was subdrilling. As shown in Figure 5, the holes closest to the duct banks were loaded as buffer holes, with 1 kg of explosive. The production holes were loaded with 2.3 kg of explosives. All loaded holes were stemmed about 1 m with 6 mm stone. Each hole was separately delayed with delay periods of

5 3

1

2

4

Figure 4. Plan view of layout of elevator pit shots.

25 milliseconds to reduce the total kilograms per delay, which subsequently minimized vibration levels. The burden to spacing ratio was approximately 1.1. Due to the confined nature of the first shot, it was anticipated to be the most critical shot and would likely produce the highest vibration levels on the duct banks. To protect the duct banks from impact by flyrock generated by blasting, the contractor covered the shot with blast mats. Shot no. 2 in the elevator pit consisted of 23 loaded holes. The outer most holes were lightly loaded smoothwall holes intended to break the rock along the excavation limits and were delayed to fire last. Production holes were fired first in the sequence. The holes were separately delayed and sequenced in an alternating pattern. Delay periods were a minimum of 8 milliseconds apart to avoid a cumulative effect on vibrations. Shot no. 2 had the benefit of relief as it was delayed to move rock into the area of shot no. 1. Similarly, shot no. 3 had the same advantage, which significantly reduced vibration levels in the duct banks. Shot nos. 4 and 5 were combined into one shot. The distance from the duct bank to the loaded holes ranged from 0.83 to 1.19 m. The limits of the elevator pit were line drilled using 30 cm spacing. Shot nos. 4 and 5 also had relief to the center of the pit, which would reduce vibration levels. 5 RESULTS Actual peak particle and strain levels measured in the field were lower than the anticipated or predicted levels. Table 1 presents a summary of the blasting results. After the first shot at the elevator pit, the blast mats hit and damaged the strain gages. Later shots were not monitored with strain gages, as they could not be replaced in time due to the blasting schedule. It was decided that the next blasts were not as critical as the initial blast and that monitoring using only the seismographs was sufficient. The first shot at the elevator pit produced the highest vibration levels as expected, due to the confined nature of the shot. However, they were much lower

50

Figure 5.

Shot no. 1 – hole layout and delay sequence. Table 1. Summary of blast monitoring readings.

Shot no.

Distance to duct bank (m)

Kilograms of explosive

Scaled distance

Predicted max. PPV (cm/sec)

Measured PPV (cm/sec)

Measured strain (microstrain)

Test shot 1 Test shot 2 1 (Buffer hole) 1 (Production hole) 2 (Production hole) 4/5 (Buffer hole) 4/5 (Smoothwall hole)

3.35 3.35 2.01 2.77 2.01 1.25 0.83

0.5 0.9 1.0 2.3 2.3 1.0 0.7

4.7 3.5 2.0 1.8 1.3 1.2 1.0

38 60 147 171 284 312 449

3.4 2.8 25 25 6.4 5.2 5.2

1.15 0.98 147 147 NA NA NA

NA  Not available.

132 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-16.qxd

2/18/04

11:58 AM

Page 133

than anticipated. Subsequent shots rendered low vibration levels even when blasting within about 1 m from the duct banks. The favorable results are due to: 1. Relief provided by previous shots, 2. Using one delay per hole, 3. Line drilling along the perimeter of the excavation, and 4. The duct bank was not directly founded on the rock in the vicinity of the blasting. The predicted vibration levels were conservatively based on the high confinement equation by Oriard due to the sensitivity of the duct banks. However, blasts were less confined due to the layout and sequencing of the shots and the vibration levels are thus lower than predicted by this equation. Even with the standard equation, vibration levels typically vary by about an order of magnitude at a given scaled distance. 6 CONCLUSIONS The criterion for strain/vibration levels for typical residential type structures cannot be applied blindly in all circumstances. Monitoring devices that could get damaged during the course of the work should be provided with standby replacements on site. Reasonable yet prudent, substantiated strain and vibration limits can be used effectively to protect sensitive structures from close-in blasting. Careful planning and communication between all parties allowed the work to move forward with satisfactory results to all parties. NJ TRANSIT will get the station constructed, PSE&G’s duct banks were not impacted, and the contractor completed the work in a timely fashion.

ACKNOWLEDGEMENTS The authors would like to thank NJ TRANSIT for allowing the publication of this paper. In addition, the authors would like to thank Jose Morales and Michael Babin for their assistance with the preparation of this paper.

REFERENCES Dowding, C., 1996, Construction Vibrations, Prentice-Hall, Inc., Upper Saddle River, NJ. ISEE, 1998, Blaster’s Handbook, 17th Edition, International Society of Explosives Engineers, Cleveland, Ohio. Parsons Brinckerhoff Quade and Douglas, Inc., January 2001, Geotechnical Report, Design Unit N-30, Weehawken, Union City & North Bergen, NJ, Hudson-Bergen Light Rail Transit System. Parsons Brinckerhoff Quade and Douglas, Inc., January 2001, Geotechnical Design Summary Report, Design Unit N-30, Weehawken Tunnel and Bergenline Avenue Station, Hudson-Bergen Light Rail Transit System. Parsons Brinckerhoff Quade and Douglas, Inc., December 2001, Engineering Specifications (Conformed), Design Unit N-30, Weehawken Tunnel and Bergenline Avenue Station, Hudson-Bergen Light Rail Transit System. Revey, G.F. and Scott, G.A., 1999, Blasting Tunnel through Folsom Dam, ISEE Conference, Nashville, TN. Tart, R.J., Oriard, L.L. and Plump, J.H., 1980, Blast Damage Criteria for a Massive Concrete Structure, in Minimizing Detrimental Construction Vibrations, ASCE, NY. Westine, P.S., Esparza, E.D., and Wenzel, A.B., 1978, Analysis and Testing of Pipe Response to Buried Explosive Detonation, Report L51378, American Gas Association, Arlington, VA.

133 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-17.qxd

2/18/04

11:58 AM

Page 135

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Subway rehabilitation – secant wall cofferdams and penetration of tunnel liner Vincent Tirolo, Jr. & Norman Hirsch Slattery Skanska Inc

ABSTRACT: In order to improve passenger access a new escalator, elevator and reconstructed mezzanine were constructed at New York City Transit’s (NYCT) Lexington Avenue/Third Avenue IND Station. The station is located on East 53rd Street between Lexington Avenue and Third Avenues in Manhattan. The station is a hub that not only provides subway access to the east side of Manhattan but also is a transfer point to other subway routes. The subway platform is about 22 meters below the street grade. The platform was constructed between two mined tunnel bores. The tunnels and station were constructed in 1928. The existing station is in an area with a three tube concrete liner. During construction, the street was decked from building line to building line using a longitudinal steel stringer decking system supporting both the precast concrete roadway and sidewalks. After excavating to the new mezzanine level 6 m below street grade, two separate secant pile cofferdams were constructed. The secant piles were drilled to the top of rock and/or the top of the existing tunnel concrete liner. The secant wall was constructed using Bauer BG-18 and BG-22 drill rigs. The secant wall cofferdams provided a watertight excavation and their rigidity minimized ground movements during excavation. After the secant wall cofferdams were completed, the process of penetrating the tunnel arch began. First temporary struts and columns were erected and loads transferred by hydraulic jacks from the existing tunnel arch to these structural elements. This allowed the tunnel arch to be penetrated for the new escalator and elevator. Demolition of the arch was accomplished without blasting using pneumatic and hydraulic tools and splitters. Deformations were monitored during jacking and penetrations of the roof arches at both the escalator and elevator using a Bassett Convergence System. Maximum deflections were less than 7 mm and averaged about 4 mm. The new escalator and elevator work was performed with only minor disruptions to station operations.

1 INTRODUCTION The purpose of this project was to connect the two existing mezzanines and to provide additional elevator and escalator service to the existing platform. New York City Transit’s (NYCT) Lexington Avenue/Third Avenue IND Station is located on East 53rd Street between Lexington Avenue and Third Avenues in Manhattan. Construction was performed by a joint venture of Slattery Skanska Inc and Gottlieb Skanska Inc (the JV). The station consists of two separate mezzanine structures, one near Lexington Avenue and the other near Third Avenue. The existing mezzanines are not connected. They are shallow structures that were constructed by cut-and-cover construction methods. The roof is about 1.5 meters below street grade. The subway platform is about 22 meters below the street grade. The platform was constructed between two tunnel bores mined through rock and mixed face. The tunnels were extensions from the East 53rd Street Tunnel that was

mined under the East River east of the station. The tunnels and station were constructed in the late 1920s. The station is a hub that not only provides subway access to the eastside of Manhattan but also is a transfer point to other subway routes. 2 SITE CONDITIONS 2.1

The ground at the site can be divided into five distinct strata. The upper stratum is a man made miscellaneous Fill consisting of fine to coarse sand, gravel, cobbles, boulders, brick, concrete, cinders and silt. The Fill depth varies from 2 to 12 m. Underlying the Fill is a loose to very compact fine Sand. The thickness of Sand varies from one foot at the east end of the site to 9 m at the west end. Below the sand is a 1 to 4.5 m stratum of Glacial Till. The Till consists of a matrix of very compact fine to coarse sand enclosing coarser materials

135 Copyright © 2004 Taylor & Francis Group plc, London, UK

Geotechnical

09096-17.qxd

2/18/04

11:58 AM

Page 136

Figure 1. Existing station tunnel liner.

including rock fragments, cobbles and boulders. Underlying the Till is a discontinuous stratum of Decomposed Rock. Below the decomposed bedrock is Manhattan Schist and Granite Gneiss. RQD values ranged from zero to 82 percent. Some of the cores included cement grout that had penetrated the rock during tunnel construction. Groundwater is approximately 11 meters below existing street grade. The presence of boulders in both the Fill stratum and immediately above bedrock had a major impact on the progress of the work. Another major influence was the undulating top of rock surface. During construction steel sets, rock bolts timbers and concrete walls were uncovered. 2.2

Buildings

Major commercial and hospital structures are on the perimeter of the project. The major buildings surrounding the site include the 59 story headquarters of CitiGroup at the northwest side of the site; 599 Lexington Avenue, a 50 story building at the southwest side of the site, and the 11 story Memorial SloanKettering Cancer Center out-patient facility at the southeast end of the site. Other buildings on East 53rd Street vary in height from 8 to 11 stories. Fortunately for construction, all of these buildings are supported by deep foundations to rock. However, underpinning piers were installed at the 8 story Meyer’s Garage and the 11 story Memorial Sloan-Kettering Cancer Center to protect those buildings’ basement slabs. 2.3

Utilities

All underground projects constructed within the streets of New York City encounter a myriad of utilities. This project was no exception. The site includes major sewer and water lines, including their service connections; and electrical, power and telecommunications duct banks. The sewer and water lines were temporarily

relocated during construction. The new mezzanine roof precluded restoring the sewer lines within the roadway. Therefore the sewers were permanently relocated under the sidewalks. The other utilities were hung below the temporary roadway decking system. The supported utilities had to be moved during secant wall construction if they interfered with the drilling operations. 2.4

3 CONSTRUCTION APPROACH A Contractor’s primary concerns are safety and schedule. A safe project completed on schedule is generally a profitable project. Therefore, when we take an initial look at means and methods we focus on schedule and safety. An example of one of the schedule issues on the Lexington Avenue Station Rehabilitation Project, was the support of the roadway and sidewalk decking system. The suggested support system involved the installation of approximately twenty-seven 54.3 Mg pipe piles. Pile load testing would be required to confirm a minimum factor of safety of 2. The preparation of shop drawings and the fabrication of steel for a decking system are a time consuming process that could not await the results of a pile-testing program. If for any reason the pile tests indicated 108.6 Mg piles were not successful, it would be impossible to modify the decking installation without a serious delay in the project

136 Copyright © 2004 Taylor & Francis Group plc, London, UK

Existing subway structure

As shown in Figure 1, the existing subway structure is a triple barrel unreinforced concrete arch. The 4.2 m wide center arch contains two longitudinal built up steel girders supported by steel columns placed 4.6 m center to center. The arch is designed for full soil and rock overburden pressure. The soil and rock is considered fully drained.

09096-17.qxd

2/18/04

11:58 AM

Page 137

Figure 2. Project cross-section.

schedule. Also, pile penetrations of the new mezzanine roof would result in formwork and waterproofing problems. Finally, there was also the unknown impact of the piles on the rock block geometry above the tunnel roof. Another issue that impacts both safety and schedule was the uncertainties associated with the variable rock surface and rock quality. The secant wall must be seated on rock to provide a groundwater cut-off. In additional to analyzing the data contained in the MRCE’s geotechnical report, the JV took 10 additional borings to further define the rock surface. These borings indicated that the top of rock surface dropped off significantly from east to west. In fact the top of rock surface dropped below the existing tunnel roof at the west end of the escalator secant wall SOE. Based on our previous experience, investing our own money to obtain additional geotechnical information almost always reduces risk and improves the performance of our temporary structures. The general layout of the revised system is shown in Figure 2. Blasting was not permitted on the project. Rock was split with jackhammers and hydraulic rock splitters. 4 DECKING SYSTEM Construction of the new mezzanine structure and the relocation of utilities required that the decking system

extend for the full width of the roadway (three 3.3 m wide lanes) and both 4 m sidewalks (building line to building line). Since the roadway must remain open during construction, the decking is installed in stages. The decking system should also minimize requirements for utility relocation. For these reasons and to minimize the potential adverse impacts on the existing tunnel and mezzanine construction discussed in the previous section, the JV elected to install a raised decking system, utilizing a longitudinal beam and header system, which would be supported on shallow spread footings. Precast decking panels are installed between the flanges of the longitudinal decking beams. Top of the W760 decking beams were set at the crown of the existing roadway. The longitudinal decking beams, coped at their ends, were supported by six transverse W360 header beams. Thus the total depth of the decking system is limited to about 760 mm. The shallower the penetration of the decking system into the existing roadway, the less the possibility of intersecting existing utilities. Supporting utilities in-place is always preferred to utility relocation which is a time consuming “schedule breaking” activity. To maintain the project schedule, an individual box sheeted pit was constructed for each footing. The footings were then poured prior to the arrival on site of the decking steel. Prior to placing the footing concrete, the subgrade was inspected to verify its New York City Building Code (NYCBC)

137 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-17.qxd

2/18/04

11:58 AM

Page 138

classification. In general the soils at subgrade were classified as compact fine to coarse sands, NYCBC 7–65 or 8–65 soils. After the footings were poured, the decking columns were installed within the pits. When designing a decking system it is important to consider construction access. The header beams were then installed at night during a complete roadway shutdown. There should be enough room between the decking support columns for equipment to operate efficiently. The JV developed a monitoring system to check the performance of the decking and footings during construction. A jacking system was also designed and installed to allow the header beams to be jacked up if footing settlements exceeded 25 mm. During the course of the project, the jacking system was implemented on two occasions. On both these occasions footing settlement was related to loss of ground that occurred during the drilling of an adjacent secant pile. 5 SUPPORT OF EXCAVATION – SECANT WALL Two secant pile cofferdams were required for the Lexington Avenue Station Rehabilitation Project. The secant wall cofferdam for the new elevator was 5.6 m square and consisted of 32 secant piles. The new escalator cofferdam, was rectangular, 28.8 m by 7.4 m, and consisted of 100 secant piles. The secant walls are a combination of all cast-in-place concrete primary piles, which are installed first, and secondary piles that are drilled between primary piles. The secondary piles were first drilled to the top of rock and a steel wide flange “core” beam was then placed into the pile before concreting. The temporary casing was withdrawn in 1.5 m sections as the concrete was placed. A total of sixty-six core beams were installed. The overlap between two adjacent drilled shafts is 152 mm. The secant piles were installed from the level of the new mezzanine structure, about 6 m below the top of the decking system, to the top of rock. 100 mm OD steel pipes were attached to the core beams to permit postgrouting of the secant pile tip/top of rock interface. The core beams in the secant walls were W310X158 A572M Grade 345 MPa members. The SOE system consists of the secant walls and three levels of wales and struts. The JV designed the details of the secant wall. Underpinning and Foundation Inc. (U&F) constructed the secant wall under a subcontract to the JV. The secant wall was constructed using crawler mounted hydraulic rotary Bauer BG-22 and BG-18 drill rigs. The mast of the Bauer BG-18 drill rig was modified for low headroom installations. The Bauer drill rig is equipped with a casing drive adapter (turntable) which rotates and advances a watertight 750 mm diameter double wall casing. The casing

Figure 3. BG-22 drill rig on roadway decking.

is advanced by attaching new 1500 mm long casing segments. Conical bolts located around the circumference of the casing connect the segments. The lead casing is equipped with a cutting shoe. Hardened metal inserts attached to the cutting shoes allow for rotary drilling through soils, soft rock and secant pile concrete. Soil inside the casing was removed by conventional earth drilling tools. In dry ground, soil was removed with an auger. In loose wet soil, a bucket auger was used. This auger has a lockable revolving bottom gate. Rock augers were used whenever boulders or the top of decomposed rock were encountered. These excavation tools are designed to work within the 750 mm double wall casing. The installation sequence for the secant wall cofferdams on Lexington Avenue Station Rehabilitation Project included:

•

•

•

138 Copyright © 2004 Taylor & Francis Group plc, London, UK

Installation of decking. The decking system was designed to not only support normal roadway traffic but also the Bauer BG-22 rig that would be used for drilling secant piles from the decking level (see Figure 3). Where the location of a secant pile interfered with the decking steel, the Bauer BG-18 rig installed these secant piles from beneath the decking. Since the BG-22 is a more powerful machine than the BG-18, we attempted to minimize the number of secant pile installations from below the decking. It was also necessary to design the decking system so that the BG-18 could be lowered down through the decking. Excavate down to the new mezzanine level. Layout controls, benchmarks, and install guide wall template. Proper installation of the guide template is critical to the success of the wall. It is also an expensive item because of the precise formwork required. Multiple use of this formwork is important to reduce costs. Remove or relocate all overhead obstructions interfering with normal movement of BG-18. These “obstructions” often were utilities that had been previously hung from the decking system. These

09096-17.qxd

2/18/04

11:58 AM

Page 139

Figure 4. Actual vs. theoretical rock line.

• • • • •

often had to be moved again to accommodate movements of the BG-18 and sometimes the BG-22 drill rig. Install the primary piles first. As the casing is advanced excavate within the casing using an auger or bucket, as soil conditions require. Confirm casing is down to competent rock before placing concrete. Concrete was placed by the tremie method or by chutes. Install the secondary piles; install core beams and place concrete Grout rock secant wall interface as required.

The average progress was one secant pile installed per drill rig per 8 to 10 hours shift. Two major impediments to construction were the presence of obstructions, e.g. timbers, concrete wall, boulders, and the irregular rock surface. Figure 4 illustrates the top of rock profile based on the boring shown in the Geotechnical Report and the additional borings taken by the JV versus the actual top of rock from the secant wall as-built data. The two primary purposes of the secant wall structure were first to act as a “bath tub” structure to allow excavation within its perimeter without external dewatering and second to provide a “rigid” cofferdam to limit settlement of adjacent structures. The water tightness of the wall is primarily a function of the secant wall construction. The functioning of the secant wall as a SOE is primarily dependent on the details of design. A secant wall design uses the same methodology as the design of a soldier pile and lagging wall. The

spacing between the core wide flange beams in the secondary secant piles is approximately 1219 mm. The thickness of the secant wall at its narrowest dimension is about 381 mm. In our design we did not assume any composite action between the core beams and the concrete. The concrete can either be designed as a concrete lagging, e.g. simple span for bending and shear, or as an inscribed arch. The inscribed arch is similar to the NYCT jack arch concept. In either case no reinforcement was necessary. The secant wall is then designed for both down stage and up stage conditions. For the Lexington Avenue Station Rehabilitation Project, we used three levels of wales and struts. The upper level was located about 1.8 m from the top of the secant wall. The span between the upper and second level of struts was 3 m. The third level was 1.8 m below the second. For the final down stage condition, we assumed the core beams were partially socketed into rock. The maximum depth of excavation below the mezzanine level was 10 m. The total deflection of the wall was minimal.

6 INSTRUMENTATION AND MONITORING After the secant walls were completed the next stage in the work was the penetration of the existing tunnel for the new elevator and the new escalator. Prior to this work, monitoring instrumentation was placed at the arch penetration locations. The instrumentation consisted of a Bassett Convergence System (BCS) to monitoring deformation of the concrete tunnel and

139 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-17.qxd

2/18/04

11:58 AM

Page 140

Figure 7. Uncovered steel sets at tunnel roof penetration.

Figure 5. Typical array deformation.

through rock was by drill and blast. The heading was normally advanced using heading and bench methods. A typical round was 2.5 to 3 m long. Progress was typically 3 to 4.5 m for three 8-hour shifts. Mined tunnels entirely in rock were normally lined with unreinforced concrete. Occasionally, in subaqueous sections mined in poor rock, the permanent tunnel liner was a combination of a primary cast iron liner and a secondary concrete liner. The permanent tunnel liner at the Lexington-Third Avenue Station is a triple barrel unreinforced arch with longitudinal built-up girders along the platform. 7.2 Figure 6. Array trend plot.

strain gages to monitor loads on the temporary struts during penetration of the middle tunnel arch. The BCS reference pins were arranged in six arrays, three at the elevator and three at the escalator. Figures 5 and 6 respectively illustrate typical array deformation data and trend charts. The maximum displacement at the springline was 7 mm and 6 mm in the crown. Average deformations were considerably lower (3–4 mm). It was difficult to protect the reference pin arrays during construction. In general jacking loads into the temporary columns and struts caused the major movements. After jacking and lock-off, movements during demolition of the arch were negligible. 7 EXISTING STATION AND TUNNEL 7.1

Tunnel construction in the 1920s

The original Lexington-Third Avenue Station was constructed as part of New York City Transit’s Route 104 Section 2 in the late 1920s. The contractor was Patrick McGovern Inc. Mined tunneling in the 1920s

As shown in Figure 4, the top of rock at the east end of the secant wall cofferdam is high while at the west end the top of rock drops below the crown of the tunnel. Therefore the easterly portion of the LexingtonThird Avenue Station was constructed as a rock tunnel, the middle portion in mixed face and the west end of the station was constructed as an open cut. The open cut was probably also used as a construction shaft for the tunnel. Figure 7 shows the temporary steel sets and timbers uncovered when the top of the subway tunnel was exposed. No steel sets were encountered at the tunnel penetration for the elevator (east end of secant wall cofferdam). Temporary 360 mm deep steel “cap” beams, typically less than 1 m center to center, were supported on the permanent built up girders, and were exposed along the entire length of the escalator penetration (west end of secant wall cofferdam). We have concluded that the mixed face portion of the station was constructed using multiple drift. We hypothesize that two drifts were excavated over EB and WB track sections using steel caps and posts. The drifts were advanced by forepoling. The invert and the longitudinal built up girders were erected in these drifts. The roof load was transferred from the interior steel posts to the build up girders and the posts

140 Copyright © 2004 Taylor & Francis Group plc, London, UK

Mixed face tunneling

09096-17.qxd

2/18/04

11:58 AM

Page 141

removed. The center drift was then constructed and the platform constructed. This methodology is similar to the illustrations in Szechy (1973) for station construction in Europe. West of the secant cofferdam the rock surface rises rapidly and the contractor was able to resume free air rock tunnel. Therefore, it is likely that the open cut section of the station was used as a work shaft. Figure 1 is a cross-section of the station’s three arch tunnel liner. 8 PENETRATION OF TUNNEL ROOF The major concern in penetrating the roof arch was the redirecting of the theoretical thrust in the roof arch around the opening. The system used to accomplish this was the sequential placement of temporary struts and strut jacking. The procedure involved first saw cutting and chipping pockets on either side of the middle (platform) arch. These pockets were approximately 1 m center to center. Immediately after a pocket was cut a strut was installed as shown in Figure 8. The struts were W310X179 members. After these initial struts were installed, the concrete between the pockets was removed and the remaining struts installed (see Figure 9). Two of these “temporary struts” became permanent. At the escalator penetration, 13 temporary struts were initially installed. Five struts became permanent. Because of the longer length of the escalator penetration compared to the elevator (10.5 m vs. 3.5 m) a secondary level of 127 mm diameter 9.5 mm wall pipe struts spaced 2.4 mm center to center were installed above the primary wide flange struts. There was also a concern that the longitudinal platform girders would carry additional vertical load when the platform arch was removed. These girders were supported by columns 4.572 m center to center. However during construction temporary columns were installed 1.524 m between the existing columns. The loads from the original arch and columns were transferred to the temporary columns and struts by jacking. The maximum vertical load in the temporary columns was limited to 900 kN and 670 kN in the temporary struts. The jack force in the pipe struts was limited to 290 kN. Figure 10 shows the penetration of a portion of the tunnel arch with all temporary struts installed. Fabricated double channels were used to permanently support the opening after the temporary struts were removed. These channels were 1 m deep with 100 mm flange plates and a 150 mm web plate. At the escalator they were over 8 m long. Placement of these members was difficult with an operating station. The channels were either lowered through the decking system or carried into the site with work trains. Figure 11 shown typical channel sections prior to lowering them through the decking.

Figure 8. Arch pockets and temporary columns.

Figure 9. Sockets expanded for intermediate struts.

Figure 10. Arch penetration.

9 CONCLUSIONS The escalator and elevator shafts successfully penetrated the roof without incident and with minimal

141 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-17.qxd

2/18/04

11:58 AM

Page 142

A raised roadway decking coupled with longitudinal decking stringers minimized interferences with existing utilities. ACKNOWLEDGEMENTS

Figure 11. Channel sections.

disruption to the public. Monitoring indicated small movements (generally less than 4 mm). The use of temporary struts and columns proved to be an efficient means of maintaining the existing middle concrete arch. The secant wall was successfully placed, within an urban environment, under roadway decking. The bottom of the secant wall was exposed during penetration of the tunnel liner to install the elevator shaft and half of the escalator. Except for two locations, all secant piles were founded on competent rock (RQD 50%). At the two locations not founded on competent rock, the secant piles were located on large boulders immediately above bedrock. At these locations, chemical grouting was used to seal the area.

We would like to acknowledge the assistance of those organizations without whose help this work could not have been accomplished. New York City Transit first, for giving us the opportunity to work on this project and later in assisting us in all our efforts. Next we would like to acknowledge Daniel Frankfurt, P.C. the prime consultant on the project and their geotechnical subconsultant, Mueser Rutledge Consulting Engineers, and Dr. Carl Costantino, the NYCT’s Geotechnical Consultant, for their fairness and open-mindedness. We wish to acknowledge our secant pile subcontractor Underpinning and Foundation Inc. Finally we would like to acknowledge the members of the JV team, both in the field and the engineering department, that have made this project both an engineering and financial success. REFERENCES Goodman, R. & Shi, G-H. 1985. Block Theory and its Application to Rock Engineering, Prentice-Hall. Mueser Rutledge Consulting Engineers, 1999. Geotechnical Report, Lexington Avenue Station Rehabilitation, New York. Szechy, Karoly. 1973. The Art of Tunnelling, Akademiai, Budapest.

142 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-18.qxd

2/18/04

11:59 AM

Page 143

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Overcoming the complex geotechnical challenges of urban construction T.J. Tuozzolo Moretrench Geotec, a Division of Moretrench American Corporation, Rockaway, New Jersey, USA

ABSTRACT: New construction within the confines of heavily developed urban settings can present designers and geotechnical contractors with complex challenges, particularly when the construction is to be developed on a constricted site. Such was the case for a Dental School addition to the University of Medicine and Dentistry in Newark, New Jersey. Earth retention and underpinning design for a site bounded by busy main thoroughfares and adjacent buildings was further complicated by the presence of adjacent underground utilities, low-headroom conditions and unanticipated soil conditions. The geotechnical contractor’s ultimate design/build solution, which required a proactive, on-site response, involved four distinct techniques – soil nailing, minipiles, conventional pit underpinning, and a drilled-in-place pipe soldier pile cofferdam – all implemented under a single contract. Extensive monitoring and testing was conducted to verify design assumptions and to provide data for future research.

1 INTRODUCTION When new construction is planned in already heavily developed urban areas, several critical factors need to be considered prior to commencement. One of the most critical factors is how the excavation for the proposed structure will be accomplished. Since most urban environments preclude open excavations, much thought must be given to the earth support system required to retain the subsurface materials during excavation for the new foundation. If the excavation will be made next to adjacent structures, considerations for underpinning these structures must also be addressed, as must the impact active utilities will have during excavation, and the limitation of the building and property boundaries. Such complexities are inherent in many of today’s urban construction projects. When project requirements, site conditions and restrictions render a conventional approach to earth retention and underpinning impractical, if not impossible, specialty geotechnical techniques can offer a viable and economical alternative. 2 CASE STUDY Expansion of the Dental School at the University of Medicine & Dentistry of New Jersey (UMDNJ) in Newark, NJ, included construction of a five-story, steel structure with a one-story, below-grade basement.

The site was bounded on one side by the existing Dental School and on the opposite side by 12th Avenue, a major access road for emergency vehicles en route to the adjacent University Hospital. Parallel, underground, electric and telephone duct banks, each 1.5 m by 0.6 m, ran along 12th Avenue. The telephone bank lay nearest the site, 1.5 m below existing grade and within millimeters of the proposed building. The proposed structure, approximately 85 m by 37 m in plan, entailed excavations to a depth of 8.2 m below grade. In order to facilitate these excavations, a temporary earth retention system was required for approximately 81 linear meters along 12th Avenue and for 38 linear meters along the west side of the site. Also, a permanent retention and underpinning system was required for 18.3 linear meters on the south side of the project along the existing 3-story Dental School where the new building would be constructed on-line and tied in. Preliminary excavation retention system concepts included driven steel sheeting and soldier beams and lagging along 12th Avenue and the west side, as well as a permanent, on-line, soil nail wall along the existing Dental School. However, the proximity of the duct banks along 12th Avenue precluded the sheeting and soldier beam options, since both systems would encroach into the new building and would require the structure to be moved and redesigned. In addition, the soil borings encountered rock at or just above the proposed subgrade elevation, which could impede the

143 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-18.qxd

2/18/04

11:59 AM

Page 144

driving of soldier piles or steel sheets and likely necessitate a more expensive, drilled-in soldier pile system. 2.1

Engineered solution

The pre-construction geotechnical investigation indicated that the subsurface conditions consisted of 1.2 m to 1.8 m of fill, underlain by natural silts and silty sands containing varying amounts of gravel, cobbles and boulders. Weathered sandstone bedrock was encountered at or above the subgrade elevation of the proposed structure. After evaluating the soil conditions, space limitations and the schedule, the geotechnical contractor offered a system that entailed a temporary soil nail wall along 12th Avenue and the west side of the site, and a combination of drilled-in minipiles and soil nailing along the existing Dental School. The minipiles would be drilled through the existing column footings to underpin the building and transfer the loads below the proposed subgrade. Soil nailing would be used to support and retain the soil between the existing columns. 2.2

Figure 1. Cross-section showing soil nail wall installation.

12th Avenue and west side

The soil nail wall along 12th Avenue and on the west side of the site entailed three to five levels of soil nails installed at 15 degrees from the horizontal on a 1.5 m grid pattern. A 75 mm thick layer of shotcrete, reinforced with wire mesh, was used to tie the system together and retain the soil between the nails. Construction began with the excavating contractor making an initial, sloped precut to locate the top of the telephone duct bank and remove the existing fill soils which did not exhibit good ‘standup’ time. The precut also allowed the first level of soil nails to be installed at a 15-degree angle below the duct bank, and did not require any redesign or relocation of the proposed structure. Once the first level was installed, and the soil nails and shotcrete had cured, the process was repeated and the remaining lifts were installed until subgrade was reached (Fig. 1). A view of the soil nail wall along 12th Avenue is presented in Figure 2. The wall was designed to withstand the temporary lateral and traffic surcharge loads. The system was also designed to withstand the surcharge loads associated with a large crane scheduled to be placed on the sidewalk directly behind the soil nail wall to aid in construction of the new addition. Since the soil nail wall for this part of the project was installed on-line, the new foundation wall was constructed using a onesided form and poured directly against the soil nail wall. A drainage composite and waterproofing membrane were placed between the soil nail wall and the new concrete wall. This 935 m2 system allowed the proposed foundation to be built ahead of schedule and without the need for any design modifications.

Figure 2. View of soil nail wall along 12th Avenue.

2.3

144 Copyright © 2004 Taylor & Francis Group plc, London, UK

Existing dental school

In order to facilitate the 8.2 m deep excavation along the existing Dental School, a combination of minipile underpinning and soil nail earth retention was designed and offered by the geotechnical contractor. The minipiles were installed through the existing column footings to underpin the footings and transfer the column loads below the newly proposed subgrade and thus eliminate any surcharge load induced on the new soil nail wall. The next step was to build the soil nail wall in lifts, finishing at the new building subgrade elevation, 8.2 m below grade. During construction of the permanent soil nail wall, unanticipated conditions were encountered during excavation of the first 1.5 m lift. A loose, cohesionless fill with poor standup time was encountered directly below the existing first-floor slab. The soil was so loose and uncompacted that voids in excess of 25.4 mm were present directly below the existing slab. To assess the actual conditions, the geotechnical

09096-18.qxd

2/18/04

11:59 AM

Page 145

Figure 3. New concrete underpinning pier tied into existing pedestal.

contractor installed a trial section of shotcrete and test nails to determine if soil nailing would still be viable. The shotcrete did not adhere to the loose fill, causing sloughing of the material and “cave-in” of the test section. Additionally, the test nails did not achieve the bond values that the design assumed, since the shear strength of the soil proved to be very low. Taking all of these considerations into account, the geotechnical contractor recommended the soil nailing system should not be installed, and offered a redesigned system. 2.3.1 Redesign at existing dental school Since the soil nail system could not be used as the permanent, earth-retention system, other methods were examined. The new system would need to:

• • •

be able to retain loose soils up to 8.2 m below the existing floor slab, be installed on-line so the new structure could abut and tie into it, and finally, be able to eliminate any lateral loads that would be induced on the existing concrete pedestals during excavation.

After examining many different options, from driven soldier piles and lagging to sheet piling and minipiles, the geotechnical contractor concluded that the only practical system was to construct concrete underpinning piers by the pit method, with treated timber lagging installed to retain the loose soil between the piers. As excavation progressed, two levels of tieback anchors would be installed to resist the permanent lateral loads. Although the minimum depth for the new basement was at 4.8 m below grade, the depth of excavation

had been designed to extend to 8.2 m below grade in order to bear the new foundation on the sandstone bedrock. In conjunction with the redesign for the earth retention system, the geotechnical contractor further proposed using a minipile foundation to support the new column loads rather than the originally proposed spread footing foundations. The minipiles would be installed from an elevation 4.8 m below grade and terminated approximately 6 m into the rock. The minipiles would be tied into pile caps at the 4.8 m elevation, thus eliminating the need to excavate to 8.2 m and significantly reducing the depth of excavation and the associated costs to the owner. 2.3.2 Production work The 0.9 m by 1.5 m underpinning pits were installed on 1.8 m to 2.4 m centers, spacings that directly related to the spacing of the existing concrete pedestals. Once the underpinning pits were formed, reinforcing dowels were installed to tie the existing pedestals to the new underpinning pier and eliminate any lateral load on the pedestals (Fig. 3). The pits were then filled with 20 MPa concrete, forming the new underpinning piers. Following completion of the piers, excavation proceeded in 1.5 m lifts as the treated timber lagging was placed. At approximately 1.5 m below grade, the first level of permanent tieback anchors was installed through the new piers to resist the lateral loads induced on both the piers themselves and the existing concrete pedestals into which they were tied. After the second-tier tiebacks were installed, tested and locked off, excavation and lagging installation continued to the new subgrade elevation, 4.8 m below the existing slab (Fig. 4).

145 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-18.qxd

2/18/04

11:59 AM

Page 146

Figure 4. Redesigned earth support and underpinning.

3 INSTRUMENTATION AND MONITORING Figure 5. Pullout test detail.

In order to evaluate design assumptions and monitor the construction, a comprehensive testing program was implemented, consisting of field pullout tests, proof testing, surveying, and strain monitoring. 3.1

Verification testing

At least one pullout test was performed on each soil nail row (lift) to verify that adhesion values used in the design program were obtained in the field. The pullout test was performed incrementally, generally to a factor of at least two times the design value on a shorter sacrificial nail (Fig. 5). In addition, a minimum of five percent of the production nails were proof tested to 1.33 times the design load to measure the bond and creep values. 3.2

Figure 6. Strain gage instrumentation detail.

Survey and monitoring

Survey monitoring was implemented to measure horizontal and vertical displacements. The deflection of the wall was generally 0.2 percent of the cut height, which is within the range of the anticipated deformation for wall construction in these soils (Byrne et al., 1996). 3.3

Strain monitoring

In addition to the normal monitoring, a strainmonitoring program was developed and implemented for the temporary soil nail wall to compare the in situ parameters of a multi-tier soil nail earth retention system with that of the original design parameters, information that would be valuable in the future. Two pairs of vibrating, wire-strain gages were installed on each of three, predetermined soil nails in a vertical plane in order to determine the average strain (Fig. 6). The nails selected for instrumentation were located on tiers 2 through 4 of the 5-tier system.

Strain readings were taken on a weekly basis for a period of two months from initial installation. Readings were taken more frequently immediately following nail installation. Readings were also taken prior to and after each new 1.5 m lift. Long-term monitoring was not feasible. However, data trends shown in Figure 7 indicate that in situ field stresses were similar to or less than the theoretical stresses calculated using the CALTRANS Design Program, SNAIL (Caltrans, 1991). As shown, the nail stress increases as the next cut is made to a new lift. Results in the upper nail are closest to the theoretical design stresses, more than likely because it was monitored longer than the lower nails. Of interest also is how the nail stress spiked in the strain gages installed 1.5 m behind the wall immediately after the crane used to set the steel structure was brought on site and erected approximately 2.5 to 3.5 m behind the soil

146 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-18.qxd

2/18/04

11:59 AM

Page 147

Figure 7. Composite in situ stress plot. Table 1. Summary of strain gage stresses.

Strain Gage Number

Bar Stress (MPa)

Average Stress (MPa)

B47-15-U B47-15-L B47-15-U B47-5-L C47-15-U C47-15-L C47-5-U C47-5-L D24-15-U D24-15-L D24-5-U D24-5-L

112.00 131.00 112.00 119.00 55.00 50.00 N/A 97.00 25.00 N/A 48.00 45.00

122.00 122.00 115.00 115.00 52.00 52.00 97 97 25 25 46.00 46.00

Theoretical SNAILZwin Stress Output (MPa) 167 167 130 130 139 139 118 118 109 109 97 97

chilled water system to provide cooling was to be installed. To facilitate this installation, an excavation with a plan dimension of 4 m by 4.9 m was required to a depth of approximately 10 m below existing grade. The difficulties of performing such a confined and deep excavation were compounded by the fact that the work had to be accomplished directly below an existing corridor, limiting the overhead clearance to 3.7 meters. Due to space limitations, the excavation could not be open cut; therefore, a method of earth retention was required. 4.1

nail wall. As previously mentioned, the geotechnical contractor’s design took this loading into account. A summary of the stresses is presented in Table 1. 4 NEW CHILLER WATER AREAWAY During the completion of the earth retention and underpinning work, the owner and construction manager were faced with yet another challenge. In order to service the new dental school addition, a new,

147 Copyright © 2004 Taylor & Francis Group plc, London, UK

Earth retention design and installation

Given this complicated condition, the geotechnical contractor had to provide a method of earth retention that could support the soil during excavation and construction of the confined areaway, and also be installed within the limited-access and overhead-clearance constraints. Working closely with the excavating contractor, the geotechnical contractor developed a plan and a design for the earth retention that entailed the installation of low-headroom, small-diameter minipiles to act as soldier piles, timber lagging, and an internal waler and bracing system. Construction began with the installation of 245 mm diameter minipiles to a depth of approximately 12 m below existing grade. The minipiles were drilled below the proposed subgrade of the chiller areaway and socketed into rock. Since the headroom was limited to

09096-18.qxd

2/18/04

11:59 AM

Page 148

Figure 8. Chiller water areaway earth retention system.

3.7 meters, the minipiles had to be installed in onemeter lengths. Once all the minipiles were installed and filled with grout, excavation began. Since the excavation was performed in such a confined location, most of the soil was removed by hand and lifted out of the areaway by conveyors. As work proceeded, an internal, three-level, ring waler system was installed around the perimeter of the earth retention system and abutted into the existing building. Large plates were bolted into the existing building wall in order to accept the ring walers. Timber lagging was bolted to the face of the drilled minipiles to retain the soil. The completed earth retention system is shown in Figure 8. This innovative approach not only allowed the system to be built, but work was performed within budget and in an accelerated time. 5 CONCLUSIONS New construction within urban settings can present complex challenges. Site conditions that could

potentially impact the execution of the excavation support and underpinning work should therefore be fully evaluated prior to the design stage. Conventional techniques such as soldier beams and lagging or sheetpiling are typical options. However, on sites where these options are not viable, specialty geotechnical techniques offer effective alternatives. Geotechnical contractors can play a vital role in the planning stage of complex, urban construction projects by offering owners, designers and general contractors experience-based input into overcoming design and production challenges.

REFERENCES Byrne et al. 1996. Manual for Design and Construction Monitoring of Soil Nail Walls. FHWA-SA-96-069, Federal Highway Administration, Washington, DC. CALTRANS, 1991. A User’s Manual for the SNAIL Program, Version 2.02. California DOT, Division of Technology, Material & Research.

148 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-19.qxd

2/18/04

12:00 PM

Page 149

Session 2 Subsurface investigations and geotechnical report preparation for design/build projects

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-19.qxd

2/18/04

12:00 PM

Page 151

Session 2, Track 1 Risk allocation

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-19.qxd

2/18/04

12:00 PM

Page 153

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Risk management in tunneling – occupational safety  health plans for drill and blast and tunnel boring machines A. Moergeli moergeli  moergeli consulting engineering, CH Schmerikon, Switzerland

ABSTRACT: Control your critical success factors by applying generally accepted best risk management practices. Establish, implement, update and document a thorough and comprehensive Risk Management System where required by law, project challenges, owner’s specifications and/or where best risk management practices are not available.

Examples for requirements:

1 WHAT CAN A RISK MANAGEMENT SYSTEM DO FOR YOU?

•

A Risk Management System (RMS) must increase your productivity – from your customer’s perspective as well – or it is useless … A scientific risk management is a relatively new approach:

• • • •

Many managers (of big companies) are talking about it these days. Most people remain pretty confused when it comes to day-to-day application. Very few people really have personal experience. There is an opportunity in combining ongoing US underground activities with previous project risk management experience, providing owners, engineers and contractors with real added value in successfully pursuing their goals.

2 WHEN SHOULD YOU USE A RISK MANAGEMENT SYSTEM?

3 HOW DOES RISK MANAGEMENT WORK? Control your critical success factors by: 1. Apply generally accepted best risk management practices. 2. Establish a thorough and comprehensive Risk Management System (RMS) where required by law, project challenges, owner’s specifications and/ or the absence of best risk management practices.

4 APPLY GENERALLY ACCEPTED BEST RISK MANAGEMENT PRACTICES

Where required

• • • •

•

by law by project challenges by owner’s specifications and/or by the absence of best risk management practices.

Control your critical success factors by establishing a pro-active Quality Assurance/Quality Management System by applying generally accepted best risk management practices.

You may choose to control your critical success factors by establishing a thorough and comprehensive RMS.

153 Copyright © 2004 Taylor & Francis Group plc, London, UK

US Military Standard 882D requires a RMS of every party within the acquisition process. Swiss Law requires all companies operating in a high risk environment (e.g. contractors in the construction industry) to perform a risk assessment.

1. Use an interdisciplinary team. Tunneling large, sophisticated and very complex projects, in partly still unknown geology and on the background of a highly sensitive environment, has to be well

09096-19.qxd

2/18/04

12:00 PM

Page 154

thought through. A well built-up interdisciplinary team is required to deliver best results. Include the owner, the designer, the construction management, all special sub-consultants etc. as well as representatives of operations and maintenance. Lead the team by an experienced Risk Manager with a substantial background in quality management and safety  health. 2. Leave the ground risk to the owner. 3. Perform a very thorough, very detailed site investigation. Be prepared to invest anything from 1% to a maximum of about 10% (as a general rule) of your Total Cost of Ownership (TCO). Be prepared to core drill at least about 0.6 times the length of your tunnel when trying to avoid dealing with successful contractor’s claims (Waggoner, Daugherty, 1985). 4. Work out a professional Geotechnical Baseline Report (GBR) at least before requesting contractor’s bids. Hand out your GBR to all bidders (Essex, 1997). 5. Plan for contingencies. 6. Establish, run and maintain (before, during and after construction) an adequate control system. Enable real-time on-site interpretation and decision-making. 7. Use a partnering approach. 8. Require Escrow Bid Documents (EBD). 9. Prepare for Alternative Dispute Resolution (ADR). Avoid settling claims before courts as this often proves to be a very expensive process. 10. Emphasize a complete documentation. Keep all available records updated and carefully stored after the project’s completion. 11. Have the contractor to enroll a project-specific Quality Management System (PQM) on site. Fully integrate Safety  Health (OSH) and environmental issues into your PQM. 12. Whenever feasible and practical, test and validate your control of all critical success factors on a small-scale first (e.g. by a pilot tunnel, a test shaft, etc).

5 HOW DOES A RISK MANAGEMENT SYSTEM WORK? A four-folded approach is recommended by the author (m  m/am): 1. Start with a Preliminary Hazard Analysis (PHA). 2. For hazards with a high potential of harm and where generally accepted best risk management practices are not available (risks in Risk-Zone 1), apply a full and thorough Risk Assessment (RA). 3. Use a Fault Tree Analysis (FTA) to verify your preventive action as an option.

Figure 1. Example of a preliminary hazard portfolio.

1 Start Risk Assessment 2 Define system’s limits

3 Identify hazards

4 Estimate risks

5 Assess risks

6 Tolerable residual risks?

7 Documentation

8 Sign list of residual risks

9 Periodic update

No

10 Risk Management

Figure 2. Example of a risk assessment procedure.

4. Establish, implement, deploy, maintain and document System Safety (SS) for risk mitigation. Obviously the recommended risk management procedures do not only apply to the construction phases of a project but – sometimes even more important, at least to the owner – to the overall system life expectancy as well. None of the enlisted procedures are new; some have even more than a fifty-year’s track record. What is really new is the

• • •

154 Copyright © 2004 Taylor & Francis Group plc, London, UK

Yes

combination sequence and extrapolation from safety  health and machinery safety systems to build up a comprehensive RMS.

09096-19.qxd

2/18/04

12:00 PM

Page 155

7 HOW DOES A RISK ASSESSMENT WORK? A risk assessment matrix, adopted to meet your requirements, let you decide which risks you can (or have to) accept. US Military Standard 882D can deliver a starting point.

8 HOW DO YOU MITIGATE YOUR RISKS? 1 Start Risk Assessment 2 Define system’s limits

3 Identify hazards

4 Estimate risks

Yes

5 Assess risk

6 Tolerable residual risks?

No

Yes

7 Documentation

8 Sign list of residual risks

12 New hazards?

9 Periodic update

No

10 Substitution?

Yes

Figure 3. Example of a risk matrix.

11 Susbstitution of hazardous procedures/ materials etc

No

6 HOW DOES A PRELIMINARY HAZARD ANALYSIS WORK?

13 Safe system/ new strategy ?

A Preliminary Hazard Analysis acts as a risk-focused filter to allocate your restricted resources (to perform thorough and scientific risk assessments) to the most urgent and important issues. It regularly follows a three step procedure:

Mitigation iteration

Yes

14 Change of strategy, Safety System (S)

No

No

16 Technical measures?

15 Yes Risk mitigation ok?

Yes

17 Technical measures (T)

18 Yes Risk mitigation ok? No

No

a) Set up a comprehensive hazard inventory (a preliminary, qualitative risk identification) b) Provide evidence of available best risk management practices c) Fill in your identified risks in the appropriate Risk-Zone of your hazard portfolio.

Yes

19 New system’s limits?

No

20 Measures to change human behavior patterns (O/P)

21 Yes Risk mitigation ok? No

Figure 4. Example of a risk management procedure.

You will end up with a fast and easy to understand graphic overview (a risk map). You may then concentrate your efforts to perform a thorough risk assessment of all risks remaining in Risk-Zone 1. But before taking off into that timely and costly task make sure that your interdisciplinary team really confirms your hazard portfolio’s allocations.

9 HOW DO YOU APPLY A RISK MANAGEMENT SYSTEM TO TUNNELING? Any underground construction is considered to be a high risk environment. This is for several good reasons:

•

155 Copyright © 2004 Taylor & Francis Group plc, London, UK

Rock/ground always remains unpredictable

09096-19.qxd

2/18/04

12:00 PM

Page 156

Start

1 (Integrate in) Mission Statement

2 Build up/ optimise organization 3 Training

Hazards with high Yes damage potential?

No

4 Available safety rules?

Yes

No

5.1a Risk Assessment/ Management 5.1b Work out a Safety Concept

Figure 7. Gotthard Base Tunnel: Geology.

5.2 Build up a Safety Program 4 Available safety rules?

No

5.3 Work out safety rules

Yes 5.4 Support company’s superiors

6 Plan+implement measures 7 Emergency planning 8 Ensure employee’s contribution 9 Build up a Health Program 10 Controlling + Audits => CIP*

Emergencies/ absences * CIP = Continuous Improvement Process

Figure 8. Gotthard Base Tunnel: Scheme of tunnel system.

Figure 5. Swiss Safety  Health 10 Points System.

Figure 6. Gotthard Base Tunnel: Construction Program.

Figure 9. Gotthard Base Tunnel: TBM Cross Section.

•

• • • •

• •

Unforeseen water in big quantities can always be a big, crucial factor anywhere anytime Available space is very limited Heavy weight, high energy transport activities

156 Copyright © 2004 Taylor & Francis Group plc, London, UK

Darkness presumes, light is rare High construction noise High temperatures, high moisture Dealing with explosives, high voltage

09096-19.qxd

2/18/04

12:00 PM

Page 157

Figure 13. Current excavation status at Amsteg.

Figure 10. Gotthard Base Tunnel: State of excavation.

Owner Project Location Lot Designer & Client Representative Contractor

AlpTransit Gotthard AG Gotthard Base Tunnel Amsteg, Canton Uri/Switzerland 252, Tunnel Amsteg ING GBTN (Engineering Joint Venture Gotthard North) Joint venture Amsteg, Lot 252, Gotthard Base Tunnel North (AGN) Tunnel length 2 tunnels by ca. 11’350 m Excavation cross section Ca. 71 m2 Construction method (Drill & Blast +) TBM Tunnel construction costs Ca CHF 627 M* Ca. USD 471 M** Construction time frame 02/2002–07/2007 Project status (11/2003) TBM operation just started*** Contractor’s support for OSH Author’s mandate

Figure 14. Control Center at the portal.

* Excl. VAT (Value Added Tax) ** 1 USD (US Dollar) ≈ CHF (Swiss Franc) 1.33 (11/2003) *** For more information please log on to the owner’s website http://www.alptransit.ch and the contractor’s website http://www.agn-amsteg.ch - thank you.

Figure 11. Selected project data Lot 252, Tunnel Amsteg.

Figure 15. Control Center: Continuous monitoring.

•

Fresh air is very limited, etc.

In addition, any underground activity normally brings with it

• •

High public profile, high capital investments Work schedules around the clock.

All these puzzles bring about owners requests for at least some kind of proven risk management procedures. 10 HOW DO YOU MANAGE SAFETY  HEALTH IN TUNNELING?

Figure 12. Gotthard Base Tunnel: Tunnel Boring Machines (TBM).

Swiss Law requires the implementation of a (Occupational) Safety  Health (OSH) 10 Points System.

157 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-19.qxd

2/18/04

12:00 PM

Page 158

Figure 19. Site Safety Officer (SS0) at work. Figure 16. HK-TBM S-229. Owner

Luzern Stans Engelbergbahn LSE Steilrampe Tunnel Engelberg Project Grafenort (near Luzern)/ Location Switzerland North + South Lot Designer & Bucher + Dillier Client Representative Ingenieurunternehmung AG + IG LSE Joint Venture Tunnel Engelberg Contractor (ATE) Ca. 4'040 m Tunnel length Excavation cross section Ca. 30 m2 Construction method Drill & Blast Tunnel construction costs Ca. CHF 72 M* Ca. USD 54 M** Construction time frame 05/2001 – ca 12/2005 Project status (11/2003) Under Construction*** Author’s mandate Contractor’s support for OSH

* Excl. VAT. ** 1 USD ≈ CHF 1.33 (11/2003). *** For more information please log on to the owner’s website http://www.lse-bahn.ch - thank you.

Figure 17. Risk assessment workshop.

Figure 20. Selected project data Engelberg Tunnel.

Figure 18. Excerpt from SSO’s routine inspection report.

11 WHERE HAS A RISK MANAGEMENT SYSTEM IN TUNNELING ALREADY BEEN USED? A random selection of several (larger) tunneling projects in Switzerland (CH), in which the author is currently involved, may illustrate the benefits of an RMS approach for controlling the construction processes and their inherent risks.

Figure 21. Annual OSH System Check at the face.

158 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-19.qxd

2/18/04

12:00 PM

Page 159

Figure 25. AlpTransit Gotthard AG (ATG).

Figure 22. Water ingress at the face.

Figure 26. Arbeitsgemeinschaft AMSTEG, Los 252, Gotthard-Basistunnel Nord (AGN).

Figure 23. Water at the portal.

Figure 27. Murer/Strabag.

Figure 24. Probing ahead from the face.

Figure 28. Herrenknecht.

For further information please feel free to visit the owner’s website http://www.alptransit.ch and the author’s website (the author’s contribution to the AUA NAT02 conference)

• •

159 Copyright © 2004 Taylor & Francis Group plc, London, UK

http://www.moergeli.com/d1doc10e.htm http://www.moergeli.com/d1doc11e.htm Thank you.

09096-19.qxd

2/18/04

12:00 PM

Page 160

11.3

Engelberg Tunnel/Switzerland

Early on, the owner took some strategic decisions with regard to risk management:

• •

Figure 29. rowa.

Systematically probe ahead for gas detection. Require the construction contractor to establish, implement, keep current and document an Integral Safety Plan.

The contractor contributed by establishing and implementing an Integral Safety Plan following suva’s (Swiss National Accident Insurance Fund) suggestions. Annual OSH System Checks keep it updated. So far the contractor dealt successfully with more than his fair share of unexpected ground conditions. 12 CONCLUSIONS

Figure 30. Swietelsky.

Control your critical success factors by: 1. Apply generally accepted best risk management practices. 2. Establish a thorough and comprehensive Risk Management System (RMS) where required by law, project challenges, owner’s specifications and/or where best risk management practices are not available.

Figure 31. Amberg.

11.1

AlpTransit, Gotthard Base Tunnel/Switzerland

Early on, the owner took some strategic decisions with regard to risk management:

• • • • • •

Where possible, cross the alpine mountains through the most favorable geology. Where possible, cross difficult ground/rock perpendicular to strata as short as possible. Build two tunnels. Provide cross passages about every 312 m . Establish, implement, keep current and document a comprehensive owner’s RMS in all phases of the project. Require the consultants, site supervisors and main construction contractors to run their own RMS, keep it updated and document it at least twice a year.

11.2

Gotthard Base Tunnel, Lot 252, Tunnel Amsteg/Switzerland

The contractor contributed by establishing, implementing, periodically updating and documenting a project-specific OSH solution:

A RMS provides a unique opportunity in combining ongoing US underground activities with previous project experience, providing owners, engineers and contractors with real added value in successfully pursuing their goals:

• • • • •

Identify all known hazards graphically on a onepage-sheet early on. Allocate limited resources for Risk Assessments where inevitable (Risk-Zone 1). Mitigate your risks systematically by System Safety. Add value to the owner’s project by minimizing costs, time and third party impacts. Provide evidence of legal compliance.

To perform an RMS, a four-folded approach is recommended by the author (m  m/am): 1. Start with a Preliminary Hazard Analysis. 2. For hazards in Risk-Zone 1, apply a full and thorough Risk ssessment. 3. As an option verify your planned preventive action by a Fault Tree Analysis. 4. Establish, implement, deploy, maintain and document System Safety to mitigate your risks. ACKNOWLEDGEMENTS

• • •

Preliminary Hazard Analysis Risk assessments in team workshops Implementation of System Safety

The author thanks and acknowledges the very competent support and kind provision of plans, schemes and

160 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-19.qxd

2/18/04

12:00 PM

Page 161

pictures by: Last but not least, my thanks goes to Mrs. S. Tschupp for her always very competent support in improving my use of the English language. Without their big help this document would not have been possible. The biggest thanks goes to all crews on site, safely coping with the unforeseeable as their daily routine. Every day they move into places where no human being has ever been before. Always just one small step for a man, but a giant leap for mankind … The author’s apologies go to the readers for any inconvenience dealing with small print, reduced tables and pictures. An update of the paper can be offered through our presentation at the AUA Conference. The original paper will be available for download on http://www.moergeli.com/dldocuebersichte.htm after the AUA North American Tunneling 2004 Conference, April 17–22, Hyatt Regency Hotel Peachtree Plaza, Atlanta, GA-USA.

REFERENCES

AlpTransit Gotthard AG (ATG), Zentralstrasse 5, CH-6003 Luzern/Switzerland (http://www.alptransit.ch). “Geologic Site Investigations for Tunnels”, USNC/TT Study, Eugene B. Waggoner, Charles W. Daugherty, Underground Space, Vol. 9, pp. 109–119, 1985. “Geotechnical Baseline Reports for Underground Construction, Guidelines and Practices”, Randall J. Essex, American Society of Civil Engineers, 1997, ISBN 07844-0249-3. Herrenknecht AG Tunneling Systems, Schlehenweg 2, D77963 Schwanau/Germany (http://www.herrenknecht.de). Joint venture Amsteg (German: Arbeitsgemeinschaft Amsteg), Lot 252, Gotthard Base Tunnel North (AGN), Grund, CH-6474 Amsteg/Switzerland (Murer AG/Strabag AG) (http://www.agn-amsteg.ch). Joint venture Tunnel Engelberg (German: ARGE Tunnel Engelberg): (ATE), CH-6388 Grafenort/Switzerland (Achermann AG – Swietelsky Bau Tunnelbau Gesellschaft m.b.H). Rowa Tunnelling Logistics AG, Leuholz 15, CH-8855 Wangen SZ/Switzerland (http://www.rowa-ag.ch). System Safety Scrapbook, P. L. Clemens, 2002, Sverdrup Technology, Inc. Swietelsky Bau Tunnelbau Gesellschaft m.b.H., EduardAst-Str. 1, A-8073 Feldkirchen/Graz (http://www.swietelsky.ch). US Military Standard 882D; 10 February 2000, US Department of Defense.

AIB, Amberg Consulting Engineers Ltd. (AIB), Trockenloostr. 21, CH-8105 Regensdorf-Watt/ Switzerland (http://www.amberg.ch).

161 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-20.qxd

2/18/04

12:01 PM

Page 163

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Managing underground construction risks in New York Nasri Munfah STV Inc.

Sanja Zlatanic & Paul Baraclough Parsons Brinckerhoff

ABSTRACT: New York Metropolitan Transportation Authority (MTA) has embarked on the construction of the largest transportation project ever undertaken in New York City. The East Side Access Project (ESA), a $6.3B project, will connect the Long Island Rail Road (LIRR) to the landmark Grand Central Terminal (GCT) on the East Side of Manhattan. In Queens, the connection will require the construction of five soft ground tunnels, using pressurized face TBM. The tunnels will be constructed under active railroad yard and the main line of the LIRR. In addition, cut and cover tunnels and underpinning of existing transit lines will be required. In Manhattan, the construction will require rock tunnels totaling over 7300 m long, and a series of rock caverns connecting to two station caverns to accommodate eight tracks with four platforms in a stacked configuration. The station caverns are approximately 20 m wide 22 m high 400 m long each and are situated about 10 m beneath the historic Grand Central Terminal (GCT) and high-rise buildings. The tunnels pass under existing subway tunnels, and a variety of historic and sensitive buildings in one of the world most expensive real estates area. In addition two underground ventilation and traction power facilities and eleven shafts of varying sizes and functions will be constructed. Several construction techniques will be used to excavate the tunnels, caverns, and shafts including tunnel boring machines, roadheaders, raise bores, and controlled blasting. Such major undertaking involves significant construction and commercial risks that must be dealt with. This paper provides a status report of the project focusing on the technical challenges in Manhattan, the construction risks and the identified measures to deal with such risks.

1 EAST SIDE ACCESS PROJECT GOALS AND DESCRIPTION 1.1

Project goals

The East Side Access project is vital for the general growth of New York Metropolitan Area by increasing the transportation facilities in the region and improving the daily commute to over 300,000 people. Presently, the Long Island Rail Road operates 36 trains per hour during the morning and evening rush periods to and from Penn Station in the West Side of Manhattan carrying about 240,000 commuters each way. Approximately half of these commuters final destination is the East Side of Manhattan. They travel within New York City using transit lines or surface transportation to reach their final destinations. This imposes added congestion on already overcrowded transit lines and

surface streets. In addition, Penn Station, which serves two other railroads, is at its capacity. The East Side Access project will improve the mobility in the region, will improve commuting for the LIRR passengers, and will relieve congestion in Penn Station. LIRR is planning to operate 24 additional trains per hour during the morning and evening rush periods to Grand Central Terminal on the East Side of Manhattan. This will increase its service by about 109,000 passengers. In addition, the project will provide a single seat ride to most riders to their final destination and will reduce their overall travel time by about 30 minutes each day. Overall, the project will improve mobility in the region and will stimulate the economic growth in New York. Recognizing these needs dates back to the 1960’s when the Metropolitan Transportation Authority (MTA), the

163 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-20.qxd

2/18/04

12:01 PM

Page 164

parent company of the LIRR, developed a plan to provide an East Midtown Terminal for the LIRR. In the 1960’s and 1970’s the MTA constructed 2560 m (8400 ft) tunnel under the East River. The tunnel extends from 63rd Street and 2nd Avenue in Manhattan under the East River to 29th Street and 41st Avenue in Queens. A four-track (two over two stacked) tunnel constructed by TBM, drill-and-blast, immersed tube and cut-and-cover tunneling methods. The Metropolitan Transit Authority (MTA)-New York City Transit (NYCT) operates a rapid transit line through the two upper level tracks. Due to fiscal constraints the two lower level tracks remained dormant until now and will be used by the LIRR. The project was revived in the 1990’s. An Environmental Impact Statement and Preliminary Engineering were completed in year 2000 and submitted to the Federal Transportation Administration. The selected alternative is a connection of the LIRR Main Line and the Port Washington Branch to a new deep underground terminal under the historic Grand Central Terminal. The project is budgeted to cost $6.3 Billion and to be completed by year 2012. See figure 1. 1.2

Figure 1. Overall plan.

Project alignment

The alignment in Queens extends about 1800 m (6000 ft) from the existing bulkhead near 29th Street and 41st Avenue across Northern Boulevard and connects to the LIRR’s Main Line and the Port Washington Branch. The existing two track tunnel branches into four track configuration which in turn connects to the Main line at the Harold interlocking complex East of 43rd Street passing across and underneath Yard A and Sunnyside Yard. In order to provide midday storage capacity, lead tracks will be provided to Yard A which will be reconstructed to provide a mid-day storage. Figure 2 illustrates the alignment in Queens. The alignment in Manhattan starts at the corner of 2nd Avenue and 63rd Street where the existing tunnels are terminated about 50 m (140 ft) underground and extends south-westerly to Park Avenue turning south. The tunnels start as single-track tunnels then separate through two wye caverns into four singletrack tunnels. The outer tunnels converge under the inner tunnels. At 51st Street the tracks converge via two crossover caverns situated over each other. They then separate again through two bi-level wye caverns into four two-level structures that tie into the end walls of the station caverns. The alignment passes under several underground transit lines and structures including the Metro North Railroad (MNR) Park Avenue Tunnel, the Lexington subway line, the 60th and the 53rd Street subway lines, and the 42nd subway shuttle. The LIRR Station will be situated under the historic Grand Central Station and it will consist of two caverns having eight tracks (two over two in

Figure 2. Alignment in Queens.

each cavern) with four island platforms. The caverns are situated between 44th and 48th Streets approximately 52 m (170 ft) below street surface. The cavern dimensions are 18.3 m (60 ft) wide by 23.8 m (78 ft) high and they are approximately 400 m (1200 ft) long and they are located 30.5 m (100 ft) on centres along the centreline of Park Avenue and the operating underground Metro North Railroad tracks. The lower level of the existing Grand Central Terminal will function as a large concourse providing for circulation, waiting areas, station functions, and retail spaces. South of the station eight tunnels extend south passing under the landmark New York Central Building, the 59-storey MetLife building, and the historical main hall of Grand Central Terminal. Each pair of tunnels merge via two three level caverns to provide four tail track tunnels extending to 38th Street. See figure 3. 1.3

164 Copyright © 2004 Taylor & Francis Group plc, London, UK

Project geology

Geotechnical considerations are the key elements of the success of the Grand Central Connection project. Most of the construction is underground and geotechnical issues will impact almost all design and construction activities.

09096-20.qxd

2/18/04

12:01 PM

Page 165

Figure 3. Alignment in Manhattan.

The excavations in Queens will be predominantly in soil. Some excavations in the Ravenswood Gneiss will be encountered in the cut-and-cover area as well as the initial portion of the bored tunnels where they are in a mixed face condition. The soil deposits in Queens represent a complex glacial depositional environment upon which a post-glacial tidal marsh environment and manmade fill deposits were superimposed. The soils consist of fill of heterogeneous mixture of coarse to fine silts and sands with building rubble and a possibly high concentration of rock fragments; glacial outwash deposits, predominantly granular soils, deposited by glacial meltwaters, consisting of clean coarse to fine sands to silty sands. Ice-rafted boulders of various sizes are anticipated in this deposit. Glacial lake deposits consisting of stratified silts, sand and clays deposited in quiet water is another major formation. Ice-rafted boulders may also be encountered in this deposit. Glacial till consisting of dense mixture of sand, silt, clay, cobbles and boulders, and nested boulders, is major layer along the tunnels in Queens. Also found along the alignment in certain areas compressible soils consisting of peat and partially decayed vegetation underlying the fill within the limits of Sunnyside Yard, along the course of a former creek. Most of the construction in Manhattan will be in a rock formation known as the Manhattan Schist. The rocks underlying Manhattan belongs to the New England Upland, and is locally known as the Manhattan Prong. This rock comprises three lithologically distinct sequences of a metamorphic assemblage of Proterozoic to lower Paleozoic age consisting of schist, gneiss and marble. The main rock types recovered from borings along the Manhattan alignment are metamorphic, dominated by schist and gneiss. The essential minerals are muscovite, biotite, quartz and feldspar (plagioclase, microcline and orthoclase). Garnet is the principal accessory mineral occurring as fine disseminated crystals and clusters. The schist frequently grades into a granofels (a fine to medium grained, equigranular metamorphic rock, in which there is no discernible foliation or banding). Amphibolite, consisting mainly of hornblende, plagioclase feldspar, quartz and biotite, is occasionally intercalated with the schist and gneiss, and lies parallel to the foliation. Also occurring within

the rock are frequent “igneous” layers normally parallel to the foliation that vary from medium-grained granite to coarse-grained pegmatite. The tectonic history of the rock has left the Manhattan Schist fractured and dislocated. Four dominant joint sets for the rock mass have been confirmed by the geotechnical investigation and exposed rock wall mapping. However, dip directions and dip angles vary widely across the Manhattan alignment. Manhattan is slightly raised above the sea level, and is bounded by the East River to the east and the Hudson River to the west. The area is heavily urbanized so infiltration of rainfall is low. More intense conductive fracturing occurs at locations of buried streams. These fractures are conduits for the groundwater with much greater hydraulic conductivity than other fractures in the undisturbed rock mass. The groundwater levels measured in observation wells range from 4.5 m (15 ft) below the street level along Park Avenue to less than 1.5 m (5 ft) below the invert of the existing lower level of Grand Central Terminal. The permeability was determined from in-situ packer tests and varies from 107 m/sec to 104 m/sec. 2 CONSTRUCTION OF THE MANHATTAN SEGMENT Building the ESA project will involve construction of a number of tunnels using a variety of construction methods. Selection of an appropriate tunneling method for each section of the alignment from a variety of tunneling techniques available in current construction practice will have a major impact on the successful completion of the project. Technical, practical, operational and economical factors affect the tunnel selection. In selecting a tunneling method a number of factors are evaluated including: ground conditions, groundwater, geometry and site constraints, impact on adjacent structures and utilities, environmental concerns, local construction practice, permitting and community acceptance, cost, schedule, contracting forms, and construction risks. In Manhattan, the construction will require rock tunnels totaling over 7300 m (24,000 ft), and a series of rock caverns (single level wyes, three-level wyes, crossovers) connecting to two station caverns. In addition, two underground ventilation facilities and traction power substations along with utility shafts, ventilation shafts and several vertical circulation shafts will be constructed. To meet the construction schedule constraints the Manhattan segment was divided into three major civil underground tunneling contracts: CM009 Manhattan Tunnels, CM012 GCT Caverns, and CM013 Ventilation Facilities. In addition a variety of preparatory contracts and finishing contracts are provided to complete the project.

165 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-20.qxd

2.1

2/18/04

12:01 PM

Page 166

Mechanized tunneling

The first tunneling contract (Contract CM009) will be let as an early contract to build most of the singletrack tunnels in Manhattan: two tunnels branching into four tunnels excavated by two tunnel boring machines (TBMs). The TBM will be 6.55 m (21-6) in diameter to accommodate an internal tunnel diameter of 5.95 m (19- 6). See figure 4. The tunnels in this contract will be provided with the initial support only consisting of rock dowels, rock bolts, and steel ribs as needed. Future underground structures include the East 55th Street ventilation plant, the East 51st Street crossover structures, the station approach wye caverns, the GCT caverns, the tail track wye caverns and the East 38th Street vent structure, will be enlarged in future contracts. This Contract also includes the construction of an underground TBM assembly chamber and two re-assembly chambers in the location of future track wyes. The TBM assembly and reassembly chambers will be constructed using controlled blasting techniques. The contract also incorporates vertical raise bores 3 m in diameter (10-0) in five future vertical shaft locations and four inclined raise bores in future station escalator shafts. A preparatory contract CM016 will be awarded earlier to excavate the approach tunnels of about 100 m (330 ft) to the assembly chamber using roadheader. In this contract the performance of the roadheader will be assessed for its suitability technically and commercially to excavate the Manhattan schist for potential use in future contracts in lieu of the drill and blast method. If successful the roadheader use will significantly mitigate issues related to excavation overbreak, noise and vibration, and impact on adjacent facilities and the public. The TBM tunnels will be excavated using one of three initial support systems referred to as support classes using the observational approach. These support systems are designed based upon anticipated ground conditions and ground behavior. The support

Figure 4. Typical single track tunnel.

systems are designed to arrest potential movement of rock blocks and wedges and to prevent loosening of the surrounding rock mass. Support elements are installed concurrent with the TBM excavation. In drill and blast or roadheader excavations, initial ground support elements include rock bolts, dowels and a reinforced shotcrete lining. The shotcrete lining will consist of a reinforced shotcrete layer 100 mm (4 inches) to 250 mm (10 inches) in thickness, depending on ground conditions, tunnel size and support class. Shotcrete reinforcement will include welded wire fabric or lattice girders. A secondary, final cast-inplace concrete lining will be provided for long-term support under a subsequent contract. 2.2

The major underground contract in Manhattan is CM012 GCT Caverns. This contract completes the civil underground work in the previously constructed tunnels. Several single level and triple level caverns will be built in this contract to accommodate track crossovers, switches, and railroad system facility spaces. In addition cross passages, vertical circulation and utility shafts, service tunnels, inclined escalator shafts, and other underground structures will be built as part of this contract. The final liner, duct bench, and other embedment for the entire structures and tunnels will also be constructed in this contract. Furthermore, rehabilitation of the existing tunnels will be done in this contract. The main feature of this contract is the twin station caverns situated between 44th and 48th Streets, with their invert about 40 m (131 ft) below grade. The caverns are spaced approximately 30 m (100 ft) on centers horizontally beneath Park Avenue, a 35-storey high landmark building and a portion of the Met Life tower. The caverns are approximately 18 m wide 20 m high 360 m long each (59 ft 66 ft 98 ft) and are situated about 10 m (33 ft) beneath the historic GCT. See figure 5: GCT caverns.

Figure 5. GCT caverns.

166 Copyright © 2004 Taylor & Francis Group plc, London, UK

Cavern construction

09096-20.qxd

2/18/04

12:01 PM

Page 167

Each cavern will have four tracks, two over two with center-island platforms. Each cavern has three levels, a lower train room, a mid level mezzanine, and an upper train room. The two caverns are interconnected with four cross passages leading to four escalator banks, two elevator banks, and emergency stairs connecting the caverns with the concourse level in the existing Grand Central Terminal. The inclined escalator shaft at 45th Street houses five escalators and is approximately 11 m (36 ft) wide by 7 m (23 ft) high. The other escalator shafts house four escalators each and are approximately 9 m (30 ft) wide by 6 m (20 ft) high. In addition, several utility and ventilation shafts and tunnels will be constructed to provide the lifelines to the station. The excavation of the rock in this contract will be done by drill and blast using controlled blasting techniques. Alternatively, if the roadheader has proven its ability to excavate the Manhattan schist economically, it can be used to excavate the top headings. The initial liner will be rock dowels, rock bolts, and lattice girders and shotcrete. The final liner will be cast in place concrete liner varying in thickness from 300 mm (12 inches) for the running tunnels to 750 mm (30 inches) for the GCT caverns. The tunnels and caverns will be waterproofed utilizing an open waterproofing system consisting of a continuous PVC membrane and a layer of drainage fabric. The waterproofing system will be designed as a drained, open system with perforated sidewall drainpipes for collection of groundwater. The architectural finishes and the electrical and mechanical works will be part of a subsequent finishing contract. 2.3

Ventilation facilities

Three underground ventilation structures will be built as a separate contract (CM013) to provide the ventilation needs in the tunnels during emergency and congested modes. The equipment of the ventilation fans and associated mechanical, electrical and control systems will be part of a project wide contract. The ventilation facilities will be located in the 38th Street, the 50th Street and the 55th Street. The latter also houses traction power substation. The ventilation facilities vary in size and equipment. The 38th Street facility consists of two fans 200,000 cfm each placed horizontally in an underground cavern and connected via a ventilation tunnel to a shaft to the surface placed in a parking lot. The shaft is 5 m 5 m (15 ft 15 ft) and is 42.5 m (129 ft) deep. The construction of the shaft will be top down using drill and blast technique. The 50th Street facility consists of four fans 150,000 cfm each placed horizontally in the mid-level of the station’s northern interlocking structures. The two sets of fans are connected via cross flues and dampers to enable ventilating any of the eight tunnels

at that location. The fans are connected to the surface via ventilation tunnel and a vertical shaft. The construction of the shaft will be top down using drill and blast. The 55th Street ventilation facility consists of four vertically placed ventilation fans 200,000 cfm each and a two-unit traction power substation. The facility is a two level underground structure approximately 12.4 m 11.7 m (40-7 38-6) and 80 m (263-6) long. The upper level consists of a plenum and mechanical electrical rooms while the lower level houses the traction power substation. The facility will be placed in the 55th Street bed and is connected with the tunnels via a vertical shaft 14.5 m 14.5 m (44 ft 44 ft) and 44 m (134 ft) deep. The shaft connects with tunnel via plenum cavern. The construction of the surface facility and the shaft will be done by the cut and cover method under street decking using drill and blast. 3 RISK MANAGEMENT The Manhattan segment will be built under and adjacent to several operating transit lines including the 63rd Street line, the Lexington Avenue line, the IND 53rd Street lines, MNR’s Park Avenue tunnels, Park Avenue viaduct, the 42nd Street No. 7 line, and the Time Square Shuttle, and the Park Avenue Tunnel. The project alignment lies beneath mid-town Manhattan, the most densely populated urban business area in the region and certainly of extreme value to the City. Due to the large number of sensitive and critical structures above and adjacent to the planned tunnel and shaft excavations and construction, it is imperative that major risks and associated mitigation measures be identified during the early stages of the project. At the early stages of the design development, the project team prepared a risk management plan. The plan emphasized that risk management is not a one-time task, but rather a continuous process throughout the life of the project to identify, track and manage risks in the planning, design, and construction stages. Generally, the process consists of four elements:

• • • •

Each area of the project, during the development of the respective construction packages, was examined and analyzed to identify potential risks. The risks identified were evaluated for their probability of occurrence and a risk response and monitoring plan was developed. It was recognized that the level of risk sharing is a major factor in deciding the type of procurement practice to be implemented. Although several potential

167 Copyright © 2004 Taylor & Francis Group plc, London, UK

Risk Identification Risk Quantification Risk Response Development Risk Response Monitoring during design development and construction.

09096-20.qxd

2/18/04

12:01 PM

Page 168

procurement approaches were evaluated it was determined that due to the high risk and exposure of this project that the conventional design-bid-build process be used for the underground construction contracts. This decision was made because the design-bid-build process provides strong owner control and better quality; it delineates clearly the roles and responsibilities between the owner, and the contractor and provides an equitable sharing of the risks. The MTA provides complete identification of the geotechnical and tunneling issues and resolutions to the environmental issues. It also provides a complete design and obtains all approvals, while the contractor executes the work with the best means and methods. Using the conventional contracting practice and to equitably share and manage risks, several provisions were provided in the contract form. These include pre-qualification of contractors, complete geotechnical disclosure, the implementation of a Dispute Review Board (DRB), the use of differing site condition clause, the use of unit prices and contingent bid items, the provision of value engineering, the provision of owner controlled insurance program, and a partnering program. 3.1

Pre-qualification of contractors

The major underground contracts required technical and financial pre-qualifications of the potential bidders to ensure their ability to perform the work effectively, economically, and to a high quality. The pre-qualification would include the company’s or the joint venture’s technical ability to perform the work. Items to be evaluated include: approach, experience with similar projects, with similar ground, and similar proposed methodology. Pre-qualification of the key staff, such as the contractor project manager, the field manager or superintendent, the TBM operator, the TBM maintenance person, etc., is critical for a successful project. In addition to the bidder’s financial ability to obtain bonding and its solvency, its history of completing projects on time and within budget is a factor in the bidder financial qualifications. Pre-qualification could also extend to major subcontractors and major suppliers such as the TBM manufacturer, liner supplier, etc. 3.2

Full geotechnical disclosure

Experience has shown that full disclosure of geotechnical information would reduce the risk to both the owner and the contractor and thus the project cost. Therefore it was important for the MTA to invest in a comprehensive geotechnical program including an extensive boring program, exposed rock face mapping, laboratory testing, and in-situ testing. The information is included in the contract documents in the form

of Geotechnical Data Report, Geotechnical Design Summary Report, Geotechnical Interpretive Report, and/or Geotechnical Baseline Report. The GBR establishes quantitative values for selected conditions anticipated to have great impact on construction. These values are established through technical interpretation of the data and commercial considerations of risk allocation and sharing. The advantages of this report are ease of administration of contractual clauses, unambiguous determination of entitlement, clear basis of contractor’s bid, and clear allocation of risk between owner and contractor. The intent of the disclosure of geotechnical information and the use of the Geotechnical Baseline Report is to allocate and share underground construction risks between the owner and the contractor. 3.3

In the DRB process, a board of independent, experienced, and impartial members is selected to hear and address disputes. Generally the board consists of three members, one representing the owner, one representing the contractor, and the third, who will act as the chair of the board, selected by the other two members. The board provides recommendations to resolve disputes that participants are unable to make. It was found that this process has resulted in lower bids, better communication and less acrimony at the job site, fewer claims, and more timely and cost effective resolutions. For the ESA project, the DRB process will be limited to underground related issues. It will involve formal and informal processes in which position statements and expert reports are made and presented to the board. The board will meet and visit the construction site regularly to familiarize themselves with the project issues. Board findings and recommendations will be made; however, acceptance of the DRB findings is not binding or admissible in a legal process. The MTA is committed to a fair and equitable resolution and committed to support the process and does not want the DRB process to be just a step in a legal process, rather it will be the basis of negotiating a fair and equitable resolution. 3.4

Differing site condition

The Differing Site Condition clause is used on the ESA Project as a measure of allocation of risk between the owner and the contractor relative to the ground condition. In exchange for lower initial bids, the owner bears some portion of the risk of subsurface condition. Bid contingencies in the low bid are paid by the owner whether adverse conditions are encountered or not. On the other hand, DSCs are paid only if they are encountered. Two categories of DSCs are identified in the contract documents: Category 1 governs when subsurface

168 Copyright © 2004 Taylor & Francis Group plc, London, UK

Dispute review board

09096-20.qxd

2/18/04

12:01 PM

Page 169

conditions differ from those indicated in the contract. This is based solely on what is stated in the contract, including geotechnical data or geotechnical interpretation, or geotechnical baselines included in the contract documents. Category 2 applies when conditions which were not known to the contractor at the time of contracting differ from those normally encountered in the area. It is generally related to unusual conditions and not based on contract documents. It is important to note that to recover on the DSC clause; the contractor must show impact on cost and time and must show causality. 3.5

Partnering

The MTA incorporated in all the contracts a formal partnering provision. The goal of this process is to minimize disputes and to prevent them from escalating in time and value by resolving them at the lowest possible level in the project organization. It attempts to establish a win-win attitude between the project participants, including the owner, the contractor, the engineer, and the construction manager. This process encourages dialogue among the various participants and relies on reasonable people to resolve disagreements reasonably. It seeks to eliminate adversarial posturing and positioning that often develop when disputes and claims arise. Through this process a series of dialogues and interactions are developed whereby the team members are encouraged to work out differences for the best interests of the project. When an issue is not resolved at the lowest level, it is brought up to a higher level for resolution avoiding legal proceedings. 3.6

•

•

Value engineering

To stimulate innovative approaches within the limitations of the contractual requirements, the MTA opted to include a value engineering clause in the contract. Efficiency is achieved by relaxing the design criteria where not critical or meeting the intent of the design more efficiently via creative approaches. The saving achieved by value engineering is shared between the owner and the contractor. It is important to assess the potential effects of differing site conditions on the design as modified by the value engineering. 3.7

Environmental issues

The selection of the tunneling method in Manhattan takes into consideration not only the technical challenges and the construction risks, but also environmental concerns and public acceptance. Therefore several unique approaches were developed for implementation of this project:

•

Access Shaft: Locating the TBM access shaft in Manhattan is difficult because of the extensive

169 Copyright © 2004 Taylor & Francis Group plc, London, UK

development near the site and the potential for shaft construction to disrupt the local community and businesses. Extensive utility relocations will be required and significant impact on traffic will occur in addition to the noise, dust and disruption. Therefore, it was decided to construct the Manhattan shaft in Queens. All materials will be brought into Manhattan via the shaft in Queens and through the existing tunnels. Similarly, the muck will be removed using conveyors or hoppers through the existing tunnels to Queens where it will be hauled away by rail or trucks. Construction of the Queens shaft proved to be a feasible, economical and environmentally friendly solution that minimizes construction risks on adjoining developments and the public. Minimizing Impact on Overlying Buildings: To assess possible impact of the tunnels and caverns excavation on the existing high-rise buildings and underground structures such as transit lines, stations, and utilities, settlement analyses were performed at critical sections along the alignment. In addition, a sophisticated system of monitoring was designed using inclinometers, extensometers, liquid level, and global optical survey. The results of these analyses showed insignificant elastic surface settlements in the range of a few millimeters. The analysis included considerations of sequential excavation followed by immediate installation of initial support. Settlements (tunnel and surface), convergence, and stress and strain measurements will be closely monitored and response measures will be implemented quickly to avert adverse impact if encountered. Noise and Vibration: Determination of the likely levels of vibration, noise, and ground borne noise during construction is necessary to provide contractors with guidelines for the explosive charge and delay configuration during blasting, selection of construction equipment for drilling, mucking and mechanical excavation, and to provide for appropriate construction sequencing. To estimate impacts of construction, design criteria were developed for peak particle velocity (PPV), extensive data collection and interpretation were made, and analyses based on the collected information were performed. The results were used to recommend measures for structures protection and instrumentation and monitoring. As part of public participation program the results of the analysis will be shared with affected third parties. The Contractor will perform test blasts before production blasting starts for each construction contract. This test blast program will confirm the applicability of the resulting vibration regression equations. Furthermore, the test blast program will aid in determining necessary adjustments to the blasting procedures, round length, and delay. Construction sequencing will be further adjusted to comply with the established criteria. Ground

09096-20.qxd

•

2/18/04

12:01 PM

Page 170

borne noise impact was assessed. Factors that influence ground borne noise include local geology, building foundation construction, number of floors above street level, and resonance of upper floors. During drilling, TBM, and raise bore excavation, the estimated ground borne noise level at sensitive receptors is estimated. The data will be used to monitor construction performance and a basis to provide good relationship with the public and affected neighbors. Coordination with the Public and Stakeholders: An important element in a successful major construction program is public opinion and the approval of citizen groups and public civic associations especially in a well-established place such as Manhattan. It was recognized the importance of obtaining consensus from the public and the facility users early in the project development stage. It is important to identify their concerns and address them. Concerns are usually related to construction noise and vibration, working hours, disruption to businesses and daily routine activities, emission of dust and pollutants, and the potential increase in pests and rodents. It is important to listen to the people’s concerns and address them early in the project developments. It is important to obtain buy-in of construction methodologies and approaches by the public in the early stages of the project and is critical for its successful completion. An extensive public outreach program was established and implemented during the design development to address the people’s concerns. This program will continue during construction to assure the people that agreements reached are being implemented and also to address new concerns that might develop after the start of construction. Coordination with the affected third parties has been an important aspect of the project. Public outreach, dissemination of timely and accurate information about the project’s different construction phases

and the impact of construction are part of the overall program and are being successfully implemented. 4 CONCLUSION The East Side Access Project, the largest transport ation project ever undertaken in New York City, will improve the overall New York metropolitan area transportation system and will stimulate economic growth. Scheduled for completion by 2012, early preparatory contracts have been under way and the award of first tunneling contract is scheduled to be early 2004. To be successful the project has been planned using state of the art design concepts, advanced construction approaches, suitable provisions for risk management and sharing, and careful attention to environmental issues.

REFERENCES Della Posta, M. and Zlatanic, S. “Manhattan Segment of The East Side Access Project: Design Evolution” RETC Proceeding 2001 Munfah, N. “Connecting Long Island Rail Road to Grand Central Terminal in Midtown Manhattan” RETC Proceeding 2001 Munfah, N. “Contracting Practices for Underground Construction” Proceeding Underground Construction British Tunneling Society 2003 Munfah, N., Zlatanic, S. and Stehlik, E. “Connecting a Commuter Railroad to Historic Terminal” Proceeding ITA Prague, 2003 Munfah, N. “The Manhattan Connection” Tunnel and Tunneling, 2001 Munfah, N. “The Application of the DRB on the East Side Access Project”, Presentation in the RETC New Orleans, 2003 Wone, M. “Rock Tunneling Challenges in Manhattan” Proceeding ITA 2003

170 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-21.qxd

2/18/04

12:29 PM

Page 171

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Risk allocation in tunnel construction contracts William R. Wildman Sutherland Asbill & Brennan LLP, Atlanta, Georgia, USA

ABSTRACT: One way to minimize claims on a tunnel construction project is to develop a fair contract that allocates risks to the party best able to manage that risk. This paper will explore various provisions found in tunnel construction contracts that allocate risks among the parties, including the differing site conditions clause; delay damages provisions that limit contractors to their extended job site conditions cost and owners to liquidated damages; indemnification for site accidents and insurance provisions, including wrap up policies covering all the parties to a tunnel project. This paper will then briefly touch on design/build as an alternative project delivery method for allocating risks on a tunnel project. The paper concludes with the general hypothesis that a contract which fairly allocates risks is a contract which minimizes the risk of litigaton.

1 THE DIFFERING SITE CONDITIONS CLAUSE A tunnel project involves great uncertainty and potential for catastrophic cost overruns. Under the common law, the contractor would bear the risk of additional expenses due to unforeseen conditions. Indeed, even if the owner decided to voluntarily modify the contract and pay the contractor the additional expenses, the pre-existing duty rule made it very difficult to do so because the new promise would be unsupported by consideration. The general rule is that no matter how onerous the burden, once a contractor has promised to perform, any added expenses caused by unforeseen conditions is allocated to the contractor. Although the common law allocated the risk of unforeseen conditions to the contractor, public bodies engaged in large civil construction projects, including tunnel projects, became mired in endless litigation over claims regarding unforeseen conditions or claims that the plans and specifications inadequately described the sub-surface conditions. Moreover, these government bodies also faced the problem of bankrupting contractors who could not afford to complete the work based on the agreed upon fixed price in the contract. A differing site conditions provision now almost uniformly appears in fixed fee contracts. [The American Institute of Architects (AIA), the Federal Acquisition Regulations (FAR), and the Engineers Joint Contract Documents Committee (EJCDC) agreements all have differing site conditions clauses.] An example of such a differing site conditions clause was

included in the Milwaukee Metropolitan Sewerage District (“MMSD”) Northshore Inceptor-Phase 1A 30-foot diameter main tunnel and access shafts contract. This 28,000 foot long, 300 foot deep, 30-foot diameter tunnel was designed to capture combined sewer overflows during storm events and then eventually pump the combined sewer overflow to a sewage treatment plant for later disposal into Lake Michigan. The differing site conditions clause provided: “A. The Contractor shall promptly, and before such conditions are disturbed, notify the Owner by written notice of: Subsurface or latent physical conditions at the site differing materially from those indicated in this Contract, or Unknown physical conditions at the site, of an unusual nature, differing materially from those ordinarily encountered and generally recognized as inherent in the work of the character provided for in this Contract. B. The Owner shall promptly investigate the conditions. If he finds that conditions materially differ and will cause an increase or decrease in the Contractor’s cost or the time required to perform any part of the work under this Contract, Owner shall, after receipt of the Contractor’s written statement under “D” below, make an equitable adjustment and modify the Contract in writing. C. No claim of the contractor under this Article shall be allowed unless the Contractor has given the notice required in paragraph A of this Article. However, the Owner may extend the time prescribed in paragraph A of this Article.

171 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-21.qxd

2/18/04

12:29 PM

Page 172

D. If the Contractor intends to assert a claim for an equitable adjustment under this article, he must, within thirty (30) days after the owners determination as to whether a Differing Site Condition exists, submit a written statement setting forth the nature and monetary extent of such claim. The Owner may extend the thirty (30) day period. E. No claim by the Contractor for an equitable adjustment hereunder shall be allowed if asserted after final payment under this contract.” This differing site conditions clause can be characterized as dealing with a “Type I” and “Type II” Condition. Under a Type I Condition, the contractor must show that it “encountered subsurface or latent physical conditions at the site that differed materially from those expressly or impliedly indicated in the contract.” To prove a Type II Condition, the contractor must prove that it encountered “an unknown physical condition at the site, of an unusual nature, differing materially from those ordinarily encountered and generally recognized as inherent in the work of the character provided for in the contract.” Even with the differing site conditions clause, the contractor continues to bear the common law risk of the usual and expected conditions at a construction site; however, the contractor no longer must gamble that the unexpected might occur. In the Milwaukee tunnel project, the contractor encountered conditions that were materially different from those represented in the contract documents. When large water inflows and poor rock support were encountered through major reaches of the tunnel construction, the engineer granted differing site conditions status to the problems and compensated the contractor on a time and material basis over and above the contractor’s lump sum price. Although the contractor’s original contract price was for $46,000,000.00, the Milwaukee Metropolitan Sewer District ended up paying the contractor $166,000,000.00 to complete the North Shore Interceptor Tunnel. 1.1

Justifications for the differing site conditions clause

[For a very detailed examination of these justifications, see Hazel Glen Beh, Allocating the Risk of the Unforeseen, Subsurface and Latent Conditions in Construction Contracts: Is There Room for the Common Law? 46 U. Kan. L. Rev. 115 (1997).] A differing site conditions clause in a tunnel construction contract will encourage bidders to submit their lowest bid rather than build cushions into their bids for contingencies that may never occur. The differing site conditions clause should save the government money over time because it allows the contractor to remove its contingency from its bid and the owner avoids overpayment on the majority of projects and is

required to pay for differing site conditions only when they occur. The differing site conditions clause should also minimize claims. Unlike the common law, a construction contract containing a differing site condition clause requires the owner to negotiate a new price for the unanticipated work. Although this is the noble goal of the differing site conditions clause, in practice, while the owner must adjust the contract if a differing site condition exists, the owner can still dispute the existence of a valid differing site condition claim or the amount of the adjustment requested by the contractor. Still, under the common law, the contractor was faced with financial ruin. The contractor would necessarily have to litigate if the differing site condition was significant. It is therefore more likely that the differing site conditions clause has reduced the adversarial nature of tunnel construction projects. The differing site conditions clause also reduces the likelihood that contractors will go out of business if they encounter a differing site condition. Absorbing the costs of unforeseen conditions protects the industries upon which the large owner depends. Additionally, if contractors have to face the costs of unforeseen conditions in tunnel projects, they may choose not to bid on such high risk projects, and in the long run there will be fewer competitors performing this type of specialized work. The different site conditions provision also keeps down the cost of preparing bids and doing business. With such a clause the contractor is not forced to do his own extensive and expensive soil boring tests. This also encourages the owner to provide the contractor with as much information about the site as possible so that all the bidders are bidding on a level playing field. Rather than having multiple soil testing programs ongoing (and hence having that cost of that soil testing program incorporated into the contractor’s bid), the government can spend the money once on its own program and keep the cost of all bids down. Another benefit of the differing site conditions clause is that it permits the contractor to recover for the additional cost of completing the contract when a true differing site condition is encountered rather than cutting costs and otherwise inappropriately attempting to make up for the potential loss. Thus, the differing site conditions provision promotes direct recovery rather than indirect and inefficient recoupment of costs. The government body planning a tunnel project usually has substantial knowledge or the ability to acquire such knowledge about the nature of the latent conditions and risks involved at the site. Moreover, the government usually has greater ability and time to conduct site exploration and investigation than does a contractor who must confine its inspection to a brief pre-bid period. The government body planning a tunnel

172 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-21.qxd

2/18/04

12:29 PM

Page 173

project also has the ability to make a decision about the extent of investigation desired versus the amount of risk it wants to assume for unforeseen conditions. The public owner can then budget for the contingencies just as the contractor did at the common law. With the differing site conditions clause, the public owner has an incentive to share its knowledge about latent conditions with the contractor so that it gets the best and most accurate bids from the contractors. Under the common law, however, the owner has little incentive to be candid concerning the site because he would be rewarded by a contractor’s inaccurate low bid. The typical differing site conditions clause does not eradicate all problems associated with site investigations. Since a contractor performing tunnel work must still conduct some reasonable site inspection and can only make claims if the conditions materially vary from those typically encountered on such projects, public bodies still face the potential for significant litigation over whether a contractor did do an adequate site investigation and whether the conditions encountered were indeed unusual. In the long run, however, the typical unforeseen conditions clause will benefit the public by allocating the risk of such conditions to the party best able to manage it, the owner. 1.2

Risk sharing versus risk allocation

Under the common law, the risk of unforeseen conditions in a tunnel contract was allocated to the contractor. The unforeseen conditions clause found in most tunnel contracts today, however, allocates the risk of unforeseen conditions back to the owner. One commentator has proposed that the risk be shared by making sure that the contractor is paid for the additional work performed but that he does not reap a profit from the unexpected condition. [See Beh, supra.] By eliminating recovery for any profit on differing site conditions work, especially the Type II kind where neither party can foresee that the condition would be different from the type of condition normally encountered, the contractor gives up the benefit it receives for an unforeseen condition (the profit) without suffering a loss, while the owner pays the cost of the unexpected condition without rewarding the contractor. Both parties bear part of the financial risk of the unforeseen condition. Although this proposal may reduce a contractor’s incentive to submit an inordinately low bid with the hope that he will make it up through change orders, it is perhaps naïve to believe that contractors will bid that much more accurately knowing that they will only be paid for the cost of doing their work and not any profit, since it is very difficult to determine exactly how contractors calculate their profit. Moreover, the owner might still litigate whether or not the condition is truly atypical, because the bulk of the costs paid are

for direct costs rather than any additional profit, and it might be in the owner’s best interest to challenge the contractor’s entitlement to the claim. On balance, however, it would be worthwhile to experiment with this proposal to see if it truly saves the public money. 2 DELAY DAMAGES Apart from the high cost of dealing with the extra work associated with encountering differing site conditions in tunnel projects, the public owner is often faced with significant delay claims from the contractor. These claims typically involve extended job site general condition costs (e.g., additional supervision, jobsite trailer, utilities, and other costs), acceleration costs in the form of overtime labor charges, unabsorbed home office overhead claims, and lost bonding capacity claims. Some public bodies have responded to these claims by including “no damages for delay” clauses in their contracts. Thus, apart from the direct costs associated with overcoming the unforeseen subsurface conditions, the public owner may require the contractor to give up his delay damage claim and settle for only a time extension. The “no damage for delay” clause has resulted in a significant amount of litigation throughout the country. [Bates & Rogers Const. Corp. v. North Shore Sanitary Dist., 92 Ill. App. 3d 90, 414 N.E.2d 1274 (1981) (enforcing the clause); Corrino Civettes Const. Corp. v. City of New York, 67 N.Y.2d 297, 502 N.Y.S.2d 681, 493 S.E.2d 905 (1986) (finding an exception to the clause).] Although most courts do enforce the clause, others have not. A more reasonable approach is to reimburse the contractor for his direct jobsite overhead costs associated with the delay and to deny recovery for any of the more subjective or “soft” type of delay damages. If a contractor is delayed by additional work ordered by the owner or he encounters unforeseen subsurface conditions, the contractor should be able to recover what he can demonstrate as his extended job site general conditions. Some owners go one step further and agree up front what the daily overhead rate will be in the event the contractor is delayed by an unforeseen condition. The owner must be sure that he does not pay a higher daily rate than what he would pay if the contractor were forced to prove what his actual jobsite general condition costs were at the time of the delay. For example, a daily rate agreed to up front might reflect higher jobsite general conditions incurred at the front end of a project versus the lower overhead generally incurred as the project is winding down. If the delay is encountered toward the end of the project, the owner might pay a higher daily rate than he would otherwise pay if the contractor were forced to produce

173 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-21.qxd

2/18/04

12:29 PM

Page 174

records showing his actual jobsite general conditions costs at the time of the delay. Conversely, the owner may wish to avoid disputes over the contractor’s actual cost and simply agree to a daily rate up front. The owner in a tunnel project will undoubtedly want to allocate the risk of contractor delay to the contractor and require him to pay his damages if the contractor fails to complete the project by the substantial completion date. The public owner must decide whether to apply a liquidated damages provision or simply make a claim for the actual damages sustained because of the contractor’s delay. Most contractors are unwilling to sign construction contracts which expose the contractor to unlimited damages for delay. In tunnel construction contracts, it is also difficult for the owner to prove its actual delay damages because such damages for a public body are often speculative. The following liquidated damage provision is therefore a good choice for addressing contractor delay in a tunnel project: “The parties agree, by initialing here where indicated (owner) ______ (contractor) _______ that it would be extremely difficult and impracticable under the presently known and anticipated facts and circumstances to ascertain and fix the actual damages the owner would incur should contractor delay in achieving substantial completion by the date set forth hereof. Accordingly, the parties agree that if the contractor fails to achieve substantial completion within such time, then the owner’s exclusive remedy for such failure shall be to recover from the contractor the sum of $_______ for each calendar day substantial completion is so delayed by contractor.” The owner must insure that the amount of the daily rate for liquidated damages reasonably compensates him for any delays while at the same time bearing a rational relationship to the amount of actual damages sustained lest it be deemed a penalty and hence unenforceable. 3 SITE ACCIDENTS Perhaps one of the greatest risks on any tunnel project is the risk of worker injury or death. Most standard form construction contracts, including public contracts for tunnel projects, require the contractor to be responsible for means and methods, including all safety programs on site. The problem arises as to whether or not other parties involved in the construction process might be responsible for site safety accidents. Specifically, the design engineer might be responsible if he plays a more active role in the day to day operations of the tunnel project. The problem becomes acute when injured workers are prohibited from suing

their employers because of workers compensation statutory immunity. The injured worker or his estate then sues the engineer and the owner. An additional threat of liability is the Occupational Safety and Health Act (“OSHA”). [29 U.S.C. §651, et. seq, (2002).] Although most contracts will usually provide immunity to design professionals against third-party claims, design professionals need to be careful about potential OSHA violations. In one famous case, the Occupational Safety and Health Review Commission (the “Commission”) held that CH2M Hill Central, Inc. (“Hill”) was liable for violations of federal construction standards that apply to employers “engaged on construction work.” [See CH2M Hill Central, Inc., 1997 USAACL Lexis 34 (No. 89-1712, April 21, 1997.] The case arose out of the deaths of three workers from a methane gas explosion while working on the tunnel project in Milwaukee referenced above. In 1977, the MMSD undertook a $22 billion construction program calling for eighty miles of sewer tunnels. The MMSD contracted with a variety of companies including Hill, which served as the lead engineering consulting firm for the project. Hill’s contract with the MMSD included a provision that stated that visits to the construction site by Hill would not relieve the contractor of its obligation for, among other things, “all safety precautions.” In May 1988, a subcontractor requested a clarification from Hill regarding whether certain types of electrical equipment were approved for excavation due to a methane gas build up. After discussing the request with the MMSD, Hill provided its subcontractor with further explanation regarding electrical equipment. An explosion resulted when the electrical equipment ignited the methane. The OSHA commissioner issued citations to Hill as a result of this accident claiming that Hill violated construction industry specific standards promulgated by the Commission. The Commission established a new test holding engineering and architectural firms liable for the OSHA construction industry standards where they (1) have broad responsibilities in relation to construction activities, including both contractual and de facto authority relating to the work of trade contractors, and (2) are directly and substantially engaged in activities that are integrally connected with safety issues, notwithstanding contract language expressly disclaiming safety. [Id at 51, 53.] Although the Seventh Circuit Court of Appeals ultimately found that Hill could not be held accountable for OSHA violations because Hill did not “exercise substantial supervising over actual construction,” it did acknowledge that a design professional could be liable under OSHA under different circumstances. [CH2M Hill, Inc. v. Herman, 192 F.3d 711, 724 (7th Cir. 1999).] This again underlines the importance of

174 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-21.qxd

2/18/04

12:29 PM

Page 175

careful contract drafting to allocate the risk of site safety to the contractor and encourage design professionals to avoid any involvement in construction safety to avoid liability under OSHA standards. One way to insure that the contractor will cover and defend against such claims is to require the contractor to sign a broad form indemnification agreement in the contract. An example of a broad form indemnity that would cover such claims is as follows: “To the fullest extent permitted by law, the contractor shall indemnify, defend and hold harmless the owner, the owner’s representative, and the engineer, and their agents and employees, from and against all claims, liens, damages, losses and expenses, including reasonable attorneys’ fees, attorneys and expert fees, arising out of or resulting from the performance of the work, the existence of this contract, or the presence on the jobsite of the contractor, its subcontractors, their agents, servants or employees, or the ownership or possession of any property or material by said parties on the jobsite, regardless of whether or not it is caused in part by any of the parties indemnified hereunder. In all claims against the owner or the engineer or any of their agents or employees by any employee of the contractor, any subcontractor, anyone directly or indirectly employed by any of them, or anyone for whose acts any of them may be liable, the indemnification obligation hereunder shall not be limited in any way by any limitation on the amount or type of damages, compensation or benefits payable by or for the contractor, or any subcontractor under workers or workman’s compensation acts, disability benefits, or other employee benefits acts.” While such an indemnification clause (and the clear assignment of site safety responsibility in the contract) will not eliminate worker injury claims against the owner and his engineer, it should reduce them and provide the owner and his engineer with strong arguments that the contractor is responsible for defending against such claims. 4 PROJECT INSURANCE One way to control risks is to insure against them. The owner will typically require the contractor to carry comprehensive general liability insurance protecting the contractor, his employees, his subcontractors, the public, the interests of the owner and engineer against bodily injury and property damage. The policy should also cover contractual, strict and negligence type liability. It is also a good idea to require a umbrella excess liability policy. The general liability policy limits will of course be related to the size of the project. In most tunnel

projects, the limits should be at least $10 million, if not more. The typical general liability policy will exclude pollution, faulty workmanship, damages to selfperformed work, and professional services. The owner will usually provide the builders risk insurance coverage for the work included in the contract documents. This policy covers the property of the owner, and the liability of the owner for property of others, consisting of all real property in the course of construction, alteration or repair by the owner. The contractor must also purchase worker’s compensation and employers’ liability insurance for the statutory limits prescribed. On large tunnel construction projects, it may indeed be best for the owner to purchase a so called wrap-up owner controlled insurance program. The owner can cover all firms working at the jobsite and purchase workers compensation, general and umbrella liability, professional liability and builders risk. These types of insurance programs can also cover professional design services. Given the complexity of tunnel construction projects and the enormous risk of claims, it makes sense to implement such a wrap-up policy to minimize the potential litigation. If a single insurance company covers all types of claims that might arise on a tunnel project, it is more likely that the claim will be paid by the insurance company rather than result in endless litigation with the parties involved in the claim. 5 PROJECT DELIVERY METHOD Most tunnel projects will follow the traditional method of project delivery. The public owner retains an engineering firm to design the tunnel and then prepares bid documents for sealed competitive bidding. The lowest responsible bidder is selected and then the general contractor builds the project, hopefully in accordance with the contract documents. Many public owners believe that this is the “only” way to build a tunnel project since the engineer will or should be loyal to the owner’s interest in seeing to it that the contractor delivers the project on time, within budget and in accordance with the contract documents. Some government bodies might be prohibited from using alternative project delivery methods, but more states are permitting alternative project delivery methods as long as competitive bidding is used. Unfortunately, today’s design community is risk averse due to the uncertainty of subsurface construction and will often be at odds with its owner/client should the contractor make a claim that the contract documents contained errors and omissions. The owner must often defend its engineer in order to resist a claim by its contractor, only later to determine that the

175 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-21.qxd

2/18/04

12:29 PM

Page 176

contractor’s claim had merit and be forced to litigate with its engineer. Under the Spearin Doctrine [United States v. Spearin, 248 U.S. 132, 39 S.Ct. 59, 63 L. Ed. 166 (1918)], the owner impliedly warrants the sufficiency of the plans and specifications. An owner is responsible to the contractor for errors and omissions in the plans and specifications even if those errors and omissions do not constitute negligence. Thus, the public owner engaged in a tunnel construction project can find that it is responsible to its contractor for plan errors, but otherwise unable to recover from its design engineer unless he proves that the design engineer was negligent. Negligence is often difficult to prove since the owner is required to get another engineer to testify that the design engineer’s performance was below the standard of care for engineering work performed under like or similar conditions. One way to minimize this risk is to hire a design/builder to design and build the tunnel. Under the design/build project delivery method, the owner hires a design/builder for single point responsibility for the design and construction of the project. Unless the design/builder can show that the owner changed the scope of the project, the design/builder must absorb the cost of any errors and omissions in the plans and specifications. Thus, the design/build project delivery method may indeed reduce litigation over design errors.

the owner since the owner is usually in the best position to evaluate the risk and absorb the costs of such unforeseen conditions. A modified proposal of risk sharing, which deprives the contractor of his profit for Type II unforeseen conditions, may be reasonable. It is also fair to share the risk of owner-caused delays by limiting contractors to the recovery of their jobsite general conditions. Owners should also be limited to the recovery of liquidated as opposed to actual damages for contractor delays on the project. The contractor is clearly in the best position to control and absorb the risk of site injuries or death. An owner is well served to require the contractor to sign a broad form indemnification provision which protects not only the owner but his design professional from site safety claims. The wise public owner would do well to purchase an owner-controlled insurance policy that covers all of the parties involved in the tunnel construction project, including the design professionals, so as to minimize the potential for litigation over claims. Finally, although unusual, the public owner embarking on a tunnel construction project should consider whether to use a design/build project delivery method as opposed to the traditional design, bid, build method. By assigning single point responsibility to the design/builder, the owner may protect itself against a significant number of design error claims by the contractor.

6 CONCLUSION A contract that fairly allocates risk will necessarily minimize the risk of litigation. The risk of unforeseen subsurface conditions should typically be allocated to

176 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-22.qxd

2/18/04

12:30 PM

Page 177

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Getting back on-track: Exchange Place Station Improvements M.F. McNeilly Golder Associates Inc., Newark, New Jersey, USA

S.A. Leifer & G.F. Slattery The Port Authority of NY & NJ, Newark and Jersey City, New Jersey, USA

ABSTRACT: The September 11, 2001 attacks on NYC tragically resulted in the collapse of the World Trade Center (WTC) twin towers. In addition to the horrific lose of life and property damage, the collapse of the towers also destroyed the Port Authority Trans-Hudson (PATH) commuter rail station located within the foot print of the WTC site, flooded the two connecting Hudson River tunnels, and closed the Exchange Place Station in New Jersey. Immediately following these events, the Port Authority of NY & NJ (PANYNJ) recognized that restoration of PATH service to lower Manhattan would be vital to recovery efforts and the regional economy, and set in motion plans to rebuild the WTC Station and restore service by November 2003. This paper focuses on the Exchange Place Station Improvements Project, which required design and construction of underground rock excavations with spans upwards of 18 m (60 ft), rock cover as low as 7.4 m (25 ft), and buildings overlying planned excavations. This project also represents the first use of road-header type excavation equipment and the first application of steel fiber reinforced shotcrete (SFRS) linings for ground support on a mass-transit project in the NYC Metropolitan Region.

1 INTRODUCTION As a result of the September 11, 2001 terror attacks on the WTC towers, nearly 2,800 lives were lost and approximately 2.79 million m2 (30 million ft2) of commercial office space was either damaged or destroyed. The collapse of the towers also destroyed the Port Authority Trans-Hudson (PATH) commuter rail station, which occupied the lowermost levels of the WTC site, flooded the twin Hudson River tunnels carrying PATH’s downtown service, and forced the closing of the Exchange Place Station, the next station on the line. Loss of commuter rail service between the WTC and exchange Place Stations severed a vital transportation link between New Jersey and lower Manhattan affecting nearly 67,000 daily commuters. Commuters that previously disembarked at the WTC Station were forced to seek alternate routes including new/additional downtown ferry services and/or train/bus services to midtown Manhattan with connections to the NYC subway system. Immediately following the September 11, 2001 terror attacks, PANYNJ recognized that restoration of PATH service to lower Manhattan would be vital to recovery efforts and the regional economy, and needed to be completed as-soon-as-possible. Hence, PANYNJ

approved an ambitious program to restore PATH service to both the Exchange Place Station and the former WTC Station by July 2003 and November 2003, respectively. To achieve these fixed completion dates, it was apparent, from the beginning, that design and construction activities would have to be “fast-tracked”, and required the formation of somewhat unconventional partnerships between PANYNJ, contractors and numerous design consultants to expedite the project and mitigate project risks. This paper focuses on improvements to the Exchange Place Station, which required design and construction of underground rock excavations with spans upwards of 18 m (60 ft), rock cover as low as 7.4 m (25 ft), and buildings directly overlying planned excavation limits. This project also represents the first use of road-header type excavation equipment and the first application of steel fiber reinforced shotcrete (SFRS) linings for ground support on a mass-transit project in the NYC Metropolitan Region. 2 STATION IMPROVEMENTS Following the events of September 11, 2001, the Exchange Place Station was forced “out-of-service”

177 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-22.qxd

2/18/04

12:30 PM

Page 178

Figure 1. Existing tunnels and crossover tunnel plan.

because its configuration was originally designed to accommodate only “through” station and not “terminal” station traffic conditions. Hence, PANYNJ elected to construct a series of six (6) new track crossovers between the five (5) existing tunnels west of the station. In addition, PANYNJ elected to extend the station platforms about 46 m (150 ft) west to accommodate its planned 10-car capacity expansion program. See Figure 1.

3 HISTORY OF STATION Construction of the PATH system, by the Hudson Tunnel Railroad Company, started in 1874. Train service commenced in 1908, but the entire system was not operational until 1911. Exchange Place Station is located in Jersey City, New Jersey on the banks of the Hudson River, and is the first station in New Jersey on PATH’s downtown service to lower Manhattan. Originally, the twin Hudson River tubes comprising PATH’s downtown service connected the Exchange Place Station to the Hudson Terminal Station located beneath the former Hudson Terminal Building in lower Manhattan. However, the Hudson Terminal Building was demolished during the 1960s to make way for the WTC development, and the Hudson Terminal Station was abandoned in favor of the former WTC Station. Exchange Place Station and its connecting tunnels were constructed using drill-and-blast techniques, and recent field observations indicate that unsupported excavations were the preferred method of construction. Upon completion of the original rock excavations, the station and its connecting tunnels were lined with unreinforced concrete of variable thickness. See Figure 2.

Figure 2. Tunnels F to H to L Crossovers Circa 1907.

In general, the station and its connecting tunnel structures remained unaltered since commencing operations during the early 1900s. However, the station did undergo a major modernization sometime around 1986 to add/upgrade a new head house, ventilation towers, escalators and elevators from street-to-station levels. 4 CHALLENGES To achieve the project’s fixed completion date of July 2003, it was apparent that design and construction efforts had to be “fast-tracked” to re-open the station on-schedule, as-promised. This meant that tunnel excavation work had to commence by April 2002 and be completed by November 2002 to allow the other design disciplines adequate time to complete their parts of the project. This required PANYNJ to solicit and procure construction contracts prior to finalizing the project design.

178 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-22.qxd

2/18/04

12:30 PM

Page 179

Therefore, PANYNJ elected to implement a unique, unconventional contracting strategy to balance its associated project risks with the Contractors by putting the entire scope-of-work for Phase I of its Downtown Restoration Program (DRP) including the WTC temporary station, restoration of the two Hudson River tunnels, as well as the Exchange Place work under a single “net-cost-plus-fixed-fee” Contract. Contractors bidding on this contract were only provided general overviews of the scope-of-work, but were apprised of the project’s mandated schedule milestones. On February 1, 2002, this contract was awarded to a joint venture of Yonkers Contracting Co., Inc. of Yonkers, New York; Tully Construction Co., Inc. of Flushing, New York; and A.J. Pegno Construction Corp. of College Point, New York. Design and construction of the Exchange Place Station Improvements Project had to overcome the following technical challenges:

developed at these review meetings is as follows:

•

Based on the findings and recommendations, tunnel excavation drawings and technical specifications were generated and issued to the Contractor. In addition, these documents were revised, modified, expanded and re-issued to the Contractor, as necessary, under a series of design bulletins, as design and construction efforts advanced.

• • • • • •

Excavation of large underground rock caverns with spans upwards of 18 m (60 ft); Shallow rock cover as low as 7.4 m (25 ft); Localized areas of poor quality rock; Lower than expected rock mass strengths; Presence of multi-story buildings adjacent to and directly above planned excavation limits; narrowness of existing tunnel structures; and Limited/restricted site access.

5 PROJECT APPROACH From the start, the project’s highest priority was placed on development of construction alternatives and generation of Contract Documents. To this end, PANYNJ divided the project into nine (9) separate work order packages, based on the various design disciplines, and accepted an approach whereby documents were prepared and issued to the Contractor based on preliminary level designs with final designs advanced and completed concurrent with construction. In particular, the project’s requisite tunnel excavation activities fell under Work Order Package #3-EP. Contract Documents for this package were generated based on a series of geotechnical engineering design review meetings, which were convened to discuss and establish acceptable design and construction concepts based on a collective pooling of knowledge, experience and precedent. These design review meetings focused on ground support requirements, sequences of construction, methods of excavation and removal of muck materials, and alternate methods of site access. Consensus opinions among the meeting participants were formed and preliminary design recommendation established. A summary of the findings and preliminary recommendations

• • • • • •

6 DESIGN INVESTIGATIONS As part of the project’s design efforts, a series of geotechnical engineering design investigations were undertaken and completed to define subsurface soil and rock characteristics, establish relevant material properties and determine thicknesses and strengths of existing concrete tunnel linings. To define subsurface conditions, thirty-eight (38) vertical and/or inclined borings were drilled from both street-to-tunnel and tunnel-to-street levels. These borings were drilled immediately adjacent to and/or between the existing tunnels, and rock cores were collected using orientated and non-orientated drilling techniques. In addition, subsurface rock mapping was undertaken within each excavation heading, as tunnel excavations advanced. To determine thicknesses of existing concrete tunnel linings, one-hundred-two (102) cores were drilled through both tunnel sidewalls and crowns, and the concrete cores were collected for further strength testing. In addition, each corehole was videoed over its entire length. To establish relevant material properties, a laboratory testing program was conducted, including twentyfour (24) unconfined compressive strength (UCS) tests and sixteen (16) direct shear tests on selected rock core samples. In addition, sixteen (16) UCS tests were completed on selected concrete liner core samples.

179 Copyright © 2004 Taylor & Francis Group plc, London, UK

Excavation and support of the new crossovers would be feasible and constructible, but it was acknowledged that highly controlled excavation procedures and sequences would be required; Long-term ground support could be effectively provided using a combination of shotcrete linings, lattice girders and rock bolts; Portions of abandoned tunnels must be infilled with concrete to reduce excavated span widths; Rock cover thicknesses must be maximized to the greatest extent possible; Narrowness of existing tunnels will limit the type, size and amount of equipment that could work in the tunnels; and Removal of excavated materials will be difficult and may control production rates.

09096-22.qxd

2/18/04

12:30 PM

Page 180

7 SUBSURFACE AND TUNNEL CONDITIONS

•

Based on the findings and observations from the completed geotechnical engineering design investigation, subsurface and existing tunnel conditions can be characterized as follows:

•

• • • • • •

•

Concrete liner sidewall thicknesses varied from 0.3 to 1.5 m (1 to 5 ft), and were generally found to be “tight”; Concrete liner thicknesses at tunnel crowns varied from 0.25 to 0.5 m (0.75 to 1.5 ft), and voids upwards of 0.25 m (9 in) were observed; Overburden soils consisted of man-made fill, siltyclay, sand, silt and gravel deposits with a total overburden thicknesses that ranged from 4.8 to 9.1 m (16 to 30 ft); Groundwater was observed in the overburden at a depth of 2.5 to 3.5 m (8 to 12 feet) below ground surface; Manhattan Schist bedrock underlies the site. Average rock cover thicknesses were about 9 m (30 ft); Rock mass foliation was the observed dominant structural feature. Foliation dip angles ranged from sub-horizontal (less than 5 degrees) to 40 degrees, and dip directions varied from northeast to northwest; and Steeply dipping (vertical and sub-vertical) joints were uncommon and no measurable dip angles or directions were recorded from the collected rock cores. However, a series of steep joints were observed and mapped at tunnel level, as tunnel excavations advanced.

8 MATERIAL PROPERTIES 8.1

Overburden

Overburden materials were assumed to be cohesionless with the following properties: (a) friction angle of 30 degrees; (b) unit weight of 1,600 to 1,900 kg/m3 (100 to 120 pcf); (c) Poisson ratio of 0.3; and (d) modulus of elasticity of 150 MPa (21.7 ksi). 8.2

Bedrock

Manhattan Schist was encountered in all core borings, and consisted of light grey to dark grey, banded gneiss, schistose gneiss and/or schist intruded by pegmatite sills and dikes scattered throughout the rock mass. Material properties were established based on inspection of collected cores, geomechanical logging and laboratory testing. Relevant design values are as follows:

•

Total Core Recoveries (TCR) ranged from 50 to 100%, average 98%;

• • •

• •

8.3

Existing tunnel linings

Existing unreinforced concrete liners were assumed to have the following properties: (a) unit weight of 2,200 to 2,400 kg/m3 (140 to 150 pcf); (b) UCS of 21 to 31 MPa (3 to 4.5 ksi); and (c) modulus of elasticity of 2,068 to 2,758 MPa (300 to 400 ksi). Material properties for new concrete backfills used in the new crossover tunnel design were developed using American Concrete Institute (ACI) empirical equations. 8.4

Proposed ground support

Stability analyses incorporated galvanized, No. 9, Grade 75, resin grouted rock bolts, and all rock bolts were prestressed to 40% of the bar yield strength. Rock bolt lengths and spacing varied from 2.4 to 4.6 m (8 to 15 ft) and 1.5 to 1.2 m (5 to 4 ft), respectively, as excavated span lengths increased from 9 to 18 m (30 to 60 ft). In addition, new tunnel linings consisting of 34.5 MPa (5 ksi), high early strength steel fiber reinforced shotcrete (SFRS) having a thickness of 0.15 to 0.28 m (6 to 11 in) were constructed. Pre-fabricated steel lattice girders spaced on 1.5 m (5 ft) centers were also embedded within the shotcrete linings of excavated spans greater than 9 m (30 ft). Stability analyses assumed unit shear strengths of 1.5 MPa (219 psi) for SFRS materials, which was

180 Copyright © 2004 Taylor & Francis Group plc, London, UK

Rock Quality Designation (RQD) values ranged from 0 to 100%, average 86%, but typical RQD values varied between 50 and 100%; Rock mass unit weight, UCS and modulus of elasticity values ranged between 2,403 and 2,723 kg/m3 (150 and 170 pcf), 13.8 and 34.5 MPa (2 and 5 ksi) and 1,379 and 2,068 MPa (300 and 400 ksi), respectively; Rock Mass Ratings (RMR), (Bieniawski, 1976), ranged between 33 to 64; Rock Tunneling Quality Indices, Q-ratings, (Barton et al., 1974; Bieniawski, 1989) ranged between 2.7 and 15; Rock mass strength parameters were derived using the Hoek-Brown strength criteria (Hoek et al., 1998), and design values were as follows: a) “m” ranged between 1.31 and 3.95; and b) “s” ranged from 0.0018 and 0.0056; Spacing of foliation joints ranged between 0.2 and 0.6 m (0.5 and 2 ft), and these joints were assumed to have zero cohesion and friction angles of 20 to 23 degrees; and Spacing of steep (vertical, sub-vertical) joints crossing foliation were assumed to be 3 m (10 ft) or greater, and these joints were assumed to have zero cohesion and friction angles of 50 degrees.

09096-22.qxd

2/18/04

12:30 PM

Page 181

derived from correlations relating compressive and shear strength for plain and fiber reinforced concrete and shotcrete (ACI, 1984 and 1988; Fernandez et al., 1979; Mahar et al., 1975). These analyses also assumed composite unit shear strength of SFRS with embedded lattice girders equal to 2.7 MPa (391 psi), which was established in a similar manner except steel cross sectional areas were added to the SFRS cross-section. 9 DESIGN ANALYSES AND EVALUATIONS Design efforts included a combination of historical precedent evaluations and analyses to assess stress conditions around existing and new tunnel structures and evaluate the effectiveness of proposed rock reinforcement elements in two-dimensions for selected design cross-sections. 9.1

Historical precedent evaluations

Design of large underground rock excavations can be based largely on precedent by extending prior experience and construction observations from similar case histories to assess expected performances of specified ground supports. However, it should be acknowledged that reliance on precedent data must be used with caution. One such design reference (Cording et al., 1971) provides a compilation of case histories for underground rock excavations, and this reference presents relationships between excavation sizes (span and height) and installed rock bolt lengths and support pressures. Using this reference, rock bolt lengths and support pressures to excavated span ratios of 0.2 to 0.3 were found to be reasonable when compared to the case history data. Therefore, rock support pressures of 0.07 to 0.14 MPa (10 to 20 psi) and rock bolt lengths of 3.6 to 4.6 m (12 to 15 ft) were selected for spans of 15 to 18 m (50 to 60 ft). 9.2

UNWEDGE analyses

The computer program UNWEDGE (Rocscience, 2002, Version 2.37) was used to evaluate and define ground support requirements for each new crossover tunnel. This program provided three-dimensional visualizations of kinematically capable rock wedges, and calculated design factors-of-safety assuming the rock wedges are infinitely stiff, homogeneous masses acted upon by gravity, friction and applied internal support. Calculations were performed considering rock mass discontinuity and tunnel orientations for each new crossover, and results were used to evaluate required rock bolt lengths and spacing to achieve specified short-term (construction) and long-term

design criteria. In addition, sensitivity analyses were conducted to assess relative importance of variations in design parameters. Based on these UNWEDGE results, it was determined that calculated factors-of-safety are sensitive to excavated span lengths, strongly controlled by presence and location of steeply dipping discontinuities relative to excavation orientations and sensitive to applied hydrostatic pressures. In addition, it was determined that short-term (construction) design conditions were more critical than long-term design conditions. 9.3

The two-dimensional software program Phase2 (Rocscience, 2002, Version 5.0) was used to evaluate states-of-stress and stress field changes around existing and new crossover tunnels by modeling the rock mass as a continuum. The purpose of these models was to evaluate the potential for stress induced failures in both excavated crown and rib pillars. In addition, these models were used to assess stress changes within existing tunnel concrete liners. The modeling approach for these Phase2 stress analyses consisted of developing models to approximate state-of-stress conditions prior to excavation of the new crossover tunnels. After these initial conditions were developed, sequential excavations and/or construction of new concrete backfills were introduced into the model to simulate planned construction sequences, and rock bolts were added in areas requiring ground support. Results from these Phase2 stress analyses indicate that imposed incremental stress changes, due to the project’s crossover tunnel excavations, are relatively small when compared to in-situ rock strengths. In addition, compressive stresses that develop in adjacent rib pillars are moderate, but are well supported by adjacent concrete backfills and/or excavation support. 9.4

UDEC modeling

The two-dimensional computer code UDEC (Itasca, 1998, Version 3.0) was used on select design cross-sections to evaluate the stability of large excavated spans with shallow rock cover in a highly jointed and foliated rock mass. Conditions evaluated by these UDEC models consisted of the development of a 17 m (55 ft) wide span. Rock mass jointing consisted of foliation partings spaced 0.6 m (2 ft) apart and steep (vertical to subvertical) joints across foliation spaced 3 to 4.5 m (10 to 15 ft) apart. Based on these UDEC results, it was determined that the new crossover tunnels will be stable and the recommended rock bolt support will provide adequate short-term (construction) support.

181 Copyright © 2004 Taylor & Francis Group plc, London, UK

Phase2 modeling

09096-22.qxd

2/18/04

12:30 PM

Page 182

ROCKBOLTS INSTALLED AS EXCAVATIONS ADVANCE

PRE- SUPPORT ROCKBOLTS

SFRS LINING WITH EMBEDED LATTICE GIRDRS EXISTING CONCRETE LINING

Figure 3. Recommended ground support system.

9.5

Recommended ground support

Based on the results and findings of completed precedent evaluations, numerical modeling and stability analyses, the project’s recommended ground support included a combination of prestressed rock bolts, steel lattice girders and SFRS linings. See Figure 3 for typical design section showing recommended ground support.

Figure 4. Typical tunnel conditions and rock drill mounted on bobcat excavator.

10 CONSTRUCTION OBSERVATIONS Large projects such as the Exchange Place Improvements Project require integrated approaches to design and construction for them to be successful, and these efforts are complex enough when Contract Documents are prepared based on final designs. Given the unique nature of this project, it was apparent that unconventional methods were necessary to overcome what appeared to be “Herculean” schedule constraints. 10.1

Project coordination

To achieve the mandated scheduled milestones, an elevated, heightened degree of communication was required between the various parties. As the project progressed, all parties worked well together to achieve a common goal. As design efforts advanced, the need to modify previously issued Contract Documents became apparent and revised documents were re-issued under a series of design bulletins. The Contractor was made aware of each design bulletin before issuance, so its input could be incorporated to expedite construction. 10.2

Confined tunnel conditions

Inside clear dimensions of the existing tunnels were on the order of 4.3 m (14 ft), which limited the Contractor’s ability to utilize standard construction equipment. Hence, the Contractor was forced to think creatively, and adapt/modify available equipment to complete its work. See Figure 4 for photo of tunnel conditions and equipment used to install the specified ground support elements.

Figure 5. AM-50 road-header on track “F”.

10.3

182 Copyright © 2004 Taylor & Francis Group plc, London, UK

Tunnel excavations

At the project’s onset, it was envisioned that planned tunnel excavation activities would be undertaken and completed using controlled rock blasting techniques. However, the Contractor was unable to adequately control excavated perimeters, which resulted in excessive rock over-breakage and potential schedule slippages. Therefore, alternate methods of rock excavation were investigated, and road-header type equipment was considered to minimize rock over-breakage and accelerate the project’s tunnel excavation activities. See Figure 5. It should be noted that road-headers had never been used to excavate the local Manhattan Schist bedrock before consideration on this project. In addition, the Contractor and other industry professionals familiar with the project expressed reservations that roadheaders might not be well suited to excavate the rock.

09096-22.qxd

2/18/04

12:30 PM

Page 183

However, laboratory test results on collected rock cores indicated that intact rock UCS values typically ranged from 27.6 to 34.5 MPa (4 to 5 ksi). In addition, manufactures’ data for medium-sized road-headers, which could fit inside the existing tunnels, indicated that this equipment should be capable of excavating rock with UCS strengths in the range of 50 to 75 MPa (7 to 11 ksi). Therefore, PANYNJ and the Contractor elected to undertake a pilot program to assess the suitability of such equipment. This pilot program determined that road-header type equipment could excavate the local Manhattan schist rock. However, the machine used in the pilot program was found to be too small and experienced frequent breakdowns. Hence, the Contractor initially acquired a heavier ABM-330 Alpine Miner to start production excavation work, and later acquired two (2) additional VoestAlpine road-headers (AM-50 and AM-75) to complete the required excavations. As excavations advanced, the various road-headers used to complete the required tunnel excavations were found to be quite effective at excavating the rock, and production rates increased substantially over the previous drill-and-blast methodology, and the Contractor started to recover time on its schedule. Rock excavation rates were no longer hyper-critical to the project’s schedule. However, these roadheaders created new, additional challenges that had to be overcome, such as: (a) ventilation and water control issues; (b) greater volume of muck being generated than could be removed; and (c) need for closely controlled and carefully coordinated material handling sequences and procedures to reach the uppermost cavern crown limits. 10.4

Material handling and removal

The site was constrained by access from either end of the tunnels with the Contractor working simultaneously in five parallel tunnel structures. The Contractor was also confronted with PATH’s on-going commuter rail operations west of the site and the other two DRP projects (Hudson River Tunnel Rehabilitation and WTC Station Projects) east of the site. Consideration was given to constructing temporary access shafts from street-to-tunnel level to provided additional access to tunnel excavation work areas. In addition, combinations of belt and flight conveyor systems were considered to remove excavated muck materials through the existing station head house and onto barges in the Hudson River. However, both of these alternate access and material handling options were rejected in favor of more conventional rail haul methods utilizing only work trains operating out of a maintenance yard about 2.4 km (1.5 miles) west of the station.

Figure 6. Application of shotcrete.

This required coordination of work train movements with PATH’s on-going commuter rail operations, and necessitated restricting train movements to off-peak commuter hours. In addition, work trains could only use two of the five tunnels, which limited the number and frequency of work trains. 10.5

183 Copyright © 2004 Taylor & Francis Group plc, London, UK

Installation of ground support

Prior to start of tunnel excavation work, the Contractor installed pre-support rock bolts and completed contact grouting behind existing liners to provide continuous contact between the lining to remain and the rock. Presupport rock reinforcement consisted of 2.4 m (8 ft) long galvanized, No. 9, Grade 75, resin grouted rock bolts spaced on 1.5 m (5 ft) centers. The rock bolts were also pre-tensioned to 40% of the bar yield strength. As tunnel excavation advanced, the Contractor installed the specified rock reinforcement, which was similar to the pre-support rock bolts. Rock bolt lengths and spacing varied from 2.4 to 4.6 m (8 to 15 ft) and 1.5 to 1.2 m (5 to 4 ft), respectively, as excavated span lengths increased from 9 to 18 m (30 to 60 ft). This project included the first application of a steel fiber reinforced shotcrete (SFRS) tunnel liner system on a mass transit system project in the NYC metropolitan region. Nominal shotcrete thicknesses varied from 0.15 to 0.28 m (6 to 11 in), depending on excavated span lengths. In addition, SFRS linings were used in combination with pre-fabricated steel lattice girders spaced on 1.5 m (5 ft) centers and rock bolts installed perpendicular to the excavated rock face. SFRS linings were applied using “wet-mix” techniques, and steel fibers were incorporated into the mix design at the concrete batch plant to offset needs for welded wire fabric. This approach saved time during construction. SFRS materials were delivered using drop pipes from street-to-tunnel levels, and applied using nozzlemen and assistants operating from aerial man-lift equipment. See Figure 6.

09096-22.qxd

2/18/04

12:30 PM

Page 184

ACKNOWLEDGEMENTS The authors wish to acknowledge that the success of this project was not attributed to any one agency, company or individual, but rather the product of a combined partnership between the Owner, the Contractor and the numerous design consultants working on the project, and all parties involved should be proud to have assisted in re-opening of the Exchange Place Station on-schedule. The authors would also like to thank Bill Fetters and George Yoggi for providing selected images used in this paper.

REFERENCES Figure 7. Completed tracks F to H to L crossovers.

11 SUMMARY Design and construction of the Exchange Place Improvements Project required the development of six (6) new crossover tunnels having spans upwards of 18 m (60 ft) in width. This project also faced numerous contractual, technical and management challenges, which directly impacted the project’s overall critical path. Tunnel designs were advance and Contract Documents prepared on a “fast-tracked” basis (concurrent with construction), and these documents included controlled tunnel excavation sequences and procedures and installation details of pre-support and final ground support elements. To achieve the mandated schedule completion date, unique, unconventional partnerships were established between the PANYNJ, the Contractor and the various design consultants. This project was an example of how genuine partnership and cooperation between the various parties can accomplish a task some viewed as insurmountable. On June 29, 2003, the Exchange Place Station reopened to commuter service, and the new WTC Station is scheduled, as of this writing, to re-open in November 2003. Completing this project in two (2) years was a major accomplishment. See Figure 7 for view of completed tunnel crossovers.

ACI Committee 506, 1984, State-of-the-Art Report on Fiber Reinforced Shotcrete, ACI 506.1R-84, American Concrete Institute. ACI Committee 544, 1988, Design Considerations for Steel Fiber Reinforced Concrete, ACI Structural Journal, September–October, 1988, pp. 563–579. Barton, N.R., Lien, R., and Lunde, J., 1974, Engineering Classification of Rock Masses for the Design of Tunnel Support, Rock Mech., Vol. 6 No. 4, pp. 189–239. Bieniawski, Z.T., 1989, Engineering Rock Mass Classification, New York, John Wiley & Sons. Bieniawski, Z.T., 1976, The Geomechanics Classification in Rock Engineering Design, Proc. 4th Int. Congress on Rock Mech., ISRM Montreax, Vol. 2, pp. 41–48. Cording, E.J., Hendron, A.J., and Deere, D.U., 1971, Rock Engineering for Underground Caverns, Symposium on Underground Rock Chambers, ASCE. Fernandez, G.D., Cording, E.J., Mahar, J.W., and Van Sint Jan, M.L., 1979, Thin Shotcrete Linings in Loosening Rock, Rapid Excavation and Tunneling Conference, Vol. 1, pp. 790–813. Hoek, E., Kaiser, P.K., and Bawden, W.F., 1998, Support of Underground Excavations in Hard Rock, A.A. Balkema, Rotterdam. Itasca 1998, UDEC Users Manual, Itasca Consulting Group, Minnesota. Mahar, J.W., Parker, H.W., and Wuellner, W.W., 1975, Shotcrete Practice in Underground Construction Report No. FRA-OR&D 75–90, Washington, D.C., Federal Railroad Administration. Rocscience, 2002, UNWEDGE Users Manual, Rocscience, Inc., Toronto. Rocscience, 2002, Phase2 Users Manual, Rocscience, Inc., Toronto.

184 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-23.qxd

2/18/04

12:30 PM

Page 185

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Influence of geologic conditions on excavation methodology E.C. Wang, L.M. Hsia & C.C. Chang Parsons Brinckerhoff Quade & Douglas, Inc., New York

A.N. Shah MTA-NYCT, New York

ABSTRACT: The existing MTA-New York City Transit’s (MTA-NYCT) No. 7 Subway was built in the early 1900s and currently terminates at Times Square near 41st Street and Seventh Avenue. As part of the City’s redevelopment of the west side of Manhattan, the MTA-NYCT’s No. 7 Subway extension project will add approximately 1.5-miles of twin tunnel alignment to the current line. The proposed alignment will extend from the existing Times Square Station west beneath 41st Street then turn south under 11th Avenue and terminate between 24th and 25th Streets. The project will include two new stations, a two-track line station near 41st Street and 10th Avenue and a three-track terminal station at 34th Street and 11th Avenue. The alignment lies in close proximity to several major underground structures. The tunnels and station caverns will be constructed in the Manhattan Schist, which has been extensively intruded by pegmatitic and aplitic veins along and across the foliations. The lithology of the bedrock as well as the orientation and condition of the rock mass discontinuities are critical to the design of the tunnel boring machines. This paper will focus on the identification and influence of geology on the selection of excavation methodologies that minimize impact to the adjacent underground structures.

1 INTRODUCTION The proposed No. 7 Subway Line Extension is configured as a two-track subway extending approximately 1.5 miles of twin tunnel alignment, primarily in hard bedrock underlying New York City, westward from existing No. 7 subway tunnel at the Times Square Station between 7th and 8th Avenues below the Port Authority Bus Terminal Bus Ramp beneath 41st Street and turning with a 650-foot radius south to follow 11th Avenue. The alignment proceeds southward along 11th Avenue and terminates between 24th and 25th Streets. The alignment lies in close proximity to the following overlying structures: Eighth Avenue Subway, Lincoln Tunnel and Amtrak’s Empire Line and Hudson River Tunnels. Two future station locations are planned at 41st & 10th Avenue and 34th & 11th Avenue (see Figure 1). Subsurface geotechnical investigation is on-going to support the design and the EIS efforts. 2 GEOLOGIC SETTING The following represents the current working knowledge of local geology, combined with field observations

from investigation borings and subsequent review of recorded rock core samples. The project area forms a part of the Manhattan Prong of New England Upland Province. The hard crystalline metamorphic bedrocks form the ridges and valleys within this province. The project alignment mainly occupies the portion of western Manhattan along 41st Street from 8th Avenue to 11th Avenue and from 41st Street to 26th Street along 11th Avenue. A valley along 41st Street from 8th Avenue has been observed along the subsurface stream and further in the south between 30th and 27th Street, another valley is observed along another subsurface stream. The orientation of both these valleys is WNW-ESE.

2.1

185 Copyright © 2004 Taylor & Francis Group plc, London, UK

Overburden

The thickness of overburden deposits varies substantially. Based on the geotechnical investigation borings, relatively thin soil cover was encountered along 41st Street and generally increased both eastward toward 11th Avenue and southward towards 25th Street. The overburden consists of glacial and postglacial soils, combined with recent manmade fill.

09096-23.qxd

2/18/04

12:30 PM

Page 186

Figure 2. The Egbert L. Viele Map of 1874.

Figure 1. Site location plan.

Generally, the soils include glacial till, modified glacial drift, sand and gravels, some glacio-lacustrine silts and clays, and manmade land. The locations of old stream channels, exposed rock and marshland before the recent human intervention are recorded in the Egbert L. Viele Map of 1874 (see Figure 2). Review of these historic data reveals an apparent ancient creek which once flowed in a westerly direction, crossing 9th through 11th Avenues between 39th Street and 42nd Street, as well as between 29th and 32nd Streets; thus crossing and paralleling the future tunnel alignment. High water inflows are typically associated with the stream location; indicating zones of weakness in the rock mass (i.e., fault zones, joints, and fractures). 2.2

Bedrock geology

The geology of New York City is complex and has been studied and well documented in numerous publications. However, some of the observations made during the on-going subsurface investigation are noted herein. The crystalline rocks of New York City are divided into two major units separated by Cameron thrust, a regional NE-SW striking structural feature whose surface dips eastward. The rocks west of this line are called New York City Group or Manhattan Formation and the rocks east of this thrust fault are known as

Hartland Formation (Hutchinson River Group). The Cameron thrust fault extends from Connecticut through the east side of East River (between Roosevelt Island and Western Queens, through Staten Island and further south into central and southern New Jersey. This regional structural feature has also been called a “Suture”. The second prominent regional structural feature is the Manhattanville (125th Street) fault within Manhattan Formation striking WNW-ESE with low to moderate dip. A number of fractures, joints and faults (major as well as minor) have been observed along this orientation in the area. The project area comprises the youngest group of Manhattan Formation i.e., the Manhattan Schists. Although commonly referred to as schists, these crystalline schists vary in composition from quartzose schists, quartz-felspar schists, quartz-garnet mica schists, biotite schists, hornblende schists to muscovite schists. Numerous pre-, post- to late-kinematic (during Cameron thrust activity) pegmatitic intrusions of varying sizes (80 feet wide to a fraction of an inch) have been emplaced within these schists along and across the foliations and along fractures and joints in the area (see Photo 1). In certain places, mainly around midtown area of Manhattan, these intrusions have locally elevated the metamorphic grade and modified the texture of these schists which almost resemble aplitic, gneissic to granitic rocks. Serpentinite, talc and chlorite schists rocks were encountered in the borings between 27th and 28th

186 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-23.qxd

2/18/04

12:30 PM

Page 187

Photo 2. Deformation D3 – Regional foliation (S2) in Schist N10°E/S190°W. Minor folds shown by Schists.

Photo 1. Schist/pegmatite contact along foliation.

Street on 11th Avenue. These rocks seem to have been intruded along the fractures and fault planes during Late Taconic to Early Acadian orogeny. They are highly sheared and are seen as talcose schists, which display a distinct pulverized nature. Structurally, the rocks of Manhattan Formation have undergone multi deformational events. The imprints of these events can be seen from various discontinuities observed in the rock cores. The most prominent fold episode during the deformational event D3 has developed the main regional macroscopic antiformal (F2) structure in Manhattan Island, which has modified the cleavage/foliation (S1) generated by isoclinal recumbent folding (F1) during the second phase of deformation (D2). The orientations of cleavage/foliation (S2) during the deformational event D3 trend NNW to NE and dips at low to high angles to either W/NW to SW or E/NE to SE, according to the orientation of respective attitude of bed rocks due to the F2 folding. The orientations of cleavage/foliation and associated fractures and joints are classified according to deformational events which are described below: (a) D1 – the first foliation/bedding formed – (S0): orientation is obliterated. (b) D2 – the foliation/cleavage trends E/W to ENE/ WSW, highly obliterated, can only be seen near the fold cores (F1). (c) D3 – represents the main foliation/cleavage in the area trending NNW/SSE to NE/SW due to large

scale antiformal folding (F2), dipping either to the East or West (see Photo 2). (d) D4 – developed large scale shearing and fracturing due to the Cameron thrust faulting. Though this fault is not encountered in the area, the effect is seen in the rocks by shearing and chloritization (see Photo 3). (e) D5 – the rocks show moderate to steep plunging folds (F3). Foliations trend WNW ESE with the development of S3 cleavages (see Photo 4). These cleavages due to the extreme stresses resulted into slickensides, joints, fractures and faults on all scales (see Photo 5). This event marks the most important fault/fracture system in the area and is associated with the Manhattanville (125th Street) fault, which had been developed in this deformational phase (Middle Devonian–Acadian Orogeny). The resulting orientation and condition of these joint sets were evaluated with respect to the opening orientation and geometry using the Unwedge software program for stability analyses of potential wedge formations within the rock mass. These data are also being used in the final tunnel liner design. 2.3

Manhattan schist is generally not saturated, as groundwater is held in the open joints within a generally tight rock mass. Groundwater flow is controlled by the random interconnection of these more open joints as observed in deep excavation in and around the New York City. Groundwater levels measured in the soil

187 Copyright © 2004 Taylor & Francis Group plc, London, UK

Geohydrology

09096-23.qxd

2/18/04

12:30 PM

Page 188

Photo 5. Slickenside and stretched quartz veins along S2 & S3 cleavages.

Photo 3. Chloritization along foliation.

The network of fractures in bedrock will control the groundwater conditions in the rock mass. The permeability of the discontinuities will be influenced by several factors including the proximity of adjacent surfaces, alteration processes that have been removed or placed minerals on fracture surfaces, and joint wall material that has been fragmented or crushed by faulting or shearing. These geological processes can increase or decrease the permeability of individual joints. Water levels were measured in the observation wells installed in selected completed boreholes. The groundwater levels in the overburden do not appear to be sensitive to seasonal variation. Groundwater inflow during construction of running tunnel, station and crossover caverns, and construction shafts are being estimated using Packer test data. Photo 4. F3 folds in schists.

2.4 tend to follow the bedrock elevation, suggesting that they are probably perched on top of the bedrock, where rock forms a ridge and within the soil above the rock in bedrock valleys. The sources of groundwater recharge in Manhattan are leaking sewers, drains and water lines, and the adjacent East River and Hudson River. Though the recharge in the bedrock mass is unlikely to be from precipitation filtration due to the relatively impermeable nature of the city streets and buildings over land, there is evidence that the natural groundwater flow may be lateral following the old buried stream beds.

During the subsurface geotechnical investigation, which is still continuing, the following geologic features have been noted: (a) Between 25th and 29th Streets, there is a depression in bedrock surface such that the tunnel has minimum cover, and bedrock may be breached. (b) Between 25th and 29th Streets, there is an apparent shear zone, which contains serpentinite, talc schist and chlorite schist. (c) Talc schist and chlorite schist appears to be much weaker than the Manhattan schist.

188 Copyright © 2004 Taylor & Francis Group plc, London, UK

Distinct geologic features

09096-23.qxd

2/18/04

12:30 PM

Page 189

(d) Substantial soft compressible organic deposits underlying the fill encountered along 11th Avenue south of 28th Street.

3 EXISTING STRUCTURES IN THE VICINITY OF THE ALIGNMENT No. 7 Subway Line Extension construction impacts on the existing structures along the proposed tunnel alignment must be minimized. Nearby infrastructure includes NYCT structures, Port Authority Bus Terminal, Lincoln Tunnel East Approach tubes, Amtrak Empire Line and Hudson River Tunnels, the 11th Ave. Viaduct, the 8th Ave. Subway A, C, and E lines, as well as privately owned properties above or adjacent to the proposed alignment. The vertical separation from these underground structures was one of the important considerations in selecting the final tunnel alignment. Construction of the existing structures dates from the early turn of the 20th century to the 1950s, with expansions completed during the 1970s. The considerable amount of blasting that was necessary to excavate bedrock during construction of these overlying structures is expected to have altered the condition of the rock mass, and hence is being ascertained in the subsurface investigation program for the tunnel support design. The existing structures which may be impacted are as follows: (a) New York City Transit (NYCT) Facilities: 8th Ave. Subway A, C, and E lines – The active subway structure beneath 8th Avenue, consisting of an upper mezzanine level over four sets of tracks, which will be modified to accommodate the connection to the future tunnel extension. The proposed tunnels will be below these facilities. (b) The 41st Street Bus Ramp – This bus ramp is actually an underpass below 41st Street between Dyer Avenue and the west face of the Port Authority Bus Terminal Extension. This ramp provides underground access for buses entering the Terminal, and accommodates overhead ventilation system. (c) Lincoln Tunnel East Approach – This heavily traveled approach has three tunnels running in east-west direction along 38th and 39th Streets. Each tunnel tube has two traffic lanes. The proposed twin TBM tunnels will pass one tunnel diameter under these tubes. (d) Amtrak Empire Line – The Amtrak Empire Line originates from Penn Station, and continues as a tunnel extending to the west and southwest direction underneath 8th, 9th, 10th and 11th Avenues. At the proposed future 34th Street Station the tunnel turns abruptly north, crossing under 11th Avenue twice, and then proceeds beneath 35th

Street prior to day-lighting north of 37th Street. This 18-ft by 18-ft reinforced concrete box tunnel is founded on bedrock, and is above the proposed tunnel and the 34th Street station cavern. (e) Amtrak Hudson River Tunnel – The existing Hudson River Tunnel continues eastward from the Hudson River below Pier 62, underneath the Long Island Rail Road West Side Yards, and terminates at Penn Station. In the vicinity of 11th Avenue, the Hudson River Tunnel is consisted of a 19-ft span cast-in-place concrete twin tunnel with an elliptical brick arch. The proposed tunnel alignment is about two tunnel diameter above the proposed running tunnel, but is close to the proposed future 34th Station Street cavern. Although minimum one tunnel diameter separation was achievable for the running tunnel portion, such criteria will not be possible for the mined tunnel portion. Sections of the tunnel alignment where geology or nearby structures favor drill-and-blast excavation techniques will require specifications for controlled blasting methods to reduce overbreak and to minimize the vibration impact on existing structures. A comprehensive pre-construction condition assessment of all existing buildings and structures located within the influence zone of the new alignment will be conducted to establish baseline conditions. During construction geotechnical instrumentation will be installed to monitor ground movement at the streetlevel and within existing railroad and transit tunnels. The instrumentation data will be used to evaluate preconstruction assessments. 4 KEY ISSUES The key issues resulting from construction of the No. 7 Subway Line Extension project include the following. 4.1

The composition of ground (rock type) being excavated greatly affects the level of vibration from TBM’s. Mitigation of TBM vibration is generally through public information, reducing night-time operation or periods of long activity where possible. The mechanical cutting heads on Tunnel Boring Machines (TBM) typically generate ground vibration at frequencies between 20 to 80 Hz. However, occasionally, TBM’s have been reported to cause very low frequency motion (5 Hz) and it is believed that “machine-shaking” while mining through blocky or faulted ground caused the low-frequency ground motion. The intensity of low-frequency motion attenuates slower than motion occurring in higher frequencies, thus measured PPV values would be accordingly higher at the ground surface.

189 Copyright © 2004 Taylor & Francis Group plc, London, UK

Tunnel Boring Machines

09096-23.qxd

2/18/04

12:30 PM

Page 190

Vibration from operating TBM machines is influenced by distance, structural rock conditions, physical rock properties, ground water and depth of overburden. From the Chattahoochie Tunnel Project near Atlanta, vibration levels created by an 18.4-ft Robbins machine in granodiorite ranged from 0.02 to 0.08 in/sec. In this case the average tunnel depth was approx 200-ft below ground surface. When monitoring data was compared to overburden depths it was observed that PPV readings were inversely proportional to the depth of overburden. This may be explained by the well-known phenomenon whereby high-frequency attenuations occur more rapidly in loose deposits and energy loss due to reflection at the overburden-rock interface. Since the majority of Manhattan buildings are founded directly on bedrock and the depth of TBM tunnels are less than 200-ft, it may be assumed that the predicted PPV values will occur in the high-range of curves based on historical data. 4.2

Rock drilling

Hydraulic and pneumatic rock drills typically generate vibration occurring within a narrow band of frequencies ranging between 70 and 125 Hz. During quiet periods of the day, the resulting drill noise may sound like a hammer drill operating across the street. Throughout the alignment and at different times of the day, the ambient noise level will fluctuate; therefore it would be much more practical to limit groundborne noise increases to ambient levels plus 5 dBA. For the rock cover in Manhattan (generally 40 to 200ft), it is unlikely that ground-borne drill noise will increase ambient noise levels by more than 5 dBA. However, during quiet periods of the day it is likely that residents on ground floors or within basement areas will hear ground-borne drill noise. Also, despite occurring at very low intensity, this steady state noise will be “tonal”, meaning it occurs in a very narrow frequency spectrum (70 to 90 Hz). Public exposure to tonal noise generates more negative response compared to broadband noise. Inhabitants of buildings overlying the tunnels will hear and/or feel noise and vibration caused by TBM mining and rock drilling. Although, TBM induced vibrations are unlikely to damage structures; the persistent vibration and noise of TBM excavation may become a public annoyance, especially during quiet hours. 5 EXCAVATION METHODOLOGY The following description of potential excavation methodology along the alignment is based on the subsurface profile developed from the geotechnical

investigation program with the intent of minimizing construction impacts to overlying structures. Excavation of the running tunnels commencing from the Site A TBM Launch Shaft located along 11th Avenue between 25th and 26th Streets, and proceeding northward under 11th Avenue, turning eastward below 41st Street and terminating east of Ninth Avenue for retrieval from Shaft L located within the 10th Avenue Station footprint. Assembly and subsequent launching of the TBMs will require the following drill-and-blast structures: adit connecting off-line shaft to the tunnel alignment, assembly/launch chamber, back-shunt tunnel extending approximately 200ft south of the chamber, and TBM starter tunnels. Protection of nearby utilities along 11th Avenue may demand a rigid shaft excavation support system within overburden due to poor subsurface conditions. As such, Contract Specifications prohibit dewatering outside the excavation walls, in order to prevent reduction in pore water pressure, and initiating consolidation process in the compressible stratum underlying the fill. 5.1

Rock tunneling methods typically include: (a) Mechanical boring by Tunnel Boring Machine (TBM). (b) Drill-and-blast excavation/mining. (c) Mechanical excavation/ mining by machine (road header). All three tunneling options may be required to complete the tunnels, passages and caverns for the No. 7 Line Extension. 5.1.1 Mechanical boring by TBM Generally designed to perform all or most of the following functions: (a) Excavation of tunnel profile. (b) Control in the tunnel face. (c) Temporary support of the working area between the face and installed permanent lining by way of a steel shield usually around the full perimeter of the tunnel. (d) Installation of the permanent support and lining suitable for the final function of the tunnel. Three main types of TBMs, which are distinguished by the excavation and support method: (a) Open face machine – A road header or backhoe type excavator to excavate the face, which remains exposed and unsupported for long periods. Generally only used in soft rock or stable cohesive ground such as stiff clays and incorporate a shield to provide temporary support behind

190 Copyright © 2004 Taylor & Francis Group plc, London, UK

Tunneling methods in rock

09096-23.qxd

2/18/04

12:30 PM

Page 191

the face area. This type of machine is unsuited for the hard rock. (b) Full-face machine – Incorporates a full cutter-head, which cuts the full face area of the tunnel but does not necessarily provide support to the face. The cutter-head includes excavation systems of rotating discs or spokes with cutters. The excavated material falls through the face discs apertures to the inside of the cutter-head where it passes via chutes to the conveyor systems for removal. This TBM is designed for stable hard rock requiring minimal rock reinforcement. (c) Closed face machines – similar to full-face machines but possess the added capability to control and support the tunnel face. They are used to bore through weak, unstable and typically saturated ground. The Earth Pressure Balance Machine (EPBM) and Slurry TBM are examples of closed face machines. Of the above, a full-face hard rock TBM is expected to handle the majority of the main running tunnels excavation. The two main types of hard rock tunneling machines are identified by their means of propulsion:

•

•

Gripper-type TBM – such machines incorporate grippers, which are expanded against the tunnel rock walls at the rear of the machine providing the required reaction for the TBM cutter-head thrust at the face. Hence, the rock wall and the pillar between adjacent tunnels must be of sufficient strength to support the grippers. The gripper machines normally only feature a short slotted mechanical roof at the tunnel crown for protection of the operating crew. The slots within the mechanical roof provide exposure to the rock for rock bolt installation. Also, the machines may incorporate an erector system for placing segmental linings for permanent support. Typically however, they are independent from the permanent lining operation. In stable rock conditions, where tunnel lining is installed separately, impressive advance rates have been recorded. Shield-type TBM – Use a full shield to provide full perimeter support behind the cutter-head. Propulsion of the TBM is provided by rams thrusting off the last ring of the installed permanent segmental lining, which is placed under protection of the shield at the rear of the machine. Since the erection of the permanent segmental lining is required for the operation of these machines, construction rates are generally lower than the gripper-type machines. But, installation of the temporary or primary support is unnecessary for shield type machines. These machines are suited for a fractured hard rock environment, which might otherwise require considerable temporary support.

Selection of the most suitable machine will depend on the quality and fracture state of the rock mass, in addition to the type of permanent tunnel support selected. If competent rock possessing good stand-up time is encountered, then a gripper TBM will be preferred. Rock bolts would provide temporary support of tunnel walls, with optional shotcrete, as necessary. In this case, a cast-in-place (CIP) permanent lining could then be placed independent of the tunneling operations – either during tunneling or upon completion of the excavation. High progress rates would be expected under these conditions. For highly fractured rock mass a shield-type machine may be preferred to avoid installation of excessive temporary support. TBMs incorporating the segmental lining erector require a relatively long train of equipment. This requirement impacts the length of the TBM launch chamber required, as well as affecting the minimum turning radius of the machine. A larger construction laydown area may be required to accommodate stockpiles of liner segments near the shaft. 5.1.2 Drill-and-blast Drill-and-blast techniques are anticipated for cavern enlargements of initial twin TBM tunnel sections, as well as excavation of cross-passages and adits/ chambers associated with TBM launching and retrieval operations. Modern controlled blasting methods attempt to minimize blast vibration and noise. Reducing the blast disturbance of the surrounding ground is not only required to avoid disturbance to existing structures and building occupants within the immediate construction zone, but is also necessary for preserving the inherent strength of the surrounding ground. Blasting loosens the rock mass immediately surrounding the blast area. Excessive blast impact on the ground surrounding the tunnel could lead to loss of rock strength and reduction of stand-up time as well as excessive overbreak, resulting in increased support requirements to stabilize the underground opening. Smooth blasting techniques, line drilling, reduced excavation size and excavation round length, sequential excavation all provide means of limiting blast vibration and disturbance to adjacent ground and structures. 5.1.3 Mechanical excavator Hard rock applications typically restricted to fractured rock as well as circumstances, which preclude alternate techniques such as drill-and-blast. For extremely hard and very strong rock, in situations where blasting is not permitted expansive chemical products may be used to split rock between pre-drilled holes. Several types of mechanical excavators include: (a) Road header excavator – over the past decade, the cutting capacity of road headers have improved

191 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-23.qxd

2/18/04

12:30 PM

Page 192

substantially. However, the successful and economic performance of road headers in the strong Manhattan Schist will be governed by the degree of rock mass jointing and fracturing. Road headers may be used for sections of tunnel located close to the top of rock where weathering and loosened rock mass conditions are anticipated. Road headers may also be favored compared to blasting in sections of tunnel near to existing structures or utilities. In addition, the construction sequencing of the tunnels may require road header use to prevent damaging previously excavated tunnel sections. Road headers may be needed to mine the tunnels beneath the Port Authority Bus Terminal (PABT) underpass where relatively shallow rock cover exists. They may also provide an effective means of excavating sections of the relatively complex network of interlocking tunnels. (b) Hoe Rams – Hoe rams mounted on excavators will be restricted to relatively weak and/or highly fractured rock. Hoe rams will not be utilized for a full-scale tunnel excavation but may be favored for trimming tight spots and profiling work.

and its design of cutters, grippers, power requirement, and exploration equipment at heading. Serpentinite, talc and chlorite schist may further influence the design and construction due to their considerably distinct properties relative to the host rocks: Manhattan schist and pegmatite.

7 LIMITATIONS Any views or opinions presented in this paper are solely those of the authors and do not necessarily represent those of their companies, their employers or their subsidiaries. ACKNOWLEDGEMENTS The authors gratefully acknowledge P. McGrade, S. Singh, and M. Naik (MTA-NYCT) for their permission and D. Donatelli and P. Das (of PB Team) for their cooperation and support to publish this paper.

BIBLIOGRAPHY 6 CONCLUSION Geology is critical to the planning of the project tunnel alignment. The geologic conditions influence the planning and design process, from selection of alignment, to initial excavation support, tunnel final lining, through to the long-term operations. The subsurface conditions may be altered from the previous underground construction in the vicinity of the tunnel alignment. The lithology and the structural discontinuities as well as existing structures located within the tunneling influence zone will influence the selection of TBM

Baskerville, C.A., 1982. The foundation geology of New York City: Geological Society of America. Reviews in Engineering Geology. Vol. 5, pp. 95–117. Fluhr, T.W., 1941. The Geology of the Lincoln Tunnel Part 4: Journal of Rocks and Minerals Association. Vol. 16, No. 7, pp. 235–239. Isachsen, Y.W., 1980. Continental Collisions and Ancient Volcanoes, the Geology 9 of Southeastern New York: New York State Geological Survey, Educational leaflet, Vol. 24, pp. 1–15. Shah, A.N. et al. Geological Hazards in the Consideration of Design and Construction Activities of the New York City Area, Environmental and Engineering Geoscience, Vol. IV, No. 4, Winter1998, pp. 524–533.

192 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-24.qxd

2/18/04

12:31 PM

Page 193

Session 2, Track 3 Non-mechanized construction

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-24.qxd

2/18/04

12:31 PM

Page 195

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Santiago’s Metro expands C.H. Mercado Metro de Santiago, Santiago, Chile

G.S. Chamorro Empresa de Ingeneria Ingendesa S.A., Santiago, Chile

K. Egger Dr. G. Sauer Corporation, New York, United States

ABSTRACT: Metro de Santiago has planned an ambitious array of new projects for the period of 2000 through 2005, which will almost double the reach of Santiago’s subway by late 2005 and bring service to an additional 1.8 million people. Metro is currently operating three rail lines with a total length of about 40 km, of which 23 km are below grade. New projects include the extension of two existing lines and construction of two new lines, which will increase the number of stations from currently 48 to 80. The subway operated by Metro S.A. is the backbone of the Urban Transportation Plan for Santiago (UTPS) which includes the extension of Line 2 to the north and to the south and Line 5 to the west, and foresees construction of Line 4 and Line 4A. All new or extended lines will be connected to bus transfer stations along the Metro system and, in addition, Line 5 will interface with a suburban railway line at a new intermodal terminal. 1 HISTORY Construction at Metro Santiago started in the mid 1970s with underground portions mainly in cut and cover boxes. By the 1980s, increasing public objection to the surface disruption forced Metro to investigate less intrusive alternatives, such as mined tunnels. The subsequent development from cut and cover construction to mined running tunnels and mined stations has to be described as rapid. In 1993 the design of the initial Line 5 Extension included the first mined tunnel experiment, a 2.0 km long running tunnel portion under Bustamante Park. The tunnel design was produced by Ingendesa, a Chilean engineering firm with tunnel expertise gained from designing hydro electric power plants. The experiment proved to be successful however, at the second Line 5 Extension in 1997, a 2.8 km continuation to the west the stakes for Metro were much higher. This time the alignment was planned beneath a heavily frequented street and adjacent to the city’s 300 year old cathedral and other historic buildings. To minimize the risk Metro required the Chilean design firm Cade Idepe to utilize foreign tunnel expertise. This expertise was provided by Geoconsult of Austria, which developed a comprehensive design according to the principles of the New Austrian Tunneling Method (NATM) for the running tunnels. While the

running tunnels for this Line 5 Extension were mined, cut and cover construction was still foreseen at the stations. With improved NATM techniques it even became viable and cost effective to mine stations which further reduced the environmental impact and added alignment flexibility. Therefore, based on a preliminary design by the Cade Idepe/Geoconsult team detailed mined tunnel designs for running tunnels and station tunnels for the extensions of Line 2 and Line 5 were developed in 2001. The mined approach was also chosen for the underground portions of the new Line 4 and a second extension of Line 2 to the north. 2 EXTENSION OF EXISTING LINES Currently Metro Santiago operates three lines, Line 1, Line 2 and Line 5. Line 4 currently under construction will be part of the expansion plan to be completed by the end of 2005. Line 3, planned to run parallel to the south of Line 1, was part of an earlier urban transportation plan which since has been revised due to the rapid development of the city. The extension of Line 5 to the west of Santiago covers an underground length of 1.9 km along Catedral Street and will feature two new stations. Quinta Normal Station, the western intermodal terminal will allow

195 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-24.qxd

2/18/04

12:31 PM

Page 196

Figure 1. Metro system map.

passengers to transfer to buses and a suburban railway line which runs in a north to south direction below Matucana Street. The neighborhoods along this line extension feature historic buildings with architectural and heritage value. Construction at Line 5 Extension commenced in May of 2001 and the completion of the work is expected by the end of 2003, with service to start by March 2004. The initial extension of the existing Line 2 to the north has an underground length of 2.2 km and two new stations. Running below Recoleta Street, the first extension crosses under the Mapocho River and Costanera Norte, a major urban highway under construction. Aside from preliminary work at shafts and access tunnels, construction for this line extension began in May 2001 and will be completed by mid 2004 with service expected to start in August of 2004. A second extension of Line 2 to the north is currently under design and will add 5.1 km of underground rail and five stations to the system. The new section extends Line 2 to Avenida Américo Vespucio, a major ring road around Santiago and construction is planned to start in late spring of 2004. Besides the addition to the north, Line 2 will also be extended 2.3 km to the south, near the merging point of Gran Avenida José Miguel Carrera and the southern

Figure 2. Area of Quinta normal station.

portion of Avenida Américo Vespucio. Two new stations will be constructed along the southern extension of Line 2. La Cisterna Station, the south terminal station, will be the link to Line 4, Metro’s future expansion line currently under construction. Construction for the extension began in January 2002 and is expected to be completed by December 2004.

196 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-24.qxd

2/18/04

12:31 PM

Page 197

The preliminary and detailed designs for the line extensions were produced by the design team of Ingendesa, Arze, Reciné y Asociados Ingenieros Consultores (ARA), two local engineering firms, and the Dr. G. Sauer Corporation (DSC) of Herndon, Virginia, as their expert foreign tunnel consultant, and the Cade Idepe/Geoconsult team. 3 NEW LINES The planned two new lines for Metro’s system will connect ten communities in the east and south of Greater Santiago through a total length of 33 km of underground, at grade and elevated rail. Line 4, which extends from the existing Tobalaba Station on Line 1 to Puente Alto in the south, includes a 7.9 km long tunnel section from Tobalaba Station to Rotonda Grecia. The 6.5 km section from Rotonda Grecia to Avenida Vicuña Mackenna will be constructed at grade along the median of Avenida Américo Vespucio and a 0.6 km long over-under tunnel section. From thereon the line will be constructed to Puente Alto mostly elevated along the alignment of Avenida Vicuña Mackenna with some at-grade and underground portions. The new Line 4 will also extend westward at Station Vicuna Mackenna to end at the termination of the southern extension of Line 2, La Cisterna Station. This branch, to be called Line 4A, will run at grade along the median of Avenida Américo Vespucio with a length of 7.9 km. Detailed designs for the underground portions of Line 4 were completed in July 2003. Three teams, Arcadis Geotecnica/Bureau de Projectos, Cade Idepe/ Geoconsult and Ingendesa/ARA/DSC were involved in the designs of mined running tunnels and underground stations.

4.2

Running tunnels

Metro’s running tunnels are single tube, double track tunnels with an average cross section size between 60 m2 and 65 m2. Tunnel size, stable ground conditions and an overburden of approximately 9 to 11 m allow for a full face excavation with a substantial earth wedge used for face stabilization as well as working platform during profiling, lattice girder and wire mesh installation and shotcrete application. The average excavation round length is 1.0 m with primary support comprising of 200 mm initial shotcrete with lattice girder and wire mesh reinforcement followed by a secondary layer of shotcrete, 150 mm thick, reinforced with wire mesh or rebar. Using data and experience gained on the Line 5 Extension the running tunnel linings have been reduced from a thickness of 500 mm to the proposed 350 mm to be used on the new line and line extensions, a savings of 30%. The shotcrete used is typically dry shotcrete, however, wet shotcrete has been used at the Line 2 North Extension and is currently in use at Line 4. Steel fibers in lieu of wire mesh or rebar reinforcement have been considered at Line 4. 4.3

Station tunnels

With a cross section size of up to 150 m2 and an overburden of as low as 7 m the new station tunnels on the

4 DESIGN AND CONSTRUCTION 4.1

similar properties as found in the second Mapocho Deposit. The groundwater table is higher at approximately 20 m depth. The sediments deposited by erosive streams from the Cordillera de Los Andes consist mainly of clay and silt with low to moderate plasticity and sand lenses of variable sizes. The material is partially saturated (30% to 75%) with the groundwater table within 20 m of the surface. This deposit is interlocked with the ripio in a saw-tooth pattern along a contact line in north to south orientation. About 4.0 km of the underground portion of Line 4 will be located in this formation.

Geology

Metro’s existing and new lines lie in quaternary sediments of gravel, the so called “Grava de Santiago” or “Ripio de Santiago” (Ripio) and locally contain deposits of over-consolidated clays. In the north those sediments originated from the Mapocho River, while in the south the sediment’s origin is the Maipo River. In the east sediments were deposited by erosive streams from ravines of the Cordillera de Los Andes. In the Mapocho Deposit a superior 4.5 to 6.5 m thick stratum of fluvial origin is followed by a stratum of fluvial-glacial deposits with similar gradation except the presence of plastic fines and a somewhat bigger compactness and cohesion. The groundwater level is variable, but generally located at approximately 80 m depth. The Maipo Deposit is of fluvial origin and with

Figure 3. Running tunnel during track work.

197 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-24.qxd

2/18/04

12:31 PM

Page 198

Santiago Metro produce some impressive underground structures. This impression is enhanced by perpendicular intersections with access tunnels of similar size. The design approaches to these massive caverns differ for each design team at the line extensions and also require an approach adapted to the clay dominated deposits at Line 4’s Plaza Egaña Station and Los Orientales Station. There the final design team, Cade Idepe/Geoconsult has developed a binocular configuration for the 140 m long station tunnels with a center pilot tunnel in which permanent central support columns are cast prior to the excavation of the station tunnels on each side. On the extensions of Line 2 and Line 5 in the good quality ripio above the ground water table, the excavation of the station tunnels follows a top heading and bench/invert sequence with two side drift tunnels and a central gallery. At the El Parrón Station on the Line 2 South Extension the Ingendesa/ARA/DSC design team applied experience gained on the Line 5 Extension and modified the design to a single sidewall drift sequence. This approach was also applied to the design of Las Mercedes Station at Line 4 and was taken over by the Arcadis Geotecnica/Bureau de Projectos design team for their station detailed designs on the new line. The access tunnels at the line extensions were designed by Cade Idepe/Geoconsult as part of their preliminary design work with a full span top heading and bench/invert sequence. The wide span however requires pre-support using grouted pipe spiling and an earth wedge providing face support. Besides the grouted pipe spiling on the full span top heading excavation and grouted pre-spiling, using self drilling bolts, for break-out situations from an access tunnel into a station tunnel, no significant pre-support is necessary in the ripio when appropriate excavation and support sequences are used. During excavation according to the designed excavation sequence, station tunnels just like the running tunnels generally receive a primary lining consisting of lattice girders, wire mesh and shotcrete of 300 mm thickness. This initial support is followed by a secondary, 200 mm thick shotcrete lining, to arrive at a combined thickness of 500 mm for the Line 2 South Extension stations. No integrated waterproofing systems have been considered by Metro so far for either running tunnel or station tunnel designs. 4.4

Figure 4. Double side wall drift station excavation.

Geotechnical instrumentation monitoring

Geotechnical instrumentation of the running tunnels comprises mainly of five point in-tunnel monitoring cross sections and seven point surface settlement sections at the same station. The in-tunnel convergence points and roof leveling points are read in three dimensions according to a specified reading schedule which considers the actual location of excavation face and

Figure 5. Single side wall drift station excavation.

198 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-24.qxd

2/18/04

12:31 PM

Page 199

work progress. The monitoring cross sections are laid out at 25 m intervals along the alignment, which with today’s extensive tunneling experience in the ripio is based on liability considerations rather than on geotechnical necessity. Monitoring points are also placed on sensitive structures within the zone of influence along the alignment. At station tunnels the monitoring point configuration within a monitoring cross section varies based on excavation approach. In addition to deformation monitoring, strains and stresses on and in the shotcrete linings are measured using strain gages, ground pressure cells and concrete pressure cells. As part of an optimization effort of the whole instrumentation program, the use of sliding micrometers, extensometers and inclinometers was drastically reduced, once sufficient information and data regarding the behavior of the ripio was available. This effort has reduced the cost for instrumentation and monitoring to approximately $250,000 per km running tunnel. The readings are taken by an independent geotechnical consultant, who transmits the processed data to the design engineers, construction supervision and Metro’s project management team for further analysis. The processed data is then compared to threshold values and trigger values established by the engineers during the design. These values, although set slightly different by the individual design teams are based on mining sequence, geology and surface developments. The set limits of approximately 15 mm for surface deformation above the station tunnel center line or 12 mm for a roof leveling point on the station tunnel center line have so far not been exceeded during the ongoing line extension works.

4.5

Metro’s administrative set-up for construction contracts is predominantly design-bid-build. The design work generally starts with a preliminary design, which for reasons of time savings includes preparatory work such as the detailed designs for shafts and access tunnels. Detailed designers pick up at an approximate 30% level and remain involved throughout the construction process by furnishing construction supervision, which includes a monitoring engineer responsible for the interpretation of gathered and processed geotechnical instrumentation data. Based on the monitoring engineer’s interpretation of the data and in coordination with the design team field modifications to the design can be made, if required. Another task of the construction

Figure 7. Quinta Normal Station excavation.

Figure 6. Surface settlements at Quinta Normal Station.

199 Copyright © 2004 Taylor & Francis Group plc, London, UK

Execution of work and supervision

09096-24.qxd

2/18/04

12:31 PM

Page 200

supervision is to verify design compliance and to respond to contractor’s questions regarding the design. To supervise the actual construction activities Metro awards construction management and inspection contracts for the individual line sections. The construction management and inspection work includes quality assurance tasks, enforcement of safety requirements as well as tracking of construction progress and construction cost. The technical documentation and coordination with the independent geotechnical consultant regarding access to the tunnel for monitoring purposes are also responsibilities of the construction management and inspection team. To date construction of Metro sections is mainly carried out by local contractors and Chilean subsidiaries of large foreign companies, such as Sacyr of Spain and Mendes Junior of Brazil. The quick adaptation by local contractors from cut and cover construction to mined tunneling methods is encouraging and current excavation rates of 3.0 to 4.0 m per 24 hour day in the running tunnels are remarkable. The ground with its extended unsupported stand-up time could allow excavation rounds beyond the current 1.0 m to expedite the construction. However, the use of specialized currently not utilized tunnel equipment, such as excavators with articulating arms, man-lifts and a more efficient mucking operation would increase productivity without compromising the current low level of risk. The use of fully mechanized excavation equipment was considered for the Line 4 construction and several studies were commissioned by Metro to investigate the technical and economical feasibility of Tunnel Boring Machines (TBM) and Earth Pressure Balance Machines (EPBM) for the clay zones and water bearing zones near the San Carlos Canal in the east and the Maipo River in the south. All studies concluded that while technically feasible there is no cost or schedule advantage using fully mechanized excavation equipment on the 7.9 km long underground stretch from the existing Station Tobalaba to Station Rotonda Grecia or elsewhere along Line 4. In fact the studies concluded that there is a considerable lower risk to the schedule using NATM. 5 CONSTRUCTION COST One of the reasons for the success of NATM on Santiago’s Metro was the cost savings realized after its introduction. While the first tunnel section under Bustamante Park proved that mined tunnels are technically feasible and caused less disruption to the city, the construction cost were approximately 20% higher compared with cut and cover structures, not considering costs for utility relocation and expropriation. Today, after further improvements to the NATM designs, the

construction cost average about $6,500 per linear meter of running tunnel and about $27,500 per linear meter of station tunnel. At present, the cost of the new Line 5 Extension is about 40% less compared to the Line 5 Extension built in the late 1990s. The reasons for the reduction in cost are more favorable ground conditions, lower rise buildings along the alignment and most important the conversion of expensive and disruptive cut and cover stations into mined stations. Metro hopes to reduce this already low tunneling cost even further during the construction of the new Line 4 by introducing value engineering as a tool for the contractors to optimize the designs. 6 CONCLUSION Not enough can be said about the fast development of Santiago’s Metro from cut and cover construction in the middle of an almost 5 million people metropolis to mined tunnels at reduced cost. Within approximately 10 years the local design and construction community with support of foreign experts has learned to develop and execute complex NATM designs for large underground spaces. There is no argument that the favorable ground conditions and a low water table have helped engineers and contractors in their swift learning process. However, one has to laude Metro’s balanced approach between their desire to reduce construction cost and their willingness to take risk. While the replacement of running tunnels with cut and cover boxes by mined tunnels in the early 1990s was a successful first step, Metro’s managers understood that with foreign tunnel expertise even bigger steps could be made. Thus in 1997 foreign NATM expertise was brought in for the design of the running tunnels during the second Line 5 Extension. This proved to be just the beginning of a prolonged success for the sequential excavation method on Santiago’s Metro, where today all underground work is carried out using this method. Advantages such as lower cost and lower risk gave NATM also the edge over mechanized excavation methods, which were considered for the new Line 4 construction. With room for even further improvement of productivity through efficient construction equipment, one can only look forward to future line extensions and new lines in Santiago. REFERENCES Wallis, S., 2003a. “Evolving NATM for Santiago’s Metro”, T&T International, March 2003. Wallis, S., 2003b. “Metro’s evolution”, T&T International, April 2003. Mercado, C., 2003. “Tecnologia sobre Rieles”, Revista Bit, September 2003.

200 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-25.qxd

2/18/04

12:32 PM

Page 201

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Benchmark for the future: the largest SEM soft ground tunnels in the United States for the Beacon Hill Station in Seattle J. Laubbichler, T. Schwind & G. Urschitz Dr. G. Sauer Corporation

ABSTRACT: Sound Transit’s Beacon Hill Light Rail Station in Seattle comprises the largest SEM soft ground tunnels in the United States to date and will be constructed in highly variable glacial soils with multiple groundwater horizons. The design is based on SEM principles, experience from previous projects, engineering judgment and numerical analyses, and provides for the flexible application of various pre-defined support measures (SEM Toolbox items) to cope with variable ground conditions, assure the most economic construction and minimize risks. A Test Shaft Program with the purpose of gaining additional geological information and confirming design assumptions was carried out. Findings in regard to ground behavior and groundwater were implemented in the design. To reduce construction risks, the owner decided to retain the SEM designers to provide SEM supervision and construction support services.

1 INTRODUCTION 1.1

Project overview

The Beacon Hill Tunnels and Station are part of the 14 mile initial segment of the Sound Transit Central Link Light Rail Line that will establish a high capacity commuter connection from downtown Seattle to Tacoma. The 4,300 foot running tunnel under Beacon Hill will be mined by Earth Pressure Balance Machine (EPB), while the deep mined Station will be constructed using slurry walls and the New Austrian Tunneling Method (NATM), referred to as Sequential Excavation Method (SEM) for this project. Figure 1. Station arrangement.

1.2

Beacon Hill Station arrangement

From the Station Headhouse, a 181 ft deep, 46 ft inner diameter Main Shaft will be constructed that will house four high speed elevators, emergency staircases, ventilation shafts and mechanical and electrical equipment. A 26 ft inner diameter Ancillary Shaft will accommodate another set of emergency staircases and ventilation shafts. From the Main Shaft, the 41 ft wide Concourse Cross Adit will provide passenger and emergency access to the Platform Tunnels. These are 380 ft long by 32 ft wide and were designed to accommodate the platforms, artwork and architectural finishes, and the light rail tracks. Two Cross Adits will connect the Platform Tunnels, and Ventilation Tunnels

will provide air flow in normal operation and for emergencies. 1.3

The Hatch Mott McDonald/Jacobs (HMMJ) Joint Venture is the lead designer for the Beacon Hill Tunnels and Station, the architectural design is carried out by Otak. The Dr. G. Sauer Corporation (DSC) provides the SEM design as a subconsultant for the Concourse Cross Adit, the Platform Tunnels and the Platform Cross Adits; Shafts, Ventilation Tunnels and Running Tunnels are designed by HMMJ.

201 Copyright © 2004 Taylor & Francis Group plc, London, UK

Design responsibilities

09096-25.qxd

2/18/04

12:32 PM

Page 202

2 GEOLOGY 2.1

Geologic setting

The Puget Sound Area is characterized by a complex mixture of glacial and non-glacial soils that have been deposited, consolidated, eroded and reworked by multiple major glaciations during the Pleistocene Epoch and numerous seismic events. Beacon Hill is an approximately 300-ft high ridge that is composed of holocene, vashon and pre-vashon deposits. An extensive subsurface exploration program has been conducted by Shannon & Wilson, Inc. of Seattle. In the course of this program, more than 70 investigation borings were drilled using Hollow Stem Auger, Mud Rotary, Triple-tube Rotary Core and Sonic Core techniques. Laboratory testing of the recovered soil samples was carried out and engineering properties were derived accordingly. The exploration program showed that most of the Beacon Hill Station will be excavated within in glacial, overconsolidated, partly fractured or slickensided clays and tills. Intermittent sand and silt layers will be present with multiple perched groundwater horizons. The Seattle Bremerton Fault zone is expected to be the cause for some of the inconsistencies, inclinations and fractures observed during the geotechnical investigation. 2.2

Ground classification and ground behavior

For design purposes, the soils were grouped into classes according to their engineering parameters and anticipated ground behavior during tunneling: Class 1: Loose to dense granular deposits This soil type consists of poorly graded sand and gravelly sand; it will be encountered when excavating the Headhouses and not be of concern for tunneling.

Class 2: Soft to Very Stiff Clay and Silt This soil type comprises normally consolidated clays, and silty clays and clayey silts; it will be encountered when excavating the Headhouses and not be of concern for tunneling. Class 3: Till and Till Like Deposits Heterogeneous mixtures of gravel, sand, and silt or clay; they will be encountered in the station shafts and in sections of the station tunnels. These soils have a compressive strength similar to very soft rock are expected to stand vertically in an excavation. Water bearing sand and silt lenses may cause local instabilities, unless properly treated. Class 4: Very Dense Sand and Gravel This soil type consists of poorly graded sand, gravelly sand and sandy gravel; it will be encountered in pockets and relatively thin layers in the excavation of the station tunnels and will likely be water bearing. This material has little to no cohesion and will show flowing behavior if charged with water or running behavior if allowed to dry out. Dewatering, pre-treatment and special consideration will be required when this material is encountered. Class 5: Very Dense Silt and Fine Sand This soil type consists of silty fine sand to sandy silt; it will be encountered over a substantial portion of the Main Shaft excavation and parts of the Concourse Cross Adit. Under hydrostatic pressure, this material will show flowing behavior. If drained, it is expected to stand well in small to medium sized openings with little face support. Class 6: Very Stiff to Hard Clay This soil type consists of overconsolidated silty clay or clayey silt, with some fine sand; it will be encountered

Figure 2. Geologic profile of station.

202 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-25.qxd

2/18/04

12:32 PM

Page 203

in the Main Shaft and most of the station tunnel excavation will be in this material. Slickensided and fractured zones will be encountered during tunnel construction; to take this into account in the design, a further subdivision into Clay A, Clay B and Clay C was considered necessary. Due to the hard consistency and cohesive nature of this material, it will stand well at excavation faces and will be relatively easy to excavate with a tunnel excavator. In the slickensided and fractured zones, spalling, raveling and wedge failures in the tunnel face and heading may occur if not properly presupported. Water bearing sand and silt lenses may cause local instabilities, unless properly treated. The high variability of the geology in the area of the future station poses the main design challenge. Special considerations and flexibility in the design are necessary to address this issue. During construction, a high degree of experience, alertness, and the proper tools to react appropriately to changing ground conditions are needed. Sound Transit therefore decided to extend the services of the design team to provide SEM supervision and construction support services. In order to get a better understanding of the soil strata and the ground behavior during an SEM type excavation, it was decided to construct an Exploratory Test Shaft and Test Adits within the boundaries of the future Main Shaft. A brief description of the program and the implications for the Station Design are provided in section 5.

3 LARGE SOFT GROUND SEM TUNNELS 3.1

General considerations

The design philosophy of SEM has been described and documented in depth in numerous publications. The original concept was adapted to be suitable for soft ground tunneling and first used in the Frankfurt Clay in 1968. Since then, means and methods have been developed further and a substantial number of large soft ground tunnels have been constructed in Europe, some of them in adverse ground conditions with shallow overburden. In the United States, soft ground tunnels of the size required for the Beacon Hill Station break new ground. Some of the key elements for large SEM tunnels in soft ground are: 1. Ovoid cross sections with rounded inverts and domed excavation faces to prevent stress concentrations. 2. Ring Closure within 1.5 times the tunnel diameter to prevent loosening of the surrounding ground and excess settlements.

3. Timely installation of sealing shotcrete/flashcrete and the initial shotcrete lining to prevent deterioration and loosening of the soils. 4. Subdivision of the faces into smaller drifts and adjustment of round lengths to be able to control and stabilize the excavation. 5. Utilization of the appropriate ground support, face support, pre-support and ground improvement measures. 6. Monitoring of the structure during construction to assure stability and verify design assumptions. 7. The ability to make adjustments in the field to deal with actual ground conditions encountered. 8. Experienced Construction Management, Site Supervision and Quality Control to ensure safety and efficiency. SEM tunnel design has to take these factors into account and relies heavily on engineering judgment and experience from previous projects, but also on advanced Finite Element Modeling Tools to determine the appropriate excavation sequences and support measures. 3.2

When there is some continuity in the geologic strata and ground conditions can be reasonably anticipated for certain reaches, different ground support classes can be predefined. These contain the excavation sequence and the required support measures, i.e. shotcrete thickness, number of spiles, soil nails, etc. However, when highly variable geology is encountered, ground types and ground behavior change within several feet and mixed face conditions are encountered over large portions of the tunnel alignment, a different concept needs to be developed and deployed which is described in the following. By using the “SEM Toolbox” approach, a conservative baseline scenario is defined, an excavation sequence is prescribed and standard support measures – e.g. shotcrete, wire mesh and lattice girders – are defined. Depending on the ground conditions encountered, additional support measures (“Toolbox Items”) are used on an as needed basis to ensure stability of the tunnel face and the surrounding ground. These include:

•

•

203 Copyright © 2004 Taylor & Francis Group plc, London, UK

Coping with variable ground conditions – the “SEM Toolbox” approach

Pre Support Measures – Rebar Spiling – Grouted Pipe Spiling – Metal Sheets – Grouted Barrel Vault/Pipe Arch Face Stabilization Measures – Face Stabilization Wedge – Pocket Excavation

09096-25.qxd

• •

2/18/04

12:32 PM

Page 204

– Reduction of Round Length – Face Bolts Ground Improvement Measures – Gravity and Vacuum Dewatering – Permeation Grouting, Fracture Jet Grouting Annular Support – Additional Shotcrete – Soil Nails – Temporary Invert

access/egress considerations, space requirements for mechanical/electrical equipment and the geometry of the junction to the Platform Tunnels. Grouting,

For estimating purposes, expected location and quantities of Toolbox Items are provided. This approach provides a high degree of flexibility during construction and makes it possible to control virtually all kinds of ground conditions, thereby greatly reducing the risks of SEM construction. However, it requires that contractors are familiar with the utilization of the mentioned support measures. Experienced site supervision is essential to ensure that the appropriate measures are taken in a timely manner. The Standard support measures are paid for on a linear foot basis for each tunnel, while the SEM Toolbox items are separate line items and paid for on a unit price basis.

4.1.1 Geology The Concourse Cross Adit will be constructed primarily in Very Stiff to Hard Clay and Till and Till like Deposits, with intermittent, cohesionless pockets of Silt and Fine Sand that may contain pressurized groundwater. Layers of Silt and Fine Sand and Very Dense Sand and Gravel are located at or near the crown of the excavation.

The cross section of the Concourse Cross Adit, the largest tunnel of the Beacon Hill Station, was developed according to architectural requirements, emergency

4.1.2 Design Due to the large size of the opening and the difficult ground conditions especially in the crown, excavation will be carried out using the dual side wall drift method. Grouted with a double packer system under high pressure (1000 psi), the Barrel Vault will provide presupport over the whole length of the tunnel and be used to improve the Very Dense Sand and Gravel. The maximum specified advance length is 3 ft 4 in., and the maximum separation between the two side wall drifts in longitudinal direction is two rounds. The stability assessment for the excavation sequence and the in-place structure of the Concourse Cross Adit Tunnels was performed using two three dimensional finite element models and the finite element program ABAQUS. The first model includes the Main Shaft, the breakout from the Main Shaft, the sequential construction of the Concourse Cross Adits and the headwall. The second model is used to assess the breakout from the Concourse Cross Adit into the Platform Tunnels.

Figure 3. Concourse Cross Adit – typical cross section.

Figure 4. Dual side wall drifts.

4 DESIGN OF THE BEACON HILL STATION TUNNELS 4.1

Concourse Cross Adit

204 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-25.qxd

2/18/04

12:32 PM

Page 205

Figure 6. ABAQUS model 2.

Figure 5. ABAQUS model 1.

The decision to utilize two models instead of one was made in order to limit model size and therefore keep running times for the finite element code within acceptable limits to facilitate an effective and flexible design process. The soils in the FE models were modeled using Mohr-Coulomb failure criteria (friction angles from 27° to 40°, cohesion from 0 to 48 kPa), the shotcrete and concrete for the primary and final linings were modeled as linearly elastic materials. The construction sequence for the Concourse Cross Adit was modeled by completing top heading construction of the side wall drifts, followed by bench and invert. The top heading, bench and invert excavation sequence of the center drift was modeled in the subsequent steps. As all anticipated construction stages were modeled in the FE analyses, the numerical results were used to assess the stability of the excavation and the excavation face as well as the structural performance of the tunnel linings.

4.1.3 Lining design For the section forces determined in the FE analysis, the structural design for the tunnels was performed to meet the requirements of ACI 318. It could be shown, that a 14 in. thick shotcrete lining (fcu  5000 psi) is capable of providing the required support for the tunnel structure. Due to stress concentrations around the openings in the Concourse Cross Adit at the junction with the Platform tunnels, a local thickening of the primary lining of 17 in., was required to avoid additional bar reinforcement. The final lining is designed for the assumption that the primary lining loses 90% of its stiffness in the course of time. Additionally, the full hydrostatic load is assumed to act on the final lining of the tunnel structures. It could be shown that a 14 in. steel fiber reinforced concrete lining (fcu  5000 psi, fiber content 70 lbs/yd3) is sufficient to withstand all the occurring loads. Additional reinforcement is only provided in the junction areas and the connection areas to the headwalls. 4.2

The cross section geometry for the Platform Tunnels was developed according to architectural requirements and train clearance. An additional requirement is the possibility of walking the TBM through the Platform Tunnel for the completion of the east section of the running tunnels. 4.2.1 Geology The Platform Tunnels will be constructed primarily in Very Stiff to Hard Clay and Till and Till like Deposits, with intermittent, cohesionless pockets of Silt and Fine sand that may contain pressurized groundwater. Layers of Silt and Fine Sand and Very Dense Sand and Gravel are expected to be located at or near the crown of the excavation in one section of the tunnel, and dry sand (“hour glass sand”) can be expected in

205 Copyright © 2004 Taylor & Francis Group plc, London, UK

Platform Tunnels

09096-25.qxd

2/18/04

12:32 PM

Page 206

Figure 7. Platform Tunnel.

Figure 8. Single side wall drift.

the invert of the Platform Tunnels in one area of the excavation.

is assumed to act on the final lining of the tunnel structures. It could be shown that a 12 in. steel fiber reinforced concrete lining (fcu  5000 psi, fiber content 70 lbs/yd3) is sufficient to withstand all the occurring loads. Additional reinforcement is only provided in the junction areas and the headwalls.

4.2.2 Design As the tunnel cross section for the Platform Tunnels is somewhat smaller than the Concourse Cross Adit, it was assessed that those tunnels can be constructed utilizing the single side wall drift method. The stability assessment for the specified construction sequence and the primary and final shotcrete and concrete structure was performed using the second three dimensional FE model. The Side Wall Drift for the Platform Tunnels is constructed using a top heading bench and invert excavation sequence; the remainder of the tunnel is excavated in the same fashion following the completed side wall drift with a minimum distance of 30. As all anticipated construction stages are modeled in the FE analyses, the analytical results were used to assess the stability of the excavation and the excavation face as well as the structural performance of the tunnel linings. 4.2.3 Lining design For the section forces determined in the FE analysis, the structural design for the tunnels was performed to meet the requirements of ACI 318. It could be shown, that a 14 in. thick shotcrete lining (fcu  5000 psi) is capable of providing the required initial support for the tunnel structure. As stress concentrations in the ground in the vicinity of the Concourse Cross Adit Tunnels could be observed, a localized thickening of the initial shotcrete lining was required in the junction area. The final lining is designed for the assumption that the primary lining loses 90% of its stiffness in the course of time. Additionally, the full hydrostatic load

4.3

For the construction of each of the SEM tunnels, prescriptive excavation sequences were developed. These contain breakout sequences, advance lengths, sizes of openings, distances to ring closure and distances between side wall drifts. In conjunction with the excavation, the standard support measures, i.e. flashcrete, wire mesh, lattice girders and shotcrete are defined. It is specified that the standard support elements for any round have to be complete prior to commencing the next excavation round in the sequence. To reduce the uncertainty about ground conditions ahead of the face, the systematic drilling of 35 ft long horizontal exploratory probe drill holes every 6 excavation rounds is specified. The results of the exploratory drilling and the assessment of ground conditions at the tunnel face will be used in the field to determine if there is a need for ground improvement or additional support measures. If so, the appropriate SEM Toolbox Items for the conditions encountered can be utilized to ensure the safety of the tunneling operation. To assist the contractor in choosing the appropriate support measure, requirements for the application of a particular item were defined in the GBR and the Special Provisions. For the preparation of the bid documents, baseline quantities for each Toolbox Item were defined according to the anticipated geologic conditions.

206 Copyright © 2004 Taylor & Francis Group plc, London, UK

Excavation and support design and utilization of NATM/SEM Toolbox

09096-25.qxd

2/18/04

12:32 PM

Page 207

5 MONITORING

7 BEACON HILL TEST SHAFT

During SEM Tunneling, monitoring, recording and interpreting deformations and stresses of the initial lining is essential to ensure construction safety and to verify the results of the design assumptions. For monitoring during construction of the mined Station, a comprehensive instrumentation scheme has been developed. Convergence Bolt Arrays will be used to monitor absolute and relative deformations. Concrete pressure cells will record the normal stresses in the tunnel lining, while earth pressure cells will be used to record the ground loads that are transferred to the tunnel lining. In addition, a surface monitoring program will utilize surface settlement points, inclinometers and extensometers to provide complete information about ground movements during the excavation.

7.1

6 WATERPROOFING The Beacon Hill Station is designed as a “tanked” structure, meaning it will be equipped with a waterproofing system to make it completely watertight. In addition, a Sectioning System is foreseen that will provide remedial repair options in case of leaks. The waterproofing system is installed between the initial shotcrete lining and the final lining and consists of the following elements:

•

•

•

Geotextile A non woven polypropylene geotextile is fastened to the initial shotcrete lining with PVC disks. It is designed to protect the waterproofing layer from sharp projections of the initial lining surface. Waterproofing Membrane The waterproofing layer is the actual sealing element of the system, designed to keep groundwater from the interior of the tunnel. It consists of flexible membrane sheets welded together to form a continuous, impervious layer. This geomembrane is made of a polymeric material, like polyvinylchloride (PVC). Its material properties allow it to adapt to the irregularities of the initial tunnel lining and it is designed to permanently withstand biological and chemical deterioration due to aggressive groundwater. Furthermore, it is fire retardant to minimize safety hazards during construction and operation. Sectioning System The waterproofing system is divided into sections by the means of water barriers. Should a leak occur at a certain location, only one relatively small section is affected, which can be repaired by grouting through preinstalled grout pipes.

In the course of the design process, the designers entertained the idea of a Test Shaft, its purpose being a more thorough understanding of the complex geology and the evaluation of the performance of the SEM construction method. The engineering team designed a 148 ft deep, 18 ft diameter SEM shaft and two Test Adits in different geologic strata within the foot print of the future Beacon Hill Station Main Shaft. By means of the Test Shaft and Adits, ground behavior of the various geologic strata, especially of the water bearing sands/silts considered most critical for tunneling, and of the hard clays, where most of the tunneling work will be performed, should be closely monitored and assessed. The value of a Test Shaft was determined to be the additional knowledge about geology and ground behavior and the resulting design optimization. Following an assessment of the experiences during construction and the results of the monitoring program, the original assumptions and the resulting design were to be confirmed or modified as required. 7.2

Test Shaft construction

The construction of the test shaft took place between April 2003 and September 2003. Deviating from the original intent to construct a shaft using the sequential excavation method (SEM), the contractor decided to excavate the first approximately 50 ft using an auger drill and subsequently install a reinforced shotcrete lining. Once ground conditions worsened and the excavation method using an auger drill could not be further utilized, SEM using a mechanized excavator and a reinforced shotcrete lining was used as prescribed in the Test Shaft design for the following approximately 60 ft. However, schedule delays and cost overruns, mainly due to more complex ground conditions and additionally employed dewatering measures necessitated the decision to terminate the construction of the Test Shaft before excavating the Test Adits. The test shaft was completed to its intended depth using a 6 ft diameter steel cased boring. 7.3

Test Shaft findings

Generally, the encountered ground conditions were well suited for SEM construction. For most of the depth the ground remained stable for the full depth of each excavation round (up to 6 ft) and for a considerable length of time (up to 5 hours and more). However, as permeable geologic layers (sands and silty sands) that had not been adequately dewatered, either by means of deep wells or vacuum well points, were encountered,

207 Copyright © 2004 Taylor & Francis Group plc, London, UK

Test Shaft purpose and design

09096-25.qxd

2/18/04

12:32 PM

Page 208

the sides of the shaft excavations locally displayed instability and the contractor could not install the required pre-support and support measures in the required time frame. During the Test Shaft excavation, it could be observed, that the geologic conditions in the Beacon Hill Test Shaft area were more diverse and variable than originally anticipated. Furthermore, it became apparent that for conditions where groundwater control is critical for the success of the construction method, the usual approach of a contractor designed dewatering system is problematic and should be replaced with an owner designed dewatering system. It has to be emphasized that the employment of a contractor capable of utilizing SEM principles and SEM Toolbox items in the required manner and time is paramount for the successful implementation of SEM design. Following the findings of the Test Shaft excavation, risk considerations lead to several design changes as outlined below. The value Test Shaft program was confirmed, as the design changes prior to the bid phase will avoid claims based on inaccurate design assumptions and save more than the costs incurred.

8 REDESIGN As the findings from the Test Shaft created concern among the designers of the shafts regarding the appropriateness of the SEM excavation in this geology, it was decided to redesign the SEM shafts and replace them with slurry wall shafts. For the construction of the tunnels, the SEM approach was maintained. The numerical analysis of the Concourse Cross Adit had to be rerun to take the changed geometry and stress regime of the slurry wall into account and the reinforcement and lining thicknesses were adjusted accordingly; the breakout sequences from the shafts were redesigned. Schemes for dewatering from the surface and from within the tunnels were added to the design package. In addition, more stringent requirements for exploratory probe drilling during construction were established and the application of SEM support elements was shown in more detail. Jet Grouting from the surface for the Ventilation Tunnels and to a limited extent for the Platform Tunnels was added. Finally, the anticipated ground conditions were adjusted in the GBR and new baseline

quantities and distributions for the SEM Toolbox items were established.

9 CONCLUSION The challenges posed by the geology and the station arrangement were systematically analyzed and addressed in the production of the design package. Concourse and Platform Tunnels, with widths of 45 ft and 34 ft respectively required the development of sequences able to cope with soft and potentially running conditions, but also very stiff and heavily slickensided soils. The Test Shaft program greatly reduced the uncertainty about the ground conditions and gave the design team the opportunity to evaluate and adjust the design approach. This will pay off as the information gained and implemented in the design package will give the contractors a better basis to bid the project. The decision to extend the designer’s services into the construction phase and to task him with the supervision of the SEM works will greatly minimize the construction risks and assure that the design intent is conveyed through construction. The design and subsequently the construction of the large Beacon Hill Station tunnels will serve as a Benchmark for the future of soft ground SEM tunneling in the United States.

REFERENCES Duddek, H. & Städing, A. 1990. Tunneling in Soft Ground and Sedimentary Rock for High Speed Double Track Railway Lines in Germany, Tunnelling and Underground Space Technology, Vol. 5, No. 3, pp. 257–263 Maidl, B. 1995 Handbuch des Tunnel und Stollenbaus. Essen: Glückauf Pacher, F. & Sauer, G. 1989. Grosse Querschnitte in nicht standfestem Gebirge. Wien: Springer Verlag Sauer, G. 2003. Ground Support and its Toolbox, ASCE Conference May 6 & 7, 2003. New York City Shannon & Wilson 2002. Geotechnical Data Report. Seattle Tunnels & Tunneling International Dec. 2002. Test Shaft to start at Beacon Hill Station: 51 Hatch Mott McDonald Jacobs 2003. Design Report, Seattle Hatch Mott McDonald Jacobs 2003. Geotechnical Baseline Report. Seattle Hatch Mott McDonald Jacobs 2003. Test Shaft Report. Seattle World Tunneling Oct. 2003. Beacon Hill Tunnel Project and Test Shaft: 314–315

208 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-26.qxd

2/18/04

12:33 PM

Page 209

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Application of the Press-In Method in East Side Access tunnel project Jinyuan Liu & Verya Nasri STV Incorporated, New York

ABSTRACT: The East Side Access (ESA) Project in New York is one of the largest tunneling projects undertaken by the US railroad. The project is divided into two major segments namely Manhattan and Queens. In this study, the potential application of an innovative Japanese construction technique, the Press-In Method is evaluated for the ESA project. This system will be used as the support of excavation for the four emergency exits and approach structures in the Queens segment. The deepest emergency exit for two adjacent parallel tunnels has an excavation depth of 85 ft and a water head of 68 ft. Soil within the excavation depth consists of typically coarse to fine, cohesionless glacial material, well sorted to well-graded, interspersed with cobbles and boulders. Tubular pile and the crush piler are selected because of the high water pressure and difficult subsoil conditions. The nonstaging method that continuously presses-in piles to construct a wall and utilizes the top of this pile wall as the platform for the equipments is selected to meet the stringent requirement of the site. The construction and design methods are presented in this paper, while the waterproofing, corrosion, and supporting details will be addressed. Based on the result of this study, the Press-In Method is a feasible and costefficient system for ESA project comparing to slurry wall.

1 INTRODUCTION 1.1

study of a new construction technique to be used in Queens part. More information about ESA project can be found in Nasri et al (2003).

East side access project

The East Side Access (ESA) Project in New York is one of the largest tunneling projects undertaken by the US railroad (Fig. 1). Its budget is around $5.2 billion. The project is divided into two major segments namely Manhattan and Queens. This paper presents a feasibility

Figure 1. Plan view of East Side Access project.

1.2

The aerial view of Queens bored tunnel is shown in Fig. 2. There will be four emergency egress shafts to

Figure 2. Aerial view of bored tunnels in Queens.

209 Copyright © 2004 Taylor & Francis Group plc, London, UK

Emergency shaft and approach structures

09096-26.qxd

2/18/04

12:33 PM

Page 210

be constructed and connected to tunnels. The deepest shaft that would be used for two overlapping tunnels will be discussed in this article. The bottom of this shaft will be approximately 85 ft below the existing ground surface. The plan dimension of this shaft will be an ellipse with two axial lengths of 44 and 58 ft, respectively. The approach structure is located on the north side of Harold Interlocking, immediately west of the 39th Street Bridge. The approach structure brings the track from the ends of the bored tunnels to the surface. It will be approximately 900 ft long and 25 ft wide. 1.3

Geotechnical condition

Per ESA Geotechnical Design Summary Report, the project site in Queens Segment is underlain by Ordovician/Cambrian age metamorphic bedrock, which is covered by Pleistocene glacial and interglacial deposits, and by postglacial materials. The various subsurface strata encountered in Queens site are not uniformly spread across the area. The soil stratum encountered at this site is a Fill with a varying thickness up to 10 ft, followed by a 20 to 45 ft thick Mixed Glacial Deposits (Stratum 2), and a 25–90 ft thick Glacial Till (Stratum 5). Bedrock is encountered at the deepest boring at a depth of 120 ft near emergency shaft and 40 ft at approach structure area. Water table is 68 ft above the bottom of excavation for the interested shaft. Water table is below the bottom of excavation in approach structure area. Stratum 2 consists of loose to dense coarse to fine sands with silts, gravels, cobblers, and boulders. Stratum 5 consists of dense to very dense sand with silts and gravels. Typical SPT N value for emergency shaft and approach structures are shown in Fig. 3, which ranged from 50 to greater than 100. 25

SPT N 50 75 100

0

0

0

20

10

40

20 Depth, Feet

Depth, Feet

0

60

SPT N 50 75 100

2.1

Press-In piles

The Press-In Method is a 30-year-old construction technique. Recently is has been further developed by research collaboration between Giken Seisakusho Co., Ltd. and Cambridge University Geotechnical Engineering Group Since 1994 (ENR 2001, Bedian 2002, and White 2002). The press-in principle is to utilize reaction force derived from fully installed piles and hydraulically press-in subsequent piles. The Silent Piler works on top of the reaction piles and self-moves to the next position gripping the pile being pressed-in. Technical details of the press-in mechanism are illustrated below in Fig. 4. In practical terms, the Silent Piler grips previously installed piles with hydraulic jaws. The next pile is hydraulically gripped by the Chuck at proper pressingin point and jacked into the ground with a static load generated by the main hydraulic rams. The Silent Piler derives reaction force from skin friction and interlock resistance of the previously installed reaction piles, which surpasses the press-in resistance during piling. Installing the pile to the designed depth by accurate hydraulic control, the Piler repeats the same press-in procedure until the last pile is put into the ground. Since the piles are pressed-in, the Silent Piler does not cause any damage to the environment including neighboring structures and local residents through noise and vibration. The Press-In Method operates at only 69 db of noise allowing pile installation in areas where environmental disruption is strictly precluded. Because reaction force is used as its basic mode of operation, the self-weight of the piling machine is rather unimportant. Moreover, the superiority of the principle enjoys great advantages with the integrated GRB System, which allows transporting of material, crane system and pressing-in to be systematically carried out from the top of fully installed piles utilizing a minimum of right of way.

30

80

40

100

50

120

25

2 PRESS-IN METHOD

Bed rock

60 Bed rock

(a)

(b)

Figure 3. Soil profile for (a) emergency shaft and (b) approach structure.

Figure 4. Schematic of Press-In Method.

210 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-26.qxd

2/18/04

12:33 PM

Page 211

3 DESIGN OF PRESS-IN METHOD As for the emergency shaft, water intends to inflow to the shaft due to the unbalanced water table. Ground modification is needed to reduce the inflow and provide a dry working environment. A jet grout plug or soil replacement plug of treated ground can be created prior to commencement of excavation within the shaft to prevent the inflow and balancing the uplift force (Coomber 1985). The bedrock is at 120 ft below the ground, which is about 35 ft deeper than the bottom of excavation. The pile will be driven through the soft ground and socketed into the bedrock. The length of tubular pipe for emergency egress shaft will be approximately 120 ft. Since the bottom of excavation is below water table, the tubular pile with P-P interlocking will be used in this shaft. As for the approach structure, tubular pile will also be used. The design length of piles will vary based on the retaining heights. 3.1

Press-In pile material

One of steel tubular pile manufacturers is Nippon Steel Corporation in Japan. Material can be procured and fabricated from American local markets. This material has a high elastic rigidity and high bearing capacity and can be constructed in soft ground. It has been used as both temporary structures and permanent retaining structures. Tubular piles provide bigger moments of inertia comparing to Z-type and U-type sheet piles with similar weight. Under the same loading condition, steel tubular piles can reduce the lateral deformation of the structure. With the introducing of inner bracing, the maximum deflection in emergency egress shafts can be reduced substantially to ensure the safety of nearby bridge pier and in-service tracks. 3.2

Physical restraints

There are many environmental constraints from nearby buildings and in-service railways. At the proposed optimum location for the emergency egress shaft at Honeywell Street Bridge, there is a dense network of surface tracks, utilities, catenaries, buildings and slopes. The design of the emergency egress shafts should minimize the impacts and relocations of these restraints. The compact equipment and its ability in limiting deformation of nearby structures makes the Press-In Method an excellent solution to these restraints. The main components of Giken’s Press-In Method consist of a piler and a power pack (Fig. 5). The PP260 piler proposed by Giken America Corporation is approximately 16 ft long, 7 ft wide, and 16 ft high, which is used for tubular piles with diameter of 31 to 36 in, The power pack for this piler is 14 ft long, 6 ft wide and about 8 ft high.

Figure 5. Components of Press-In Method equipment.

The compact and lightweight Silent Piler limit workspace to just the area ultimately required and it minimizes the effects on the environment. Under normal working conditions, the Silent Piler can operate with one crane to pitch piles. When a pile being pressed-in is sufficiently stable, the Silent Piler releases the clamps from the reaction piles and uses the pile to raise itself and travel forward. This “self-moving” system eliminates the need for support by a crane during the piling operation. In other words, even where a site requires a large jib radius for pitching, a relatively lightweight crane can be used. 3.3

For the approach structure, the press-in pile can be used for the permanent structures. Corrosion control should be considered. The corrosion control for steel tubular sheet piles is the same as that for any other type of steel piling. This is accomplished either by adding an extra pile wall thickness, which is technically called corrosion allowance, at the time of material production at the mill to make up for possible corrosion, or by spreading an adequate coating over the steel pipe surface. Nippon Steel Corporation has developed a number of corrosion-resistant steel and coated steel sheets using covering materials. As for the corrosion allowance, a corrosion rate of 0.03 mm/year can be selected for corrosion in soil above water level and 0.02 mm/year for the part in soil below water table. These rates can be refined in the detailed analysis.

4 CONSTRUCTION 4.1

Alignment

Alignment of the wall depends on the design. In general, straight line, circle line and their combination are often used. As the interlocking junctions can be welded at the designed place, tubular pipes can provide the curved alignment as well as the straight line, shown in Fig. 6 (Nippon Steel Corporation). For the case with curve and straight combination, the bracing

211 Copyright © 2004 Taylor & Francis Group plc, London, UK

Corrosion control

09096-26.qxd

2/18/04

12:33 PM

Page 212

Figure 6. The alignment of tubular pipe in a curved wall.

Figure 7. Mortar filling for waterproof.

and struts are normally used. In the case of the emergency shaft wall, braced supports will be used. 4.2

Waterproofing

Environmental site investigation finds the following minor environmental effects: soil with semi volatile organic compounds, petroleum-impacted soil, and groundwater with chlorinated volatile organic compounds near the emergency egress shafts. Waterproofing has to be guaranteed in order to not trigger the movement of the plumes. Steel tubular sheet piles are provided with interlocking junctions that have enough room to fill mortar concrete for water sealing of the junctions. There are two types of connections, P-T and P-P type interlocking connections. From waterproof point of view, pipe-pipe (PP) type will be used for its better waterproofing performance. The tubular sheet pile interlock should be fabricated to about 10 ft above the water table. The principle of waterproofing is to empty the interlocking connection by water jet and air lifting after piling work and then filling the room with mortar as shown in Fig. 7. Other than the retaining wall, the steel tubular sheet piles are often used as wall type foundation in water areas. That is because the steel pipe type sheet pile can serve as temporary water sealed cofferdam as well as the wall type foundation. Concrete work can be done in dry conditions inside the cofferdam of the steel tubular sheet pile wall. For instance, the maximum cut-off level was 89 ft in Tama River ventilation tower construction project in Japan. The construction completed successfully with only a few oozing out zones. 4.3

Misalignment

Based on the site condition in the ESA project, tubular sheet pile wall would be an excellent option for the shafts and approach structures. Giken’s equipment has been used successfully adjacent to live rail traffic in such a safe manner that construction was preformed without traffic interruption during peak traffic

conditions. This will provide the cost savings from having to work during strict working times, usually during the weekends at night. The pile installation is guided by a laser beam resulting in a remarkable alignment tolerance. The tolerance was less than 3 mm (1/8) in Long Island Expressway project. Giken’s system exhibits no perceived vibration during installation, which was proven by the U.S.A.C.E., New Orleans, so settlement to the adjacent ground is very limited. Settlement at the immediate adjacent area is generally limited to less than 1 inch in very dense gravel. Survey data recorded to date indicates that there was little to no settlement adjacent to the piles. Based on the soil conditions at the ESA project, the maximum settlement would be expected to be within 1/2 inch adjacent to the piles. More detailed analysis is needed to evaluate settlement of the bridge pier. 4.4

Due to the pile length of 120 ft and the soil conditions, a crush piler from Giken America Corporation is recommended for installing the tubular pile. This crush piler, which has been successfully used in similar soil, consists of a Silent Piler equipped with an integral augering method to advance the piles to the anticipated tip elevations. This system advances the tubular pile as far as it can through pressing techniques, at which time the Silent Piler can no longer press the pile any deeper, a continual flight auger is pitched and lowered on the pile. The auger is seated through a conventional crane on the pile with a series of hydraulic clamps. Once the auger is properly seated on the piler, and the head is lowered to the tip of the pile, the auger head is engaged. By having the auger fixed on the top of the pile, the auger head pulls the pile into the ground. This is done to the final tip elevation. The spoils generated from the augering are carried up a continual flight auger to the top of the pile where it is delivered to a chute, which drops the spoils into a spoil bucket. The spoils bucket is held in place with

212 Copyright © 2004 Taylor & Francis Group plc, London, UK

Tubular pile installation

09096-26.qxd

2/18/04

12:33 PM

Page 213

the crane. Once the bucket is full, it is lowered to the ground surface, when it is emptied into a stockpile, or directly backfilled into previously pressed piles. Once the pile is at sufficient depth, the auger motor is reversed and extracted with the crane and placed in a holder until the next pile. For a pile with a length of 120 ft, if a conventional crush system were used, it would require a continual auger system of over 120 ft in length and a crane with a minimum boom clearance of over 180 ft. A crush system newly developed by Giken will be proposed for this project, which will be similar to conventional crush system, however a mast leader system is used to allow for smaller spliced sections. This reduces the crane boom length to that of a medium sized hydraulic truck crane. This system is equipped with an auger lead system on which the auger motor is attached, similar to conventional augering rigs. This system, however, does not use a telescopic Kelly bar, sections are attached to the auger as the pile is advanced further in the ground. This system would allow for 40 ft long sections to be driven without requirement of a large crane. The 40 ft sections will be loaded with an auger in a horizontal position, prior to being hoisted to the piler. Once the sections have been inserted to the leader system and advanced for the next section to be attached to the auger. The pile section will be welded in the field. This system will be more suitable for the ESA project due to the presence of the catenary and compact construction space and will have a slightly better production rate in comparison with the conventional crush system. 4.5

Figure 8.

Inner brace system for tubular pipe.

Support system

In order to reduce the required modular of the wall and the lateral deformation of soil, brace system will be implemented in the excavation for emergency shafts. Tubular sheet piles can be fully equipped for additional bracing, struts, tiebacks, Wales, etc. Typically less bracing is required due to the large section properties of the tubular piles. Additional aesthetic facing can be fixed to the piles as either pre-cast units, or cast-in place. Studs will be required to be welded to the piles if the facing is cast in place. As to the inner support at the construction stage, steel braces and struts of H-Shaped will be used for the tubular pipe and sheet pile wall. In the tubular pile case, the space between the wall and brace should be filled with cast-in-situ concrete to distribute the pressure evenly to the pipe and avoid excessive deformation of the steel pipes (Fig. 8). As for the inner support used as the permanent structure, in general, reinforced concrete type braces and struts, or reinforced concrete slab are used. In these cases, shear connectors and steel tension bars are welded on the surface of the tubular piles. These bars

Figure 9. Brace system for emergency shaft.

are designed to transfer external loads between the reinforced concrete slab and tubular pile wall. The brace system for emergency shaft is schematically shown in Fig. 9. 4.6

213 Copyright © 2004 Taylor & Francis Group plc, London, UK

Cost efficiency

As for the emergency shaft, the unit cost for Giken system will be around $60 per SF, which does not include the cost of inner bracings and Wales. This is almost half the cost of slurry wall. The estimate includes Giken SCP-260 Crush Piler system and GRB system along with a 50-Ton clamp crane and pile runner. The cost estimate is based on a daily rental of Giken system with an assumed production rate from projects in similar soil conditions. The cost also includes the material including the shipping from Nippon Steel Corporation. Actual material price will likely vary depending upon the final design and material selected. Tax may vary depending on local tax rates.

09096-26.qxd

2/18/04

12:33 PM

Page 214

The second project was Long Island Expressway project, where a 35 ft cantilever wall was constructed on a steep slope with similar soil conditions to the ESA project (Fig. 11). 6 CONCLUSIONS Press-In Method is an advanced piling technique. Based on the analyses regarding the design and construction, press-in pile method can be used in the ESA project for supporting the construction of emergency shaft and approach structure. It is also cost efficient comparing to the slurry wall system.

Figure 10. Renovation project near railway in Japan.

ACKNOWLEDGEMENTS The authors are grateful for the helps of Mr. John Santos and Mr. Michael Carter of Giken American Corporation and Dr. Kazushige Tokuno, Mr. Shigeki Terasaki, and Mr. Takeshi Katayama of Nippon Steel Corporation. We also appreciate the editorial helps from Victor Shey and Ahmed Firoz.

REFERENCE Figure 11. Long Island Expressway extension project.

The cost estimate for the approach structures is based on the cantilever tubular wall. The embedded depth was determined from Plaxis analyses. The length of piles varies from 15 to 61 ft. The unit cost is similar to that in the emergency shaft, which is not economical because of high cantilever height near TBM reception pits. The cost can be reduced by introducing tie back system in the final design. Material cost can also be reduced by sourcing local market.

5 SUCCESSFUL PROJECTS Two successful case histories using press-in pile are introduced below. The first project, shown in Fig.10, was for a railway project in Aomori, Japan. It demonstrates the environmental friendly characteristics of Giken Press-in piling system, where the clearance was only 309 mm. Press-In Method will provide an excellent solution due to the physical restraints in the ESA, including live railway service.

PB/STV, ESA geotechnical design summary report for Queens segment, 2000. White, D. 2002. An investigation into behavior of pressed-in piles. PhD dissertation, Cambridge University, England. Coomber, D.B. 1985. “Groundwater Control by Jet Grouting.” Proc. 21st Reg. Conf., Eng. Group of Geolog. Soc., Sheffield, 485–498. Bruce, D.A., Boley, D.L. and Gallavresi, F. (1987). “New Development in Ground Reinforcement and Treatment for Tunneling.” 1987 RETC Proceeding, 2, 811–835. Bedian, M. 2002. “ ‘Value engineering’ in United States of America.” The 9th int. conf. on piling and deep foundations, Nice, France. Engineering News Record, 2001. “Japanese system quietly breaks ground on highway job.” July 9, 2001, 16. Giken American Corporation, Construction revolution guide, http://www.giken-smp.com/. Japanese Association of Steel Pipe Piles 2002, Steel Tubular pile foundation: Design and Construction, Nippon Steel Corporation. Nasri, V., Jafari, R. and Wone, M. 2003. East Side Access Project in New York, hard rock and soft ground tunneling. 12th PanAmer. Conf. SMGE, MIT Cambridge, MA, June 22–25. Nippon Steel Corporation 1987, Nippon steel’s steel pipe pile construction methods. Nippon Steel Corporation 1988, Steel pipe piles.

214 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-27.qxd

2/18/04

12:34 PM

Page 215

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Shotcrete for tunnel final linings – design and construction considerations V. Gall & K. Zeidler Gall Zeidler Consultants, LLC

N. Munfah Parsons Brinckerhoff Quade & Douglas, Inc.

D. Cerulli Parsons Brinckerhoff Construction Services

ABSTRACT: The use of shotcrete for tunnel final linings has gained increased popularity on a national and international basis. The high quality of the shotcrete material, flexibility in application and workability, as well as the ability to adapt to complex tunnel geometries have contributed to this popularity. When evaluating if shotcrete should be utilized as the final tunnel lining, several aspects should be carefully evaluated to determine the final product’s quality and durability, as well as cost and construction schedule implications for a given tunnel configuration. Among others, geometric complexity, tunnel length and size, staging of a multi-layered application, finish requirements and type of waterproofing will play a major role in the decision. This paper establishes and discusses aspects and criteria that should be considered in the evaluation process for, or against, a tunnel final shotcrete lining. This discussion is supported using recent case histories, in particular the Pedestrian Walkback Tunnel at Washington Dulles International Airport in Dulles, Virginia, and the Weehawken Tunnels in New Jersey for New Jersey Transit to demonstrate the decision process.

1 SHOTCRETE As reported in many documents, the material shotcrete has undergone significant developments during the past decade. Improvements of the material as well as the application method have been achieved. Intensive research in the material quality led to a better understanding of the interaction between the various constituents of a shotcrete mix, to the development of a series of new admixtures and better quality control of cement types. In particular, the use of wet mix techniques, the development of new low/non alkali accelerators, water content reducing admixtures and continuous cement quality resulted in improved final shotcrete quality. But also the use of fiber reinforcement and high-end concrete pumps and guns have furthered the shotcrete quality. The new materials have allowed better slump control, which did not only contribute to a more steady flow with the new pumps and therefore continuous shotcrete application, but much more to a more controlled and uniform compaction and, consequently, shotcrete density. The reduction of the W/C ratio, now

enabled by the use of plasticizers and partial replacement of cement, dramatically reduced the overall pore volume and, hence, improved the durability of shotcrete. With the help of the admixtures, the quantity of rebound was reduced to acceptable values, eliminating one economic disadvantage of shotcrete. With today’s shotcrete mix designs and application equipment, high final strengths of up to approximately 70 MPa (10,000 psi) are achieved in standard applications. Together with the use of shotcrete as permanent support material, requirements for the surface quality became more demanding. The improved workability, smaller aggregate grain sizes and better hydration heat control (cracks) enabled the contractors to satisfy these requirements. Trowel finished shotcrete surfaces (Varley 1998, Eddy & Neumann 2003) or architectural ornamental finishes (Gall et al 1998) are examples for shotcrete finishes achieved on past projects. The compressive strength of sprayed concrete is only an indirect indicator for the shotcrete durability. Durability and water tightness are intimately interconnected. Crack development and dispersion control

215 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-27.qxd

2/18/04

12:34 PM

Page 216

Figure 1. Trowel finished shotcrete lining.

and the volume limitation of the effective pores reduce the permeability of shotcrete. Values of 1012 m/sec, desired minimum values for sufficiently water tight and durable concrete, are achieved or even surpassed. If concrete is exposed to groundwater and no water flow exists across the concrete section, water absorption is of greater concern than permeability. The control of the volume of permeable pores within the concrete section and limitation to a maximum value of 14 to 17%, as recommended by various documents, is achievable in standard shotcrete applications. Fibers are not only used to better the behavior of shotcrete during fire, but also to increase the ductility of shotcrete and shrinkage crack control and dispersion. Above improvements combined with the inherent flexibility of shotcrete application resulted in a high acceptance of shotcrete within the industry and authorities. Shotcrete can be compared to high quality castin-place concrete and, in some fields, even proved to have superior characteristics. 2 LINING DESIGN PHILOSOPHIES During the history of tunnel lining designs, different lining philosophies have been developed. Dependent on the assumption, whether or not the initial lining will have sufficient quality and durability under the project specific conditions, the initial shotcrete lining has been taken into account for the long-term support, or has been considered sacrificial. In the latter, a secondary lining had to carry all expected ground and groundwater loads in the long term. The different water tightness criteria implemented at various projects under specific project conditions led to diverse waterproofing solutions, including the use of shotcrete for water tight linings, or the installation of membrane waterproofing systems sandwiched between initial and secondary lining.

In Europe, various authorities developed their preferences with respect to tunnel waterproofing systems. For example, most of the railroad and metro authorities in Germany and Austria tend to utilize shotcrete/ concrete to control the desired degree of tunnel water tightness, while the road and highway authorities prefer membrane waterproofing systems. The decision whether or not to use and be able to achieve a water tight concrete/shotcrete is also driven by the project specific environmental conditions, such as hydrostatic pressure conditions, chemical attack potential of the groundwater, and construction complexity. In some projects, the shotcrete initial lining has been considered sufficiently durable to withstand the longterm loads over the design life. The designers of several access shafts and stub tunnels for the upgrade project of London Electricity’s power supply network (London, UK) have opted to use the sprayed concrete lining, which was placed after excavation, for the long term support of these structures (Field et al 2000) as the so called Single Pass Lining. Specially detailed construction joints and high quality shotcrete were required to meet the client’s water tightness criteria. Damp patches were acceptable. The lining design thickness was considered appropriate to provide sufficient long-term stability, even when a certain portion of the shotcrete lining exposed to ground and groundwater will degrade. Similar to the classical two-pass lining systems with water tight cast-in-place concrete secondary linings, sprayed concrete has been used in lieu of cast-in-place concrete. At the Jubilee Line Extension, Contract C104 – London Bridge Station (London, UK), the complex geometry and alignment of the ventilation tunnels and the step-plate-junction housing a track bifurcation instigated the contractor to install a shotcrete lining on the inside of the initial lining (Varley 1998). The design was based on the assumption that the initial lining would deteriorate over the years and would lose its support capacity. The secondary lining has to carry all ground and hydrostatic loads expected to act during the design life. The water tightness criteria, where damp patches were permitted, were met by a high quality, steel fiber reinforced shotcrete and specially designed construction joints. A finishing layer of plain, small size aggregate shotcrete was applied to cover the steel fiber reinforced shotcrete. To meet the smoothness criteria for the ventilation tunnels, the finishing layer received a trowel finish. Similar principles have been applied at the ventilation chambers for DART’s City Place Station Project in Dallas, TX (Ugarte et al 1996). An early application of composite shotcrete linings was the lining system installed at the Heathrow Airport Transfer Baggage System Tunnel (Arnold & Neumann 1995). The shotcrete initial tunnel support was designed to provide the long-term ground support, while a

216 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-27.qxd

2/18/04

12:34 PM

Page 217

secondary layer will provide support to the hydrostatic loads. Both shells are interlocked by a rough, prepared joint surface and cross reinforcement and are expected to act as a composite structure with load sharing between the shells, effectively forming a single shell lining. Water tightness criteria, a dry tunnel had to be supplied, were achieved by high quality shotcrete and the continuous secondary layer of approx. 100 mm (4 in) thickness. Requirements for the composite function of the shotcrete layers and the shotcrete product itself have been identified by, among others, Kusterle & Lukas (1990) and Kupfer (1990). The more traditional two-pass lining system, combined with a membrane waterproofing system, is currently being applied at the Russia Wharf Segment in Boston, MA for MBTA’s Silverline Extension (Zachary 2003). There, the initial shotcrete lining is expected to deteriorate over time under the onerous environmental project conditions. A secondary shotcrete lining is being installed to provide long-term support to full overburden ground loads, surcharge and hydrostatic loads. A full-round membrane waterproofing system completely wraps the twin tunnels to provide a dry tunnel environment and to protect the secondary lining from potentially adverse groundwater affects. High quality shotcrete is used for the long-term support. Similar principles have been applied at WMATA’s Contract B10, Washington, DC for the construction of the double cross over and ventilation chambers in the mid 1980’s. Detailed design and practical considerations are described below based on a similar application at the Pedestrian Walkback Tunnel (PWT) at Washington Dulles International Airport (Hirsch et al 2003) and the Weehawken Tunnel project, in Weehawken, New Jersey (Ott & Jacobs 2003). These also include aspects of a layered shotcrete lining application. The PWT is approximately 240 m (800 ft.) long with a springline diameter of ca. 12 m (42 ft.) and features a double lining system, whereas a continuous PVC waterproofing membrane separates the initial and final linings. The Weehawken Tunnel involves the re-construction (enlargement) of a 1,269 m (4,156 ft) long, existing railroad tunnel into a two-track light rail tunnel with an underground station and a large passenger access and ventilation shaft. The widening of the tunnel to the station structure comprises a widening from an 8.4 m (28 ft) wide tunnel to an 18 m (60 ft) wide station tunnel structure to both sides of the future center platform station. Based on a Value Engineering Change Proposal submitted by the contractor, this transition, designed in a step plate junction configuration per contract, will be carried out using shotcrete for the arch final lining in a bifurcation as shown in plan in Figure 2. Another concept of lining design is currently being applied at the King’s Cross Station Redevelopment

Figure 2. Concrete vs. Final shotcrete lining geometry in plan and longitudinal section (schematic).

Project, London, UK (Cox et al 2003). The complex geometrical and alignment conditions, as well as the multiple tunnel junctions and intersections proved castin-place concrete secondary lining an uneconomical solution. Hence, the lining system will comprise a steel fiber reinforced shotcrete initial lining, a full round membrane waterproofing system (for completely dry tunnels) and a steel fiber reinforced shotcrete secondary lining. Rebar or welded wire fabric reinforcement may be required around tunnel junctions. Due to the rather benign environment offered by the surrounding London Clay and the groundwater contained in it, it has been decided to take some benefit from the initial shotcrete lining for the long-term support. The initial lining is not expected to completely deteriorate and lose its support capabilities. This is made possible in part by new shotcrete technologies, producing highdensity shotcrete, steel fiber reinforcement and a better understanding of the ground and groundwater impact on sprayed concrete. Part of the initial lining is expected to deteriorate over time, while the remaining portion will contribute to the ground support in conjunction with the secondary lining. Due to a requirement by the owner, all steel reinforcement forming parts of the permanent tunnel support must be located inside the membrane waterproofing system. Therefore, no benefit can be taken from any steel reinforcement located within the initial lining. The initial lining is taken into account as mass concrete material that will contribute to the support in confinement. The shotcrete secondary lining will, protected by the waterproofing system, provide the long

217 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-27.qxd

2/18/04

12:34 PM

Page 218

term support for the hydrostatic loads and the remaining part of ground and surcharge loads. The waterproofing membrane, sandwiched between the initial and secondary lining, is expected to permit radial load transfer only with negligible shear transfer between the linings.

3 GENERAL APPLICATION CRITERIA Shotcrete final linings are typically utilized where one or more of the following conditions are encountered:

•

• •

The tunnels are relatively short in length and the cross section is relatively large and therefore investment in formwork is not warranted, i.e. tunnels of less than 150–250 m (400–600 ft) in length and larger than about 8–12 m (25–35 ft) in springline diameter. The access is difficult and staging of formwork installation and concrete delivery is problematic. The tunnel geometry is complex and customized formwork would be required. Tunnel intersections, as well as bifurcations qualify in this area. Bifurcations are associated with tunnel widenings and would otherwise be constructed in the form of a step plate junction configuration and increase cost of excavated material (see Figure 2).

If the above conditions characterize a tunnel structure then a shotcrete final lining is likely to provide for flexibility in production, schedule advantages, savings in formwork and possibly savings in excavation. Therefore, a detailed shotcrete final lining cost analysis is warranted.

4 FINAL LINING EQUIVALENCY CONSIDERATIONS 4.1

Structural calculations

Structural calculations for final shotcrete linings follow the same principles and are based on the same structural codes as concrete linings. With current high shotcrete product quality and knowledge of application procedures, shotcrete is internationally viewed as concrete applied by different placement means. Due to the application process however, the reinforcement may, and in most cases will, be different in a shotcrete application. Whereas in a regular concrete section two layers of rebars at a wide spacing are sufficient, the shotcrete section will utilize welded wire fabric for better embedment within the shotcrete and to facilitate the shotcrete application. Where the loading conditions for the lining are well established, the same loadings are used in a structural calculation to arrive at reinforcement needs. Alternatively, equivalency considerations

may be applied, equating the given concrete section and its reinforcement to a proposed new section with a different reinforcement arrangement. The PWT shotcrete final lining reinforcement needs were a result of equivalency considerations, i.e. the reinforced shotcrete lining had to provide the same capacity as the castin-place concrete lining. An exception was the complex three-dimensional section between the mechanical room tunnel and the main tunnel where additional reinforcement beams were installed at the intersection along the groin lines (Figure 4). When considering the application of a final shotcrete lining, the following aspects should be addressed prior to acceptance and execution in the field. 4.2

In principle, there is no structural difference between a sprayed or cast-in-place concrete lining. However, when the sprayed lining is applied in multiple layers with distinct time intervals, which include installation of reinforcing steel, the bond between the different layers has to be adequate to qualify as a monolithic member in the structural sense. Limitations and requirements are therefore imposed on application sequencing, curing techniques, cleaning of surfaces and adapted concrete technology (Hoehn 1999). Keeping the time lag between shotcrete applications short aids this process. For verification, minimum tensile and shear strengths between the layers (in the joint) shall therefore be achieved. For example and to assess the requirements for these values at the PWT project, finite element calculations were carried out that considered a representative three-layer composite system with two joint surfaces in the final lining section (see Figure 3). The model investigated the capacity of the 30 cm (12 inch) layered shotcrete final lining for the long-term condition, when the initial support is assumed to be deteriorated and overburden and live loads are imposed onto the final shotcrete lining. From this model, minimum tensile and shear strength requirements in the joints were derived to be 0.69 MPa (100 psi) and 1.38 MPa (200 psi) respectively. Hoehn, 1999 for example calls for minimum values for strength for both tension and shear of 1.5 MPa (217.5 psi). Kusterle and Lukas, 1990 rather report ranges of values to account for statistical characteristics of sampling and testing. A review of these ranges, combined with the fact that the literature reports 1.5 MPa for tensile strength as a “universal number” and the availability of detailed calculations led to the conclusion that the above minimum values for tensile and shear were plausible. 4.3

Testing

Testing requirements for a final lining shotcrete resemble very much those of an initial shotcrete lining,

218 Copyright © 2004 Taylor & Francis Group plc, London, UK

Multi-layered vs. Monolithic

09096-27.qxd

2/18/04

12:34 PM

Page 219

Figure 3. FE model of shotcrete lining layers.

Figure 5. Final lining shotcrete application at PWT.

Figure 4. Shotcrete final lining installation at PWT intersection.

however with modified requirements, in particular to test for the bond capacity of the layered shotcrete. The shotcrete mix design is often developed based on historical data available from the initial lining application. At the PWT project pre- and during production testing requirements involved testing of tensile and direct shear tests on samples taken from test panels sprayed according to application and curing conditions resembling the site application, considering that the full thickness of the final shotcrete lining was to be achieved in panels not to exceed 10 m (30 ft) in length. Tensile strength was tested according to ACI 506R, whereas the shear tests were carried out according to Michigan DOT’s shear test. Minimum test requirements were as per the above, 0.69 MPa (100 psi) for tensile and 1.38 MPa (200 psi) for shear

strength. During pre-construction, testing time intervals between applications of 24-hours and 72-hours were tried and led to strength developments yielding a minimum of 2 MPa (290 psi) in tensile strength and 4.70 MPa (680 psi) in shear after ten days. During construction, a total of four tests with two samples each were required for the entire tunnel, again time lag and application to simulate application and site conditions. The minimum tensile strength developed at three days was recorded as 0.8 MPa (116 psi), with an average of 1.47 MPa (213 psi). The minimum shear strength at three days was 5.03 MPa (730 psi), with an average of 6.83 MPa (990 psi). Therefore, test results showed that the minimum bonding requirements of the composite final shotcrete layer were well achieved by the selected construction process. Application of the shotcrete final lining is shown in Figure 5. 4.4

The use of a dedicated waterproofing layer between the initial and final shotcrete linings creates a debonding effect. The degree of de-bonding depends on the type of waterproofing selected. In particular when using a loosely laid, continuous, flexible membrane type waterproofing (PVC) for complete water tightness (Gall 2000), special attention has to be given to membrane attachment, reinforcement installation and

219 Copyright © 2004 Taylor & Francis Group plc, London, UK

Waterproofing and contact grouting

09096-27.qxd

2/18/04

12:34 PM

Page 220

to contact grouting. A frequent use of attachment disks will achieve a tighter fit of the membrane to the initial shotcrete lining and reduce the amount of void space otherwise created by sagging membrane sections. For the spraying of shotcrete against the membrane, a carrying layer of welded wire fabric will be required. Spacers may be used between the welded wire fabric and the membrane to push the membrane further against the initial shotcrete lining. Despite these measures, a void space will exist between the membrane and the initial shotcrete lining. For proper contact between the initial and final shotcrete linings, systematic contact grouting is essential. This contact grouting, unlike the one in roof sections in cast-in-place final lining installations, is not limited to roof sections only, but a radial and more frequent distribution of grouting ports and pipes around the lining perimeter should be considered for this purpose. By injecting low viscosity cementitious grouts between final shotcrete lining and the membrane will assure a tight contact between the initial and final lining. Where water barriers have been utilized for the purpose of enhanced membrane repair (compartmentalization) a re-injectable grouting hose should be installed in the centerline of the barrier, between the ribs. Injection of grout through this hose will assure a tight embedment and contact between the ribs and shotcrete, and thus prevent leakage water to migrate across water barrier ribs. 4.5

Surface finish

There are various aspects of surface finish requirements that strongly depend on the tunnel’s intended use. These include, but are not limited to, reflectivity (in vehicular tunnels), ease of maintenance (washable), smoothness (in ventilation tunnels), appearance (general), and frost resistance (exposure to cold climates). For all of the special applications solutions exist and include screeding and trowel finishing, use of special mix shotcrete, and very fine aggregates for the finishing layer, yielding surface finishes that, by appearance and function, very well compete with the cast-in-place concrete. However, such surface finishes are often not required and omission of special finishes provides for further economy. At the PWT, for example, an internal architectural finish will be used. Therefore only limited requirements for the surface were established for ease of maintenance and facilitate installation of embedments and a flatness/smoothness criterion, which called for a deviation of not more than 2.5 cm (1 inch) in 1.5 m (5 ft.), was established. 4.6

Fire resistance

Recent fire incidents, in particular in European tunnels, have initiated numerous investigations in adequate

fire testing and the improvement of the fire resistance of concrete and sprayed concrete. One prime element contributing to spalling and subsequent section thickness loss has been identified: The free water contained within the concrete section leads, when evaporating due to rapidly increased temperatures, to explosive spalling of the concrete. Tests have proven that the addition of microfilament fibers to the shotcrete mix significantly improves the fire resistance of shotcrete. The fibers melt under the influence of heat and provide escape channels for the vapor, allowing the pressure to dissipate (Tatnall 2002). A detailed review of fire resistance needs at the Weehawken Tunnel led to the application of 1.9 kg/m3 (3 lbs/cy) of microfilament fibers for the inner 10 cm (4 inch) of the shotcrete final lining in transition sections.

4.7

• • • • • • •

Execution of Work (Installation of Reinforcement, Sequence of Operations, Spray Sections, Time Lag) Survey Control and Survey Method Mix Design and Specifications QA/QC Procedures and Forms (“Pour Cards”) Testing (Type and Frequency) Qualifications of Personnel Grouting Procedures

5 SUMMARY AND CONCLUSION Based on general trends in the application of shotcrete for final linings and as demonstrated on recent case histories, it is apparent that shotcrete presents a viable alternative to traditional cast-in-place concrete. The product shotcrete fulfills cast-in-place concrete requirements, or sometimes can even surpass those. Design and engineering, as well as application procedures, can be planned such as to lay the basis for a high quality product. However, excellence is needed in the application itself. Skilled nozzlemen have to ensure a high degree of workmanship through formalized training, experience and quality assurance during application.

220 Copyright © 2004 Taylor & Francis Group plc, London, UK

Method statement/application procedures

Probably the most important factor that will influence the quality of the shotcrete application is workmanship. While the skill of the shotcrete applying nozzlemen (by hand or robot) is at the core of this workmanship, it is important to address all aspects of the shotcreting process in a method statement. This method statement becomes the basis for the application procedures, the applicator’s and the supervision’s Quality Assurance/ Quality Control (QA/QC) program. Minimum requirements to be addressed in the method statement are as follows:

09096-27.qxd

2/18/04

12:34 PM

Page 221

ACKNOWLEDGEMENT The authors would like to acknowledge several firms and individuals for the information that forms the basis for the two projects discussed in detail. The Pedestrian Walkback Tunnel: Metropolitan Washington Airports Authority (Project Owner), Parsons Management Consultants (Construction Manager). The Weehawken Tunnel: New Jersey Transit Authority (Project Owner, Mr. Anthony Murtah, Mr. Ty Dickerson), Parsons Brinckerhoff Quade and Douglas (Prime Design Consultant), and Frontier-Kemper/ Shea/BuM Joint Venture (Contractor, Mr. Vincent Sambrato, Mr. Leon Jacobs). REFERENCES Arnold, J. & Neumann, Ch. 1995: Umsetzung eines innovativen NÖT-Konzeptes im Zuge eines “Know-howTransfers”. Felsbau 13 (1995), No.6, 459–563. Cox, R., Dulake, Ch. & Eddie, C. 2003: Complex redesign for London link. Tunnels and Tunnelling International, Vol. 35, No. 4, April 2003, 50–52. Eddie, C. & Neumann, Ch. 2003: LaserShell leads the way for SCL tunnels. Tunnels and Tunnelling International, Vol. 35, No. 6, June 2003, 38–42. Field, G., Legge, N. & Liew, B.S. 2000: Optimizing Shaft Design and Construction Using Sprayed Concrete. Our World in Concrete & Structures, Proc. 25th Anniversary Conference, Singapore. Gall, V., Zeidler, K., Predis, T. & Walter, J. 1998: Rehabilitation concepts for brick lined tunnels in urban areas. Tunnels and Metropolises, Proc. World Tunnel Congress Sao Paulo, Vol. 1, 539–546, Rotterdam.

Gall, V. 2000: Three Pillars for an Effective Waterproofing System. Proceedings, North American Tunneling 2000, Boston, Massachusetts, June 6–11, 2000. Hirsch, D., Moran, P. & Patel, A. 2003: Tunneling Under Washington Dulles International Airport. Proceedings, Rapid Excavation and Tunneling Conference 2003, 648–656. Hoehn, K. 1999: The Single-Shell Shotcrete Method Applied at Two Tunneling Sites – Concrete Technology and Economic Viability. Proceedings, Spritzbetontechnologie ’99, BMI 1/99, 255–270. Kupfer, H. & Kupfer, H. 1990: Statical Behavior and Bond Performance of the Layers of a Single Permanent Tunnel Lining, Proceedings, Spritzbetontechnologie ’90, 11f. Kusterle, W. & Lukas, W. 1990: High-Grade Shotcrete for the Single Permanent Shotcrete Lining Method, Proceedings, Spritzbetontechnologie ’90, 29–40. Ott, K. & Jacobs, L. 2003: Design and construction of the Weehawken Tunnel and Bergenline Avenue Station. Proceedings, RETC 2003, 936–946. Schreyer, J. 1999: Constructive and Economical Suggestions for the Lining of Single Shell Tunnels. Proceedings. Spritzbetontechnologie ’99, BMI 1/99, 271–281. Schwarz, J. 1999: Structural Design and Quality Assurance of the Joint between Outer and Inner Layer when Using the Single Shell Shotcrete Lining Method. Proceedings. Spritzbetontechnologie ’99, BMI 1/99, 237–240. Tatnall, P. C., Shotcrete in Fires: Effects of Fibers on Explosive Spalling. Shotcrete, Vol. 4, No. 4, Fall 2002, 10–12. Ugarte, E., Gall, V. & Sauer, G. 1996: Instrumentation and its Implications – DART Section NC- 1B, City Place Station, Dallas, TX. Proceedings, North American Tunneling ’96, April 21–24, 1996. Varley, N. 1998: Concrete tunnel linings at London Bridge. Concrete, Feb. 1998, 13f. Zachary, W. 2003: The Cold War: Boston’s Uncommon Dig. AUA News, Vol. 18, #3, 9–11.

221 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-28.qxd

2/18/04

12:35 PM

Page 223

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Robotic shotcrete applications for mining and tunneling Michael Rispin Underground Construction and Allentown Equipment Master Builders, Inc.

Chris Gause Underground Construction Master Builders, Inc.

Thomas Kurth Meyco Equipment

ABSTRACT: While shotcrete has evolved as a means and method for ground control, so too have the demands for faster and safer placement. Spraying manipulators, or robots as they are commonly referred to, have become the rule rather than the exception both in mining and larger tunnel projects. Even after the capital investment of a robotic shotcrete machine, the benefits can be measured and returns achieved by:

• • • •

Increased production Reduction of rebound Higher quality shotcrete in-situ Improved safety for shotcrete crews.

This paper discusses state-of the-art mechanized shotcrete machines and provides case histories describing the benefits in mining and tunneling.

1 INTRODUCTION If a Robotic Applicator is mentioned in connection with mining or tunneling and sprayed concrete, what is basically meant is an apparatus used to hold and control a spraying nozzle. Why should this be necessary when a man can do the same work? Tunneling and mining development are intrinsically hazardous forms of construction, when sprayed concrete is used as initial temporary support after blasting, using a mechanical arm to extend into an unsupported area is a great enhancement to personnel safety. A spraying manipulator is a hydro-mechanical, remote-controlled spraying unit for mechanizing and automating the application of sprayed concrete. It is suitable for use anywhere substantial quantities of wet or dry shotcrete will be applied, and offers significant advantages in construction applications where conditions are such that manpower might be exposed to potentially unstable, unsupported ground, rebound or dust. Mounted on various kinds of carrier vehicles and able to achieve a reach of up to 14.5 m (47 ft), a

robotic applicator will save the cost and time of erecting scaffolding, where due to the very size of the working area it would otherwise be needed. As this paper will show, there are many combinations and permutations of configurations of robotic applicators in use today around the mining and tunneling worlds. 2 A HISTORY OF ROBOTIC APPLICATORS Thirty years ago, the first manipulators really were just nozzle holders. Over the next twenty years, innumerable variations appeared in all parts of the world based upon cranes, drill jumbos and lifts with a device enabling the nozzle to be attached. These assemblies were not designed for quick and nimble nozzle and arm movements, so efficient placing of quality shotcrete with a smooth finish on difficult substrate was extremely difficult, if not impossible. Specialized spraying manipulators began to appear in the early 1980s, by which time sprayed concrete had become an acceptable form of construction (if only

223 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-28.qxd

2/18/04

12:35 PM

Page 224

by the dry method). The most suitable of these dedicated units had features that are still standard today: spraying heads with universal movements, eight fields of motion, and the “parallel-lance” with great extension possibilities. These arms were generally mounted onto existing carrier vehicles; used trucks and excavators were favorites. Features such as automation of movements of lance and nozzle holder were integrated into remote controls to help the nozzleman to produce a better spraying pattern. By the mid 1990s, with the proliferation of large scale wet-spraying, the spraying manipulator had firmly established itself as a piece of equipment to be found on almost all large construction projects where sprayed concrete was used as temporary or permanent support. But the demands made upon the manipulators had changed: the “bar” had been raised! A mechanical device to hold a nozzle was no longer enough. New standards and economic constraints demanded more speed and efficiency in placement. This meant that a manipulator had to be able to hold and control a nozzle and hose with diameters of up to 80 mm to enable the full capacity of the shotcrete pump to be used, in order to save time and therefore money. This required not just robustness, but also operational dexterity to allow large amounts of concrete to be placed quickly and accurately, typically impossible with a converted placing boom. Remote controls also developed, from hydraulic levers to electric operation with cables, and later radio remote control became a standard option. This period in time also saw the development of the autonomous spraymobile. These vehicles were trimmed from top to bottom with all the equipment necessary and with one aim in mind: quality sprayed concrete. Manipulators became very diverse and specialized, as construction was customised to be exactly suited to application conditions, be it for large civil projects, tight mining tunnels, shafts or even integrated into a TBM. The present day sees the demand for more quality and accountability in both results and the application process on site. Even more automation is required. There is only one way these attributes can be assimilated into a robotic applicator and that is through the use of computer technology. By 2000 the first computer-controlled robots had appeared. Able to be programmed to spray an area automatically and keep records of the work, this advance opened up vast new possibilities in improving tunneling and mining safety, economy and efficiency. Computer control eliminated the need for a nozzleman to work continuously close to the “danger” area. The required finished surface accuracy increased as the machine, coupled with the computer through laser measuring technology, worked much more precisely than a human. The fatigue and skill factor variables were removed from the equation.

Automation holds great advantages. In deep mines, for example, long travelling times and short shifts can be replaced with full employment of resources by a nozzleman who sits safely on the surface controlling processes through his MMI (ManMachine-Interface). Simpler units can be equipped with “teach-in” features that repeat various patterns. Work in hostile environments, such as a uranium mine, can now be tackled with much less risk. The future will belong to these types of robotic applicators, but there will also be place for the dedicated standard hydro-mechanical manipulator. 3 ROBOTIC SPRAYING VS. HAND SPRAYING The benefits of mechanized shotcrete application can be evaluated by three categories: 1. Increased production 2. Higher quality shotcrete in-situ 3. Improved worker safety. 3.1

A multitude of reasons exists which allows increased production with the use of a shotcrete robot, most of which are due to the elimination of the human fatigue factor. The predominant reasons are as follows:

•

•

•

224 Copyright © 2004 Taylor & Francis Group plc, London, UK

Increased production

Increased concrete hose diameter – Some spraymobiles are equipped with 102 mm (4.0 in) hoses. Hand nozzling will typically use a 51 mm (2.0 in) hose diameter. The weight of shotcrete in the 102 mm (4.0 inch) line is equal to 18.3 kg/m (12.3 lbs/ft). When you multiply this by 1–2 m (3–6 ft) of hose length often being supported by the nozzleman, combined with the compressed air supply, any person would quickly become exhausted. Fatigue also carries over to pumping rates or pump output. As the shotcrete output is increased, the nozzleman must also resist the increase in line surge that comes from temporary interruption of pumping while the swing tube changes cylinders and begins the next stroke. The nozzleman in a sense must act as shock absorber. In addition to pump surges, the compressed air (5–7 m3/min for hand spraying, whereas robotic spraying involves 10–14 m3/min with 7 bar pressure) delivered to the nozzle body also applies a backward pressure that must be compensated for by the nozzleman. This additional fatigue factor is of course eliminated with mechanized spraying equipment. With the human fatigue factor eliminated, shotcrete volumes can increase dramatically. Hand nozzling volumes can range from 7–9 m3/hr (9–12 yd3/hr), while mechanized spraying can easily reach volumes

09096-28.qxd

2/18/04

12:35 PM

Page 225

of 20 m3/hr (26 yd3/hr). This is particularly beneficial in larger diameter tunnels, stations, galleries or when shotcreting for the final lining is being utilized. 3.2

Higher quality shotcrete in-situ

There are combinations of capabilities with mechanized spraying that allow shotcrete to be placed with improved in-situ properties. Some of these are: 1. Dedicated maximum air volume for optimum compaction 2. Lance mounting is automatically held parallel to the axis of the tunnel 3. New robotic manipulating capabilities also allow for automated nozzle adjustments to be made to maintain proper standoff distance as well as nozzle angle to the substrate. 3.3

Improved worker safety

The contributions to a safer working environment via robotic spraying are clear. With use of a remote control, crews are able to remain in supported areas while letting the reach of the spraymobile apply shotcrete in the newly excavated areas. In areas that require a combination of rock bolts and shotcrete, the bolting crews can take advantage of working in a supported environment where an initial layer of shotcrete has been sprayed for temporary support.

4 STATE-OF-THE-ART ROBOTIC APPLICATORS In producing top quality sprayed concrete, the best manipulator is still only one component of a system. The complete system is imperative if the manipulator is to be used to its full potential. On large construction sites such as tunnels, it is imperative that the spraying set-up is installed and ready to start performing within minutes of the heading being ready for it. As soon as the spraying operation is finished, the equipment has to be removed so that the next work cycle can begin. Furthermore, it is a common trend to execute different jobs simultaneously, which demands complete, self-contained equipment. For example, because a central air supply is seldom large enough to supply all site demands, the complete mobile therefore carries its own compressor. 4.1

Meyco Potenza

The Potenza is one of the better examples of a complete mobile unit for the spraying of concrete. This type of spraymobile has been setting the standard for sprayed concrete in tunnels and other areas of application

using the wet-mix shotcrete method. They have become commonplace on many of the world’s most important sites where sprayed concrete must be applied in large quantities without compromising quality. The standard components of the complete mobile unit are:

• • • • • • • • • • • • • • •

225 Copyright © 2004 Taylor & Francis Group plc, London, UK

Robojet spraying manipulator Potenza sprayed concrete pump for wet-mix process Integrated Dosa TDC accelerator dosing unit MEYCO Data for compiling operating and performance information Central power unit Chassis, 4 wheel drive and steer, with stabilisers Cable reel Air compressor Nozzle system Liquid accelerator tank Water storage tank Working lights Water pump High pressure water cleaner Release oil pump.

09096-28.qxd

4.2

2/18/04

12:35 PM

Page 226

Allentown MBS-02E

The MBS-02E features most of the sprayed concrete capabilities outlined for the Potenza but is built as a smaller and more robust package designed for the smaller mining headings and the rigors of the North American underground mining environment. Most importantly, while its primary purpose is to spray concrete, it is also capable of quickly and efficiently travelling the underground tunnels and ramps in a typical mine in order to be efficient in its utilization in multiple headings. The spraying manipulator, designated Meyco Minima, boasts a folding boom that retracts for tramming, yet is unfoldable in a 3 m 3 m heading (10 ft 10 ft), and offers a maximum spraying range of 9 m in height (29.5 ft), 7 m lateral (23 ft), and 8 m forward (26.2 ft). Due to frequently encountered, unexpected conditions in a mine, the spraymobile is also equipped with a shotcrete pump and hopper assembly that can be hydraulically positioned at various heights to adjust to any type of feed, typically from a transmixer, even when parked on uneven surfaces.

enable movement, dexterity and ease of handling for the nozzleman. A large ring, as shown above, would be similar to equipment used in the Lötschberg Transalpino Project in Steg and Raron.

4.4

4.3

TBM ring construction sprayer

A TBM can also be viewed as a carrier vehicle. Manipulators on TBMs should be, of course, part of an integrated system designed with the quality of end product, i.e. sprayed concrete, in mind. TBMs vary greatly in their individual construction, dictated by the geological environment where they will be employed. This in turn influences the design of the manipulator. Space is at a premium and the logistics are difficult so a manipulator must have the greatest possible movement but not clutter up the already crowded back-up rig. MEYCO has manufactured both ring construction type manipulators and centrally placed lance units according to local requirements. These units are always tailor made, but should contain all the basic principles and components to

226 Copyright © 2004 Taylor & Francis Group plc, London, UK

Shaft robo

09096-28.qxd

2/18/04

12:35 PM

Page 227

Shaft manipulators actually have a lot in common with ring constructions made for TBMs. The big difference is obviously the angle and direction in which the carrier vehicle, in this case a Galloway stage, is either lowered or raised within a vertical excavation. Again, the manipulator must be an integral part of a coordinated sprayed concrete system. Depending on the diameter of the shaft, a centrally mounted lance or a ring running around the stage would be used. 4.5

Meyco Oruga

The Rama is a range of manipulators manufactured by MEYCO. Their common features are that they can be mounted on almost any type of carrier vehicle or mounting stage. They are all of robust and simple construction and they vary in that each model has a different maximum spraying range, derived from their physical dimensions. They all have a spraying head with two hydraulic oscillating motors with nutation device transmitting the required wobble movement to the spraying nozzle; adjustable speeds allowing optimum nozzle positioning. 5 CASE STUDY – MINING – KIDD CREEK

The Oruga is small and compact when driving around on its tracked carrier. It also has a reach of up to 8 metres and reliable stability when spraying. The Rama 4 manipulator operation is through electric remote control. They are ideal for slope protection and are compact enough to work within a TBM back-up rig. 4.6

Meyco Rama

The Kidd Creek Mine is located in Timmins, Ontario, Canada, where copper–zinc–silver deposits were discovered in 1963. Owned by Falconbridge Limited, it was put into production in 1965 with an open pit mine, which was excavated from 1965–1977. Subsequently, the ore body has been mined through three separate shafts known as the No.1, No.2, and No.3 mines. For years, Kidd Creek used bolt and screen construction for primary ground support. Dry shotcrete was used as secondary reinforcement and for repair where needed. At the end of the 1990s the company began searching for a better, faster, safer ground support method. In approaching the search for a new ground support protocol, the challenge was to develop a system that would be safe and economically feasible to apply in a complex and deep mining environment, and would be accepted by workers in the mine. They needed a viable new ground support method that would reduce exposure to unsafe working conditions, and meet stringent government and company regulations. Early in the process, the workers were focused simply on finding a better, faster way of applying dry shotcrete – they were not considering wet shotcrete. The mine had tried wet shotcrete in the early 1980s and it was not a success, so they were reluctant to explore that alternative. In 1999 the mine explored

227 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-28.qxd

2/18/04

12:35 PM

Page 228

new technologies to help them meet their goal. They recommended an innovative wet shotcrete system using the latest automatic delivery and spray equipment as a solution that would meet all of the Kidd Creek goals – increased productivity, enhanced safety and improved structural performance. In July of 2000 the Falconbridge Board of Directors approved the development of a new mine on the Kidd Creek site. Known as Mine D (Deep), it will extend the mine from a depth of 2070 M (6800 ft.) to 3050 M (10,000 ft.). When completed it will be the deepest base metal mine in the world, and it requires a significant infrastructure, including a new internal underground shaft, hoisting facilities, and approximately 15 kilometers (9 miles) of development. Started in 2001, the project is estimated to take four years to complete and approximately 100,000 m3 (130,000 yd3) of construction concrete and 60,000 m3 (78,000) of shotcrete will be used. As a result of the success of the SFRWS in field trials and subsequent use, Kidd Creek Mine management decided to use the system in the construction of Mine D. The mine commissioned a state-of-the-art on-site batch plant with capacity to feed two 200 mm (8) diameter cased boreholes to depths of 1400 m (4600 ft.) and 1460 m (4800 ft.) respectively. Five wet shotcrete spray mobiles and seven transmixers were acquired to meet the needs of the mine. Tenders were let and the mine chose the MSV shotcrete sprayer as supplied from MBT’s Allentown Equipment manufacturer. The MSV was designed especially to handle the underground environment. It features a robust carrier and utilizes some of the most effective and efficient drive components on the market. It has an overall tramming length of only (7.3 m) 24 ft. and a height of (228 cm) 90 inches. The sprayer was not only capable of higher and safer tramming speeds, it was also able to cover numerous headings in one shift. As of late summer 2003, the Mine D project had reached a level of 2438 m (8000  ft), with more than 14,000 m3 (18,200 yd3) of shotcrete applied using the MBT–MSV spraying units. An indicator of the improved safety is the fact that since fully implementing SFRWS as primary ground support in early 2002, there has not been one loose related injury. 6 CASE STUDY – CIVIL – BERGEN TUNNEL Economics based upon productivity will vary based upon the tunnel size, mining cycle and purpose of shotcrete application. A comparison between productivity of mechanized spraying vs. hand spraying can be extracted from the Bergen Tunnel Rehabilitation Project, North Bergen, NJ and the Cameron Run Tunnel Rehabilitation Project, Alexandria, Va.

During rehabilitation of the Cameron Run Tunnel, the contractor, Merco, Inc., hand sprayed using a Reed B30 concrete pump. The Reed B30 has a theoretical output of 22.8 m3/hr (30.0 yd3/hr.). Typical actual volumes applied were 3.6 m3 /hr.(6.0 yd3/hr.) or 30.4 m3/shift (40.0 yd3/shift). On the Bergen Tunnel Rehabilitation Project, Merco/Obayashi, JV utilized a self-contained robotic shotcreting machine known as the Meyco Potenza (described above). The Potenza is equipped with a Suprema shotcrete pump with a theoretical output of 20.0 m3/hr (26.0 yd3/hr). With the use of the Potenza shotcrete robot, the shotcrete volumes on the Bergen Tunnel project reached an hourly average of 14.0 m3, (18.2 yd3/hr) with the best day (two shifts) being 168.0 m3, (218.4 yd3). In addition to the increased output, the shotcrete crew size for robotic spraying was reduced to 3 men vs. 5 men for hand spraying. Although specific dollar values were not applied to the cost reduction in comparison, the increased output and shotcrete manpower reduction made some obvious contributions to an in-place cost savings.

7 THE FUTURE 7.1

228 Copyright © 2004 Taylor & Francis Group plc, London, UK

Logica

Meyco Logica is a relatively new machine, based on the well known kinematic principles of the Robojet, and has been developed in cooperation with industry and academia. This manipulator with 8 degrees of freedom has a new automatic and human oriented control system. The new tool enables an operator to manipulate the spraying jet in various modes, from purely manual to semi-automatic and fully automatic, within selected underground areas. It is also able to measure the tunnel profile with a laser scanner. In one of the modes the operator uses a six directional joystick (Space Joystick). The calculation of the kinematics is done by the control system. A laser scanner sensor measures heading geometry and this information is used to control automatically the standoff distance and the angle of the spraying jet. The aim of this control is not to automate the whole job of spraying but to simplify the task and enable the operator to use the robot as an intelligent tool, and to work in an efficient way with a high level of quality. With a correct angle of application and constant spraying standoff distance, a remarkable reduction in rebound and therefore savings in cost is achieved. Further, if the heading profile is measured after spraying as well, the system will relay information on the thickness of the applied shotcrete layer, which up to today was only possible with core drilling and measurement. If an exact final shape of the heading profile is required, the control system is being developed

09096-28.qxd

2/18/04

12:35 PM

Page 229

The Westerschelde Project was an example of a successful project using 2 Logicas to spray a 50 mm lining of passive fire protection mortar with tolerance 4 . The total tunnel length was 2 6 km. In the Netherlands, Meyco supplied equipment used for the application of fire protection mortar to the Groene Hart tunnel project. To fulfill the standards to spray apply a defined, constant and homogenous layer of a passive fire protection mortar, preference was given to a jobsite tailored solution. The entire spraying equipment was placed on gantries allowing trucks supplying the TBM to pass underneath. Mechanical engineers designed and built a spraying nozzle mounted on a wagon travelling on a guide rail along the tunnel. The whole construction moves on a ring beam along the tunnel wall. All these movements can be conducted with “teach-in” functions to allow automatic spraying within a defined area. After a 4 m longitudinal length is sprayed, the gantry will be moved and the next spraying phase can be repeated. By the time the whole set up was commissioned a spraying accuracy of 2 mm by 35 mm thickness was being achieved! 8 CONCLUSION Sprayed concrete is an economic, efficient, and versatile means of ground support for modern mining and tunneling operations. As we learn more about the benefits that shotcrete technology can bring to our underground industries, its use will proliferate. Robotic applicators have already proven to be a useful, and sometimes indispensable tool, in the application of shotcrete. The advances chronicled above have been built on systematically developed experience, and as each case study is completed and analyzed, further developments and efficiencies will ensue. The new frontier is automated shotcrete application, with a very high degree of applied thickness control. While the technology is here today, it needs to be employed on a larger scale where its benefits will be brought to bear for tunnel owners and contractors, and mine operators across the globe.

currently to manage the robot to spray to these defined limits automatically. The system shows that increased productivity in shotcrete application is doubtlessly possible without increasing danger to personnel or without huge increases in cost.

REFERENCES Mergentime, S.: personal communications. Melbye, T., Dimmock, R. and Garshol, K.: “Sprayed Concrete for Rock Support”, 2001. Master Builders article: “Shotcrete Developments at Kidd Creek Mine.”

229 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-29.qxd

2/18/04

12:36 PM

Page 231

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Development of the LaserShell method of tunneling C.M. Eddie Morgan Est, Tunneling Division, Rugby, UK

C. Neumann Beton-und Monierbau, Innsbruck, Austria

ABSTRACT: This paper describes the development of a new sprayed concrete tunneling process known as LaserShell™. The system has been developed by Morgan Est and Beton-und Monierbau to improve the safety, quality and efficiency of underground works using sprayed concrete and is currently being used on major road, rail and sewerage tunnels in the UK. The profile of both the excavation and spraying operations is controlled using an innovative real time survey system known as TunnelBeamer™. The developments of the TunnelBeamer™ and LaserShell™ technologies are described in the paper. The paper also describes the development of a permanent, ultra high quality steel fiber reinforced sprayed concrete mix which has enabled lattice girders and bar and mesh reinforcement to be effectively eliminated from the sprayed concrete tunneling process. In recent years a concerted effort has been made by the UK tunneling industry to put in place robust risk mitigation measures to further improve the safety of tunnel workers at the face. Exposure to unsupported (or inadequately supported) ground is an area of undoubted risk, which needed to be addressed in relation to sprayed concrete lining methods (SCL). Although the UK Health and Safety Executive (HSE) regards tunneling with sprayed concrete lining as an effective and viable method of construction, the residual risk of exposure to unsupported ground is considered unacceptable. A recent major project utilizing a sprayed concrete lining was the North Downs Tunnel for the Channel Tunnel Rail Link constructed by Eurolink (Morgan Est, Dumez GTM and Beton-und Monierbau). Following a serious incident involving a fall of material from a tunnel roof, Morgan Est made an undertaking to the UK Health and Safety Executive, that developments would be made for future projects whereby tunnel workers would not be required to enter an area of inadequately supported ground. This required therefore that sufficient support, as determined by analysis, had to be in place prior to entry.

lattice girders to provide profile control of the lining. Lattice girders also provide a mechanism for securing mesh during the application of the primary lining (Figure 1). It is the installation of the girders and mesh together with profile checks, which place the tunnel workers in the exposed vault to an unacceptable risk. To eliminate this risk, a method of controlling the lining shape, thickness and position remotely has

1 REQUIREMENTS Traditionally, sprayed concrete lined tunnels in soils; unstable ground or shallow tunnels require the use of

Figure 1. Mesh and Lattice Girder Installation on the North Downs Tunnel part of the Channel Tunnel Rail Link, UK.

231 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-29.qxd

2/18/04

12:36 PM

Page 232

Figure 4. Single Shot TunnelBeamer™ instrument fitted to a tunnel excavator.

Figure 2. Single Shot TunnelBeamer™ instrument.

Figure 5. Single Shot TunnelBeamer™ instrument fitted to a tunnel excavator.

Figure 3. Multi Shot TunnelBeamer™ instrument.

been developed by Morgan Est, Tunneling Division and Beton-und Monierbau. The brief was to develop a user friendly system of tunneling (methods and equipment) to enable the real time control of profiles of both the excavation and sprayed concrete lining without the use of lattice girders or mesh. The tunneling system was also required to provide comprehensive documentation of “as-built” work for quality and certification reasons.

With the removal of the lattice girders, the excavation and spraying operations have no existing orientation line or physical profile control mechanism. The “TunnelBeamer™” system consists of either a single laser (Figure 2) or a number of lasers (Figure 3) grouped together to act as a distometer which are directed at the excavation or sprayed concrete lining faces as required. The information from these lasers is linked continuously to a tunnel computer (situated in the tunnel), which contains the 3D-tunnel geometry information and which produces a comparison to the theoretical position. This comparison information is continuously displayed on a monitor in the operators cab (Figure 6). The TunnelBeamer™ instrument does not need the level and stable platform necessary for standard survey

232 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-29.qxd

2/18/04

12:36 PM

Page 233

Figure 6. Driver’s continuous display.

Figure 8. Driver’s continuous display screen. Figure 7. TunnelBeamer™ system integration.

instruments and has been designed to withstand heat, dust and vibration levels commensurate with being fitted directly to excavation or spraying equipment. A servo theodolite is used to locate the TunnelBeamer™ in three dimensions to allow the tunnel computer to relate the TunnelBeamer™ information to the theoretical tunnel alignment/profile (Figure 7). The system utilises existing tried and tested hardware and software components. The TunnelBeamer™ system has been developed by Morgan Est and Beton-und Monierbau around a construction method known as LaserShell™. This method employs an inclined face excavation for increased stability and improved safety for tunnel workers. To minimise the number of construction joints and improve productivity, for tunnels up to 6.5 m diameter it is proposed that LaserShell™ will be constructed full face. For the larger tunnels, only the crown or pilot excavation would be undertaken using full face LaserShell™ techniques. The LaserShell™ method can be used for the construction of “One-Pass”, “Composite” or “Traditional” sprayed concrete tunnel linings, although clearly the One-Pass approach delivers maximum economy. In tunnels adopting the One-Pass philosophy, the initial 75 mm layer is considered to be sacrificial and is therefore not considered in the permanent load case.

To provide a robust face and vault support measures at all times, whilst allowing access to clean and prepare the invert prior to the construction of the structural lining, the LaserShell™ construction sequence can be visualised as follows (Figures 9–15). 2 BENEFITS As previously stated, the main purpose of this development programme was to eliminate the risk to tunnel workers in an exposed ground situation. However there are many other benefits to be derived from such systems. Speed of advance and improved ring closure times, in conjunction with an inclined face significantly reduce surface settlement. Removal of lattice girders and the replacement of mesh with High Carbon Steel Fibre substantially improve the quality and durability of sprayed concrete linings by eliminating shadowing. Systematic capture of profile data relating to both the excavation and the sprayed concrete lining gives absolute confidence with respect to lining shape, thickness and position (Figures 16 & 17). Compared to traditional SCL construction methods, an assessment on cost and time has shown that savings of up to 50% will be achieved in certain applications. This can be demonstrated by the fact that the often three-stage excavation process of

233 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-29.qxd

2/18/04

12:36 PM

Page 234

Figure 9. LaserShell™ stage 1 (excavation commences).

Figure 10. LaserShell™ stage 2 (excavate top 70% of face).

Figure 11. LaserShell™ stage 3 (spray 75 mm initial layer on top 70% of face).

234 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-29.qxd

2/18/04

12:36 PM

Page 235

Figure 12. LaserShell™ stage 4 (excavate invert and carefully clean).

Figure 13. LaserShell™ stage 5 (spray 75 mm initial layer in invert).

Figure 14. LaserShell™ stage 6 (spray structural primary lining 360 degrees).

235 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-29.qxd

2/18/04

12:36 PM

Page 236

Figure 15. LaserShell™ stage 7 (extend temporary fill in invert and repeat stages 1 to 6).

Figure 18. Crown, Bench and Invert, three-stage excavation process. Figure 16. Profile data for the finish layer (sprayed concrete layers and excavation profile data similar).

Figure 19. LaserShell™ excavation commencement development trials). Figure 17. Visualization of the Multi TunnelBeamer™ instrument during spraying.

(Pre-

Shot

crown, bench, and invert (Figure 18) can be replaced with a one-stage excavation process (Figure 19). For larger tunnels or tunnels in unstable conditions where some sub-division of the face is considered to be

prudent, the savings relate to the elimination of the lattice girders and mesh. To meet the requirements of LaserShell™, Morgan Est, Tunneling Division, Research and Development (in co-operation with Beton-und Monierbau and Prof. Dr. Wolfgang Kusterle of Innsbruck University) has recently completed extensive pre-commencement

236 Copyright © 2004 Taylor & Francis Group plc, London, UK

process.

09096-29.qxd

2/18/04

12:36 PM

Page 237

Figure 22. Morgan Est R&D facility, Permeability testing. Figure 20. Morgan Est R&D facility, Core bath.

Figure 21. Morgan Est R&D facility, Bump table testing.

development trials for an ultra high quality permanent sprayed concrete mix. The purpose of the trials was to develop a sprayed concrete mix and application methods that would meet all the requirements of a single pass, permanent sprayed concrete lining system. The key objectives of the trials was to prove the structural integrity from application (15 minutes strength) up to 120 years, whilst also providing a mix with sufficient workability retention to enable efficient application in real tunnel environments.

Detailed and onerous performance criteria in respect of strength gain; flexural toughness, permeability, bond characteristics between layers and durability were set prior to the commencement of trials and benchmarked against comparable high quality cast in place structural concrete (Figures 20, 21 and 22). Early age strength development relative to workability retention times was a key factor and despite claims of all the admixture suppliers, the trials failed to support the view that the sprayed concrete could be heavily retarded without significant loss of early age strength. Following extensive laboratory testing, a total of 24 field mixes were tested. A combination of hand spraying and robot spraying techniques were used to replicate real construction conditions. Overhead spraying was undertaken on a purpose built frame incorporating a 5 m radius to simulate construction of a large diameter tunnel (Figure 23). In addition to the measurement of performance against the criteria set prior to the trial, an extensive investigation into the effects of high early age loading on immature sprayed concrete samples was performed (Figure 24). Samples were subjected to loading from 15 minutes to simulate. Utilisation Factors of between 30% and nearly 100% [Utilisation factor is the ratio of induced compressive stress to strength and is time dependent]. Even where samples had been subjected to stress states nearly equivalent to failure for up to 40 hours, no loss of integrity was measured. In order to prove the flexural toughness of the MF24 mix, a series of tests were performed by the British Building Research Establishment (BRE).

237 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-29.qxd

2/18/04

12:36 PM

Page 238

Figure 23. Overhead Spray Trials at Morgan Est R&D facility.

Figure 24. Early age testing of Sprayed Concrete.

Early testing with Dramix standard sprayed concrete fibres identified potentially brittle failure in certain circumstances due to the high strength and bond of the MF24 mix. Following consultation with Bekaert’s Engineers, a decision was taken to use High Carbon fibres. Tests on beams reinforced with High Carbon fibre (even when sprayed in separate layers) have shown excellent post crack toughness (Figure 25).

Figure 25. Testing of a layered MF24 Mix Beam Reinforced with High Carbon Steel.

control samples of in situ concrete and samples deliberately overdosed with accelerator. The control samples were taken from a mass concrete block cast and vibrated in a shutter. The sprayed concrete samples were recovered from panels sprayed overhead and the samples were cured in air up to 28 days. After 28 days, the samples were divided into three curing environments; under water at 20 degrees Celsius, in air at 65% Relative Humidity at 20 degrees Celsius and in cycles, one week in water at 40 degrees Celsius and 3 weeks in air at 40 degrees Celsius. Testing was performed at 1 month, 6 months and 1 year to determine compressive strength, Modulus of elasticity and Porosity. In addition, at 1 year, thin samples were taken and examined under a scanning electron microscope (SEM) to look for signs of deleterious behaviour. On completion of the trials, no adverse strength or stiffness values or trends have been recorded on any sample and the SEM/Petrographic analyses have shown no deleterious processes in any samples. In summary, it has been demonstrated that the sprayed concrete, even when overdosed with accelerator, is stable and can be categorised as highly durable.

4 PROOF OF CONCEPT TUNNEL 3 DURABILITY TESTING In order to understand the differences (if any) between sprayed concrete and traditional in situ concrete, three types of sample were prepared by Morgan Est at their facility in Rugby, UK. These samples were then carefully transported to Innsbruck University for curing and testing. The samples of sprayed concrete with the designed accelerator dosage levels were tested in addition to

In order to demonstrate the safety and accuracy of the system, a full scale proof of concept tunnel was undertaken. A 4.5 m diameter tunnel with one diameter of clay cover was constructed using the LaserShell™ method. The tunnel was extensively monitored and tested during construction and valuable data was recovered relating to soil and tunnel displacements, profile control and concrete quality. An attempt was also made to

238 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-29.qxd

2/18/04

12:36 PM

Page 239

Figure 26. Proof of concept tunnel showing DIBIT survey to confirm accuracy of TunnelBeamer™.

Figure 27. Sprayed concrete accreditation course – Overhead Robot Spraying.

measure stresses within the lining, but as with many previous attempts to capture stress data, results proved disappointing. To demonstrate the accuracy of TunnelBeamer™ system, an independent survey using the camera based DIBIT system was undertaken on each advance (Figure 26). Correlation between profiles recorded by TunnelBeamer™ and DIBIT was very good, giving a high degree of confidence that the system was reliable. 5 SPRAYED CONCRETE ACCREDITATION TRAINING The use of sprayed concrete is highly dependent upon the avoidance of human failure. Morgan Est therefore recognise the critical importance of training of all staff and operatives involved in the construction,

Figure 28. A completed tunnel driven by TunnelBeamer™ and excavated by the LaserShell™ method of tunneling.

inspection and testing of sprayed concrete tunnel linings. Accordingly, Morgan Est is the only organisation in the UK to run a sprayed concrete accreditation course suitable for the production of permanent sprayed concrete (Figure 27). The course is held at Morgan Est, Tunneling Divisions Research and Development facility near Rugby, UK. The experienced Nozzlemen and Pump Operatives are given a minimum one-week course involving a combination of theoretical and practical work. Only if all of the operatives sprayed concrete test pieces pass the performance criteria (hand spraying and robot spraying) and only if the operative passes a written examination, will a certificate be awarded. No man is permitted to spray concrete unless they are in possession of a current certificate of competence.

6 CONCLUSIONS The key conclusions are: – The developed sprayed concrete mix delivers a high performance rating in terms of medium to long term strength and provides excellent joint integrity. – Claims of admixture manufacturers that concrete can be retarded for long periods without detriment to the performance of the sprayed concrete have been proved to be wrong. Retardation of over 3 hours has been shown to reduce the early age performance to dangerously low levels (in respect of block retention and fall of newly sprayed concrete). – Early age strength gain from the sprayed concrete with low alkali accelerator is comparatively low (approximately J2), even when retarded to give only 2 hours life.

239 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-29.qxd

2/18/04

12:36 PM

Page 240

– High Carbon steel fibres required to prevent brittle failure of the high strength concrete. – High integrity joints and layers can be formed which show high levels of structural integrity and low permeability. Tensile splitting tests and beam tests on samples with and without joints showed very little difference. – Results from durability testing to date show no signs for concern. The concrete is classified as

“highly durable” and can meet all criteria set in respect of concrete that is “comparable to cast in place” can be met. – Rapid advance rates, coupled with systematic face support and early ring closure delivers excellent settlement control capabilities. – The LaserShell™ method of tunneling, utilising TunnelBeamer™ delivers unparalleled levels of safety, quality and efficiency.

240 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-30.qxd

2/18/04

12:37 PM

Page 241

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Ground support design and analysis: Exchange Place Station Improvements M.R. Funkhouser & M.F. McNeilly Golder Associates Inc., Lansing, Michigan & Newark, New Jersey, USA

ABSTRACT: The Exchange Place Station Improvements Project was one of three parts to the Port Authority of NY & NJ (PANYNJ) Downtown Restoration Program (DRP), which was implemented following the destruction of the World Trade Center (WTC) twin towers and the loss of commuter rail service to the Port Authority TransHudson (PATH) WTC Station. The project involved design and construction of six (6) new tunnel crossovers between five (5) existing 90-year-old concrete lined tunnels, and extending the existing station platforms approximately 46 m (150 ft) west. Design of this project involved excavating cavern spans upwards of 18 m (60 ft) in the underlying Manhattan Schist bedrock formation with rock cover as low as 7.5 m (25 ft) and multi-storey buildings directly above planned excavation limits. Design of ground support alternatives were completed using both review of historical precedents and UNWEDGE software analyses to evaluate load carrying capacities of various composite support systems. The design called for staged excavations with support installed at each stage. Ground support consisted of pre-stressed, resin grouted rock bolts, pre-fabricated steel lattice girders and steel fiber reinforced shotcrete (SFRS) liner systems.

1 INTRODUCTION

2 PRE-EXISTING CONDITIONS

Exchange Place Station (EPS), on the PATH commuter rail system, is located adjacent to the Hudson River in Jersey City, New Jersey. The project involved constructing six (6) new tunnel crossovers between five (5) existing tunnels, and extending both of the stations existing platforms to the west. The purposes for these improvements were to reconfigure EPS to allow operation as a “terminal” station, provide PATH with greater operational flexibility, and re-open EPS by July 2003. This work was undertaken as part of PANYNJ’s overall DRP, and was performed under an extremely tight schedule. Exceptional communication and coordination among the Owner, the Contractor and the numerous design consultants was key to the successful completion of the project. Exploratory drilling, core logging, and laboratory testing was still in progress when preliminary ground support designs were being drafted and construction procurement processes were being finalized. With needs to develop initial designs before site specific data was available, it was necessary to rely heavily on the experience of the project design team and use of precedent evaluations to assess the suitability of the project’s preliminary design, with final design analyses advanced concurrently with construction.

The west end of the Exchange Place Station and the locations of the project’s new tunnel crossovers and platform extensions (shown as dark shaded areas) are shown in Figure 1. The station was originally constructed between the late 1890s and 1908, and became operational around 1910. Very little about the station has been modified since its original construction except for the addition of a new head house and escalator banks, which were added at the station’s east end during 1986. Much of what was known about the station and its connecting tunnel configuration was based on available drawings termed “the 1908 Drawings”, which did not appear to be either construction plans or as-built drawings. As originally configured, EPS could operate only as a “through” station, and could not accommodate “terminal” station operations. This is the reason EPS was forced to close following the tragic events of September 11, 2001. Immediately above the station area, there are four (4) buildings ranging in height from 5 to 30 stories immediately overlying planned excavation limits. Foundation information for most, but not all, of these buildings was available at the time when the project’s new crossover locations were being finalized. The pre-existing tunnels and station are approximately

241 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-30.qxd

2/18/04

12:37 PM

Page 242

Figure 1. New crossovers and platform extensions.

18 m (60 ft) below ground surface and approximately 9.0 m (30 ft) below top of rock on average. More specific details of subsurface conditions are described in the following sections.

3 SUBSURFACE INVESTIGATION Subsurface investigations were planned based on a general knowledge of the bedrock conditions known to exist in the area, the existence of buildings, utilities that precluded drilling exploratory holes, and on the need to get as much geotechnical information as quickly as possible. Ultimately, the exploratory program consisted of the following:

• • • •

29 vertical borings drilled from street level; 2 inclined borings drilled from street level; 3 vertical borings drilled from tunnel level; and 4 inclined borings drilled from tunnel level.

Drilling was performed with truck-mounted drilling equipment from the ground surface and with skidmounted drills from inside the tunnels. Holes drilled from the ground surface were wash-bored and cased down to the top of rock, and then cored with NX sized double-tube core barrels. All collected rock core was preliminarily logged at the drill rig at the time of coring, and was then boxed and transported to PANYNJ’s materials laboratory, where it was logged in greater detail and photographed. See Table 1 for a summary of core recoveries and Rock Quality Designation (RQD), (Deere, 1963), values for each core boring. See Figure 2 for plot of RQD versus elevation for each collect core run. As can be seen in Table 1, average core recovery for the project was approximately 96 percent, and RQD was also reasonably high and averaged 84 percent.

Table 1. Summary of core recovery and RQD. Core recovery (%) Boring

No. Runs

Min

Max

Avg

Min

Max

Avg

450 451 452 454B 455 456 457 458B 459 460A 461 462 465B 467 468C 470A 471A 472 473 474C 475 476 477 478 479B 480D 481 482 483 484 485 486 487 488 489

6 14 12 8 12 1 16 15 1 12 14 17 1 15 14 4 15 11 11 11 13 1 12 14 6 12 7 12 5 10 7 9 8 7 8

86 79 98 96 90 100 90 93 100 92 78 67 96 80 95 95 93 97 85 90 55 88 80 60 90 93 88 50 66 75 85 70 65 68 78

100 100 100 100 100 100 100 100 100 100 100 100 96 100 100 100 93 100 100 100 100 88 100 100 100 100 100 100 100 93 100 100 95 100 100

95 97 100 99 99 100 98 98 100 98 96 95 96 98 99 98 99 99 98 97 93 88 97 93 98 98 96 95 84 86 91 87 87 90 92

42 0 98 24 38 66 34 80 82 77 78 33 84 46 92 65 85 85 0 50 40 75 0 57 77 62 58 0 21 42 23 63 60 48 56

96 100 100 100 100 66 100 100 82 100 100 100 84 100 100 96 85 100 100 100 100 75 98 100 96 100 100 96 93 88 96 94 84 100 97

79 78 100 82 78 66 76 92 82 91 93 85 84 87 98 85 94 95 70 84 83 75 81 85 88 91 81 71 59 69 70 75 74 83 81

242 Copyright © 2004 Taylor & Francis Group plc, London, UK

RQD (%)

09096-30.qxd

2/18/04

12:37 PM

Page 243

Table 2. Summary of rock core UCS testing.

Boring

Figure 2.

455 455 455 455 455 458B 458B 458B 458B 458B 460A 460A 460A 460A 468C 468C 468C 468C 468C 474C 474C 474C 474C 474C Min. Max. Avg.

Distribution of RQD with depth.

Virtually all of the discontinuities encountered in the drilling program were foliation partings. Depending on plan locations, boring were advanced to approximately 1.5 m (5 ft) above existing tunnel crown, or advanced to below tunnel invert elevations, if they were located in rock pillar areas. Select borings were oriented by use of carbide tip scribes fixed to the inside of the core barrels. Orientation of these scribes were measured before the core barrel were lowered down the hole, and the resulting lines scratched down the sides of the collected cores were used to orient the core after it is out of the hole. The presence of buildings directly above planned tunnel excavation limits precluded drilling exploratory holes in these areas from street level. In addition, only limited drilling from within the existing tunnels was allowed because construction was underway and the presence of an exploratory drill rig in the tunnels would hinder the Contractors progress. Also, as a part of the exploratory program, cores were obtained from the existing concrete tunnel linings. A total of 102 cores were obtained to evaluate the thickness and character of the existing concrete lining, and to detect and measure voids between the lining and rock. In general, concrete liner thicknesses were found to be thicker than shown on the “1908 Drawings”. Tunnel sidewalls were found to be “tight” against the rock, and voids upwards of 230 mm (9 in) were

10.21 13.41 16.46 19.51 22.25 11.58 14.63 17.37 20.42 21.95 11.58 15.85 18.90 22.25 10.06 13.11 16.15 18.90 23.77 10.67 13.72 16.76 19.81 23.47

Qu (MPa)

Es (MPa)

Es/Qu

18.6 27.8 22.6 30.9 29.9 26.0 20.6 33.7 20.7 74.6 25.4 51.5 38.7 29.5 38.0 49.5 45.2 26.3 42.2 10.2 9.7 12.2 13.6 39.7 9.7 74.6 30.7

966 1,156 894 1,016 6,012 3,172 3,516 4,544 3,668 7,028 5,013 5,511 4,541 4,572 789 1,049 1,477 3,804 663 2,461 158 1,590 2,508 4,566 158 7,028 2,945

52 42 40 33 201 122 171 135 177 94 197 107 117 155 21 21 33 145 16 241 16 130 185 115 16 241 107

observed at the tunnel crowns. In addition, no steel reinforcing was encountered in the concrete lining. The project elevation datum is elevation 300.0 ft (90.0 m) equals mean sea level at Sandy Hook, New Jersey.

4 LABORATORY TESTING A laboratory testing program was developed to measure rock and concrete strength characteristics. This testing program included 24 unconfined compression strength (UCS) tests and 16 direct shear tests on selected rock core samples, and 16 UCS tests on selected concrete liner core samples. Summaries of UCS test results and direct shear test results from the rock core testing are included in Tables 2 and 3, respectively, and test results for the concrete liner tests are included in Table 4. Compressive strengths for the rock samples ranged from 9.7 to 74.6 MPa (1.4 to 10.8 ksi) and averaged 30.7 MPa (4.5 ksi). However, UCS test results appeared to be highly influenced by foliation, even though the foliation dip angles were typically less than 20 degrees. In addition, strength test results were lower than

243 Copyright © 2004 Taylor & Francis Group plc, London, UK

Sample depth (m)

09096-30.qxd

2/18/04

12:37 PM

Page 244

Table 3. Summary of rock core direct shear testing.

Sample No.

Test orientation

DS1-R7 Parallel DS2-R6 Parallel DS3-R10 Parallel DS4-R7 Parallel DS5-R8 Parallel DS6-R8 Parallel DS6-R9 Parallel DS7-R3 Parallel DS1-R7 Perpendicular DS2-R6 Perpendicular DS3-R10 Perpendicular DS4-R7 Perpendicular DS5-R8 Perpendicular DS6-R8 Perpendicular DS6-R9 Perpendicular DS7-R3 Perpendicular Avg. Parallel Avg. Perpendicular

Cohesion (kPa)

Residual  (deg.)

156.5 98.6 46.2 126.2 48.3 39.3 57.2 22.8 87.6 82.7 77.9 117.2 6.9 2.1 48.3 20.7 74.4 55.4

14.9 25.0 36.3 25.0 31.0 23.4 21.8 26.6 29.5 40.9 29.5 16.7 21.8 13.1 14.9 18.4 25.5 23.1

5 GEOLOGIC CHARACTERIZATION The project is excavated in the Manhattan Schist formation, which can be described as gray, schistose gneiss and schist with occasional pegmatite intrusions. Banding and foliation is locally pronounced, and the dominant structural discontinuity found at the site is foliation. Foliation dip angles range from near horizontal (less than 5 degrees) to approximately 40 degrees, and dip directions varied from northeast to northwest. However, recent geologic deformations have imparted high angle jointing that is overprinted onto the rock mass foliation. Faults or major shear zones were not encountered, and foliation partings were found to be fresh to slightly weathered. At the project site, top of rock slopes towards the Hudson River, and top of rock elevations vary from 90.2 to 83.5 m (296 to 274 ft) from west to east across the site. Available information indicates that the top of rock continues to drop off steeply to the east, as you enter further into the river channel. Overburden thicknesses range from about 4.9 m (16 ft) in the western portion of the site to about 9.1 m (30 ft) in the eastern portion of the site, and consist of man-made fill overlying organic silt, sand, and gravel deposits, which are part of the Hudson River estuary. In general, overburden characteristics were of relative minor concern for the project’s tunnel design analyses, and are discussed no further in this paper. A groundwater table was observed in the overburden at 2.5 to 3.5 m (8 to 12 ft) below ground surface. However, water pressures within the rock mass at tunnel level were not found to exist, based on observed “dry” rock surface conditions upon demolition and removal of the existing concrete tunnel linings.

Table 4. Summary of concrete liner core UCS testing. Track

Station

Qu (MPa)

Es (MPa)

Es/Qu

E E F G G G G/E G/L H H H H H L L L Min. Max. Avg.

1228+00 1229+25 1089+35 1228+75 1229+10 1231+85 n/a n/a 1084+25 1085+45 1088+15 1088+45 1088+90 1229+30 1229+90 1231+70

56.1 36.4 35.7 37.8 48.9 26.0 16.2 13.9 27.4 10.1 41.0 42.2 47.3 24.4 6.7 23.1 6.7 56.1 30.8

4,831 3,325 3,635 4,288 4,211 3,150 1,453 1,388 2,787 1,554 3,831 3,940 3,888 2,104 853 2,410 853 4,831 2,978

86 91 102 113 86 121 90 100 102 154 93 93 82 86 128 104 82 154 102

foliation parting. No direct shear tests were run for joints across foliation, because few were observed in the collected cores and adequate samples could not be obtained for testing. Compressive strength testing of the existing concrete liner cores indicated a range of strengths from 7 to 56 MPa (1 to 8 ksi) with an average strength of 31 MPa (4.5 ksi).

expected based on comparisons with available, published test results for Manhattan Schist and similar schist rock formations (Deere and Miller, 1966; Baskerville, C.A. 1992). Direct shear test results indicated that foliation shear strengths have an average of 65 kPa (9.4 psi) cohesion and an average friction angle of approximately 24 degrees. Furthermore, test results did not appear to depend on the orientation of shearing across the

6 GEOTECHNICAL CHARACTERIZATION Design parameters used in the stability analyses were derived from testing core samples, or in the absence of test data, based on available published information. In addition, analyses were performed using both mean and low range values to assess variable design conditions and concerns.

244 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-30.qxd

2/18/04

12:37 PM

Page 245

Table 5. Summary of geotechnical design parameters. Design values Description

Mean

Low

Rock mass classification RQD RMR (Bieniawski, 1976) GSI (Bieniawski, 1976) Q (Barton et al., 1974; Bieniawski, 1989) Classification from RMR Description

90% 64 74

40% 33 43

15 II Good

2.7 IV Poor

Foliation joints Joint spacing (m) Friction angle (deg.) Cohesion (MPa)

0.6 23 0

0.2 20 0

Steep joints across foliation Joint spacing (m) Friction angle (deg.) Cohesion (MPa)

3 50 0

3 50 0

2,723 34.5 2,068 0.3

2,403 13.8 1,379 0.3

3.95

1.31

Rock mass characteristics Unit weight (kg/m3) UCS (MPa) Modulus of elasticity (MPa) Poisson ratio “m” Strength parameter (Hoek et al., 1998) “s” Strength parameter (Hoek et al., 1998)

7 ROCK SUPPORT ANALYSIS AND DESIGN

0.0056

0.0018

Overburden characteristics Unit weight (kg/m3) Friction angle (deg.) Cohesion (MPa) Poisson ratio Modulus of elasticity (MPa)

1,922 30 0 0.3 149.6

1,602 30 0 0.3 149.6

Existing concrete lining Characteristics Unit weight (kg/m3) UCS (MPa) Modulus of elasticity (MPa)

2,403 31.0 2,758

2,243 20.7 2,068

91.44

91.44

0

6

Water conditions Water elev. in overburden (m) Height of water Above tunnel inverts (m) Rock bolt properties Bar No. Steel grade (MPa) Pre-tension load (kN) Lengths (m) Spans  9.1 m 9.1 m  Spans  12.2 m Spans  12.2 m Spacing (m) Spans  9.1 m Spans  9.1 m In-situ stress ratio

See Table 5 for a summary of design parameters ultimately distilled from available data, published literature and past experience. Table 5 also includes design parameters for the rock bolts ultimately used as part of the project’s initial and final support systems. Ultimate punching shear strengths of 1.51 MPa (219 psi) were used in the stability analyses for plain SFRS materials, and were derived from correlations relating compressive strength and shear strength for plain and fiber reinforced concrete and shotcrete (ACI 1984 and 1988, Fernandez et al., 1979, and Mahar et al., 1975). Composite unit shear strengths for SFRS with embedded steel lattice girders were computed in a similar manner with exception that lattice girder steel cross sectional areas were added to the unit cross sectional area of the lining. For SFRS with embedded lattice girders, ultimate punching shear strengths of 2.70 MPa (391 psi) were used in the stability analyses.

9 517.1 133.4 2.4 3.7 4.6 1.2 1.5 1.5

0.5/2.0

The general approach for stability analyses consisted of evaluating design conditions for a select set of design cross sections that represent expected ranges of conditions. Stability analyses were performed using the software program UNWEDGE (Rocscience, 1997/2002). To analyze stability and develop ground support for discrete wedges using UNWEDGE, the following input parameters were used: (a) dip, dip direction, spacing and strength for three (3) discontinuities; (b) excavation geometry; (c) SFRS unit punching shear strength and thickness; and (d) rock bolt strength, length, spacing and orientation. In addition to these UNWEDGE analyses, evaluation of ground support systems was also performed by use of precedent, and further by use of rock mass rating systems Q and RMR. For each modeled span, multiple sets of discontinuity orientations were used to model potential failure modes, and it was found that sizes (weights) of potential wedge failures were most sensitive to the existence and orientation of high angle joints within the modeled spans. Ultimately, foliation dip angles were set at 15 to 20 degrees and high angle joint orientations were rotated to generate potentially unstable wedges. Hence, UNWEDGE was used to analyze worst case credible wedges generated during this phase of the analysis. For several reasons, not the least of which was the mandated construction completion date, it was decided to use SFRS tunnel lining systems to provide both the initial, short-term ground support and the final tunnel lining. Short-term and long-term design conditions were modeled with various shotcrete strengths (strength gain with time to model curing) and with two different

245 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-30.qxd

2/18/04

12:37 PM

Page 246

water conditions: (1) fully drained (zero head); and (2) 3 m (10 ft) of water head above tunnel invert elevations. These variations in loading conditions and support conditions were intended to model to expected extremes. The proposed new tunnel crossovers consist of excavated spans ranging from approximately 6 to 18 m (20 to 60 ft), with median spans in the range of 12 to 15 m (40 to 50 ft). During the project’s early design phase, much discussion revolved around the effective spans for each new crossover and how much three-dimensional (3-D) effect could be relied upon. Ultimately, it was decided that the rib pillar terminal ends would not be stiff enough or strong enough to provide effective support, because they were too narrow and any 3-D effect would be diminished until the reach under consideration was extended several meters further into the rock mass. It was also concluded that reliance on 3-D arching effects would not be prudent due to the relatively thin rock cover and resultant low confinement, and any 30-D benefits were considered a bonus and not relied upon in design. One method used to provide preliminary assessment of rock bolt lengths for underground support is by use of the “Q” System (Bieniawski, 1989). As described in this reference, preliminary rock bolt lengths can be based on the following formula: (1) where L  rock bolt length (m); B  excavated span Length (m); and ESR  effective span ratio. Using Equation 1 and a range of ESR values from 1.3 to 1.6, preliminary rock bolt lengths would be in the approximate range of 3.0 to 3.7 m (10 to 12 ft) for an excavated span of 18 m (60 ft). Another precedent type evaluation was made using the reference (Cording et al., 1971), which provides rock bolt length and equivalent support pressure data for numerous case histories. Review of several key figures from this reference indicates the following: (a) for excavated spans on the order of 15 to 18 m (50 to 60 ft) rock bolts lengths of roughly 4.6 m (15 ft) have been used on other projects (see Figure 3); and (b) for excavated spans on the order of 15 to 18 m (50 to 60 ft) roof support pressures on the order of 70 to 140 kPa (10 to 20 psi) have been used in the past. From the perspective of equivalent rock loads, 70 to 140 kPa (10 to 20 psi) would be equivalent to 2.6 to 5.2 m (8.5 to 17 ft) of rock having a unit weight of 2,723 kg/m3 (170 pcf). However, this precedent evaluation undertaken with particular caution because, unlike the “Q” System and experience reported by (Cording et al., 1971), shallow excavation depths and thin rock cover conditions were known to exist at the EPS project site.

Figure 3. Rock bolt length vs. excavated span (cording et al., 1971), reprinted with permission of ASCE.

The computer software program UNWEDGE (Rocscience, 1997/2002) provided 3-D visualizations of underground excavations with potentially unstable wedges formed by intersecting discontinuities. The software also allowed for the installation of ground support in the form of rock bolts and shotcrete linings of specified strength, length and/or thickness. UNWEDGE considers rock wedges as infinitely stiff, homogeneous masses acted upon by gravity, water pressure, friction, and applied internal supports. Calculations were performed considering planned new tunnel crossover excavations and assumed rock mass discontinuities to evaluate rock bolt length and spacing, shotcrete thickness and strength, and lattice girder strength and spacing to achieve the desired factorsof-safety for the various design conditions. Sensitivity analyses were performed using UNWEDGE to evaluate the relative importance of the different input parameters including planned support elements (rock bolts, shotcrete, and lattice girders), and rock mass strength and loading parameters (discontinuity strength and orientation, rock unit weight, and water pressure). In these UNWEDGE analyses, rock bolts were assumed to be 28.6 mm (1.125 in) diameter, 517 MPa (75 ksi) yield strength steel (US No. 9, Grade 75) “allthread” bars that were readily available in the NYC metropolitan region. These rock bolts have yield strengths of approximately 334 kN (75 kips), and in these stability analyses, 67% of yield strength or 223 kN (50 kips), was used as the maximum design capacity. Ultimately, using the different evaluation techniques, three (3) different design categories were developed for the project, and these categories were located based on the spans within specific areas of known rock mass

246 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-30.qxd

2/18/04

12:37 PM

Page 247

Figure 4. Typical ground support detail.

quality, and are summarized as follows:

•

•

•

For excavated spans greater than 12 m (40 ft): use 4.6 m (15 ft) long rock bolts at 1.5 m (5 ft) spacing each way in the crown, 2.4 m (8 ft) long rock bolts at 1.2 m (4 ft) spacing each way in the sidewalls, lattice girders at 1.5 m (5 ft) spacing, and 280 mm (11 in) thick SFRS lining. For excavated spans between 9 and 12 m (30 and 40 ft): use 3.7 m (12 ft) long rock bolts at 1.5 m (5 ft) spacing each way in the crown, 2.4 m (8 ft) long rock bolts at 1.2 m (4 ft) spacing in the sidewalls, lattice girders at 1.5 m (5 ft) spacing, and 280 mm (11 in) thick SFR shotcrete. For excavated spans less than 9 m (30 ft): use 2.4 m (8 ft) long rock bolts at 1.2 m (4 ft) spacing each way in the crown and sidewalls, and 150 mm (6 in) thick SFRS lining.

See Figure 4 for a composite section showing the recommended ground support for a typical new tunnel crossover. In the design, pre-existing concrete linings to remain were maintained to the greatest extent possible by installing pre-support rock bolts through the existing concrete tunnel linings, because it was important to minimize excavated spans during construction. Because it was known that the pre-existing lining had voids behind it at some locations, it was necessary to contact grout behind these existing linings before installing and tensioning the specified pre-support rock bolts. Contract Documents were developed and issued to the Contractor specifying prescriptive excavation and support sequences, which were necessary given the project’s critical nature, existence of buildings overlying planned excavation limits and thin rock cover conditions. These construction sequences reduced the sizes of the unsupported spans to the greatest extent practical.

8 CONSTRUCTION OBSERVATIONS During construction it was possible to visually observe the actual rock mass conditions and the size and extent

Figure 5. Completed tracks F to H to L crossovers.

of voids behind the pre-existing concrete lining, and in general, conditions encountered were as expected. The dominant rock mass feature was foliation, but there were also high angle joints encountered in the new crossover excavations, which were anticipated but not encountered in the exploratory boring program. Rock mass joints were tight and unweathered, and isolated water seeps were encountered in a few of the headings, but the excavations were typically dry at times of excavation. Drill-and-blast excavation techniques were initially chosen by the Contractor, but were found to be difficult to implement for several different reasons, and were abandoned in favor of mechanical excavation techniques using road-header type equipment. Mechanical excavations were found to be favorable from both the standpoint of perimeter control and excavation rate of progress. It should be noted that use of road-header equipment to excavate the local Manhattan Schist bedrock is not common practice, and this was just one of the unexpected changes that happened during the work. Installation of rock bolts was completed using small rock drills mounted on skid-steer type equipment and handheld jackleg drills. Installation and tensioning of the resin grouted rock bolts proceeded as anticipated. Application of shotcrete materials were performed only during dedicated night shifts, due to logistics. SFRS materials were batched at remote concrete plants, delivered to the project site by ready-mix transit trucks, and delivered to the station level via slick lines drilled and installed from street level. SFRS was applied with handheld nozzles, and access to upper sidewall and crown areas was completed using small man lift equipment. Steel lattice girders had to be individually fabricated because of each new crossover’s changing crosssection. Each lattice girder was individually identified and no two lattice girders had the same radius, so

247 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-30.qxd

2/18/04

12:37 PM

Page 248

fabrication and timely delivery was a significant logistical challenge. Ultimately, all of the lattice girders were fabricated, delivered, and installed as the excavations were opened up in the predetermined order. See Figure 5 for a representative photo of the completed ground support system. Geologic rock mapping was also performed as excavations progressed to verify design assumptions, and confirm that the specified rock support was adequate, and it was found that the design had an adequate factor-of-safety when considering the actual conditions encountered. 9 CONCLUSIONS The Exchange Place Station Improvements project was an extremely complex underground excavation, support, and infrastructure refurbishment undertaking. It was conceived, planned, and executed under an extremely short schedule, and implementation of innovative techniques, such as the use of road-headers to excavate the Manhattan Schist, and use of a SFRS tunnel lining systems, helped to get the project completed on time. A high level of coordination and communication by PANYNJ and its various design consultants and contractors was also critical to making this project a success. ACKNOWLEDGEMENTS The authors would like to thank PANYNJ for providing the artwork used in Figure 1, reviewing this paper and allowing publication for this conference. The authors would also like to thank their colleagues who worked on the Exchange Place Improvements Project and reviewed and provided comments to this paper. We would also like to thank George Yoggi for providing the photo shown in Figure 4. REFERENCES

ACI Committee 544. 1988. Design Considerations for Steel Fiber Reinforced Concrete. ACI Structural Journal/ September–October 1988: 563–579. ASTM. 2001. Designation D 2938-95 Standard Test Method for Unconfined Compressive Strength of Intact Rock Core Specimens. Annual Book of ASTM Standards 2001, Section four Construction, V04.08, (I): 312–314. ASTM. 2001. Designation D 5607-95 Standard Test Method for Performing Laboratory Direct Shear Strength Tests of Rock Specimens Under Constant Normal Force. Annual Book of ASTM Standards 2001, Section four Construction, V0408, (I): 1353–1364. Barton, N.R., Lien, R. & Lunde, J. 1974. Engineering Classification of Rock Masses for the Design of Tunnel Support, Rock Mech., Vol. 6, No. 4, pp. 189–239. Baskerville, C.A. 1992. Bedrock and Engineering Geologic Maps of Bronx County and Parts of New York and Queens Counties, New York, U.S. Geologic Survey, Miscellaneous Investigations Series, Map I-2003, 2 sheets, Scale 1: 24,000. Bieniawski, Z.T. 1976. The Geomechanics Classification in Rock Engineering Design, Proc. 4th Int. Congress on Rock Mech., ISRM Montreax, Vol. 2, pp. 41–48. Bieniawski, Z.T. 1989. Engineering Rock Mass Classification. New York: John Wiley & Sons. Cording, E.J., Hendron, A.J. & Deere, D.U. 1971. Rock Engineering for Underground Cavers. Symposium on Underground Rock Chambers Herndon, Virginia: ASCE. Deere, D.U. 1963. Technical Description of Rock Cores for Engineering Purposes, Rock Mech. and Eng. Geol., Vol. 1. Deere, D.U. & Miller, R.P. 1966. Engineering Classification and Index Properties for Intact Rock. Urbana, Illinois: University of Illinois. Fernandez, G.D., Cording, E.J., Mahar, J.W. & Van Sint Jan, M.L. 1979. Thin Shotcrete Linings in Loosening Rock. Rapid Excavation and Tunneling Conference. V1: 790–813. Hoek, E., Kaiser, P.K. & Bawden, 1998. Support of Underground Excavations in Hard Rock, A.A. Balkema, Rotterdam. Mahar, J.W., Parker, H.W. & Wuellner, W.W. 1975. Shotcrete Practice in Underground Construction Report No. FRAOR&D 75–90. Washington, D.C. Federal Railroad Administration. Rocscience. 1997/2002. UNWEDGE Users Manual, Rocscience, Inc., Toronto, Ontario.

ACI Committee 506. 1984. State-of-the-Art Report on Fiber Reinforced Shotcrete ACI 506.1R-84. Detroit: American Concrete Institute.

248 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-31.qxd

2/18/04

12:38 PM

Page 249

Session 2, Track 4 Ground modification for underground construction

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-31.qxd

2/18/04

12:38 PM

Page 251

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Cantilever frozen ground structure to support 18 m deep excavation Dong K. Chang, Peter W. Deming & Hugh S. Lacy Mueser Rutledge Consulting Engineers, New York, NY

Peter A. van Dijk Interbeton, Inc., Boston, MA

ABSTRACT: Artificial ground freezing was used to create a massive block of frozen ground to support an open excavation for a cast-in-place tunnel segment connecting a jacked tunnel with an immersed tube tunnel. This work was done for the Central Artery/Tunnel Project in Boston, Massachusetts. Freezing was performed adjacent to active railroad tracks. The massive block of frozen ground was 9 m wide, 26 m long, and 45 m deep. It worked as a cantilever retaining structure to support an 18 m deep excavation in soft ground (Boston Blue Clay). This frozen ground structure performed effectively for more than 18 months, with only minor deformation. The railroad adjacent to the frozen block operated throughout construction without service interruption. This paper describes the frozen ground structure design and construction, and documents ground performance using field data obtained before, during and after construction.

1 INTRODUCTION Part of the Massachusetts Turnpike Authority’s Central Artery/Tunnel (CA/T) project required three tunnels below active railroad tracks. These tunnels (Ramp D, I-90 West Bound, and I-90 East Bound) were constructed by tunnel jacking through artificially frozen ground that provided a stable material and groundwater control. These jacked tunnels were to be connected to immersed tube tunnels, installed across the Fort Point Channels, by cut and cover tunnel segments. A lean concrete gravity wall and a braced soldier pile and tremie concrete wall were successfully utilized to construct Ramp D cut and cover construction. For I-90 WB cut and cover excavation, localized freezing and soldier pile and lagging earth support wall at the northwest corner were successfully used while the remaining excavations were performed in the soil–cement stabilized ground. Developing an excavation support system for I-90 EB cut and cover tunnels was a more challenging task. Numerous obstructions consisting of abandoned timber pile supported masonry piers, just east of the I-90 EB jacked tunnel, made it impossible to install the planned soil–cement ground stabilization and T-shaped slurry wall excavation support system. After evaluating several options, SIWP (Slattery, Interbeton, J.F. White, Perini, the Contractor) proposed and B/PB (Bechtel/ Parsons Brinckerhoff, the Authority’s engineer)

accepted that the most effective design approach to establish the excavation support system for the 18 m deep cut and cover excavation was to freeze a massive block of ground, approximately 9 m wide, 26 m long and 45 m deep. Mueser Rutledge Consulting Engineers, who was the SIWP’s freezing consultant for the jacked tunnels, was hired by SIWP to develop the ground freezing design. Figure 1 shows the site location plan. Since the freezing operation was on-going for the I-90 WB and I-90 EB jacked tunnels, installing new freeze pipes and providing additional frozen ground mass within the same vicinity of the tunnel freezing areas were a significant advantage and a time saving. The ground freezing subcontractor (FreezeWall, Inc.) indicated that the proposed additional freezing would not require additional refrigeration units. The currently operating refrigeration units had sufficient capacity for this additional freezing work. However, there were several design concerns for implementing the proposed frozen ground design. Safely providing a stable and durable 18 m deep vertical frozen ground face was a technical challenge because frozen ground has rarely been used to support a deep excavation, especially when the excavated frozen ground would be exposed for at least 6 months, including the hot summer season. Another design challenge was to evaluate potential heave and the subsequent thaw settlement resulting from the ground freezing, especially at the adjacent active railroad tracks.

251 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-31.qxd

2/18/04

12:38 PM

Page 252

Figure 1. Site location plan.

2 SUBSURFACE CONDITIONS The project site, bounded by active commuter railroad tracks to the north and west and by the Fort Point Channel to the south and east, was formerly a tidal estuary. Subsurface conditions (in descending order from the ground surface) consist of approximately 10 m of granular fill, 4 m of compressible organic clay/silt, 26 m of marine clay (locally known as Boston Blue clay), 8 m of glacial deposits (till), and argillite bedrock. Groundwater levels generally exist 4.5 to 6 m below the ground surface and are influenced by the tidal fluctuations. However, a confined aquifer exists in the till. The fill contains numerous obstructions, reflecting a century of waterfront construction that had been abandoned and demolished and covered with fill. These obstructions include granite and concrete bridge pier foundations, timber piles, bricks, rubble, and buried abandoned railroad track structures. 3 DESIGN OF THE FROZEN GROUND STRUCTURE The initial frozen ground design approach for the I-90 EB cut and cover tunnel excavation was to install a retaining earth support system consisting of a shallow frozen ground mass supported on deep H piles.

Freezing the shallow ground mass to the tunnel subgrade would provide stable earth support for excavation and the deep H piles would provide vertical and horizontal support to preclude both local and deep stability failures. However, it was determined that driving deep H piles in this area would not be feasible because of numerous obstructions existing in the ground. Thus, use of deep freezing was substituted in lieu of the deep H piles. 3.1

The ground freezing design was subdivided into two parts. One is shallow ground freezing to a depth of about 18 m and the other part is the lower freezing to a depth of 45 m. Figure 2 shows the ground freezing area designations and the locations of the freeze pipe installation. Area B covered the area south of the I-90 WB jacked tunnel and northeast of the I-90 EB jacked tunnel end. This was the primary ground freezing area to support the I-90 EB cut and cover excavation. Both shallow and deep freeze pipes were installed. Areas A and C covered the footprint of the I-90 WB and I-90 EB tunnels and were original freezing areas for the jacked tunnels. Freezing in this area was re-established after the tunnel jacking was completed to provide stable ground above the tunnels facing the cut and cover

252 Copyright © 2004 Taylor & Francis Group plc, London, UK

Ground freezing design

09096-31.qxd

2/18/04

12:38 PM

Page 253

Figure 2. Ground freezing area designations and the locations of the freeze pipe.

253 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-31.qxd

2/18/04

12:38 PM

Page 254

excavation. Deep freeze pipes were added in Area C because the cut and cover excavation would be performed prior to the arrival of the I-90 EB jacked tunnel. The shallow frozen ground in Area B was connected to the Areas A and C frozen ground and provided additional frozen ground continuity to support the excavation. 3.1.1 Shallow freezing The shallow freezing was established with 18 m deep freeze pipes. The pipes were typically placed on a 2 m by 2.3 m spacing. But, at the perimeter of the excavation they were placed on 1.5 m spacing and extended 1.5 m deeper to provide an early frozen groundwater barrier. They also brought the face to the colder temperature to provide sufficient coverage of the exposed frozen ground face for temperature maintenance during the summer months. A total of 65 shallow freeze pipes were installed in Area B. 3.1.2 Deep freezing Deep freezing was established with 43 m deep freeze pipes, penetrating through the marine clay (Boston Blue Clay) and bearing into the till. They were installed in rows perpendicular to the excavation face to provide “barrette” shapes. The deep frozen barrettes were intended to support the shallow frozen ground mass and provide lateral stability. The deep pipes were spaced 2.2 m within a barrette, and each barrette was spaced 4.5 m along the excavation support. A total of 24 deep pipes were installed in the primary excavation support area (Area B). In the eastern end of the EB jacked tunnel area (Area C), a total of 18 deep barrette freeze pipes were added to resist ground movement into the excavation. 3.2

Design analyses

Step by step design analyses were performed to evaluate the proposed ground freezing system. The sections below describe these analyses. 3.2.1 Thermal analysis A thermal analysis was performed to evaluate the ground freezing system. The 2-D TEMP/W FEM computer program from Geoslope, Inc. was used to determine the rate of freezing, the extent of the frozen ground, and the frozen ground temperature profiles. The shallow and deep freezing were modeled independently using the freeze pipes as constant temperature sources. An average brine (freezing) circulation temperature of 25°C was used for both shallow and deep freezing analyses. The freezing influence from both the I-90 WB and I-90 EB jacked tunnel freezing that started much earlier than the present freezing area were included in the analysis as additional cold boundary conditions.

The results of the thermal analysis indicated that the shallow freezing with freeze pipes spaced on a 2 m by 2.3 m should provide sufficient shallow frozen ground mass after 90 days of continuous freezing. Additional freezing would make the frozen ground colder and stronger. The results of the deep freezing analysis indicated that at 90 days of continuous freezing, the ground would be frozen between the deep freeze pipes in each barrette and the area of frozen ground would cover about 45 percent of the area below the shallow frozen ground. At 120 days, the barrettes would merge to become a deep frozen mass and the frozen ground area would cover about 65 percent of the area below the shallow frozen ground. Figure 3 shows a typical output (shallow frozen ground temperature profile after 90 days of freezing) from the TEMP/W FEM analysis. 3.2.2 Strength of frozen ground It was very important to estimate the frozen ground strength to perform realistic stability analyses at various construction stages. In general, frozen ground strength increases as the frozen ground temperature decreases, but the strength at a constant temperature decreases with time because the frozen ground creeps under a constant loading. Table 1 shows a summary of the frozen ground strength properties estimated for the stability analyses. These strengths were based on actual frozen soil laboratory tests performed on undisturbed samples obtained from the jacked tunnel box areas (Deming 2000) and available frozen soil test results obtained from the nearby Russia Wharf frozen ground tunnel project (Lacy 2000). The frozen soil laboratory tests included short term unconfined compressive strength tests and creep strength tests. The creep tests were performed under a constant stress level while measuring strains with time. These test results were then used to develop time-dependent constitutive equations, which define the relationship between strain and applied stress at various creep times (Andersland 1994). 3.2.3 Stability analyses The frozen ground stability analyses were performed to make sure that the frozen ground provides the following stability requirements: (a) the available frozen ground strength should be greater than the stresses induced in the frozen ground; (b) the deep rotational slope stability failure through the frozen ground barrettes after the full excavation should have sufficient safety factors; (c) the local bearing capacity failure through the shallow frozen ground should have sufficient safety factors; (d) the sliding failure at the top of till stratum should have sufficient safety factors. The results of the various stability analyses as described above indicated that after 90 days of continuous freezing, the frozen ground (both the shallow and deep) would provide sufficient strengths to support

254 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-31.qxd

2/18/04

12:38 PM

Page 255

Figure 3. Temperature contours after 90 additional days of freezing (TEMP/W FEM thermal analysis). Table 1. Summary of frozen ground strengths used for stability analyses. Assumed frozen ground conditions Average temperature (°C) 7.5 10

Frozen ground strength

Creep age (days)

Allowable strain level (%)

150

2

150

5

the first 10.5 m of excavation. At 120 days of freezing, the excavation could safely advance an additional 7.5 m of excavation to the final subgrade level. Figure 4 shows a typical cross section with the proposed excavation stages and various stability failure modes that were analyzed. In order to evaluate the most critical stability condition, the stability analyses assumed that the excavation would remain open for 150 days for the cut and cover tunnel construction and the brine circulation temperature would be switched to maintenance temperature after the initial 120 days of freezing to hold an average frozen ground temperature of 7.5°C. The results of stability analyses indicated that the safety factors were about 2.0. 3.2.4 Heave and thaw settlement Heave prediction was an important design issue because the nearby railroad operation would be

Compressive (kPa)

Shear (kPa)

Fill Organic clay Marine clay Fill Organic clay Marine clay

1440 770 510 1630 960 620

720 380 260 810 480 310

Ground Surface (Elev. +35)

FILL

El.+25 El.+21

I-90 West Bound Jacked Tunnel (in-Placed)

ORGANIC SILT MARINE CLAY El.-4.5

TILL

SHALLOW FROZEN GROUND

I-90 East Bound Cut-and-Cover Tunnel Excavation El.+24.5

Local Bearing Capacity Failure DEEP “BARRETTE” FROZEN GROUND

Stage 1 Excavation Stage 2 Excavation

El. +17

Deep Rotational

MARINE CLAY

Deep cement Mix Stabilized Ground

Sliding Failure TILL

Figure 4. Typical section across deep frozen soil mass between jacked tunnels (facing east).

influenced by the heave resulting from the ground freezing. It was assumed that the fill and till strata, which exhibit higher permeability, would not produce volume expansion because the excess pore

255 Copyright © 2004 Taylor & Francis Group plc, London, UK

Frozen soil type

09096-31.qxd

2/18/04

12:38 PM

Page 256

water would drain away quicker than the advancement of the freezing front. The organic clay and marine clay having lower permeability were expected to produce heave. It was determined that the shallow freezing produces about 130 mm of heave directly above the freezing area and would not produce heave in the adjacent railroad area because the ground beneath the railroad had been frozen to the same depth for a long time for the jacked tunnels and would resist (hard boundary) heave in these areas. However, the deep barrette freezing would have the potential to produce heave by lifting up the shallow frozen ground. It was determined that the deep barrette freezing would produce about 250 mm of heave directly above the freezing area and would produce up to 75 mm of heave in the railroad area. However, it was determined by the railroad that the slow rate of heave and the relatively small magnitude of heave would not influence the railroad operation because the railroad would have enough time to make track adjustments. The thaw settlement would start after the completion of the I-90 EB cut and cover tunnel construction and the freezing circulation was turned off. The process of thawing would be a slow process and take more than a year. It was determined that thawing of the shallow frozen ground would produce ground surface settlement of up to 200 mm at the center of the frozen ground. The deep barrette frozen ground would produce thaw settlements of up to 130 mm within the frozen ground area and produce about 50 mm of thaw settlement at a 20 m radius from the edge of the frozen ground. 4 INSTALLATION OF FREEZE PIPES AND INSTRUMENTATIONS The freeze pipes (both shallow and deep) were 114 mm OD steel closed-end pipes. They were installed using an ultrasonic drill rig and installation did not encounter any significant problems. Two types of instrumentation devices, temperature probe pipes and inclinometers were installed to monitor the performance of the frozen ground during the freezing and maintenance periods. Three shallow temperature probe pipes (18 m) were installed in the shallow frozen ground and three deep temperature probe pipes (43 m) were installed in the deep freezing area. Temperature readings were used to evaluate the freezing progress of the shallow and deep barrette. Two deep inclinometers were intended to be outfitted with probe extensometer “Sondex” magnets to evaluate elevation changes due to heave and thaw. However, the installation of the Sondex system was not successful. The locations of these instruments are shown on Figure 2.

5 MONITORING, CONSTRUCTION, AND PERFORMANCE After the freeze pipes were installed and connected to the brine circulation system, the freeze plant was switched on. Frozen ground temperatures were continuously monitored in the deep and shallow temperature probe pipes. The data was initially used to calibrate the thermal FEM computer models and this refined the prediction of the freezing progress. During the freezing and excavation, the use of the inclinometers was unsuccessful. The inclinometers did not survive the freezing environment because water infiltrated into the inclinometer casings froze within the casings. Several attempts were made to melt the ice in the casings, but they were not successful and the use of the inclinometers was abandoned. Because of this, optical survey points were installed on the exposed frozen ground faces, in an approximately 3 m by 3 m grid. The optical survey measured three movements (in and out, up and down, and left and right) at each point. The optical survey was performed twice weekly. After 90 days of continuous freezing, it was determined from the temperature data that the frozen ground had sufficiently low temperatures and it would be safe to excavate the first 10.5 m. The excavation of the frozen ground was sloped back at approximately 20 V: 1 H to prevent overhanging. The exposed face was insulated by spraying on polyurethane foam within 12 hours of excavation. The polyurethane was directly attached to the face with chicken wire mesh, which was pinned to the frozen ground by nails. The nails were installed in an approximately 2 m by 2 m grid. A 10 cm thickness of polyurethane foam was sprayed on and painted a reflective white color to minimize heat absorption from the direct sunlight. After 120 days of freezing, a similar assessment as above was made based on the temperature data and the movements of the exposed frozen ground wall face. It was determined that the frozen ground would provide sufficient excavation support to advance the excavation to the final subgrade. The Contractor performed the additional 7.5 m of excavation to the final subgrade and insulated the exposed face with the polyurethane foam. After the excavation was completed, the brine circulation temperature was cut down to a maintenance temperature. This was intended to save freezing energy, especially because the cut and cover tunnel construction was significantly delayed due to delays in other areas of the immersed tube project. However, the brine temperature was switched back to a lower temperature when the temperature data and movement data showed that the frozen ground was showing a reduction in strength due to warming up or accelerated creep rates. The optical survey data indicated that the frozen ground wall has been stable throughout the

256 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-31.qxd

2/18/04

12:38 PM

Page 257

support an 18 m deep excavation without any bracing system was a success. It provided its primary design objectives without encountering any significant problems and demonstrated that this technology can be used in other difficult construction sites. Installation of freeze pipes was much quicker than the original design (T-shaped slurry elements) because freeze pipes were drilled through the obstructions. The frozen ground structure performed effectively. Deformation of the exposed frozen ground walls was minor and the walls were stable for more than 18 months. Whenever the deformation rate of the wall increased, the freezing system was quickly switched back to the lower brine temperature to gain additional frozen ground strength. This flexible strength control system was one of the advantages of using ground freezing. There were no stability problems for the frozen ground mass. Conservative design assumptions used for the stability analyses demonstrated the effectiveness of the frozen ground mass. The railroad adjacent to frozen block operated throughout construction without service interruption. Both the shallow and deep freezing did not generate significant heave and lateral deformation. Thaw consolidation settlement from the organic clay and marine clay was not significant and did not impact the adjacent railroad tracks.

Figure 5. Foam insulation on frozen walls – areas B and C.

ACKNOWLEDGEMENTS

Figure 6. 18 m high Area B wall – west bound at right.

construction period. During the 18 months of the construction period, the maximum cumulative deformations of the wall were less than 15 mm. It was difficult to establish whether there was any contribution of heave at the railroad tracks from the deep ground freezing because of other freezing and thawing activities that were ongoing at the same time. However, the insignificant vertical movements at the face of the frozen ground wall suggest that the deep freezing had no significant impact on the adjacent railroad track areas. Figures 5 and 6 show an overall view of the frozen ground support walls and the completed base slab of the cut and cover tunnels. 6 SUMMARY Utilizing artificial ground freezing technology to create a massive shallow and deep frozen ground mass to

The authors would like to thank the Massachusetts Turnpike Authority and the Federal Highway Administration for their support in publishing this paper, which summarizes our valuable experience and knowledge gained through this project. The author also would like to acknowledge the contributions of their colleagues at the CA/T project, B/PB, SIWP, and FreezeWall, Inc. who installed and operated the freezing system.

REFERENCES Andersland, O.B, & Ladanyi, B. 1994. An Introduction to Frozen Ground Engineering. Chapman & Hall. Deming, O.W., Lacy, H.S., & Chang, D.K. 2000. Ground Freezing for Tunnel Face Stabilization. In L. Ozdemir (ed.), Proceedings of the North American Tunneling 2000 Conference in Boston, Massachusetts, USA: 383–392. Rotterdam: Balkema. Lacy, H.S., Arland, F.J., & Chang, D.K. 2000. Supporting Historic Buildings While Tunneling Below. In L. Ozdemir (ed.), Proceedings of the North American Tunneling 2000 Conference in Boston, Massachusetts, USA: 383–392. Rotterdam: Balkema.

257 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-32.qxd

2/18/04

12:39 PM

Page 259

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

New chemical grouting materials and delivery equipment technologies Gary N. Greenfield Tunneling Director (Retired) Minova USA

Anthony C. Plaisted M.R.S.C. (UK), M.A.H.P. (USA) Technical Manager (Retired) Minova USA

ABSTRACT: The authors present a critical survey of grouting materials technologies that combine “new or new to America” products and their delivery systems for ground water management, backfill grouting and ground support during mining and recent advances in non-explosive demolition in confined spaces. Emphasis is placed on the significance of the necessary and related delivery systems. This paper reviews the changes that have transpired over time in the industry, which have increased the delays to “timely introduction” of novel materials and their appropriate and recommended delivery or application methods for both mining and tunneling projects in North America. Examples sited include several underground mining and tunneling projects that have encountered severe ground and water infiltration problems. The unintended consequences have included, loss production time or the marginal loss of a mine section. Tunnel advance has been halted due to excess water infiltration, affecting standard backfill grout placement with the potential for contaminating nearby fresh water resources. Alternatives that have been considered to alleviate these problems include modification materials for cement and chemical grouts, allied to grout delivery and placement systems, engineered for the job. Increased limits placed on the permitting and use of explosives for demolition both rock and concrete have resulted in evaluating chemical rock splitting mortars as an alternative procedure. Modification of these materials has provided for wet ground and low temperature applications, making it user friendly. To anticipate that any new technologies can be considered and applied within the context of current Contract Documents, the following fundamentals will be offered. Initial presentation of (a) documented field trial data or related site specific application histories, so that (b) materials application methods and equipment are clearly understood, leading to (c) a timely, initial cost/benefit study and (d) followed by evaluation trials properly supervised and witnessed.

1 INTRODUCTION Products for rock support and ground water management that were introduced in the early 1970s as relatively “new or new to America” with application to the tunneling and underground mining industries saw their acceptance by these industries in varying degrees. Over the past thirty years, beginning with polyester resins for rock bolting through trials and targeted project application, they have helped in contributing to the awareness of a new “language”. Words such as rheology (“quality or state of being, to be deformed or to flow”) and thixotropy (“the property of various gels becoming fluid when disturbed as by shaking”) as well as phrases such as grout containment have appeared in product literature and the technical press. The purpose of this terminology is to properly

describe the function of selected modifiers or polymers for example, that alter the properties of “traditional” cement and chemical grouts. The North American underground mining and tunneling industries face many challenges, not the least of which are environmental regulations placed on current products as well as those being developed to address more effective control of “difficult” ground and manage high rates of water ingress. As this paper is being written, a current tunneling related article mentioned, “that the means of constructing this valuable infrastructure are ever changing”. We suggest that in meeting those “means” with new technologies, we cannot always apply the same old methods and materials in our battle with the ground. We have been reminded more than once over the last several years that there is no good ground anymore! Major metropolitan markets

259 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-32.qxd

2/18/04

12:39 PM

Page 260

in the United States do not have the luxury of “offering up” ideal ground conditions under their metropolis for expanding needed transportation, water and waste water facilities. This paper seeks to bring back into focus the need for planning a return to a more thorough investigative “process” for evaluating and applying “new technologies” that are proven or being developed. 2 PRODUCT INTRODUCTION AND COMMENTARY 2.1

Rock reinforcement

2.1.1 Polyester resins Early in 1970, the word “polymers” took on a new meaning for the North American mining and tunneling industry with the introduction of polyester resins using a peroxide catalyst. Produced in paste form, placed in a sealed, thin poly film, the package resembled an oversized “hot-dog” and was offered in several diameters. These resin cartridges were initially developed and placed in British and French coal, metal and non-metal mines as a fully grouted, un-tensioned rock bolt. Initially bolts that complimented the resin, were threaded, deformed steel (rebar type) rods. Application methods, installation equipment and resin product development continues to evolve, currently offering tensioned, un-tensioned and “resin assisted” mechanical anchorages to complement both steel, fiberglass and wood “bolts”. Commentary Polyester resin bonded rock bolt trials made their initial evaluation in the coal fields of Western Pennsylvania, Virginia and West Virginia under the watchful eye of both the operating companies and the U.S. Bureau of Mines (USBM). Underground metal mines, notably White Pine Copper undertook extensive application trials of resin bonded rock dowels in difficult ground conditions. With additional published technical data from Europe and the results of on-going trials in the United States, the tunneling community began to apply this rock support system to projects, such as the first Straight Creek (Eisenhower) Tunnel in Colorado and the Washington Metropolitan Area Transit Authority (WMATA), at Washington, DC’s DuPont Circle Station. Testing with close proximity blasting to installed bolts was carried out for Virginia Electric Power (VEPCO) at their North Anna nuclear station. This resin system was approved for rock foundation reinforcement on two additional reactor bases. Long term trials, evaluation and a report in 1973 by the U.S. Army Corps of Engineers was also published, following completion of a twelve month trial. The Corps then issued an approval document for use of polyester resin bonded rock bolts in all of their districts in the United States.

The rock abutments in Hannibal Shale on the Clarence Cannon Dam (USACOE) in Missouri saw the drilling sub contractor design and build drilling and resin bolt insertion rigs to maneuver on the tiered rock benches. A detailed Corps document was published on this work, in 1974. These independent documents afforded civil and geotechnical engineers a technical reference on which to base their decision for resin bonded rock bolt placement in subsequent project documents as a permanent rock support system. It may also be noted that today, two US based companies supply the American underground coal market with approximately 90% of the industry’s total roof (rock) support requirements with fully bonded and resin assisted mechanical anchorages. The early performance success of this support method enabled development in the 1970s of small diameter carbide bits and related drill steel, providing one inch (25 mm) holes to be drilled. Mine roof bolt manufacturers soon began producing headed rebar for placement in the standard drive head on mine roof bolters. The principal benefits to the mining companies? Improved roof control safety and reduction in maintenance costs over time, when compared to mechanical anchorage systems. It was the design and implementation of polyester resin bonded bolts to pre-bolt and support overcast site locations (for ventilation) in coal mines that provided experience for applying pre-support methods at a Corps of Engineers (USACOE) tunnel site. Work on the tunnel began as the first heading was mined, initiating a “two pass” drill and blast excavation method from the portal. The technique involved drilling the crown for a twenty (20) foot, coupled rock bolt. The upper most ten (10) feet of the coupled rock bolt was fully resin grouted. The lower portion of the bolt was left un-grouted and tensioned from the crown of the first heading. Pulling rounds to begin the second ten feet or final excavation pass in the horizontally bedded rock exposed the bolt at the bolt’s coupling “horizon” for subsequent placement of a plate, washer and nut. In effect, the rounds were shot against a pre-supported back, minimizing rock over-break and providing tunnel mining safety. 2.1.2 Polymer modified cements Applications in “weak”, water saturated rock and underwater concrete structures offered an opportunity for the development, trial and application of a prepackaged, neat cement based, polymer modified grout to be pumped into bore holes prior to equipment (mechanical) or hand insertion of several types of deformed steel rock bolts. Of significance, the thixotropic properties enabled the cement grout to inhibit dilution from water in a “marine” environment. Equally important, grout was not lost to adjacent fractures in the rock structure, due to grout “stop and stay” thixotropic characteristics (resembling stiff, gritty mayonnaise) when grouting pressure ceased.

260 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-32.qxd

2/18/04

12:39 PM

Page 261

Commentary A request for a calculated quantity of pre-placed non-shrink, non-bleed, controlled expansion grout in thirty (30) “plus” feet of drilled hole with manual insertion of a rock dowel into a “foundation” of fractured rock again proved the effectiveness of a formulated thixotropic cement grout. Grout design enabled conventional grouting equipment to pump grout through a two (2) inch line over 200 feet in length to the collar of the holes. The grout was pumped to the mouth of the grout line, the line then pushed to the back of the individual holes. The rising column of grout gave the contractor a visual indication as to the location of the grout column as the hose was ejected by the rising column of grout. Any ground water that migrated into the drilled holes had no diluting effect on the grout. A pull test to 60% of yield of the Grade 60 rock bolt was carried out in 20 to 24 hours when the grout reached 5000 psi (34.5 MPa). Applications on foundation and rock “high walls” in the Colorado Mountains near Central City provided a proven alternative, offering rapid grout pre-placement and insertion of several hundred rock bolts of up to twenty (20) feet in length. On a project in Kentucky short starter dowels for “keying” shotcrete on four (4) adjacent tunnel portals and rock abutments were quickly placed after the drilled holes were “charged” with the thixotropic grout from a grout pump. Several hard rock mines in Canada found this grout material suitable for pregrouting cable bolts. A mined section could be pre-bolted ahead of the advancing face in the stope. The grout was pumped by hand upwards into the drilled hole. The cable bolt of a predetermined length was then fed by hand into the hole, completing the installation. 2.2

Chemical grouting

2.2.1 Sodium silicates, hardeners and fillers Sodium silicates have a time and job proven history for providing ground stabilization and minimizing ground water infiltration in major tunneling contracts. Combined with a compatible hardener and where severe ground water infiltration is encountered, with the addition of reinforcing fillers, effective reduction of water ingress on tunnel and mining sites has been achieved. Such fillers may include pulverized limestone. A sprayable system was developed for coating and reinforcing specific rock types (St. Peters Sandstone and Austin Chalk) to minimize ground sloughing in utility and transit tunnels. Commentary Critical to performance of sodium silicate based grouts is the selection of the hardener, which should be used at the right stoichiometric ratio to the silicate to insure longevity of the set grout. In an Austin Chalk

formation, a sprayed silicate (40% silicate concentration) coating was applied with a hand “back pack” sprayer during the “off ” mining cycle of the tunnel boring machine (TBM). This procedure allowed the chalk to be protected until shotcrete was applied clear of the trailing gear. Initial trials were first conducted with the contractor, which demonstrated excellent bonding at the shotcrete-“treated” chalk interface. Sodium silicate and a hardener system, comprised of specifically selected organic esters was chosen to shut off water while sinking two shafts to a depth of over 3300 feet in New Mexico. The shafts penetrated seven discrete aquifers. The grout was tested at both an independent laboratory and by the owner prior to approval and was found to exceed Environmental Protection Agency (EPA) standards. Approximately 1,500,000 gallons of grout was placed at the site.

2.2

2.3.1 Non-bleed grouts The parameters of a non-bleed and non-settling grout were seen as essential for improving smaller annuli grout placement for tendon grouting of post-tensioned structures, notably nuclear reactor vessels and cable stayed bridges. North Sea sited oil well drilling rigs were one of the first structures to utilize this type of grout. Application has also involved the grouting of steel tendons in fractured rock to stabilize an underground hydroelectric turbine chamber in Latin America. Critical to product quality and performance in eliminating water bleed was the use of high sheer mixing equipment during the introduction and dispersion of a powdered polymer into the grout. Commentary Testing modified grout anti-bleed performance required introducing a new method. A Gelman filter was chosen to test this thixotropic mixture. Grouts, both control and polymer modified, were placed under pressure (80 psi) to quantify bleed rates. As an example, under then current Post Tensioned Institute (PTI) guidelines, a maximum tendon grouting distance of 125 feet was mandated to insure minimum water loss from a cement grout during grout placement between the protective sheath and tendon. The distance was more than doubled, without bleed occurring with the application of this “new technology” grout on three cable-stayed bridges in the United States. On an existing bridge, located in Mississippi, a protective grease was initially placed within the cable stayed sheaths. During a routine inspection, it was determined that the grease contained sufficient moisture to have initiated corrosion on the cables. The thixotropic, non-bleed cementitious grout was selected to be pumped into the annulus, displacing the grease.

261 Copyright © 2004 Taylor & Francis Group plc, London, UK

Cementitious grouting

09096-32.qxd

2/18/04

12:39 PM

Page 262

This material technology was “transferred to tunneling” with the owners requirement on the Channel Tunnel (UK Drives) for a different approach to backfill grouting behind the segmental lining. An expanded explanation of this development is offered within the “Backfill Grouting” segment of this paper. 2.3.2 Ultrafine cements Micro fine cements received a great deal of attention when first introduced to the civil engineering community. A number of papers presented at the ASCE conference on Grouting and Soil Improvement, New Orleans, 1992, reported on introduction, performance evaluation and continuing investigation of both physical and chemical properties of microfine cements. Included, manufacturing processes which were then being evaluated. One concern that was expressed, the necessity of adding a dispersant to the cementitious powder during grout preparation, to insure that the fine particles were wetted out to counter agglomeration, which would otherwise restrict penetration of the mix during injection into soils. Commentary In subsequent investigations into resolving the procedure for mixing two separate components in preparation for grouting, one US based manufacturer introduced an ultrafine cement grout integrating a dispersant into the “base” product during the manufacturing process. Initial site evaluation and application of the combined ultrafine/dispersant blend was carried out within the concrete lined Air Intake Shaft for the Department of Energy (DOE) at the Waste Isolation Pilot Plant (WIPP) site in New Mexico. Reported testing of the various grouts submitted confirmed that it was the only ultrafine grout of those tested that “met established set characteristics and ease of preparation” necessary in the confined space within the shaft. This grout was also chosen for grouting on the Inter Island Tunnel in Boston, through forward probes on TBM driven tunnels in California near Los Angeles and in underground mining applications in Canada. The availability of sub micron or nano range particles is on the horizon. This new technology will open up a new era in cement grouting which will allow us to tackle severe ground conditions with nano fine particles. Evaluation studies we feel will give us the confidence to exploit this breakthrough at the earliest opportunity. 2.4

Backfill grouting

2.4.1

Control modules and polymer modified grout The development of a grout modifier for improving the properties of a cement based grout, specifically designed for backfill grouting in tunneling is the result of technology transfer, based on work previously

carried out to seek elimination of grout bleed in tendon grouting. Construction of the Channel Tunnel between England and France presented the initial opportunity to provide a polymer modified cement to control ground water dilution of ordinary cement grouts during mining. The grout modifier used on the United Kingdom (UK) seaward and landward drives was derived from original work, previously carried out on cement grout modification for tendon grouting by M. Schupack and A. Plaisted. The result was a cement based grout that could be mixed and held (retarded) for up to six hours in grout cars and lines, before hydration began. The grout was thixotropic, did not bleed and resisted ground water washout during placement behind the lining. Controlled set times were carried out with sodium silicate, introduced at the point of injection on the segments. The existing onboard grout pump on each of six TBMs was connected to a purpose built hydraulic powered control module and catalyst pump. The grout line received a grout “gun” fitted with an in-line mixing element and the gun designed for connection to the individual grout ports in the segmental concrete lining. The owner established a maximum grouting pressure of 75 psi (5 Bar) that was “dialed in” and maintained by the control module. The module also provided for variable control of sodium silicate volume to provide management of selected grout setting times. Commentary The “Chunnel” project provides an example of the collaborative effort that resulted in providing an effective backfill grout program that involved an equipment manufacturer specializing in hydraulic controls and pumps, a mining and tunneling specialty chemicals firm and the owners engineering staff. Tests were witnessed, involving all the parties and carried out in as close to actual backfill grout placement conditions as was practical, prior to formal approval. 2.5

262 Copyright © 2004 Taylor & Francis Group plc, London, UK

Chemical rock splitting

2.5.1 Rock splitting mortar Recent developments in non-explosive demolition agents have permitted controlled demolition of rock and concrete structures, enabling the expansive cementitious mortars now to be placed in bore hole diameters from 1–1/8 inches (28 mm) to 3 inches (76 mm) in diameter. The material can be loaded into deep holes and pumped for underwater (marine) applications. Recent applications where explosives are prohibited involve placing demolition mortar in rock adjacent to utilities and surface structures. Currently these applications include creating access for tunnel construction and clearing rock in preparation for new surface construction. As this paper is being written, product development work continues and

09096-32.qxd

2/18/04

12:39 PM

Page 263

innovative drilling tool accessories are being fabricated for evaluation in the field to improve demolition productivity. Commentary Current construction industry interest in the use of nonexplosive demolition mortars continues to increase. In recent months, as engineering firms and municipalities have identified the growing need to limit the disturbance to adjacent businesses during new construction for surface and sub surface facilities, non-explosive demolition mortars are becoming an alternative method “of interest”.

3 CONCLUSION The need today for a more thorough investigative process in evaluating and accepting new technologies was mentioned in the Introduction segment of this paper. The authors identified materials and referenced several projects that benefited over the past thirty years from these “new or new to America” products. Today, products formulated and designed to focus on ground and ground water control for tunneling and foundation engineering must meet stringent environmental regulations, if they are to be considered. Can we reasonably anticipate that any new materials technologies and their related delivery or application methods that are brought forward have the opportunity to be evaluated? It is hoped that as projects are funded for design and construction, these technologies will be considered as solid “candidates” for inclusion in the contract documents. Listing these additional product options (prescriptive approach) in contract documents have the potential for diminishing any unintended consequences that may occur during our “battle with the ground” on a specific project. What about evaluating product performance involving ground support and ground water management? We urge owners, their retained and in house engineering staff and consultants to consider the following steps be taken during planning and the preparation of contract documents. Request potential suppliers who have come forward to provide documented field trial data or related, site specific application histories. Second, that the supplied documentation provide application methods, clearly identified, including required application equipment. An initial cost/benefit analysis should be carried out after evaluation trials that are properly supervised,

witnessed and performed to closely replicate anticipated site conditions. An Engineering News-Record (ENR) Editorial (October 20, 2003) contained a concluding paragraph that provides a final, fitting comment. “The potential is there for healthy innovation in (tunneling) construction with positive contributions from the technical, experiential and financial sides. Now we just have to use it”. REFERENCES Albritton, J.A. 1974. Rock Bolt Field Tests, Clarence Cannon Project. U.S. Army Corps of Engineers District, St. Louis, MO. 1974. American Society of Civil Engineers (ASCE). Developments in Geotechnical Engineering. 1988. Avery, T.S. Optimizing the Performance of Polyester Resin Grouted Rock Bolts. Hershey, PA. 21p. Annett, M.F. and Stewart, J. 1989. Development of Grouting Methods for Channel Tunnel United Kingdom Segmental Lining. British Tunnelling Society, London. 1989. pp.173–178. Annett, M.F. 1994. Grouts in Tunnelling. SCI Conference, Grouts and Grouting. London, England, 1994. Avery, T.S. and Daemen, J.J.K. 1994. The (In?) Significance of Creep in a Prestressed Polyester Resin Grouted Rock Anchor. Rock Mechanics (NARM) Nelson, P.P. & Laubach, S.E. (eds.) Balkema, A.A., pp. 953–960. Brierly, G. 2001. Tunneling: a Battle Against The Ground. TBM: Tunnel Business Magazine. February, 2001. pp. 32–33. Burke, J. 2001. Building a Third Stage for Carnegie Hall. World Tunneling (WT On Site) 2001. pp. 183–185. Greenfield, G.N. and Plaisted, A.C.1994. New Perspectives in Grouting Materials for the 90s & Beyond. Trenchless Technology, 1994. pp. 35–37. Haywood, H.M. 2000. Contractor Outreach and PreQualification Program. North American Tunneling 2000. Ozdemir, L. (ed.) Balkema, A.A., pp. 125–128. Kendorski, F.S. 2000. Rock Reinforcement Longevity. 19th Conference on Ground Control in Mining, Morgantown, West Virginia, 2000, 6p. Moss, T.A., Phillips, S.H.E. and Smith, D.F. 1993. Saline Tolerant Grout Use at a Nuclear Waste Facility. Conference on High Level Nuclear Waste Storage Facilities, Las Vegas, Nevada. 1993. Nelson, C.R. 1977. Spray Grouting for Tunnel Support and Lining. Underground Space. Vol. 1. Pergamon Press 1977. pp. 241–246. Reed, J.J. and Ortlepp, W.D. 1969. Grouted Bolts for Faster Rock Stabilization. The Mines Magazine, March 1970. (Reprint). Reilly, J. 1997. Owner Responsibilities in the Selection of TBMs. International Tunnelling Association (ITA). Vienna, April 1997.

263 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-33.qxd

2/18/04

12:39 PM

Page 265

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Jet grout bottom seal for cut and cover tunnel Thomas M. Hurley Layne GeoConstruction, Bridgewater, Massachusetts, USA

ABSTRACT: The Massachusetts Bay Transportation Authority, Silver Line transitway is a new transit service from Boston’s South Station to the fast developing South Boston Waterfront. The Silver Line will utilize trolley busses within a tunnel section slightly larger than the bus itself. The tunnel alignment is particularly difficult. The first obstacle to the tunnel alignment is the two seven-story buildings, the second the Fort Point Channel crossing and the third the East Cofferdam. Three distinct tunnel methods are employed for this project. NATM with micro-pile and frozen ground underpinning is utilized at the Russia and Graphic Arts buildings, immersed tube tunnels are employed at the Fort Point Channel Crossing, and cut and cover tunneling with a jet grout base seal are employed at the East Cofferdam. The east side of the project utilized a sheet pile cofferdam to construct a cut and cover tunnel. The cofferdam was approximately 140 long by 50 wide with an excavation depth of 60. Layne GeoConstruction designed and constructed a system of secant and tangent jet grouted columns to stabilize and seal the bottom of a sheet pile cofferdam, for the construction of a cut and cover tunnel. The jet grouting was used in lieu of the contract design that specified deep soil mixing. The jet grouting acted, additionally, as a sub-surface strut, providing toe support for the sheet pile system. Moreover, jet grout columns improved weak Boston Blue Clay soils and provided support for the setting of immersed tube tunnel segments outside the cofferdam.

1 PROJECT OVERVIEW The Massachusetts Bay Transportation Authority (MBTA), Silver Line is a new transit service from Boston’s South Station to the fast developing South Boston waterfront, the epicenter of commercial development in the city of Boston. Former commuter parking areas and empty lots have been transformed into a new Federal Courthouse at Fan Pier, and the Seaport Hotel at the World Trade Center Boston. The City has made strides to improve access to the area by constructing a new Northern Avenue bridge, and rehabilitating the deteriorating Congress Street and Summer Street bridges. The Silver Line will utilize the most advanced Bus Rapid Transit (BRT) system, within a tunnel section only slightly larger than the vehicle itself. The alignment of the new tunnel (Figure 1), begins at South Station, integral with the Central Artery/Tunnel (CA/T) contract C11A1, continues north through CA/T contract C17A1, until reaching the beginning of the MBTA Silver Line contract E02CN15, at the intersection of Congress Street and Atlantic Avenue. It is at this juncture that the tunnel passes beneath two seven-story, historic buildings then crosses the Fort Point Channel

Figure 1. Silver line transitway tunnel alignment.

265 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-33.qxd

2/18/04

12:39 PM

Page 266

en route to its termination at the World Trade Center Boston. The tunnel alignment through MBTA Contract E02CN15 is particularly challenging, presenting several obstacles. The first obstacle is the seven-story buildings, the second the Fort Point Channel crossing, and the third the East cofferdam. Three distinct tunnel methods are employed for this project. NATM with micro-pile and frozen ground underpinning is utilized at the Russia and Graphic Arts building, Immersed Tube Tunnels are employed at the Fort Point Channel Crossing, and Cut and Cover Tunneling with a jet grout base seal are employed at the East Cofferdam.

A4 size paper

Letter size paper

Setting

cm

inches

cm

inches

Top Bottom Left Right All other Column width* Column spacing*

1.2 1.3 1.15 1.15 0.0 9.0 0.7

0.47 0.51 0.45 0.45 0.0 3.54 0.28

0.32 0.42 1.45 1.45 0.0 9.0 0.7

0.13 0.17 0.57 0.57 0.0 3.54 0.28

2 EAST COFFERDAM CUT-AND-COVER TUNNEL The east side of the project utilized a cofferdam to construct the cut and cover tunnel. The cofferdam was approximately 140 long by 50 wide with an excavation depth of 60. Steel sheet piling was driven to depths of 80 and the excavation was internally braced with steel wales and struts, and soil stabilization at the toe of the sheet piling, see Figure 2. 3 SUBSURFACE PROFILE The subsurface conditions (Figure 3) at the East Cofferdam are primarily 10 to 20 of urban fill, underlain by

Figure 3. East Cofferdam subsurface conditions.

Table 1. Cost comparison.

Figure 2. Sheet piling soil stabilization.

Technique

Cost

Soil mixed cross walls at 8 on center and perimeter walls extending down through the clay Jet grout cross walls at 8 on center and perimeter walls extending though the clay Slurry wall (unreinforced) cross walls at 20 on center with reinforced perimeter slurry walls extending through clay

$2,000,000

266 Copyright © 2004 Taylor & Francis Group plc, London, UK

$3,400,000 $3,400,000

09096-33.qxd

2/18/04

12:39 PM

Page 267

25 of organics then 50 to 80 of Boston Blue Clay to reach glacial till. The organic and clay materials were soft to very soft as indicated by average N values between 0 and 4 blows per feet, as shown in Figure 3. Soil stabilization was required, as a result of the Engineer’s base heave stability analysis of the soft clay, to ensure a stable bottom and safe toe support for the excavation support system.

using soil mixing, jet grouting or slurry walls. The Engineer’s preliminary cost comparisons (Table 1) indicated a soil mixed cross wall as the most economical approach. The results of the cost analysis are as follows: The soil mix design recommended by the Engineer is shown in Figures 4 and 5. The design specified 3 wide cross and perimeter walls, using 3 diameter triple augers to ensure continuous walls as they were to

4 SOIL STABILIZATION ALTERNATIVES The Engineer evaluated three methods to alleviate base heave failure. 1. Reinforced concrete slurry walls installed vertically from the bottom bracing level to the till or rock. 2. Stabilization of the clay with individual elements using soil mixing, jet grouting, or stone columns. 3. Extension of the excavation support system down through the clay using cast-in-place cross walls constructed with soil mixing, jet grouting or reinforced slurry walls. Reinforced concrete slurry walls spanning through the clay (approx. 50) along the perimeter of the excavation would result in expected lateral movements of up to 6 and was deemed unacceptable. Soil reinforcement using individual elements of soil mixing, jet grouting, or stone columns was discounted due to the complexity and uncertainty of the interaction between the individual elements and the clay. In addition, tension forces would most likely develop in the elements from the basal heave forces. The Engineer recommended extending the excavation support system downward through the clay with interconnected cross walls acting as subsurface struts,

Figure 4b. Soil stabilization detail.

Figure 4a. Contract specified soil stabilization plan.

267 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-33.qxd

2/18/04

12:39 PM

Page 268

Figure 5. Contract specified soil stabilization profile.

function as structural elements. The minimum unconfined compressive strength was 300 psi at any individual location with a 450 psi average. 5 SOIL STABILIZATION ALTERNATIVE EMPLOYED – JET GROUTING Although the Engineer considered three alternatives for soil stabilization on the project, the cost analyses reflected the conservative nature of the Engineer and specialty contractors when queried for budget pricing prior to bidding. In actuality, the jet grout cross and perimeter wall alternative was the most cost effective at bid time, nearly 50% less costly than the Engineer’s estimate for soil mixed cross and perimeter walls. Therefore the jet grouting option was proposed. The jet grout alternate was submitted using the same geometric pattern as the soil mix design above, except individual 5 diameter jet grout columns would be constructed using the double fluid method, in lieu of the 3 diameter triple auger panels. Figure 6 shows a detail of the double fluid method of jet grouting. The double fluid method is a mix in place jet grouting method where neat cement grout is injected at high pressure (up to 8000 psi) and the mixing process is assisted by a cone of air injected though a co-axial nozzle. The engineer expressed concerns regarding the alignment of the individual jet grout columns for the cross walls due to the potential deviations in the jet grout drill sting at the required depths of 105. The potential deviations were a concern due to the structural

nature of the cross wall acting as a subsurface struts. Although Layne GeoConstruction was confident that a deviation from vertical for the drill string of less than 1/100 could be achieved due to the soft soils underlying the fill, there remained questions about obstructions contained within the fill layer which could adversely affect vertical alignment. After further consideration, a complete bottom seal was proposed to satisfy the vertical alignment concerns. The full bottom seal was economically feasible as jet grouting allows for the creation of discrete columns at any starting and stopping elevations, unlike soil mixing where equipment limitations require treatment from the ground surface to the design tip elevation. 6 JET GROUT TEST SECTION Figure 6 illustrates the geometric layout of the jet grout test section consisting of six 5 diameter columns, spaced at 5 center-to-center. Coring of test columns would be performed to verify geometric and mechanical properties of the improved soil. Test jet grout columns were installed from a depth of 98 to 60 below working grade. Prior to installing the columns all pilot holes for jet grout column and core hole locations were provided a steel casing for the purpose of measuring the deviation from vertical. The pilot hole deviations were measured using a gyroscopic survey which produced rectangular coordinate deviations. The data was used in conjunction with coring data to evaluate the jet grout test section.

268 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-33.qxd

2/18/04

12:39 PM

Page 269

Figure 6. Double-fluid jet grouting detail. Figure 7. Jet grout test section – geometric layout.

Five core borings were taken at the locations in shown in Figure 6 using a size P triple barrel wire line coring system and a diamond coated face discharge bit manufactured by Christensen Products.

• •

• •

Core 1 was sampled at the interstitial space and only recovered traces of grouted soil. Cores 2 and 3 were sampled at the tangential point of two columns and recovery was mostly loose fragments of jet grout that exhibited signs of excessive grinding – however drilling resistance was significantly greater than samples recovered in clay above the treatment zone. Core 4 was sampled at the center of the column and produced recovery of full cylindrical cores. Core 5 was located between columns, but 2.1 from the center, slightly less than tangent. Recovery was jet grout and only small amounts of clay, however, the lower 20 of core had no recovery.

The core borings at the column interface (tangent) exhibited poor recovery due to the non-homogeneity of the jet grouted soil at the outer limits of their intended column diameter. This is particularly evident with the double and triple fluid systems of jet grouting where the assistance of the air turbulence has much less concentration farther from the point of origin (nozzle). It is also important to note that coring systems are designed for hard formations and it is difficult to obtain good recovery of jet grouted clays and silts due to their soft nature. In this case it appears that seams of good jet grouted soil, layered between softer clays, were present and caused the harder jet grouted soil to delaminate from the softer material and shift within the core barrel during the coring process. This produces a grinding effect at the bit, and resulted in poor recovery. Hence the drill resistance indicated the presence of jet grouted soil, although the core recovery did not confirm this fact. A split spoon sample SPT 1 was performed as an alternate method to verify the presence of jet grouted

soil at the tangent interface of two 5 diameter columns. The samples recovered exhibited 8 of jet grout with significant clay with average N-values of 21 blows/ft and 16 of jet grout with little clay, and average N-values of 26 blows/ft. The N-value of the clay prior to jet grout soil stabilization was 0–6 blows/ft. The results of the core borings and SPT borings indicated significant jet grout was present at the tangent between two 5 diameter columns, although the homogeneity of the columns at their extreme limit of 5 diameter was variable. The results were deemed acceptable by the Engineer and 5 diameter columns spaced at 5 tangents were recommended for base heave resistance. Due to the variable nature of the jet grout at the extreme limit of 5 diameter the Engineer recommended a 4 center-to-center spacing along the sheet pile walls to ensure adequate toe support. 7 JET GROUTING PRODUCTION The final geometric layout for production work is detailed in Figure 7. The top and bottom elevations of each column were calculated and tabulated based upon the profile grade line of the cut and cover tunnel, in addition to the approximate elevation for the top of glacial till. Detailed shop drawings were approved and work commenced. Two drill rigs were procured specifically for this work with emphasis on the deep treatment zone up to 105 below ground surface. Figure 8 illustrates a drill rig working on production columns. Each drill rig is fully automated to control the retraction rate of the drill rods, ensuring consistent quality. The grout plant consisted of a bulk cement storage silo, a colloidal cement grout mixer, and a high pressure grout pump. To ensure consistent quality, automated features were employed at the grout plant. The grout

269 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-33.qxd

2/18/04

12:39 PM

Page 270

Figure 9. Jet grouting drill rig. Figure 8. Jet grout production work – geometric layout.

mixer accurately measures the mix proportions of cement and water to achieve the desired specific gravity. This feature was pre-programmed to the required mix proportions and batching was performed using the automated features of the mixing plant. The grout plant technician periodically verified the grout specific gravity, by mud balance, as a check to our automated plant function. Jet grout injection pressure was continuously monitored at the internal pressure gauge of the grout pump, and grout volume was measured using the automatic stroke counter at the grout pump control panel. The information was recorded and submitted for inclusion into the project record. In total over 300 columns were installed to treat the Boston Blue Clay between depths of 60 and 105. The work was accomplished in 2–3 months. Core sampling of production jet grout columns was performed at 10 locations near the center of the columns. Unconfined compressive strength testing was performed by an independent laboratory and yielded average strength of 840 psi. Excavation for the cut and cover tunnel proceeded upon completion of the jet grouting work, yielding a dry and stable bottom surface (Figure 9). The cut and cover tunnel was constructed without any dewatering and the structure was placed on top of the prepared jet grout bottom seal. 8 CONCLUSIONS Jet grouting proved to be the most economical alternative for soil stabilization on this project. Cost estimates for the soil stabilization performed by the Engineer prior to the bid proved to be very conservative. Keen observations by the Engineer and Contractor during the test section uncovered the limitations of core sampling jet grouted soils. SPT borings confirmed the presence of jet grouted soil, while the coring operation observed drilling resistance but variable recovery.

Figure 10. Base of excavation, 60 below sea level.

The SPT method proved to be the more reliable method of confirmation of the presence of jet grout when treating soft soils. The jet grouting was successfully completed in Boston Blue Clay. The jet grouting was performed to depths of over 105, with specialized hydraulic drill rigs and grout mixing and pumping equipment with automated features for optimum quality control. ACKNOWLEDGEMENTS The author would like to acknowledge the support and technical contributions of David Shields of GEI Consultants, Inc. of Winchester, Massachusetts, as well as Pier Luigi Iovino, Division President of Layne GeoConstruction, in the preparation of this paper. REFERENCE GEI Consultants, Inc. July 1998. Final Geotechnical Data Report, MBTA Contract E02CN15, South Boston Piers Russia Wharf and Fort Point Channel Tunnel. Prepared for Frederic R. Harris, Inc. Boston, MA.

270 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-34.qxd

2/18/04

12:40 PM

Page 271

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

North airfield drainage improvement at Chicago O’Hare International Airport: soil stabilization using jet grouting D.A. Lewis & M.G. Taube Nicholson Construction Company, Cuddy, Pennsylvania, USA

ABSTRACT: Drainage improvements at O’Hare International Airport, Chicago, IL included the installation of a drainage and storm water system to control the overflow from nearby Willow Higgins Creek. This involved construction of a weir structure at the creek and the channeling of water from the weir to a newly constructed reservoir via three, 3.7-m (12-ft) diameter, underground storm sewer lines. The sewer lines passed beneath an existing, 2.3-m (90-inch) diameter high-pressure water main, which was to remain in service throughout sewer line installation. The soil composition around the water main ranged from medium stiff clays to silty sands and sandy silts. Triple-fluid jet grouting was used to stabilize the variable soil profile beneath the water main in preparation for tunneling and installation of the sewer lines. The varying soil strata presented a challenge to the project team to establish a constant set of jet grouting parameters throughout the stabilized zone while keeping within the specified unconfined compressive strength range for the stabilized zone of 690 to 1380 kPa (100 to 200 psi) at 28 days.

1 INTRODUCTION Willow Higgins Creek runs within the boundaries of Chicago-O’Hare International Airport. During periods of heavy rain, the creek regularly overflows and floods airport property and nearby businesses and homes. To control storm water overflow, the City of Chicago planned to install an intake weir at the creek to channel the overflow to a newly-constructed reservoir via three, 3.7-m (12-ft) diameter storm sewer lines. These lines passed beneath an existing, 2.3-m (90-inch) diameter high-pressure water main. This water main provides service for Chicago’s western suburbs. The invert elevation of the water main was 193.0 m (633.25 ft), approximately 4.5 m (14.75 ft) below existing grade. Soils in the area of the water main crossing generally consisted of medium-stiff clays from existing grade to elevation 191.6 (628.5 ft). The clay stratum was underlain by silty sand to elevation 190.0 (623.5 ft), beneath which sandy silt extended to the storm sewer invert at elevation 186.8 (613.0 ft). Silty clay was encountered between elevations 186.8 m and 185.5 m (613.0 ft and 608.5 ft), overlaying a silt stratum. Given the flowable nature of the silty sand and sandy silt through which the storm sewers were to be advanced, measures were required to eliminate ground loss or heave during tunneling operations that would threaten the integrity of the water main.

Options considered for protecting the 2.3-m (90-inch) diameter water main included temporarily diverting the flow from the water main, or installation of structural framing to support the water main during installation of the 3.7-m (12-ft) diameter storm sewer lines. These two options were determined to be more expensive and more risky than a soil treatment option. Triple-fluid jet grouting was specified to achieve the required improvement. 2 JET GROUTING TECHNOLOGY The fundamental principle of the jet grouting technique is a high-speed erosional jet acting under a nozzle pressure of up to 50 MPa (7250 psi). The soil structure is destroyed and the soil is eroded, and typically mixed in situ with cement-based grouts to form various geometries, depending on the application. The design of the jet grouting work is often the prerogative of the specialty geotechnical contractor and is based on empirical considerations that take into account the specific subsurface conditions, equipment characteristics, and the specification requirements. Given that the jet grouting technique is virtually independent from the soil texture and structure, it can be applied to a variety of conditions. In the majority of applications, a neat cement-water mix is used, with

271 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-34.qxd

2/18/04

12:40 PM

Page 272

Figure 3. Limits of treatment: Elevation view perpendicular to crossing.

parameters was selected in order to provide treated soil that was stiff enough to control ground movement, but not too hard to make excavation difficult. The geotechnical contractor was faced with two challenges:

Figure 1. Limits of treatment: Plan view.

• •

To establish a constant set of jet grouting parameters throughout the varying soil strata while keeping within the unconfined compressive strength range specified for the stabilized zone. To avoid detrimental impact to the high-pressure water line as a result of jetting operations in close proximity.

4 TEST PROGRAM Figure 2. Limits of treatment: Elevation view in direction of crossing.

initial rheological properties typically characterized by low viscosity and low rigidity. Grout additives can be used in particular applications. Jet grouting has been used in the U.S. since the 1980’s and is now a well-accepted ground treatment technique that can provide innovative solutions to very difficult engineering problems such as those at O’Hare Airport. 3 PROJECT OBJECTIVE The intent of the jet grouting program was to treat the potentially unstable soils in order to eliminate ground loss or heave, and thus minimize movement of the water main, during tunneling operations for the sewer lines. The defined treatment area was 19.5 m by 20.3 m (64 ft by 66.5 ft) in plan and extended from elevation 194.5 to elevation 183.9 (638 to 603.5 ft), approximately three meters (10 ft) below existing grade (see Figures 1–3), providing a stabilized soil mass with an unconfined compressive strength of 690 to 1380 kPa (100 to 200 psi) at 28 days. The range of strength

Test columns were installed between elevations 188.1 m (617 ft) and 192.0 m (630 ft) to encompass the full range of soil conditions encountered during the production work. The geotechnical contractor designed the test program to achieve a target unconfined compressive strength of 690 kPa (100 psi), meeting both seven-day and 28-day strength requirements. During test column construction, cement/water ratios, replacement ratios, and lift and rotational speeds were varied while fluid pressure and flow remained constant. For the initial tests columns, the geotechnical contractor selected experience-based parameters of cement/water ratios of 0.5 and 0.6, and replacement ratios of 40, 45 and 50 percent. The replacement ratio is defined as the theoretical volume of soil replaced by grout. Bentonite was added to the mix for both test and production columns to control grout bleed to reduce unconfined compressive strength. Following the initial series of tests performed with cement/water ratios of 0.5 and 0.6, the owner requested that additional test columns be constructed using higher cement/water ratios in order to assess how workable stronger jet grout columns would be. To that end, test columns using 0.8 and 1.0 cement/water ratios were constructed. The strengths of these columns ranged from 627 kPa to 4447 (91 to 645 psi), and it

272 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-34.qxd

2/18/04

12:40 PM

Page 273

Table 1. Test program strength results. Sample retrieval elevation m (feet) 192.9 (633) 192.0 (630) 190.8 (626) 189.0 (620)

Soil description

Range of strength at 5 days: Cement/water ratios 0.5 & 0.6 kPa (psi)

Range of strength at 5 days: Cement/water ratios 0.8 & 1.0 kPa (psi)

Stiff brown clay, dry Gray sandy silt, wet Gray sandy silt, wet Gray silty sand, wet

152–958 (22–139) 283–1,200 (41–174) 469–848 (68–123) 889 (129)*

2,654–3,771 (85–547) 1,882–4,447 (273–645) 627–3,020 (91–438) N/A

*One test only.

was agreed by the owner and engineer that the treated soil would be too difficult to excavate at these higher strengths. 4.1

Verification

The verification program included:

• • • • •

Continuous monitoring of the main grouting parameters over the full length of the column, including fluid pressure, fluid flow, rate of rotation and rate and withdrawal of the monitor. Excavation and visual examination of test columns for diameter measurement and continuity of construction. Wet grab samples obtained at four elevations and strata for future testing. Spoil sampling. Batch plant sampling.

Core samples were also retrieved and independently tested in the laboratory for unconfined compressive strength. In order to shorten the duration of the test program, the curing time for the jet grout samples was reduced from seven to five days. Table 1 presents the range of unconfined compressive strengths obtained at five days during the test column program. These strengths represent approximately 60 percent of the 28-day strengths based on the interpretation of strength versus maturity relationships for concrete (MacGregor, 1988). 4.2

Production parameters

From the results of the test program, optimum jetting parameters of 0.6 cement/water ratio and 50 percent replacement ratio were determined to achieve 1.68-m (5.5 ft) diameter columns for the production work. 5 PRODUCTION WORK Prior to jet grouting, three monitoring points were established on the crown of the water main within the zone where the pipe passed through the area to be stabilized. These points were installed by hand-excavating

Figure 4. Jet grouting in progress.

to the top of the pipe, placing sleeves through which survey rods could be inserted, and backfilling around the sleeves. Pipe elevation was continuously monitored during jet grouting operations. For the majority of the work, jet grouting locations were designed to be on a 1.52 m (5 ft), center-to-center triangular pattern to construct overlapping, 1.7-m (5.5-ft) diameter columns. Adjacent to the water main, columns were to be constructed on a 15 degree batter along a straight line offset 305 mm (12 inches) from the outer extent of the pipe. Battering of the jet grout columns was necessary to treat the soils beneath the 2.3-m (90-inch) diameter water main. Grouting work was initiated along the water main alignment to create a system of primary columns to support the pipe as the intermediate columns were installed. Drilling and grouting operations were accomplished using a 90-mm (3.5-inch) diameter, triple-walled drill rod mounted on a hydraulic track rig (Figure 4). Cement grout was colloidally mixed at an on-site batching plant and stored in agitator tanks before being fed to the jet grout monitor via highpressure pumps. Given the nature of the jet grouting process, ground disturbance is inevitable. The degree of disturbance is a function of the jet grout contractor’s experience and knowledge. Sequencing of column installation,

273 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-34.qxd

2/18/04

12:40 PM

Page 274

adequate spoil return through the drill holes and venting of the pressures introduced into the soil will typically control, but not necessarily eliminate, heave potential. The specifications required that the contractor exercise extreme caution during grouting operations to avoid causing movement that might result in structural damage. No limit of magnitude of allowable movement was specified. To minimize the potential of pipe movement, the contractor implemented the following controls:

• • •

Sequencing of column installation. Close monitoring of spoil return. Continuous monitoring of the top elevation of the water main.

Despite these measures, during the first phase of column installation 12.2 mm (0.48 inches) of movement was recorded at the center of the section being monitored and 9.14 mm (0.36 inches) of movement was recorded at the east and west monitoring points. Work was halted under the water main, while column installation continued at other locations. When work resumed under the water main on a revised installation schedule, heave of the water main once again occurred during the construction of two columns. The observed movements were judged by the geotechnical contractor to be within limits commonly experienced in the industry for similar applications. The movement was also determined by an independent expert to be within the allowable deflection for the water main as designed. The owner, however, mandated zero movement of the pipeline after the initial movement had occurred. The approach to the installation of columns was modified as follows:

• • • •

The method of top of pipe surveying was changed from an automated laser to an optical leveling system. Drill hole diameter was increased from 152 mm (6 inches) to 200 mm (7.88 inches) to increase the annulus size for increased spoil return. Tighter requirements limiting the proximity of jet grout columns to be installed within the area immediately beneath and adjacent to the water main within the same shift were imposed. Vent holes were installed within treated areas to relieve the build up of pressures beneath the water main.

Completion of the support system under the water main and the remaining non-support columns was successfully accomplished in accordance with project specifications. 6 MONITORING AND TESTING During production jet grouting, on-board instrumentation provided continuous, real-time monitoring of

Figure 5. Horizontal borehole.

Figure 6. Installed sewer lines.

fluid pressure, fluid flow, rate of rotation, and rate of withdrawal of the monitor. Wet grab, spoil and batch mix samples were retrieved and evaluated for unit weight, flow and bleed. Following completion of the jet grouting and installation of the jacking and receiving pit, the tunneling contractor bored three, 406-mm (16-inch) diameter horizontal holes along the center of the alignments of the three, 3.7-m (12-ft) diameter storm sewer lines. These pilot holes were advanced to allow the tunneling contractor to assess the performance of the treated soil. Video inspection of the east and west alignments of the horizontal bores demonstrated that the bores remained open within the stabilized mass after the casings were removed (Figure 5). Extraction of the casing on the west alignment beyond the stabilized zone resulted in soil collapse into the borehole. Once exposed in preparation for sewer line tunneling operations, the stabilized vertical face on the north side of the treated area remained intact and stable. Tunneling was accomplished without inflow of soils into the excavation (Figure 6) and with zero movement of the water main.

274 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-34.qxd

2/18/04

12:40 PM

Page 275

Many details of the sampling and treatment requirements were not adequately addressed in the project specifications. For example, for the pilot holes that were advanced the full length of the three sewer lines, the specifications required that the grout installation be verified by video inspection and the areas of inadequate or ineffective grout be re-grouted and re-inspected at no additional cost to the owner. The methods by which the treated soil was to be sampled were not specified. During advancement of the augered pilot hole, samples of treated soil that had been bored through, churned up, wetted and transported via two sets of augers were collected. Upon collection and visual inspection of the highly disturbed samples, it appeared to some that the soil treatment was not adequate, even though the pilot boreholes remained open and the video inspection showed continuous treatment. It was soon realized, however, that the sampling techniques utilized yielded very low quality and highly unrepresentative samples. The excellent treatment obtained was further demonstrated during the subsequent tunneling operations.

often one of educator in order to furnish owners and their engineers unfamiliar with the technology with a level of confidence in the technique itself and with the performance of the product. The geotechnical contractor should be involved not only with the planning of the jet grouting treatment, but also with the establishment of reasonable sampling techniques and verification requirements, which should come under the specialist’s scope. ACKNOWLEDGEMENTS The authors would like to thank the following for their cooperation on this unique and challenging project:

• • • •

Nicholson Construction Company, Cuddy, PA: Jet Grouting Contractor. Plote Construction, Inc., Elgin, IL: General Contractor. Airport Owner’s Representatives, Chicago, IL: Owner’s Representative. Brunzell Associates, Ltd., Skokie, IL: Pipeline consultant for Nicholson.

7 CONCLUSIONS Over its approximately 20-year U.S. history, jet grouting has gained steady acceptance, particularly as an underpinning and excavation support technique, with many successful applications reported in the technical literature. However, soil stabilization by jet grouting to facilitate tunneling through soft, mixed face or flowable conditions has seen much fewer applications. The role of the specialty geotechnical contractor therefore is also

REFERENCES MacGregor, J.G. 1988. Reinforced concrete mechanics and design. New Jersey: Prentice Hall. Pellegrino, G. 1999. Soil improvement technologies for tunneling: selected case histories. Proceedings of stateof-the-art technology in earth and rock tunneling; ASCE metropolitan section spring geotechnical seminar, New York, 26–27 May 1999.

275 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-35.qxd

2/18/04

12:40 PM

Page 277

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Ground freezing and spray concrete lining in the reconstruction of a collapsed tunnel S.J. Munks & P. Chamley Ove Arup and Partners, London, United Kingdom

C. Eddie Morgan Tunneling, Rugby, United Kingdom

ABSTRACT: During the construction of 14 ft diameter flow transfer tunnel for Yorkshire Water in Kingston upon Hull, UK, a 350 ft length of completed tunnel collapsed within a few days of construction. The tunnel was constructed using a full face Earth Pressure Balance (TBM) and a segmental concrete lining in alluvial and glacial soils with a high water pressure. Artificial Ground Freeze with Liquid Nitrogen as the freeze medium was used to stabilize the ground to permit the reconstruction of the tunnel along its original alignment. Excavation and reconstruction was carried out within the frozen ground using sequential excavation techniques and sprayed concrete primary and secondary lining. This paper describes the design of the Artificial Ground Freeze, the reconstruction of the tunnel and a description of the lessons attained from constructing a major civil engineering structure in frozen ground. Details on the extensive monitoring system and safety measures implemented are also given.

1 BACKGROUND 1.1

Introduction

The flow transfer tunnel was being constructed for Yorkshire Water to direct sewage flows to a new wastewater treatment works being constructed to allow compliance with the European Union’s Urban Wastewater Treatment Directive (UWWTD), which was commissioned and implemented through UK legislation to clean coastal waters. Miller Civil Engineering Ltd (MCEL), now Morgan EST Tunneling, was commissioned to design and construct the Flow Transfer Works, and Ove Arup and Partners was appointed to act as Project Managers. The total scheme budget for the transfer tunnel was £67 million. The conditions of contract for the construction of the flow transfer tunnel were based on the Institution of Chemical Engineers (IchemE) Model form for Process Plant, ‘Greenbook’. This provided a Target Cost reimbursable contract incorporating ‘painshare’ and ‘gainshare’ incentives. The 5.2 mile transfer tunnel was located along the north bank of the River Humber and was designed to provide gravity flow from connections to the existing sewerage system to the new treatment works. The 12 ft internal diameter tunnel was constructed using

an Earth Pressure Balanced tunnel boring machine (EPBM). The tunnel was lined with a pre-cast segmental lining comprising a 6 piece reinforced, tapered ring of 1 ft thickness, fitted with EPBM gaskets. The ground conditions surrounding the tunnel comprised alluvial and glacial deposits. The alluvial deposits consisted of clay, silt, sand, gravel and peat, which lay conformably on the glacial deposits comprising clay, fine to medium sand and gravel. The Upper Chalk was at depth beneath the tunnel and rested unconformably with the glacial deposits. Two aquifers were present along the route. The upper aquifer was approximately 2 m below ground level and was hydrostatic. The second aquifer was tidal and was beneath the laminated clays, situated in the lower glacial deposit and chalk. 1.2

277 Copyright © 2004 Taylor & Francis Group plc, London, UK

The collapse

At 00.30 on 16th November 1999, a locomotive driver reported signs of water inflow at a segment joint which was carrying fine sand into the tunnel, at a point some 650 ft behind the TBM. Despite efforts by the tunnel gang to stem the flow of material, water and sand inflow levels rapidly increased and the tunnel became destabilized such that at 03:00 the same day,

09096-35.qxd

2/18/04

12:40 PM

Page 278

Figure 1. Location of collapse.

Figure 2. Extent of crater.

the tunnel was evacuated and subsequently collapsed. The collapse was focused some 20 ft to the east of a maintenance shaft, known as T3, which was located within a car park close to Hull town centre. Immediately following the collapse the A63 in the proximity was closed and the nearby properties evacuated; both as a precautionary measure. The location of the collapse is illustrated on Figure 1. At the location of the epicentre, the tunnel and ground surface sank by some 8 ft within a ‘crater’ spome 200 ft in diameter and it was calculated that approximately 6560 ft3 of ground entered into the tunnel. The extent of the crater is illustrated on Figure 2, an aerial photograph of the site. In total, approximately 350 ft of tunnel was affected by the collapse and required reconstruction. 1.3

Summary of ground and ground water conditions at the collapsed section

Following the collapse, a detail ground investigation was conducted to provide information for the remedial works and to determine the length of the collapsed tunnel section. The ground conditions surrounding the tunnel alignment following the collapse are summarized on Figure 3. 1.4

Cause of failure

An intensive investigation into the cause of the failure was launched and following extensive physical and numerical modeling, it was concluded that longitudinal differential movement had initiated the collapse. A significant factor which influenced the speed and magnitude of the collapse, was the presence of a thin layer of fine, single size Aeolian (Wind blown) sand

Figure 3. Summary of ground conditions following collapse.

resting on top of the Alluvial Glacial Deposits. This material proved to be highly mobile in the presence of high water pressure and was able to exploit a very minor leak and eventually cause total failure of the tunnel. 2 ARTIFICIAL GROUND FREEZE 2.1

Several options, including Tunnel Diversion, Cofferdam Construction, Jet Grouting and Artificial

278 Copyright © 2004 Taylor & Francis Group plc, London, UK

Selection of remedial works

09096-35.qxd

2/18/04

12:40 PM

Page 279

Ground Freeze (AGF) were reviewed for the reconstruction of the tunnel. With the exception of tunnel diversion, all options considered involved stabilizing the ground at the collapsed section and reconstructing the tunnel along its original alignment. Supporting the ground by AGF and forming the tunnel lining with Sprayed Concrete Lining (SCL) was adopted as this was deemed to provide the optimum solution when considering the local ground conditions, safety, program, buildability and cost. Liquid Nitrogen (LIN) was chosen as the freeze medium as it exists at a lower temperature than other freeze mediums, i.e. brine, and had sufficient cold energy to freeze the saline and moving ground water at the location of the remedial works. In addition, horizontal ground freezing using LIN was the only method that would enable the undamaged tunnel to be captured safely. 2.2

Figure 4. Designed horizontal freeze.

A cross section of the designed horizontal freeze structure is given in Figure 4.

Design of freeze system

Three phases existed for the life of the ice structure at Hull, which is common to many Civil Engineering Projects that implement AGF. The first phase was termed the Primary or Active Freeze Period and comprised the development of the freeze until it reached its design thickness. The second phase was termed the Secondary or Passive Freeze Period and comprised maintaining the ice structure at design thickness during the excavation. The final phase, termed the Thaw Period, occurred after the tunnel lining had reached its design strength and the freeze system became redundant. The freeze system used at Hull was an open system, with LIN, which exists at approximately 196°C, being pumped into a series of freeze tubes and released at an exhaust. As the LIN passed through the freeze holes, the LIN warmed and boiled as heat was removed from the ground and the resulting gas was released from the system at the exhaust. The temperature of the gas at the exhaust formed the control for the system and was termed the Set Point. Extensive testing of frozen soils taken from the post collapse recovery zone was undertaken to establish suitable parameters for the design of the frozen ground support structures. The design of the ice structure took into account the highly disturbed ground conditions that existed surrounding the collapse. For this project, Finite Element analysis (FE) was used to design the ice structure, using data from the extensive testing. The FE demonstrated that each freeze tube had to provide a 2.5 ft freeze radius giving a 5 ft diameter cylinder of ice around each tube. On this basis, the freeze tubes were installed at a maximum of 2.5 ft spacing.

2.3

2.3.2 Pressure Relief Hole A pressure relief hole was used to ensure that the excavation area was enclosed from the surrounding ground and ground water before excavation commenced. The Pressure Relief Hole comprised a slotted

279 Copyright © 2004 Taylor & Francis Group plc, London, UK

Development and maintenance of ice structure

2.3.1 Control of Set Points During the Primary Freeze, the Set Point was placed at a low temperature and once the ice structure reached its design thickness, the Set Point was increased. Various temperatures were used for the Primary Freezes at Hull. Initially, low temperatures in the region of 140°C were used as these achieved the designed ice thickness in the shortest time. Although using such low temperatures allowed excavation to commence quickly, within 14 days of switching on the freeze system, the disadvantage was that it allowed high accumulation of cold energy within the excavation zone; this resulted with high compressive strengths of the soil matrix. For the subsequent excavation stages, the Set Point was raised to the region of 120°C during the Primary Freeze; although this increased the lead-time for excavation to approximately 30 days, the accumulation of cold energy in the excavation area reduced which prevented such high gains of the compressive strength of the soil matrix. This reduced the effort required to excavate the frozen ground. Once the Primary Freeze was accomplished, the exhaust temperature was raised to the region of 90°C and then raised in 20°C increments. The Secondary Freeze exhaust temperature was never raised above 40°C.

09096-35.qxd

2/18/04

12:40 PM

Page 280

pipe inserted the full length of the excavation and fitted with a ball valve and pressure gauge. Operation of the Pressure Relief Hole was simple. Prior to commencing excavation, the Pressure Relief Hole was opened and if no water was emitted, the excavation area had to be sealed from the surrounding groundwater. The Pressure Relief Hole also provided an indication on how the freeze was developing, because as the ice structure developed, the pressure of the water trapped within the excavation area increased; the increase of pressure be monitored on the pressure gauge. 2.3.3 Monitoring of ice thickness To monitor the thickness of the ice structure during the Primary and the Secondary Freezes, monitoring holes, which consisted of cased holes with thermocouple arrays, were positioned along side the freeze tubes. Typically, one thermocouple for every 16 ft3 to 66 ft3 of frozen soil was used. Each thermocouple had a Threshold temperature, which when reached indicated that the ice wall had reached its design thickness. During excavation, it became apparent that the ground being excavated was much colder than predicted and therefore the ice structure was thicker than designed. A graphical means of projecting the ice thickness was devised which comprised formulating graphs by plotting the temperature readings in the thermocouples against the distance from the nearest freeze tube. Assuming a conservative linear approach, best-fit lines were plotted on the graphs, extending from the source temperature in the freeze tube, through the known temperature at the thermocouple, to zero. The predicted ice thickness was the resultant at zero. To maintain a conservative approach, the source temperature was taken as LIN’s injected temperature 196°C. This assumed the greatest temperature gradient and therefore projected a conservative ice thickness. An example of a graph used to predict the ice thickness, is given on Figure 5. This graph demonstrates an ice thickness in the region of 1 m. Using a combination of the graphical and the original method for calculating the ice thickness achieved a greater control over the freeze and reduced the compressive strength of the soil matrix and the consumption of LIN.

Figure 5. Predicted ice thickness.

ground within the excavation was excessively cold, a temperature of 105°C having been recorded, it was determined that the ground had begun to shrink and formed cracks. The cracks were not treated as they did not provide a means of water ingress as the ground was sufficiently cold that water entering the system would freeze before it entered the excavation area. The cracks were visually monitored for movement, as block failure had been identified as a low risk. As previously mentioned, a Set Point of 140°C had been used for the initial construction stage to allow the freeze to develop in minimum time; this caused a significant accumulation of cold energy in the excavation area that was not able to dissipate. The cracks were not encountered on the remaining construction stages when a warmer Set Point was used during the Primary Freeze. 3 RECONSTRUCTION OF COLLAPSED SECTION 3.1

2.3.4 Cracks in first excavation phase During excavation of the first construction phase, cracks were observed in the tunnel face, radiating both further into the excavation and outwards towards the freeze pipes. The reasons for the cracks are given below. As ground freezes, it initially expands and once a certain temperature has been attained, which is dependent on the ground conditions, it shrinks. As the

280 Copyright © 2004 Taylor & Francis Group plc, London, UK

Introduction

The reconstruction of the collapsed section of tunnel was conducted in five stages, with two to the west, and three to the east of Shaft T3. The five construction stages were referred to as West 1 (W1), West 2 (W2), East 1 (E1), East 2 (E2), and East 3 (E3). The length of the construction stages was governed by drilling constraints and ranged from approximately 65 ft to 80 ft in length. The tunnel axis was at a depth of 65 ft below ground level.

09096-35.qxd

2/18/04

12:41 PM

Page 281

Figure 6. Construction sequence.

The construction sequence to the east of the shaft is illustrated on Figure 6. The construction sequence to the west mimicked that to the east, but only had two construction stages. Each construction stage was supported and closed to the surrounding ground and ground water horizontally with a circular ice wall and vertically with a frozen bulkhead. The bulkheads were constructed by drilling typically 23 vertical holes from the surface to below the level of the tunnel invert and freezing the surrounding ground. The horizontal freezes were constructed by drilling typically 33 horizontal or slightly inclined holes to the vertical bulkheads and freezing the surrounding ground. Prior to drilling the horizontal freeze tubes, the freeze bulkheads were replaced with a 16 inch reinforced concrete structural bulkhead with a characteristic strength of 25 N/mm2. To allow the excavation of the first set of horizontal freeze tubes, bulkheads were formed at the tunnel eyes within the shaft. The horizontal drill holes for all construction stages, except W2 and E3, splayed away from the tunnel centre line to allow an increase in the excavated diameter from 18 ft to 25 ft. The increase in diameter was to allow for the construction of a drill chamber within the tunnel to facilitate the drilling of the next advance of horizontal freeze pipes. As the final construction stages (W2 and E3) did not require a drill chamber, the freeze tubes remained at a constant diameter around the tunnel. To prevent the LIN flowing too quickly along the inclined freeze holes, they were fitted with weirs, typically at 20 ft intervals where the gradient was steep and 10 ft intervals where the gradient was gentle. The freeze holes were drilled using directional drilling techniques. Once drilled, the holes were surveyed and as an built plot, to 0.01 inch, accuracy was compiled. If necessary, additional freeze holes were drilled. In general, between 1–3 freeze holes had to be redrilled for each construction stage. Once drilled, the 4 inch cased holes were fitted with 2 inch copper tubing and were plumbed into the nitrogen supply. Copper tubing was used as it is ductile and would not fracture in the cold conditions. Where cold energy was not

Figure 7.

Figure 8. Drill chamber.

required to be dissipated, the freeze pipes were insulated, normally with rockwool or foam. An example of a competent ‘as built’ survey is given on Figure 7 and a photo illustration of a drill chamber is given on Figure 8. 3.2

Excavation plant

The choice of plant was limited as it was governed by the size of Shaft T3. Shaft T3 had been constructed as a maintenance shaft and had a 20 ft internal diameter. A Schaeff Roadheader, a 17 tonne tracked machine with the capability of using either a rotary cutter or a

281 Copyright © 2004 Taylor & Francis Group plc, London, UK

‘As Built’ horizontal freeze.

09096-35.qxd

2/18/04

12:41 PM

Page 282

hot water, in addition a geocomposite insulating layer was attached to the side of the tunnel prior to spraying the shotcrete. These measures ensured that the concrete gained sufficient strength before the cold halted the cement hardening reaction. The primary lining had a design strength of 25 N/mm2 at 28 days. Prior to spraying the secondary lining, the tunnel was heated and a waterproof membrane was applied to the primary lining. The 16 inch shotcrete secondary lining comprised a steel reinforced shotcrete layer and was applied following the completion of the construction stage. The secondary lining had a design strength of 45 N/mm2 and was formed from the same mix as the primary lining without the steel fibres. The steel fibres were omitted as structural reinforcement was included in the secondary lining. The secondary lining was tested at 3, 7, and 28 days for strength, with 5 cores for every 300 ft3 sprayed, providing results were constant. If the results were not constant, 5 cores were tested every 150 ft3, in accordance with the Method Statement.

Figure 9. Schaeff.

pneumatic breaker was used for excavation. The Schaeff was lowered into the shaft in pieces and assembled in the pit bottom. In areas where the Schaeff could not be used, a smaller tracked machine with a breaker was utilised. Hand mining was used to trim the edge of the excavation to obtain the profile yet was intentionally limited due to Hand Arm Vibration Syndrome, a condition which is aggravated by the cold conditions. A plate illustrating the Schaeff being used during the excavation stage is given on Figure 9. 3.3

Excavation and tunnel lining

The two excavation stages adjacent to the shaft (W1 and E1) could not be conducted concurrently due to a lack of space. Once the tunneling operations had progressed away from the shaft, excavation proceeded concurrently to the east and west. Prior to the excavation of any of the construction stages and at 6 ft intervals along the tunnel drive, 4 probe holes were drilled to identify if there were any isolated spots of saturated ground. The tunnel was advanced in 3 ft sections, split into crown, bench and invert, with steel lattice girders used to maintain the tunnel profile. Following the completion of each 3 ft advance of the crown, bench or invert, a 12 inch shotcrete primary lining was applied to the tunnel. The shotcrete mix had been designed following extensive trials in cold conditions and comprised an accelerated mix with steel fibres. The steel fibres were used to act as reinforcement to aid in crack prevention. To aid in overcoming the cold conditions, the shotcrete was batched using heated aggregates and

3.4

Challenges faced during the excavation of the tunnel comprised the removal of a locomotive and the removal of the plates from the collapsed tunnel. The locomotive abandoned in the vicinity of Shaft T3 during the collapse. The locomotive had to be cut and removed in sections as excavation progressed as the ground surrounding it was frozen. All excavation that took place around the locomotive was manual to ensure that there was no leakage of hydraulic oils, diesel or battery acid. The segments from the collapsed tunnel were removed as excavation progressed. In some instances, such as in the epicentre of the collapse, the collapsed tunnel segments extended beyond the excavation zone and had to be carefully removed to ensure that the freeze wall was not ruptured. Occasionally, freeze holes pierced through the collapsed tunnel segments. In such instances, the segment had to be removed by cutting along the excavation profile with a diamond saw to avoid dislodging the freeze pipe.

4 MONITORING 4.1

Ground surface

The monitoring regime at the ground surface was designed to confirm stability of the works from the excavation and to monitor the ground movements post collapse. The monitoring was conducted on a 24 hours basis during construction and gradually reduced to monthly readings.

282 Copyright © 2004 Taylor & Francis Group plc, London, UK

Anomalies in excavation

09096-35.qxd

2/18/04

12:41 PM

Page 283

Within 24 hours of a freeze system being switched on, heave was noted at the ground surface in the order of 0.07 inch. Total heave above frozen sections was of the order of 0.5 to 1.5 inches. 4.2

In-tunnel

Monitoring of the in-tunnel deformation was performed using precise three-dimensional surveys of arrays of Bioflex targets spaced along the tunnel at intervals of between 16 ft and 52 ft. Accuracy of the readings, which was affected by the quality of the targets and the atmospheric conditions in the tunnel, in particular the temperature, was 0.02 inch. The results of the deformation monitoring were processed using the Dedalos tunnel deformation program. This allowed the presentation of the vertical, transverse and longitudinal movements at each array to be plotted against time. The targets were positioned with one in the crown, two at shoulder and two at the knee. Following the installation of the secondary lining, the targets were repositioned and subsequently monitored. The results from the monitoring were reviewed daily against the trigger, action, and evacuation levels, set at 0.3%, 0.7% and 1.3% strain respectively. These levels were devised from the initial design of the ice structure and related to the creep and the loss of compressive strength of the ground over time. Only in one instant did the convergence readings exceed 0.3%. Monitoring of heave at tunnel level was conducted by monitoring the segmental tunnel adjacent to the collapsed section. A heave in the region of 1.2 inches was recorded, which correlated with the predicted values. 5 SAFETY A comprehensive safety system was installed, which included the normal rigorous safety procedures for working in confined spaces, in conjunction with incorporating systems for working in close proximity to LIN. A summary of the safety systems installed specifically for working in close proximity to LIN in a confined space, are given below:

•

•

Emergency Stop Buttons – these were installed both in the pit bottom and at ground surface. Once

• •

• •

6 CONCLUSIONS A part of the success of the remedial works at Hull is attributed to the collaborative effort under taken by all those involved on the project, which comprised Yorkshire Water, Ove Arup and Partners, Miller Civil Engineer and their sister company BeMo. The project used challenging civil engineering techniques and it was essential that all parties involved worked together as a team to resolve the problem. ACKNOWLEDGEMENTS The authors would like to thank Yorkshire Water for providing approval to submit this paper and www.petersmith.com for usage of photographs.

283 Copyright © 2004 Taylor & Francis Group plc, London, UK

pressed, the LIN would be cut off from the system, however the LIN would not stop being emitted until the LIN already in the system was released. Ventilation – forced ventilation was implemented at all times. Furthermore, the vent system was linked to the Nitrogen Sensor so that in the event of a leak, the vent system would switch to double speed to force all the LIN fumes from the excavation. Protection – the freeze pipes were protected by the installation of steel stools fitted with ambi-decking in the shaft to prevent damage from the materials being hoisted. Personnel and PPE – An intensive site safety induction was put in place informing personnel of the risks of working in close proximity to LIN. The self-rescue kits for working at Hull were specially designed to work in the temperatures that would be experienced from a LIN leak. Numerous ‘Tool Box Talks’ were given to the operatives to inform them of safe working practices when within close proximity to LIN. All personnel that worked in the tunnel were provided with specialist PPE. The nitrogen pipes were located within covered concrete trenches, with removable steel covering. Others – the pit bottom was decked out to allow access to the freeze pipes at all times. Also, as a preventative measure, the shotcrete equipment was constantly on standby.

09096-36.qxd

2/18/04

1:45 PM

Page 285

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Ground freezing for urban applications P.C. Schmall, D. Maishman, J.M. McCann & D.K. Mueller Moretrench American Corporation, Rockaway, New Jersey, USA

ABSTRACT: Ground freezing occupies a small, but nevertheless important, niche in underground construction where complete groundwater cut-off and in-built ground support are required, but difficult or disturbed ground is present and/or deep excavation is planned. Moreover, since ground freezing is a non-invasive technique that does not disturb the ground, it is particularly well suited to the challenges encountered in urban environments. In conditions that preclude ground displacement techniques, or where the vibration or settlement potential associated with conventional shoring techniques is a concern, ground freezing offers a viable, safe alternative. This paper presents a brief introduction to ground freezing technology, together with case studies of ground freezing performed for a range of applications, including the most extensive urban ground-freezing project to date, on Boston’s “Big Dig.”

1 INTRODUCTION Many urban construction sites are characterized by restricted access, existing structures adjacent to the proposed excavation, active utilities proximate to or within the work zone, and a groundwater table above the bottom of the proposed excavation. Ground support options become limited if vibration or potential settlement associated with conventional shoring techniques are unacceptable, urban subsurface obstructions limit the use of displacement techniques, and/or access restrictions do not permit large construction equipment. The presence of difficult or disturbed ground may also preclude the use of other techniques. When the project calls for a watertight excavation, the ground support options become even more limited, particularly at greater depths. However, the use of ground freezing overcomes these challenges. Ground freezing is accomplished through smalldiameter, closed-end pipes placed in pre-dilled holes. The ground remains undisturbed, eliminating the difficulties associated with displacement techniques. Ground freezing can be accomplished in the full range of soils, from clays to cobbles and boulders, and in pervious or fissured rock, and frozen walls can be formed around underground structures or obstructions. Compared to other groundwater cut-off or excavation support methods, a frozen wall is easily connected to the underlying bedrock and will also adhere to adjoining subsurface installations, if necessary, to provide a composite cut-off structure. By providing a

complete groundwater cut-off, freezing does not impact the surrounding groundwater regime, which is often contaminated in urban areas. 1.1

Simply put, the principle behind ground freezing is the use of refrigeration to convert in situ pore water into ice through the circulation of chilled calcium chloride brine, or, in some circumstances, the evaporation of liquid nitrogen. However, the process is actually quite complex and requires a specialist contractor skilled in refrigeration, thermal analysis, groundwater flow and geotechnical engineering. To create a frozen earth cofferdam, or frozen soil mass, the closed-end freeze pipes are inserted into drilled holes in a pattern consistent with the shape of the area to be stabilized and the required thickness of the wall or mass. A frozen shaft may require freeze pipes hundreds of meters deep. As the brine moves through the pipes, heat is extracted from the soil causing the ground to freeze. The brine is returned to the refrigeration plant through an insulated header and, after re-cooling, is re-circulated within the closed system. The ice acts as a bonding agent, fusing together particles of soil or rock to increase the strength of the mass and render it impervious. Ground freezing is primarily used for shaft and tunnel construction or for the construction or excavation of other underground structures. Once the structure is completed, refrigeration is discontinued and in most cases the ground returns to its normal state.

285 Copyright © 2004 Taylor & Francis Group plc, London, UK

Ground freezing technology

09096-36.qxd

2/18/04

1:45 PM

Page 286

2 MASS FREEZING Although typical applications of ground freezing involve the creation of a peripheral frozen structure for excavation support and groundwater control, there are circumstances under which massive volumes of soil need to be stabilized to facilitate excavation within the frozen, stabilized ground. Such circumstances include projects where a broad spectrum of subsurface conditions exist, water control is paramount, and minimal intrusion or disturbance to the subsurface stratigraphy is required. Ground freezing meets all of these requirements.

2.1.1 Ground freezing design The ground-freezing program was designed by the ground-freezing subcontractor to provide multiple geotechnical functions including groundwater cutoff, encapsulation of the fill debris within a matrix of frozen ground, and improvement in the strength of the organics and marine clay along the tunnel alignment. The freeze system was composed of:

• • •

2.1

Case Study: Central Artery/Tunnel, Contract 9A4, Boston, MA

Jacking of three, massive tunnels just beneath the seven, active Amtrak lines serving Boston’s South Station and the financial district, and through what has been described as “… the most difficult soil conditions imaginable,” is acknowledged as the most demanding component of the overall Central Artery/ Tunnel Project (Rogers and Taylor 2003). The jacked tunnel method was selected to allow full rail service to the station during tunnel construction. The soil profile through which tunneling would take place included 6.1 m of historic fill containing building debris, granite seawalls, piles, wharf structures, and abandoned brick structures (Fig. 1). This was underlain by soft, organic material and a marine deposit of Boston Blue Clay. Groundwater was encountered approximately 3 m below ground surface. Preliminary stabilization concepts included dewatering, grouting the fill stratum and organics, and soil nailing the marine clay. With just 2 m of cover between the box and the rail tracks, the General Contractor was concerned with the potential for unacceptable heave generated by grouting, and also settlement, with removal of the numerous obstructions within the fill, and elected to use ground freezing to stabilize the excavation face.

Figure 1. Excavation of frozen, open-work rubble fill in Boston.

• •

Accurate spacing and freeze pipe location was critical to the success of the stabilization program. Finite element modeling was performed to determine the transient heat flow from the ground to the freeze pipes in order to evaluate freeze pipe disposition, freeze formation period, and freeze plant capacity. With the majority of the excavation work undertaken in the Boston Blue Clay, which also required the greatest refrigeration effort, all design work and detailed thermal analyses were focused on this stratum (Donohoe et al. 2001). 2.1.2 System installation and operation Freeze pipe installation was accomplished from the rail tracks using sonic drilling techniques. Highly maneuverable, high-rail mounted equipment was utilized to precisely locate the pipes among the complex track switchgear, and the sonic drill head could readily penetrate many of the obstructions. This was invaluable in allowing the contractor to maintain the required pipe configuration (Donohoe et al. 2001). Chilled brine was circulated through the freeze pipe system for three to four months prior to tunneljacking. A computerized instrumentation system, also designed by the ground-freezing contractor, was installed and remotely monitored to ensure frozen ground conditions prior to the start of tunneling. The ground-freezing program provided a stable and predictable, completely unsupported, 24.4-m wide by 12.2-m high vertical face that allowed the use of an open-faced tunnel shield (Fig. 2). The tunnel was jacked in place without incident, or without disruption to rail service, and the use of ground freezing was estimated to have resulted in a $4 M saving (Angelo 1999). Ground freezing was utilized for all three tunnel jacks.

286 Copyright © 2004 Taylor & Francis Group plc, London, UK

A massive, centrally-located refrigeration plant, A brine circulation system capable of pumping 15,000 liters per minute to any/all of the three jacked box areas, An insulated brine supply and return manifold system installed completely within the rail track structure, Heat pipes to control the lateral growth of frozen ground beyond the box perimeter, and In excess of 2000, 115-mm O.D. steel freeze pipes installed from 14 m to 18 m below ground surface

09096-36.qxd

2/18/04

1:46 PM

Page 287

Figure 2. Excavation of frozen, free-standing face in Boston.

3 DEEP SHAFT EXCAVATION Ground freezing has been used to allow shafts to be sunk in water-bearing ground to considerable depths. The technique is ideally suited for mixed ground conditions varying from highly permeable sands and gravels to clay and rock, and difficult ground conditions commonly encountered at the soil/rock interface where the geology is generally the most challenging and where displacement or improvement of the ground by other techniques is impractical. With ground freezing, the difficult ground does not need to be replaced and is simply incorporated into the frozen structure. The frozen ground conforms perfectly to the contours of the rock surface, and provides a continuous, high-strength, impermeable material through the soil/rock interface. On any deep shaft project in water-bearing ground, ground freezing is the most practical solution at depths greater than 30 m, and becomes more economic with more difficult ground and with increased depth. 3.1

Case Study: New York City Water Tunnel No. 3, Shaft 22B, Brooklyn, New York

Access Shaft 22B for the New York City Water Tunnel No. 3 is located in Lower Brooklyn, and is bounded on two sides by multi-story apartment buildings and on the other two sides by busy thoroughfares. Subsurface conditions through which the shaft was to be excavated consisted of approximately 9 m of fill, with groundwater encountered 3 m below the surface. The fill was underlain by a 1.5-m thick organic peat layer, beneath which granular sands and gravels extended to a depth of approximately 44 m below grade. The remains of the Gardiner’s Clay formation was

encountered from 44 m to 46 m, effectively creating an upper and lower aquifer within the shaft. Granular soils extended from below the Gardiner’s Clay to decomposed rock at 50 m, with sound bedrock encountered at a depth of 67 m. The shaft was designed with an excavated diameter of 13 m from ground surface into competent bedrock and a decreased diameter for the remainder of the shaft, terminating at approximately 213 m beneath the surface. Ground freezing was used to provide a complete groundwater cut-off and structural support for shaft excavation in the overburden. Freeze pipes were installed using rotary sonic drilling methods. The 2.5-m thick frozen wall was formed using 45, vertical freeze pipes on a 16-m circular pattern, extending 3 m into bedrock. Since the completed frozen wall would extend within 6 m of the adjacent apartment buildings, foundations were carefully monitored for movement during ground freezing operations. Freeze work was completed within eight weeks without damage to the adjacent structures, allowing the General Contractor to begin excavation (Fig. 3). 3.2

Shaft 23B, the break-out shaft for the New York City Water Tunnel No. 3, Stage 2, is located in a heavily trafficked area of South Brooklyn near the mouth of the Brooklyn-Battery Tunnel and at the junction of the Gowanus and Brooklyn-Queens Expressways. At the break-out location, the water tunnel lay more than 105 m below the top of biotite gneiss bedrock, which was overlain by 40 m of soft overburden soils. These consisted of 6 m of surficial fill overlaying

287 Copyright © 2004 Taylor & Francis Group plc, London, UK

Case Study: New York City Water Tunnel No. 3, Shaft 23B, Brooklyn, New York

09096-36.qxd

2/18/04

1:46 PM

Page 288

Figure 3. Shaft 22B, New York City Water Tunnel No. 3.

1.5 m of peat bog (the original shore line), beneath which glacial sands extended to top of rock. Near the soil/rock interface, the sands were interspersed with boulder beds. Groundwater was encountered at approximately 4 m below ground level. Ground freezing was selected to provide watertight support for excavation of the 12-m diameter shaft into bedrock. Forty five, vertical freeze pipes, were installed in an 18-m diameter, circular pattern around the shaft location, and seated 1.5 m into bedrock. During drilling for freeze pipe installation, previously unknown, old building foundations were encountered at a depth of 2.5 m in four of the drill holes, but were successfully penetrated. With the frozen cofferdam in place, the outside edge of the frozen ground extended within 3 m of a multistory apartment building. The use of ground freezing eliminated the potential for foundation subsidence and superficial structural damage, and the technique was later used for the construction of a further four shafts along the route of Water Tunnel No. 3. 3.3

Northeast Ohio Regional Sewer District, Mill Creek Tunnel, Phase 2

During periods of heavy rainfall, the capacity of the Southerly Sewage Treatment Plant in Cuyahoga

Heights, Ohio, is insufficient to handle the additional volume. The Northeast Ohio Regional Sewer District therefore elected to construct a 6-m diameter, 13-km long tunnel through bedrock at 60 m beneath the surface to divert and hold rainwater and residential/industrial wastes until periods of low demand at the plant. The cost of constructing such tunnels to act as reservoirs is more economical than increasing sewage plant capacity. Two of the four access shafts required for tunnel construction lay in an area that was a river valley some 10,000 years ago. Groundwater was present at approximately 18.3 m below the surface. Soils below the groundwater table were primarily glacial, highly permeable, residual, silty sands. A coarser sand and gravel layer was encountered immediately above top of rock. Dewatering to the top of the bedrock was rejected because of the potential for ground loss within the saturated, unstable soils at the soil/rock interface. The Construction Manager for the shaft excavation therefore specified ground freezing to avoid the potential for soil loss in this difficult stratum while at the same time providing groundwater control and support of excavation for the full length of the shafts. Vertical freeze pipes were installed around the proposed, 9.75-m diameter shaft locations and seated 3 m

288 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-36.qxd

2/18/04

1:46 PM

Page 289

Figure 4. Frozen shaft, Mill Creek tunnel. Figure 5. Frozen ground excavation support for tank removal at Huntington Hospital.

into rock. Calcium chloride brine circulating through the freeze system created a 3-m thick, watertight cofferdam through which excavation was able to proceed without incident (Fig. 4). 4 LIQUID NITROGEN FREEZING On smaller projects where the ground is maintained in a frozen state for a relatively short period of time, or in emergency situations, liquid nitrogen is often used as the freezing agent. Although a day of freeze formation with liquid nitrogen is more expensive than with brine, colder temperatures are achievable, allowing the liquid nitrogen to freeze the soils much more rapidly, typically in days rather than weeks, thus offsetting the unit cost. Nitrogen is typically cost effective when the freeze must be formed quickly and remain in place for a limited period, and little or no maintenance of the freeze is required.

ground freezing to a depth of 6 m to provide the structural support needed during excavation and tank removal. Liquid Nitrogen was selected as the freezing agent rather than calcium chloride brine because the excavation would only be open for several days. The freeze wall would be constructed with closely spaced freeze pipes in the form of an arch surrounding approximately two thirds of the excavation zone, leaving one portion open. This would be sloped back during excavation for ease of tank removal. To increase the strength of the freeze wall, water was added to the sands by sprinkler hoses during the initial freezing period. The system was installed with a minimum amount of disturbance. The freeze was formed in a few days, and the tanks were successfully removed in one day (Fig. 5). 4.2

4.1

Huntington hospital, Huntington, New York

Removal of two, abandoned 76000-L oil tanks at Huntington Hospital presented a number of difficulties. The tanks were buried 4.6 m below the surface, immediately in front of one of the hospital’s building walls and close to operating room facilities. Furthermore, a 9500-L diesel underground storage tank was in the immediate vicinity of the tanks to be removed. Prevailing soil conditions consisted of moist sands, with the natural water table well below the bottom of the proposed excavation. A conventional earth support system for the proposed excavation generated significant concerns with the hospital administration. Given the proximity of the highly sensitive operating rooms, vibrating or driving steel sheet piling or H-beams was not permitted and, if conventional drilled in soldier beams and lagging were to be used, there was a potential for ground loss. The earth retention contractor therefore proposed

At the Newtown Creek WWTP, a new 1.5-m diameter transmission line was to be installed using microtunneling techniques. The access shaft to launch the microtunneling operation was located in an extremely congested area of the plant, adjacent to active tanks, and was supported by tight steel sheeting on all four sides. Subsurface soils were a mix of native silty sands and organic silts, with groundwater present approximately 3.0 m below ground surface. The invert of the 1.5-m diameter “eye” through which the microtunneling machine would be launched was at 6.0 m below groundwater level. During excavation of the shaft, a split was uncovered in the sheeting below the eye to be cut. Due to the groundwater pressure outside of the sheets, approximately 15 to 20 m3 of material blew into the excavation, displacing the ground behind the eye and replacing it with loose, sloughed-in material. Although slurry

289 Copyright © 2004 Taylor & Francis Group plc, London, UK

Newtown Creek Wastewater Treatment Plant, Brooklyn, New York

09096-36.qxd

2/18/04

1:46 PM

Page 290

grouting was performed to fill the voids in the now highly unstable, sloughed-in soil, the ground on the backside of the sheets was still an indefinable, loose, water bearing mixture of the various sloughed-in soils. Ground freezing was selected as the most cost-effective and practical approach to stabilizing the soils, because the freezing would provide assured results more or less independent of the soil type and consistency. Since cutting the eye and installing a sacrificial concrete liner in preparation for future tunneling was a one day operation, liquid nitrogen was selected as the freezing agent. The freeze pipes were installed vertically from ground surface to 1.2 m below the tunnel invert on a irregular pattern due to a myriad of subsurface obstructions. Freezing was accomplished in three days and the eye was cut and the liner installed against a vertical free-standing frozen face in one day without incident. 5 CONCLUSIONS Ground freezing is a niche technology, fulfilling a geotechnical need that is difficult, and in some instances impossible, to fulfill through other groundwater control

and excavation support methods. For deep excavation, difficult or disturbed ground conditions, or for the stabilization of complex subsurface profiles, ground freezing offers the assurance critical to successful and timely urban construction. ACKNOWLEDGEMENTS The authors extend their appreciation to the management of Moretrench American Corporation, and its ground freezing division, freeze WALL, for their input and cooperation in the preparation of this paper. REFERENCES Angelo, W.J. 1999. Crucial Link Nears Completion With The Aid Of Soil-Freezing. Engineering News Record, December 13, 1999 Donohoe et al. 2001. Ground Freezing for Boston Central Artery Contract Section C09A4, Jacking of Tunnel Boxes. Proc. Rapid Excavation and Tunneling Conference, San Diego, California: 337–344 Rogers, C.R. & Taylor, S. 2003. The Big Dig’s Big Dig. Civil Engineering, September, 2003: 40–49

290 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-37.qxd

2/18/04

1:47 PM

Page 291

Session 3 Design/build contracting practices

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-37.qxd

2/18/04

1:47 PM

Page 293

Session 3, Track 1 Predicting and controlling cost and schedule

Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-37.qxd

2/18/04

1:47 PM

Page 295

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

An economic approach to risk management for tunnels Bashar Altabba HNTB Corporation, Boston, MA, USA

Herbert Einstein Massachusetts Institute of Technology, Cambridge, MA, USA

Hugh Caspe HNTB Corporation, Boston, MA, USA

ABSTRACT: Comprehensive and proactive management of risk is vital to the success of tunneling projects. A well planned and executed approach to managing risk is essential for making informed decisions, evaluating options and developing contingency plans to deal with financial and contractual risks and to improve safety on tunneling projects. Prudent and responsible management requires a broad view of all risks, their effects, the probable reaction of other stakeholders and the possible down-the-line consequences. Potential lawsuits, lack of insurance, equipment failure, errors and omissions, results of third party involvement and changed conditions are all risks that need to be considered when creating risk management plans and strategies. This paper presents an economic approach to risk management using qualitative and simplified quantitative measurements to foster confidence that a less risk prone project can be constructed. This approach will help reduce, to manageable levels, the uncertainty and its associated costs, to all parties, that are common to tunneling projects.

1 INTRODUCTION Tunneling projects are affected by unique uncertainties. Numerous judgments and decisions that need to be made in the development of tunneling projects are made without complete information, and therefore give rise to some degree of uncertainty in the outcome. Risk management techniques can be employed to identify project risks and develop strategies to effectively address them. This paper proposes an approach based on a simplified economic model for evaluating tunneling uncertainties and successfully adjusting the tools available to effectively control those uncertainties. The paper will focus primarily on defining the overall process, rather than the detailed application of each tool, to establish a rationale for a comprehensive project risk management approach. The purpose of this process is to carry out a rough preliminary appraisal of the anticipated exposures and decide how to best address them. The proposed approach starts with the identification of major potential sources of exposure and the various scenarios anticipated for their occurrence. The possible consequences of each event are then described in terms of a rough gross monetary range or

non-monetary impact. By assigning probabilities to each occurrence, one can then evaluate the project risks prior to any response. Once consequences of exposure have been identified a decision has to be made on how to best address each element of exposure. If all risks were to be mitigated during design, it would prohibitively drive-up the cost of construction. Various alternate risk management tools are available to maintain a reasonable control over the expected cost of construction while reducing the degree of exposure. These include, Risk Mitigation, Risk Allocation and Risk Absorption. These alternates all have costs associated with them that can be estimated and included in the total cost of the project. Once a strategy is identified, the cost of its implementation is estimated together with the cost should that strategy not be fully successful. This approach is not complex, and is a simple rational process to dealing with uncertainty and insuring that proper account is taken of foreseeable risks. The aim is to allow proactive management in advance, instead of allowing risks to mature, requiring crisis response. Effective project risk management should include the following elements in a thorough and systematic approach.

295 Copyright © 2004 Taylor & Francis Group plc, London, UK

2/18/04

1:47 PM

Page 296

Risk Identification

I. Risk Assessment

09096-37.qxd

Probability of Occurrence P

Consequence of Occurrence C

III. Contingency

II. Risk Response

Risk Evaluation R=PxC

Risk Mitigation

Design Mitigation

Risk Allocation

Performance Mitigation

Residual Risk Contingency

Risk Sharing

Risk Transfer

Residual Risk Contingency

Risk Absorption

Unforeseen Risks

Contingency Planning

Experience

Residual Risk Contingency

Unforeseen Risks Contingency

Figure 1. Risk management approach.

• • •

Risk Assessment. Determine and evaluate the major risk elements involved. Risk Response. Decide on an appropriate measure to deal with the risk. Contingency Planning. Make appropriate allowance for retained risks. See Figure 1.

2 RISK ASSESSMENT 2.1

Definitions

Risk combines the uncertainty and the consequences of an event. This can be expressed using the expected value as: (1) where R  Risk, P[E]  Probability of the event, and CE  Consequence of the event. Events can vary widely from equipment failures, to slowdowns because of different geologic conditions, to traffic delays. In principle, events could also have positive consequences but when used in conjunction with risk, one considers only negative consequences. In conjunction with construction, the consequences will be usually expressed in terms of cost and time, possibly also in terms of accidents and environmental effects among others. The probability term P[E] expresses the uncertainty with which an event can occur and ranges from 0 to 100%. Such probabilities can be estimated

subjectively or they can be drawn from objective statistical (relative frequency) records. An example of the former is the estimation of the probability that a major piece of equipment will fail. This can also be obtained from experience. An example of the latter can be the occurrence of traffic delays where one might have records on the number of occurrences per duration of construction. The definition of risk as given above is only one of many possibilities. Insurance companies for example, use the term risk to express the component CE in the above expression. In most civil engineering applications the definition as given above is used. The following additional points should be made: 1. It is possible that even if an event occurs its consequences may not occur. For instance, a settlement under a building may occur without the expected building damage. This means that there is a additional “conditional” probability P[CE|E] that needs to be applied to adjust the probability of occurrence. Expression (1) then becomes: (2) 2. It was assumed that risk has to be expressed in monetary terms (cost). This is usually done by transforming other consequences (time delays, injuries) into costs. This transformation is, however, not necessary as is done in multiattribute utility analysis, were consequences can be expressed by utilities (non-monetary values) and these in turn can be

296 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-37.qxd

2/18/04

1:47 PM

Page 297

related to different attributes (non-monetary objectives like schedule or safety). 3. The expected risk value model on which this analysis is based assumes that an event will occur a number of times. Should an event not be repetitive the results of this approach may not be applicable. 2.2

Scenario analysis

This is the central aspect of risk analysis and is also where professional experience comes to bear. Scenario analysis consists of a detailed description of the project and, most importantly, the process through which it is implemented. This starts with planning and goes via design to construction and could include operation if so desired. Additionally, aspects where the project implementation process can vary are also identified. Usually these are aspects where the process can go as planned (normal) or where it deviates from the normal. Depending on the phase for which the scenario analysis is conducted, there may be a varying number of different scenarios to consider. These can range from political demands, to changes in the regulatory requirements, legal actions, alignment changes, including many construction changes. These scenarios are the risk events that need to be considered. There are many ways in which the scenario analysis can be conducted. One approach is to have a group of experienced professionals involved in the project brainstorming and documenting the different scenarios with emphasis on identifying the events. A facilitator, one of the experts or an outsider with encompassing understanding, can then go through the scenarios developed, check them for consistency and document a final set of scenarios and associated events. A particularly good example of this process is the Cost Estimation Validation Workshops conducted by the Washington State DOT for all its major projects. Scenario analysis is not only essential as a basis for risk assessment but also encourage a thorough thinking about the specific project early in its planning and sets the stage for further analysis. Given the process, it is quite clear that the individuals involved in scenario analysis have to be those who have major responsibilities on or a stake in the project and are experienced in the specific issues involved in tunneling. 2.3

Consequence estimation

The consequences of the events can be described in terms of, cost, time or other quantitative or qualitative descriptions and specific values associated with event occurrence. The values can be described in different ways. Cost consequences are usually associated with monetary values. If time can then be transformed into cost, it is possible to express risk in purely monetary

terms. Similarly, accidents and environmental failures may also be transformed into costs. Another approach is to create classes of consequences, which can be verbally defined and/or each assigned with a numerical scale. The worst (most severe) class would then be associated with an event that has consequences severe enough to present a likelihood of preventing the project construction. Examples include catastrophic water inflow into a tunnel, major settlements or unacceptable environmental consequence. In any of these methods, it is possible to associate different weights with the consequence. 2.4

Probability in terms of fractions between 0 and 1 or percentages between 1 and 100 can be assigned to each event subjectively. More formal ways of achieving this has originated in economic decision theory. An example is the comparison of the estimated probability with the flip of a coin. This approach has been successfully applied in a number of recent major tunnel projects. Yet, it is important to apply so-called consistency checks. For instance if a probability of 30% is assigned to one event and 60% to another one, it is necessary to make sure that event 2 is really twice as likely to occur as event 1. Another approach of estimating probabilities is through the relative frequency method in which records of past construction projects may shed light on the frequency of certain events. An example being breakdowns of a typical piece of construction equipment such as a TBM machine. 2.5

Risk

Risk can be calculated with the process just described and can be performed using a spreadsheet. As mentioned earlier, of major importance is that risk assessment, scenario analysis, consequence estimation and probability estimation have to be done by experienced personnel.

3 RISK RESPONSE APPROACH Risk response development is perhaps the most delicate part of the risk management process, and it is here where many projects could benefit from a coherent and comprehensive approach. In general for each risk identified, see Table 1, there are a number of options or tools available to deal with that risk. These include mitigating the risk by avoiding or reducing it through design or construction measures, allocating the risk by sharing it or transferring it to another party, or absorbing the risk in a careful contingency

297 Copyright © 2004 Taylor & Francis Group plc, London, UK

Probability estimation

09096-37.qxd

2/18/04

1:47 PM

Page 298

Table 1. Risk identification. General risk types

100%

Potential risk categories High Risk

f. Environmental

Availability and pricing of materials, skilled labor and specialized equipment Soil/water condition Soil/water contamination

b. Schedule delays

c. Cost overruns

III. Technical a. Design

b. Construction performance

IV. Legal a. Contractual

b. Third party

40%

20% Low Risk

g. Utilities II. Organizational a. Management

60%

I. Mitigation

Probability

80% Unknown, unclear, delayed, unassigned

III. Absorption

I. External a. Regulations and permits b. Natural hazards c. Terrorism d. Bankruptcy e. Market condition

II. Allocation Incompetence Lack of organizational structure Inadequate planning Unrealistic schedule Lack of coordination Labor shortage or productivity Material shortage Unforeseen site conditions Owner scope change Lack of site access Schedule delays Inappropriate procurement strategy Contractor claims Under estimating Inadequate data/ground characterization Lack of experience Design inadequacies Last minute design changes Constructability Complexity Inadequate alternates analysis Construction safety Construction quality Rate of production Reliability Differing site conditions New complex technology Third party impact Misinterpretations Misunderstandings Inappropriate contractual strategy Measurement of pay items Variations in quantity Mobilization cost payment method Environmental Personal property

0

4 6 Moderate Consequence

8 High

0% 10

Figure 2. Risk evaluation and response matrix.

plan. The choice will depend on the particular project, the risks involved and the specific circumstances. Each response option needs to be evaluated, assessing its likely effect upon the risk, the feasibility and cost of implementing that option, and the impact of each option on the overall cost of the project. The most cost effective response option for each risk is typically selected, with the realization that the overall risk potential is rarely completely eliminated. Significant judgment should be exercised in determining which particular risk response measures is adopted. The effect of adopting such measures will generally be to reduce the risk but at a given cost to the project. When carrying out the planning and subsequent selection of the primary responses, it is useful to identify the parameters influencing the selection of the appropriate response. In order to facilitate a systematic approach to risk response selection a risk evaluation matrix is prepared representing the expected value of risk, see Figure (2). Equal risk value curves are then plotted and the matrix divided into two zones, high risk events occupying the top right corner and low risk events occupying the rest of the matrix. The clients risk tolerance would then need to be understood to select the appropriate risk value curve above which risks are considered to be high. Some owners may not want to accept any more risk than absolutely necessary. Owners who have a low threshold for risk often choose to specify more costly, conservative designs that can be constructed within lower bounds of risk. The overall shape or the risk assessment and response matrix is the same for all projects. However the values and scales of the axis will change depending on the specifics of the project.

298 Copyright © 2004 Taylor & Francis Group plc, London, UK

2 Low

09096-37.qxd

2/18/04

1:47 PM

Page 299

The following key issues should be kept in mind: 1. The probability estimation, expressed in verbal or numeric terms, carries with it the potential for misjudgment and the risk assessment may suffer from a biased or mistaken probability judgments. As such it may be important to pay more attention to the event impact than to their expected probability. 2. Knowing the client and his or her risk preference is of major importance. The equal value assumption of high probability and low consequence with low probability and high consequence may not be accurate. 3.1

Risk mitigation

Risk mitigations are approaches that recognize the risk and present solutions that either reduce an event probability of occurrence or reduce the probability that the consequence occurs, or both. Such reductions, i.e., mitigations, would lead to a modified residual risk (R) after implementation and would typically involve some cost to the project. One can thus evaluate if the mitigating measures are cost effective. This would occur if the reduction in risk is greater than the cost of mitigation. (3) were R  Risk, R  Residual risk after mitigation, R  Net reduction in risk and CM  Cost of mitigation. As such, for high risks, occupying the top right end of the Risk Assessment and Response Matrix, where the cost of mitigation is lower than the expected value of risk reduction, the rational economic response is to Mitigate. There are two basic mitigation strategies, design mitigation and performance mitigation. 3.1.1 Design mitigation Design mitigation is the basic tool utilized for mitigating risks that can be economically addressed by design. Options available for design mitigation are broad and may include among other approaches:

• • • • • • • • • • • •

Provide adequate time and resources for ground investigation Conduct investigation on the nature and condition of the surrounding structures and utilities in advance Hire experienced engineers Improve communications among all parties Utilize familiar proven technology Encourage peer reviews and use Technical Advisory Board Increase the level of conservatism of the design Use performance specifications were appropriate Specify a construction monitoring program Use the observational method Have regular design reviews Utilize value engineering sessions.

In applying these and other design mitigation measures care must be taken to ensure that the design does not become unduly conservative. Risk avoidance, by taking actions so that the risk event no longer impacts the project objectives, is tantamount to complete elimination of the probability of occurrence or the consequence of an event. This approach is often used when it can be achieved at no or little cost. An example of this approach is changing the alignment to avoid tunneling under a structure to eliminate third part impact. 3.1.2 Performance mitigation The contractor is ultimately responsible for the selection of the means of construction, the equipment and methods for prosecution of the work. Some of the options available that may help enhance performance include:

• • • • • • • • • • •

Performance mitigation is a delicate balance and care must be exercised not to become overly prescriptive in applying performance mitigation measures. 3.2

Risk allocation

For low risk events where the reduction in the expected risk value does not justify the cost of mitigation, the rational response depends on the severity of the expected consequence. High severity level events that the Owner may not want to absorb could be allocated to another party in a better position to bear them. In order to do this one must develop an understanding of the Owner’s risk tolerance. Risk tolerance is the point beyond which the event severity becomes intolerable to the Owner and need to be allocated in some manner.

299 Copyright © 2004 Taylor & Francis Group plc, London, UK

Use rigorous technical and financial prequalification procedures Disqualify bids with un-priced conditions, qualifications and disclaimers Utilize a pre-bid meeting to confirm understanding of what is required and offered Provide contractor value engineering incentives Have a sufficiently long bid period Evaluate contractor’s bid means and methods early in the project to ensure that they meet the project requirements Disclose all available subsurface information to bidders Establish and operate an owner controlled monitoring system with trigger limits and procedures throughout construction for critical elements Execute early site work contracts Pre-select excavation disposal area and construction sites Pre-purchase long lead items where appropriate.

09096-37.qxd

2/18/04

1:47 PM

Page 300

Risk allocation implies shifting, totally or partially, the impact of the risk event to a third party. With this approach risks can either be shared or completely transferred so as to spread the consequences through the use of contract language or by insurance. Its main use is typically limited to financial risk exposure. In allocating risk among the principal parties, experience has shown that the most cost effective strategy is to allocate risk to the party that is closest to its source and hence has control over its occurrence. One has to also ask: Which party would be most able to carry the risk if it cannot be controlled? Which party is best positioned to manage the risk effectively? The client may, however, deem it in his interest to retain some control over the management of the risk after it occurs, as in the case of hazardous materials. If the distribution of risk is not clearly understood, or is patently unfair, then disputes become almost inevitable. It is further important to keep in mind that risk allocation does not remove the risk, but simply gives another party the responsibility of managing it. It is therefore essential that recipients of allocated risks are able to manage the risks allocated to them, otherwise the project will remain exposed. 3.2.1 Risk sharing Tunneling differs from general construction contracting. Unexpected events, including those resulting from unforeseen site conditions, have a greater impact on the progression of work than in other types of construction. Differing contract provisions employed in tunneling projects have a material effect on the distribution of risk among the parties involved. The ideal tunnel contract is one that clearly defines the responsibilities, duties and obligations of each party. Yet standard contract documents typically imply a specific risk allocation, which may not be appropriate for tunneling projects. Good contracting practice will distribute equitably the risk of construction among the parties, to reduce the overall cost of construction and reduce the uncertainties by apportioning the risks. Various types of contracts can be utilized in which the degree of associated risk is shared differently between the parties. The choice of type of contract appropriate for a specific tunnel project is an integral part of the project risk management process. The specific approach for each identified risk and the determination of how they should be shared between the parties should be defined with the insertion of clear language in the contract documents that allocate the responsibility. The following contract recommendations are generally accepted approaches for sharing risk:

• •

Changed Conditions Clause Use of Geotechnical Baseline Report as a benchmark for bidding

• • • • • • •

3.2.2 Risk transfer Risk transfer involves shifting all of the risk to another party. However, risk transfer on a project should not remove ones incentive to managing the risk. The rewards should remain associated with successfully managing the risk. There are many risk transfer tools available including among others:

• • • • • • • • • • •

General Commercial Liability Insurance, which covers third-party bodily harm, property damage or faulty workmanship Professional Liability Insurance, which covers damage and injury resulting from negligence Workers’ Compensation Insurance, which covers injury to workers Builder’s Risk Insurance, which covers damage to work during construction Construction Equipment Insurance, which covers damage to contractor’s equipment Environmental Liability Insurance, which covers injury, damage and clean up cost from release of pollutants Bonding. Bid and performance bonds should be balanced between the rights and the obligations of the parties. The size of the performance bond could be reduced by prequalification if appropriate Guarantees and Warrantees Subcontracting. Highly specialized components are often better subcontracted to a specialty firm Contract Type. Lump sum type contracts can be used to transfer the potential variation in cost of construction risk to the bidder Contract Language. Certain risks can be explicitly transferred in the contract i.e., Construction Safety.

3.3

Risk absorption

Risk absorption entails the acceptance of the risk consequences and planning to deal with them should they occur. This is the rational economic response

300 Copyright © 2004 Taylor & Francis Group plc, London, UK

Means and Methods. The contractor is ultimately responsible for the selection of the construction method, equipment and prosecution of the work Employ performance penalties as well as incentives, such as Liquidated Damages Define types of Third Party Impact and responsibilities Include variations in price clause and suitable mobilization bid items Utilize Escrow Bid Documents and include specific guidelines about what details or backup data the owner wants to be able to track Incorporate a Dispute Review Board for technical issues and not for interpretation of contractual matters Establish an equitable change order process.

09096-37.qxd

2/18/04

1:47 PM

Page 301

when the expected reduction in the value of risk is lower than the cost of mitigation and the consequence severity is relatively low. This option requires the careful estimation of an adequate budget and schedule flexibility for those risk elements the Owner elects to absorb and manage through funds allocated to a contingency, a form of self-insurance. 3.4

Unforeseen risks

Problems do occur on projects which are unexpected simply because they were not anticipated or foreseen during planning and design. Examples may include political or third party involvement, which may delay a project implementation. It is therefore prudent to carry some contingency reserves, both in budget and schedule, for situations that cannot be accounted for or cannot be quantified, to cover unforeseen events. The contingency may be based on experience or historic data and should be modified by special project circumstances. This contingency also covers market fluctuations that are difficult to determine early in the process i.e., market condition, labor supply, etc. 3.5

Contingency

Contingency is the combined cost of all residual risks and unforeseen risks. Residual risks are all the risks remaining after response implementation. This includes risk remaining after mitigation, sharing and risks absorbed. As such contingency allowances are typically included for any remaining risks, which are not specifically covered in the scenarios analyzed, as well as an allowance for unforeseen and unmeasured risks. In general the higher the level of confidence in one’s risk analysis the smaller the contingency needs to be. Two types of contingencies are therefore appropriate, one to cover the expected cost of residual risk and another based on past experience to cover unforeseen risks.

a rational manner. This involves systematically identifying the risks and evaluating their probability and impact, a method to evaluate risk through conceptualizing it as a potential monetary loss, and a procedure to respond to these risks depending on the expected risk value and the severity of the consequence. Where appropriate, response measures are identified to mitigate, allocate or absorb the risk. Residual risks are then handled through a contingency plan. Lastly, an important part of risk management is the continued development of a risk aware culture that accepts known costs now in order to avoid the possibility of unknown and significant costs in the future, while creating confidence that risks are being managed effectively.

REFERENCES ANSI/PMI 99-001-2000. A Guide to the Project Management Body of Knowledge (PMBOK) 2000. Atlanta: Project Management Institute. Abdel Salam, M.E. 1995. Contractual Sharing of Risks in Underground Construction: ITA Views. Tunneling and Underground Technology Vol 10 No 4: pp 433–437. Chernoff, H. & Moses, L.E. 1987. Elementary Decision Theory. New York: Dover Publications. Duncan Luce, R. & Raiffa, H. 1989. Games and Decisions: Introduction and Critical Survey. New York: Dover Publishers. Hatem, D. 1998. Subsurface Conditions Risk Management for Design and construction Management Professionals. New York: John Wiley & Sons. Hilson, D. 1999. Developing Effective Risk Responses. Proceedings of the 30th Annual Project Management Institute Seminars & Symposium. Philadelphia. Institute of Civil Engineers/Institute of Actuaries. 1999. Risk Analyses and Management for Projects (RAMP.) London: Thomas Telford. Parkin, J. 2000. Engineering Judgment and Risk. London: Thomas Telford Publishing. Piney, C. 2002. Risk Response Planning: Selecting the Right Strategy, Fifth Europeans Project Management Conference. Cannes: PMI Europe.

4 CONCLUSION Tunneling is considered to be a risky business. A simple economic approach to risk management is presented to assist in dealing with those uncertainties in

301 Copyright © 2004 Taylor & Francis Group plc, London, UK

09096-38.qxd

2/18/04

1:48 PM

Page 303

North American Tunneling 2004, Ozdemir (ed) © 2004 Taylor & Francis Group, London, ISBN 90 5809 669 6

Top down construction of Ramp L, Value-Engineering Change Proposal for the Massachusetts Turnpike Authority, Contract CO9A4 W.D. Driscoll & G.A. Almeraris Slattery Skanska, Inc. New York, New York

ABSTRACT: This paper describes a Value Engineering Change Proposal (VECP) to change one of Boston’s Central Artery tunnels from a drilled shaft supported and soil mix restrained ‘Cut and Cover’ tunnel into a ‘TopDown Construction’ slurrywall tunnel. This VECP was conceived and submitted by the Joint Venture of Slattery Skanska, Interbeton, JF White and Perini (SIWP). The schedule and cost benefits of the VECP were recognized with an award in 1999 from AASHTO for ‘Outstanding Value Engineered achievements in Teamwork, Cost Savings, Streamlined Construction Sequencing, and Reduction of Schedule’. Discussed in this paper are many of the technical challenges that had to be overcome both in the design process and during construction. It will describe how the revised construction sequence allowed at-grade and elevated structures to be constructed as soon as the ‘top’ was complete, shaving months off the project construction schedule. This construction method especially limited the risk related to unknown subsurface conditions, which could have caused high cost overruns and project delays.

1 INTRODUCTION The Ramp L Tunnel Structure was a critical section of the Central Artery/Tunnel (CA/T) Project’s CO9A4 Contract. With the overhaul of the City of Boston’s traffic system came an extension of the Massachusetts Turnpike to Logan Airport. For the first time, Interstate traffic traveling from either south or west of Boston could go directly to Logan Airport through a new system of tunnels. One of these tunnels, Ramp L, is the focus of this paper and was constructed in perhaps one of the most challenging locations of the CA/T Project. The Ramp L Tunnel was designed to carry Interstate I-93 traffic from south of Boston into the Ted Williams Tunnel and on to Logan Airport. The most complex portion of the CO9A4 Contract was the extension of the Massachusetts Turnpike (Interstate I-90), which had previously ended at the junction with Interstate I-93 at the South Station railroad tracks. The Contract called for the installation of tunnels under the live tracks at South Station by jacking pre-cast tunnel structures from thrust pits on the west side of the tracks to the east side of the tracks. Ramp L is an off-ramp from Interstate I-93 Northbound that carries traffic to a junction with the new Massachusetts Turnpike extension that merges with traffic heading east to the airport. The extension of the Massachusetts Turnpike to

the airport is the single largest aspect of the CA/T Project, involving billions of dollars in contracts. The opening of this extension was dependent first on the completion of the jacked tunnels and the subsequent cut and cover tunnels and second on the completion of Ramp L from the south. When the Owner accepted the contractor’s Value-Engineering Proposal for the construction method of Ramp L, much of the responsibility for the on-schedule completion of the structure switched to the contractor.

2 2.1

Contract solution for Ramp L

The CA/T Construction Contract CO9A4 called for the cut and cover construction of a new 706-ft long tunnel. The cross-sectional views of the proposed tunnel indicated no reason for concern over the risk associated with the actual construction of this tunnel. However, what was not detailed in the drawings of the tunnel was exactly how to reach the point where the actual tunnel construction could commence. Many months of work were required before the first concrete could be placed for the tunnel invert. Moreover, the work required to get to this point was laden with risk

303 Copyright © 2004 Taylor & Francis Group plc, London, UK

CONTRACT CO9A4, RAMP L

09096-38.qxd

2/18/04

1:48 PM

Page 304

and uncertainty and the potential for huge delays was a reality that could not be overlooked. The construction sequencing as suggested in the Contract Documents required that the tunnel be constructed in three segments, starting from the south and working northward. It is important to note that none of the activities described below could have been accomplished in a single mobilization stage. Some could be completed in two mobilization stages but most required three, one for each segment. The