Editors

Società Italiana Gallerie

Italian Tunnelling Society
Peila
TUNNELS AND UNDERGROUND CITIES:
Viggiani ENGINEERING AND INNOVATION
Tunnels and Underground Cities: Engineering and Innovation meet Archaeology, Celestino
Architecture and Art. Volume 7: Long and Deep Tunnel contains the contributions MEET ARCHAEOLOGY, ARCHITECTURE AND ART
presented in the eponymous Technical Session during the World Tunnel Congress
2019 (Naples, Italy, 3-9 May 2019).

The use of underground space is continuing to grow, due to global urbanization, Volume 7
public demand for efficient transportation, and energy saving, production and
Long and deep tunnels

INNOVATION MEET ARCHAEOLOGY, ARCHITECTURE AND ART

TUNNELS AND UNDERGROUND CITIES: ENGINEERING AND
distribution. The growing need for space at ground level, along with its continuous
value increase and the challenges of energy saving and achieving sustainable
development objectives, demand greater and better use of the underground space to
ensure that it supports sustainable, resilient and more liveable cities.

The contributions cover a wide range of topics, from studying tunnels in squeezing
ground conditions, via case studies on the Brenner Base Tunnel, the second Gotthard
Tunnel, CERN (HL-LHCand the Dubai Strategic Sewerage Tunnel, to TBM steering
difficulties.

The book is a valuable reference text for tunnelling specialists, owners, engineers,
archaeologists, architects, artists and others involved in underground planning,
design and building around the world, and for academics who are interested in
underground constructions and geotechnics.

7 Editors:
Daniele Peila
Giulia Viggiani
an informa business Tarcisio Celestino
TUNNELS AND UNDERGROUND CITIES: ENGINEERING AND
INNOVATION MEET ARCHAEOLOGY, ARCHITECTURE AND ART
PROCEEDINGS OF THE WTC2019 ITA-AITES WORLD TUNNEL CONGRESS,
NAPLES, ITALY, 3-9 MAY, 2019

Tunnels and Underground Cities:

Engineering and Innovation meet
Archaeology, Architecture and
Art
Volume 7: Long and deep tunnels

Editors
Daniele Peila
Politecnico di Torino, Italy

Giulia Viggiani
University of Cambridge, UK
Università di Roma “Tor Vergata”, Italy

Tarcisio Celestino
University of Sao Paulo, Brasil
Cover illustration:
View of Naples gulf

CRC Press/Balkema is an imprint of the Taylor & Francis Group, an informa business

© 2020 Taylor & Francis Group, London, UK

Typeset by Integra Software Services Pvt. Ltd., Pondicherry, India

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 publishers.

Although all care is taken to ensure integrity and the quality of this publication and the
information herein, no responsibility is assumed by the publishers nor the author for any
damage to the property or persons as a result of operation or use of this publication and/or
the information contained herein.

Published by: CRC Press/Balkema

Schipholweg 107C, 2316XC Leiden, The Netherlands
e-mail: [email protected]
www.crcpress.com – www.taylorandfrancis.com

ISBN: 978-0-367-46872-9 (Hbk)

ISBN: 978-1-003-03164-2 (eBook)
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

Table of contents

Preface xi
Acknowledgements xiii
WTC 2019 Congress Organization xv

Long and deep tunnels 3503

Long and deep single shield TBM in very complicated geology under the Alps.
Saint Martin La Porte 3505
C. Acquista, G. Giacomin, F. Martin & G. Comin
Experimental setup for studying tunnels in squeezing ground conditions 3515
K. Arora, M. Gutierrez & A. Hedayat
Brenner Base Tunnel – challenges of gripper TBM application for the 15 km
long exploratory tunnel Ahrental in challenging rock mass 3525
K. Bäppler & M. Flora
Risk management for the Brenner Base Tunnel 3531
K. Bergmeister
The second Gotthard tunnel tube 3540
G. Biaggio & V. Kumpusch
Dubai Strategic Sewerage Tunnel – challenges to infrastructure at unprecedented
depths in the region 3551
D. Brancato, A. Ayoubian, M. Joye, G. Monks, F.A. Al Awadhi & J. Tharamapalan
The hugest and more complex belt conveyor system in the longest tunnel under
construction in the world: Brenner Base Tunnel 3559
M. Bringiotti, S. Portner, V. Grasso & E.R. Vitale
CERN (HL-LHC): New underground & surface structures at Point 1 & Point 5 3569
A. Canzoneri, J. Amiot, F. Rozemberg, D. Merlini, F. Gianelli, G. Como, F. De Salvo,
C. Helou, L.A. Lopez & P. Mattelaer
Specificities of the underground structure design of Cigéo. Presentation
of the constrains and construction phasing stages of “Phase 1” 3579
G. Champagne De Labriolle, H. Ouffroukh, E. Boidy & H. Miller
Engineering challenges of the Snowy 2.0 pumped storage project 3588
B. Chapman, M. Yee & A.R.A. Gomes
Brenner Base Tunnel, construction lot Mules 2–3. Production management and site logistics
organization 3598
S. Citarei, M. Secondulfo, D. Buttafoco, J. Debenedetti & F. Amadini
Challenging mixed face tunneling at India’s Sleemanabad Carrier Canal 3605
J. Clark

v
Experimental determination and plausibility proof of the longitudinal displacement profile
for deep tunnels for case studies at the Brenner Base Tunnel 3616
T. Cordes, C. Reinhold, K. Bergmeister, B. Schneider-Muntau & I. Bathaeian
Performance of hard rock double shield TBM in tailrace tunnel of Uma Oya project,
Sri Lanka 3625
H. Darabi Kelareh, A. Rahbar Farshbar, A.H. Hosseini, F. Foroutan & D. Dodangeh
Lining stresses in a TBM-driven tunnel: A comparison between numerical results
and monitoring data 3634
V. De Gori, A. De Lillis & S. Miliziano
Numerical back-analysis of the Fréjus road tunnel and of its safety gallery 3644
M. De la Fuente, J. Sulem, R. Taherzadeh & D. Subrin
Influence of the excavation rate on the mechanical response of deep tunnel fronts
in cohesive soils 3654
C. Di Prisco, L. Flessati, G. Cassani & R. Perlo
Optimum tunnel system with regard to the entire lifecycle for long rail tunnels 3664
H. Ehrbar, C. Tannò & H.-P. Vetsch
Mechanized tunnel excavation of lot KAT2 of the Koralm-Tunnel, Austria: Achievements
and lessons learned 3674
D. Fabbri, R. Crapp, H. Hölzl & H. Wagner
The Susa-Bussoleno Interconnection Tunnel 3685
A. Farinetti, P. Elia, M.E. Parisi & E. Gueli
TBM steering difficulties. Innovative equipments: Strand jack and self retaining systems 3694
A. Finamore & G. Bellizzi
Risk assessment in the unlocking of the double shield TBM within the Los Condores
project, Chile 3704
A. Focaracci & M. Salcuni
Back-calculation model for instantaneous TBM cutter wear 3714
H.I. Frostad, A. Bruland, P.D. Jakobsen & F.J. Macias
Brenner Base Tunnel, Italian Side: Construction methods for a railway 22 km long tunnel 3724
S. Fuoco, R. Zurlo, M. Moja & E.M. Pizzarotti
Presentation of the successful crossing by the “Federica” TBM of a geological accident in
Saint-Martin-la-Porte construction site 3734
F. Gamba, E. Hugot, P. Gilli, C. Salot & G. Giacomin
Use of expanded clay as annular gap filling. Design and application at the Brenner Base
Tunnel 3745
F. Gasbarrone, A. Oss, L. Ziller, D. Buttafoco & J. Debenedetti
The Brenner Base Tunnel: Back-analyses of excavation data and TBM parameters
in the Aicha-Mules exploratory tunnel 3755
G.M. Gaspari & K. Bergmeister
Constrains, design and construction methods of the Cigéo five deep shafts 3765
C. Gaudry, H. Ouffroukh, L. Moscone & R. Taherzadeh
Special intersection formworks for cast in situ with cross vault shape 3775
M. Granata

vi
Extreme ingress: Managing high water inflows in hard rock TBM tunneling 3783
B. Grothen & D. Kough
Numerical investigations on the system behaviour of a ductile shotcrete lining with yielding
elements 3792
A.-L. Hammer & M. Thewes
Optimizing TBM cutterhead design for application in very strong and abrasive rocks, case
study of Kerman Water Tunnel 3803
J. Hassanpour, J. Rostami, Y. Firouzei & H.R. Tavakoli
Key construction technology of Long Metro Subsea Tunnel in complex site conditions 3814
W. He, C. Song & B. Du
Excavations of the Povazsky Chlmec highway tunnel in Slovakia 3824
M. Hilar & M. Srb
Requirements and methods for geometric 3D modelling of tunnels 3830
M. Hofmann, R. Glatzl, K. Bergmeister, Š. Markič & A. Borrmann
Brenner Base Tunnel – Key aspects of the guide design 3839
R. Insam, W. Eckbauer, T. Gangkofner & K. Matt
Brenner Base Tunnel – an example of planning and implementation of project optimizations 3849
R. Insam, W. Eckbauer, D. Zierl & M. Ebner
Brenner Base Tunnel – interaction between underground structures, complex challenges
and strategies 3857
R. Insam, R. Wahlen & G. Wieland
Logistic management in the longest drives of the Mont Cenis Base Tunnel 3866
M. Janutolo Barlet, G. Seingre, P. Bourdon, M. Zampieri, E. Humbert & C. Pline
Varies challenges for a new railway connection between Divača and Koper, Slovenia 3876
P. Jemec, D. Dvanajščak & E. Škerbec
The feeder 9, River Humber, replacement pipeline project, United Kingdom 3884
S. Jukes
Simulation of TBM operation to assess the impact of geology on the muck transportation 3895
A. Khetwal, J. Rostami & O. Frough
Analysis of micro slurry TBM excavation parameters in subsea tunnel 3904
K.Y. Kim, H.H. Ryu, D.S. Bae & S.A. Jo
Countermeasure of jammed TBM in rock tunnel excavation: Feedback from two cases study 3911
C.H. Lee, Y.C. Chiu & T.T. Wang
Site observations and physical model tests of the friction coefficient between concrete-type
element and foundation layer for immersed tunnel 3921
M. Lin, W. Lin & X. Wang
The Montcenis base tunnel: How to turn project environmental constraints into
opportunities 3930
S. Lione, L. Brino, P. Grieco, L. Pinchiaroglio & A. Mordasini
Anisotropic convergence of tunnels in squeezing ground: The case of Saint-Martin-la-Porte
survey gallery 3940
Y. Liu, J. Sulem, D. Subrin & E. Humbert
Overcoming extreme tunneling conditions on Vietnam’s longest tunnel 3948
S. Log, B. Li & P.N. Madhan

vii
Maly Lubon road tunnel: The case of a full-face excavation of a very large tunnel
in the Carpathian Flysch, Poland 3958
A. Lombardi, A. De Pasquale & V. Capata
Milan to Genoa high speed/capacity railway: The Italian section of the Rhine-Alpine corridor 3968
P. Lunardi, G. Cassani, A. Bellocchio & N. Meistro
Cravasco jobsite optimization for the management of narrow spaces: Underground crushing
plant and suspended conveyor belt 3979
P. Mancuso, U. Russo, A. Fossati, O. Urbano, L. Russo & A. Filice
Brenner Base Tunnel, Italian Side: Mining methods stretches – design procedures 3989
D. Marini, R. Zurlo, M. Rivoltini & E.M. Pizzarotti
Lessons from adverse geological occurrences during excavation in Rohtang tunnel, India 3999
P. Mehra
Hydrogeological, environmental and logistical challenges for TBM excavation in the longest
tunnel in the Italian territory 4006
N. Meistro, S. Caruso, A. Mancarella, M. Ricci, L. Di Gati & A. Di Cara
Surveying methods and geodetic control in long railway tunnels construction 4016
N. Meistro, L. Surace, R. Maseroli, A. Mancarella & M. Vantini
The Gronda Motorway by-pass at Genoa: Underground works east of the Polcevera River 4026
P. Mele, A. Raschillà, L. Mancinelli, C. Silvestri, A. Damiani & M. Agosti
Ismailia: The first tunnels to be constructed under the Old and New Suez Canals – a case
history 4039
G.L. Menchini & M. Liti
Brenner Base Tunnel, Lots Mules, 2–3 (Italy): The emergency stop in Trens 4049
D. Merlini, M. Falanesca, D. Stocker & A. Voza
Experiences of tunnels subject to earthquake in central Italy 4059
A. Micheli, L. Cedrone, A. Andreacchio & S. Pelizza
Tunneling challenges in Himalayas: A case study of Head Race Tunnel of 720 MW
Mangdechhu Hydro-Electric Project, Bhutan 4070
A.K. Mishra & I. Ahmed
Study on cracks in concrete lining based on inspection records on tunnel 4080
T. Miyaji, H. Hayashi, M. Shinji, S. Kaise & S. Morimoto
Study on deformation and failure behavior of mountain tunnel linings which consist of
various materials 4086
M. Mizutani & K. Yashiro
Developing cross passages and safety niches in a rationalized way using remote controlled
demolition robots 4095
C. Montorfano & R. Giti Ruberto
Mixed Ground TBM tunneling 4104
R.M. Myrlund, H.I. Frostad, P.D. Jakobsen & F.J. Macias
Design of countermeasure for squeezing and swelling in mountain tunnel 4114
Y. Okui, S. Kunimura, H. Ota, K. Maegawa, T. Ito, S. Kaise, H. Yagi,
K. Nishimura & N. Isago
Tunnel T-48: A challenge in India 4124
A. Panciera

viii
Assessments on the TBM performance at the Minas-San Francisco Hydropower
Project in Ecuador 4132
K.K. Panthi & J. Encalada
Shotcrete lining installed from an open gripper TBM remedial works 4142
G. Peach, B. Ashcroft, R. Amici & J. Mierzejewski
Long and deep hard rock double shield TBM under the Alps. Base Brenner 4152
P. Pediconi, M. Maffucci & G. Giacomin
Empirical and numerical analysis of the blast - induced structural damage in rock tunnels 4159
P. Pediconi, M. Maffucci & G. Giacomin
BBT, Lot Mules 2–3. Management of data gained by the pilot tunnel drive
for the twin main tubes 4169
M. Pescara, M. Spanò, N. Della Valle, R. Sorge, G. Giacomin, D. Buttafoco,
M. Masci & E. Nuzzo
Measurements of long time lining strain in creeping rock mass and modellingwith
viscoplastic laws 4179
R. Plassart, O. Gay, A. Rochat, A. Saïtta & F. Martin
Polcevera Adit – getting unstuck on EPB-TBM in squeezing rock 4188
F. Poma, U. Russo, F. Ruggiero & L. Lampiano
Sealing of huge water ingress in headrace tunnel of Uma Oya Project, Sri Lanka 4197
A. Rahbar Farshbar & A.H. Hosseini
Challenges of designing a Tunnel Boring Machine (TBM) for development
of underground structures on the Moon 4206
J. Rostami, C. Dreyer, R. Duhme & B. Khorshidi
Problems associated with an EPB-TBM in a complex geology with serpentinites
and peridotites in Turkey 4218
M. Sakalli, D. Talu, N. Bilgin & I.H. Aksoy
A new underground laboratory for exploring the deep universe: The design
of a third generation of a gravitational waves observatory 4225
L. Schiavinato, P. Mazzalai, G. Gemme, G. Losurdo, M. Punturo, A. Paoli,
F. Ricci, E. Calloni, G. Oggiano & M. Carpinelli
Cost analysis of mine roadways driven by drilling and blasting method and a roadheader 4235
O. Su & M. Akkaş
A softening damage-based model for the failure zone around deep tunnels
in quasi-brittle claystone 4242
E. Trivellato, A. Pouya, M.-N. Vu & D.M. Seyedi
The use of Integrated Big Data software management systems on a complex
TBM project. The Follo Line Project data analysis 4252
F. Vara Ortiz de la Torre
Brenner Base Tunnel, Italian Side, Lot Mules 2–3: Risk management procedures 4262
A. Voza, R. Zurlo, S. Bellini & E.M. Pizzarotti
Evaluation of excavation method for silo-shaped deep large waste repository
underground cavern 4272
K. You, C. Park & J. Park

ix
Real behavior of shotcrete primary lining in squeezing rock mass: The experience
at the Brenner Base Tunnel 4279
L. Ziller & S. Fuoco
Tunnel costs related to the quality of the rock mass 4289
R. Zurlo, S. Fuoco, M. Loffredo & A. Marottoli
The role of construction logistics for long and deep tunnels: The model of the Mules
2-3 construction lot for the Brenner Base Tunnel 4299
R. Zurlo, R. Di Bella & E. Rughetti

Author index

x
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

Preface

The World Tunnel Congress 2019 and the 45th General Assembly of the International Tun-
nelling and Underground Space Association (ITA), will be held in Naples, Italy next May.
The Italian Tunnelling Society is honored and proud to host this outstanding event of the
international tunnelling community.
Hopefully hundreds of experts, engineers, architects, geologists, consultants, contractors,
designers, clients, suppliers, manufacturers will come and meet together in Naples to share
knowledge, experience and business, enjoying the atmosphere of culture, technology and good
living of this historic city, full of marvelous natural, artistic and historical treasures together
with new innovative and high standard underground infrastructures.
The city of Naples was the inspirational venue of this conference, starting from the title
Tunnels and Underground cities: engineering and innovation meet Archaeology, Architecture
and Art.
Naples is a cradle of underground works with an extended network of Greek and Roman
tunnels and underground cavities dated to the fourth century BC, but also a vibrant and
innovative city boasting a modern and efficient underground transit system, whose stations
represent one of the most interesting Italian experiments on the permanent insertion of con-
temporary artwork in the urban context.
All this has inspired and deeply enriched the scientific contributions received from authors
coming from over 50 different countries.
We have entrusted the WTC2019 proceedings to an editorial board of 3 professors skilled
in the field of tunneling, engineering, geotechnics and geomechanics of soil and rocks, well
known at international level. They have relied on a Scientific Committee made up of 11 Topic
Coordinators and more than 100 national and international experts: they have reviewed more
than 1.000 abstracts and 750 papers, to end up with the publication of about 670 papers,
inserted in this WTC2019 proceedings.
According to the Scientific Board statement we believe these proceedings can be a valuable
text in the development of the art and science of engineering and construction of underground
works even with reference to the subject matters “Archaeology, Architecture and Art” pro-
posed by the innovative title of the congress, which have “contaminated” and enriched many
proceedings’ papers.

Andrea Pigorini Renato Casale

SIG President Chairman of the Organizing Committee WTC2019

xi
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

Acknowledgements

REVIEWERS

The Editors wish to express their gratitude to the eleven Topic Coordinators: Lorenzo Brino,
Giovanna Cassani, Alessandra De Cesaris, Pietro Jarre, Donato Ludovici, Vittorio Mana-
ssero, Matthias Neuenschwander, Moreno Pescara, Enrico Maria Pizzarotti, Tatiana
Rotonda, Alessandra Sciotti and all the Scientific Committee members for their effort and
valuable time.

SPONSORS

The WTC2019 Organizing Committee and the Editors wish to express their gratitude to the
congress sponsors for their help and support.

xiii
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

WTC 2019 Congress Organization

HONORARY ADVISORY PANEL

Pietro Lunardi, President WTC2001 Milan

Sebastiano Pelizza, ITA Past President 1996-1998
Bruno Pigorini, President WTC1986 Florence

INTERNATIONAL STEERING COMMITTEE

Giuseppe Lunardi, Italy (Coordinator)

Tarcisio Celestino, Brazil (ITA President)
Soren Eskesen, Denmark (ITA Past President)
Alexandre Gomes, Chile (ITA Vice President)
Ruth Haug, Norway (ITA Vice President)
Eric Leca, France (ITA Vice President)
Jenny Yan, China (ITA Vice President)
Felix Amberg, Switzerland
Lars Barbendererder, Germany
Arnold Dix, Australia
Randall Essex, USA
Pekka Nieminen, Finland
Dr Ooi Teik Aun, Malaysia
Chung-Sik Yoo, Korea
Davorin Kolic, Croatia
Olivier Vion, France
Miguel Fernandez-Bollo, Spain (AETOS)
Yann Leblais, France (AFTES)
Johan Mignon, Belgium (ABTUS)
Xavier Roulet, Switzerland (STS)
Joao Bilé Serra, Portugal (CPT)
Martin Bosshard, Switzerland
Luzi R. Gruber, Switzerland

EXECUTIVE COMMITTEE

Renato Casale (Organizing Committee President)

Andrea Pigorini, (SIG President)
Olivier Vion (ITA Executive Director)
Francesco Bellone
Anna Bortolussi
Massimiliano Bringiotti
Ignazio Carbone
Antonello De Risi
Anna Forciniti
Giuseppe M. Gaspari

xv
Giuseppe Lunardi
Daniele Martinelli
Giuseppe Molisso
Daniele Peila
Enrico Maria Pizzarotti
Marco Ranieri

ORGANIZING COMMITTEE

Enrico Luigi Arini

Joseph Attias
Margherita Bellone
Claude Berenguier
Filippo Bonasso
Massimo Concilia
Matteo d’Aloja
Enrico Dal Negro
Gianluca Dati
Giovanni Giacomin
Aniello A. Giamundo
Mario Giovanni Lampiano
Pompeo Levanto
Mario Lodigiani
Maurizio Marchionni
Davide Mardegan
Paolo Mazzalai
Gian Luca Menchini
Alessandro Micheli
Cesare Salvadori
Stelvio Santarelli
Andrea Sciotti
Alberto Selleri
Patrizio Torta
Daniele Vanni

SCIENTIFIC COMMITTEE

Daniele Peila, Italy (Chair) Georgios Anagnostou, Switzerland

Giulia Viggiani, Italy (Chair) Andre Assis, Brazil
Tarcisio Celestino, Brazil (Chair) Stefano Aversa, Italy
Lorenzo Brino, Italy Jonathan Baber, USA
Giovanna Cassani, Italy Monica Barbero, Italy
Alessandra De Cesaris, Italy Carlo Bardani, Italy
Pietro Jarre, Italy Mikhail Belenkiy, Russia
Donato Ludovici, Italy Paolo Berry, Italy
Vittorio Manassero, Italy Adam Bezuijen, Belgium
Matthias Neuenschwander, Switzerland Nhu Bilgin, Turkey
Moreno Pescara, Italy Emilio Bilotta, Italy
Enrico Maria Pizzarotti, Italy Nikolai Bobylev, United Kingdom
Tatiana Rotonda, Italy Romano Borchiellini, Italy
Alessandra Sciotti, Italy Martin Bosshard, Switzerland
Han Admiraal, The Netherlands Francesca Bozzano, Italy
Luisa Alfieri, Italy Wout Broere, The Netherlands

xvi
Domenico Calcaterra, Italy Alberto Meda, Italy
Carlo Callari, Italy Davide Merlini, Switzerland
Luigi Callisto, Italy Alessandro Micheli, Italy
Elena Chiriotti, France Salvatore Miliziano, Italy
Massimo Coli, Italy Mike Mooney, USA
Franco Cucchi, Italy Alberto Morino, Italy
Paolo Cucino, Italy Martin Muncke, Austria
Stefano De Caro, Italy Nasri Munfah, USA
Bart De Pauw, Belgium Bjørn Nilsen, Norway
Michel Deffayet, France Fabio Oliva, Italy
Nicola Della Valle, Spain Anna Osello, Italy
Riccardo Dell’Osso, Italy Alessandro Pagliaroli, Italy
Claudio Di Prisco, Italy Mario Patrucco, Italy
Arnold Dix, Australia Francesco Peduto, Italy
Amanda Elioff, USA Giorgio Piaggio, Chile
Carolina Ercolani, Italy Giovanni Plizzari, Italy
Adriano Fava, Italy Sebastiano Rampello, Italy
Sebastiano Foti, Italy Jan Rohed, Norway
Piergiuseppe Froldi, Italy Jamal Rostami, USA
Brian Fulcher, USA Henry Russell, USA
Stefano Fuoco, Italy Giampiero Russo, Italy
Robert Galler, Austria Gabriele Scarascia Mugnozza, Italy
Piergiorgio Grasso, Italy Claudio Scavia, Italy
Alessandro Graziani, Italy Ken Schotte, Belgium
Lamberto Griffini, Italy Gerard Seingre, Switzerland
Eivind Grov, Norway Alberto Selleri, Italy
Zhu Hehua, China Anna Siemińska Lewandowska, Poland
Georgios Kalamaras, Italy Achille Sorlini, Italy
Jurij Karlovsek, Australia Ray Sterling, USA
Donald Lamont, United Kingdom Markus Thewes, Germany
Albino Lembo Fazio, Italy Jean-François Thimus, Belgium
Roland Leucker, Germany Paolo Tommasi, Italy
Stefano Lo Russo, Italy Daniele Vanni, Italy
Sindre Log, USA Francesco Venza, Italy
Robert Mair, United Kingdom Luca Verrucci, Italy
Alessandro Mandolini, Italy Mario Virano, Italy
Francesco Marchese, Italy Harald Wagner, Thailand
Paul Marinos, Greece Bai Yun, China
Daniele Martinelli, Italy Jian Zhao, Australia
Antonello Martino, Italy Raffaele Zurlo, Italy

xvii
 Long and deep tunnels

C. Acquista & G. Giacomin

Ghella S.p.A., Rome, Italy

F. Martin & G. Comin

Spie Batignolles Genie Civil, Saint Martin La Port, France
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

Long and deep single shield TBM in very complicated geology

under the Alps. Saint Martin La Porte

C. Acquista & G. Giacomin

Ghella S.p.A., Rome, Italy

F. Martin & G. Comin

Spie Batignolles Genie Civil, Saint Martin La Port, France

ABSTRACT: The paper describes the extreme geological conditions encountered during the
excavation of the first stretch of the Mont-Cenis base tunnel in the new high speed railway Lyon-
Turin, SMP4 site, between Saint Martin La Porte and La Praz in France. The total length of the
tunnel is approximately 8700 m realized using the mechanized boring through a single shield hard
rock TBM. The geology is characterized by the presence of various geological formations, fault
zones with high coal content, high fractured rock mass and constant lack of a vertical and stable
face. This heterogeneity has affected in a severe way the TBM advance. The aim of the paper is
to highlight all TBM requirements foreseen by the contract and all needed improvements applied
to overpassed or cope with different geological accidents. All installed equipment on the TBM
was designed with special devices assuring to manage particular exceptional conditions. Eventu-
ally, the TBM performance is detailed referring to a series of milestones where a series of technical
improvements were introduced to regain a normal excavation trend.

1 FRAMEWORK OF THE PROJECT

Saint-Martin-La-Porte 4 is the first lot of the new high speed railway link Lyon-Turin that
includes the excavation of the first 1,5+9 km of the future double 57 km long tubes Mont
Cenis base tunnel between France and Italy. The project is a key-element of the new European
transport network RTE-E.
The Client is the public promoter TELT sas, the construction supervision is entrusted to
Egis/Alpina and the tender was awarded by the temporary grouping of companies formed
by 6 contractors: Spie batignolles Genie Civil, Eiffage Génie Civil, Ghella SpA, CMC di
Ravenna, Cogeis SpA and Sotrabas.
The contract is divided in four parts: three starting from the 2,33 km long Saint-Martin-
La-Porte access tunnel (1 and 3A-3B in conventional method and 2 in mechanized method) and
one developing from the 2,47 km long La Praz access tunnel in conventional method (4).

2 GEOLOGY

The overall geology of the project is enclosed in two principal formations: “Zone Subbrian-
connaise” [ZSB] and “Zone Houillere Brianconnaise” [ZHB]. From east to west, so from
Saint-Martin-La-Porte to La Praz, ZSB gradually changes in the ZHB passing from an over-
lapping tectonic contact commonly named “Front Houiller” [FH]. A short geological descrip-
tion of these encountered formations is detailed below:
– ZSB => marly limestone, massive limestone and calcschists
– FH => anhydrites, gypsums and cornelians (even dolomites as symptom of the geological
contact)

3505
Figure 1. Saint-Martin-La-Porte site.

Figure 2. Geological profiles of the SMP4 part 2 site (left: till to PK14+000, right: from PK14+000 to
the La Praz access tunnel).

– ZHB => divided from east to west in 4 geological unit: “Encombres”, “Brequin-Orelle”,
“La Praz” et “Forneaux”, substantially from arenaceous and black shales with coal levels
to sandstones and conglomerates. Singular geological events as coal faults and shales with
a coal fraction up to 15 % were been foreseen in the preliminary project.
Part 2 is bored inside the ZHB passing from “Brequin-Orelle” unit to “La Praz” one. Part 2
starts from the TBM assembly cavern at PK11+793 to its disassembly cavern at PK20+588,
so a 9 km long tunnel with a nominal excavation diameter of 11,25 m. Part 2 is an exploratory
tunnel necessarily obliged causing the lack of vertical surface drills (approximately one probe/
km) justified by the high overburden, between 800 and 1100, and the complicated environmen-
tal conditions for its access.
From the geological synthesis remains several uncertainties along the tunnel alignment
concerning:
– Discretization of faults and coal levels
– Presence of mixed faced characterized by intensive fracturing and heterogeneity
– Variability of the inclination of the schistosity compared with the theoretical face plan
– Presence of high hydraulic loads and localization of water in-lets.
The geognostic aim of the SMP4 site is correlated to the mandatory improvements that will
extended to all future TBMs designs in order to more easily face any geological singularity
already encountered.
All synthesis reports, geological and hydrogeological, including the geotechnical model are
not considered as contractual scope. This paper is focalized on the SMP4 part 2.

3 CONTRACTUAL PRESCRIPTIONS AND TBM SPECIFICATIONS

As contractual basic prescription, the TBM design had to be suitable to all potential geo-
logical risks such as convergences, mixed faces, presence of instable blocs and faults in poor
materials (coal levels). In the following the most significant contractual prescriptions [CP] are
coupled with the contractor’s technical choices [TC] for the chosen TBM.

3.1 Exploratory probing

[CP] => Shield designed with top, bottom and face openings (as shown in figure 3 left) to
permit the realization of exploratory destructive surveys or face treatments (such as injections
of bentonite, resin. . .) in fractured rock masses. TBM mandatorily equipped with two twin
drilling machines mounted on a structure separate from the erector ring, capable to perform

3506
Figure 3. Positionning of drill openings from the contract (left) to the TBM design (right).

100 m up to 200 m long boreholes with a minimum diameter of 76 mm. Admissible overlap of
probing is from 20 to 25 m.
[TC] => top, bottom and face openings as shown in figure 3 right. 2 drilling machines
installed on a drilling ring placed on the erector bridge in front of the erector. The main drill-
ing tools normally used are: 108x82 mm casing pipes 2 or 1 or 0,8 m long, T38 M/F bars 3 m
or 2,4 long and bottom drill bit of 76 mm. Depending on registered TBM parameters the
probing is scheduled in advance to respect a minimum overlapping of 10 m. The overlapping
of probes was reduced from 25 to 10 m in order to increase their frequency and perform
shorter ones (between 30 to 50 m) to prevent the over-deviation problems. A normal average
net advance speed at the drill rig is 30 to 50 m/h but the normal drilling time to 10-15 m/h
even considering all preparatory works.
[CP] => TBM back-up equipped with: a coring machine and a bolting machine. The first
one to investigate radially and vertically, while the second one to reinforce radially the seg-
mental lining. Radial cores for annular void checking 1,5 m long at a distance of 30 m behind
the shield and vertical ones 15 m deep carried out each approx.. 200m for rock sampling and
gas measurements.
[TC] => a special mobile ring is designed on the TBM gantry 1 for either the coring machine
or the bolting one. During normal conditions, except for particular events, the coring machine
is exploitable on the TBM and the bolting one is stored on the site yard. Both machines are
able to drill following a rigorous pattern. The coring is carried out through the wire-line
method with NQ pipes and a 1,5 m long double corer and with a net cores diameter of 47,6
mm. A down vertical coring of min. 4,5 m long is performed each filling of the belt storage
(approx. each 250 m). The bolting machine was employed one time during the overpassing of
the first fault at ring 200 to drill a series of self-drilling bolts of 32 mm exploited to inject a
filling cement mixture.

3.2 Overcutting, thrust force, unlocking torque and annular gap detector
[CP] => a minimum radial overcutting equal to 150 mm using interchangeable disc cutters,
cutting wheel radial displacement and reamer cutting tools. The suggested nominal diameter
of the cutting wheel was 11140 mm with worn discs. So with the overcutting: 11290 mm.

Figure 4. 3D view of the probing phase and in-situ performing of the survey with the left drilling
machine.

3507
Table 1. TBM parameters: thrust force and cutter head torque.
[CP] – minimum
accepted value [TC]

Maximum thrust force 180000 kN 123800 kN at 350 bar => mode normal
185700 kN at 525 bar => exceptional mode
Cutting wheel contact 25000 kN 32000 to 10000 kN depending on the resulting center of
force gravity of 3 cutting wheel displacement sectors
Torque during excava- 9000 kNm 9000 kNm at 5 rpm
tion (normal mode) 25228 kNm at 1,76 rpm
Unlocking torque 35000 kNm 29050 kNm at 1,76 rpm => exceptional mode
35000 kNm at 0,5 rpm => unlocking mode

[TC] => the available overcutting is a combination of cutting wheel eccentricity, shim-
ming of cutting tools, introduction of new cutting tools on free tracks passing from 19” to
20”. Considering worn disc cutters, the nominal diameter of 11210 could be enlarged to
11410 mm passing from an intermediate stage at 11310 mm. The vertical displacement of the
cutting wheel is adjustable from 20 mm to 130 mm respect to shield horizontal axis.
Concerning the thrust force and the cutter head torque see table 1.
[CP] => TBM equipped with devices to comprehend over-excavation profiles
[TC] => 7 hydraulic cylinders with a stroke from 0 to 400 mm on the front and intermediate
shield. These one are activated during each ring building or standstill to measure the annular
gap between shield and real excavation section.

3.3 Backfilling: from mortar to bi-component injection

[CP] => the mortar mix design for filling the annular void has to avoid any movements of
the segmental ring and to assure an homogeneous contact between the excavation profile
and the extrados of the lining. In presence of soils involving high thrust forces, a complete
injection of the annular void, normally demanded at a distance of 15 m from the tail skin,
have to be continuously finalized at 360° during the TBM advance assuring the complete
filling of ring N-1.
[TC] => in the technical tender offer the first proposition was the using of two mortar mix
designs respectively applied on the basis of the TBM thrust parameters. These mix designs
obliged the use of lime filler, sand and gravels resulting in the constraint to transport the
mortar in tanks to the TBM. Operationally, the first mix design assured the filling of the annu-
lar void at 120° at N-1 and at 360° at N-10, while the second one obliged the installation at N-
1 of a special ring equipped with a bullflex. This last one is a segmental expansive injectable
joint installed directly at the segments extrados on the site yard.
This offer approved solution was on-going replaced by the bi-component system. The bi-
component mix design permits to transport their components (A and B) inside pipelines
assuring a long-term stability. Operationally, the filling of the annular void is assured at
180° at N-1 and at 360° at N-5. Actually on site, the A-component is made by a particular
weight controlled mixture of water, bentonite, cement 52,5 R, slag and retarding agent. The
B-component, a liquid activator alkaline base solution, added with A-component with a
variable B/A ratio between 5,5 and 7% permits the creation of a gel in the correspondingly
range between 8 and 11 seconds. The viscosity at the Marsh cone is variable between 37 and
40 seconds. The bleeding at 3h and 24h are respectively 1 and 5 %. The short-term compres-
sive strengths at 1h and 3h are respectively 0,3 and 0,8 MPa. The long-term compressive
strengths at 24h and 28d are respectively at 1,6 and 2,5 MPa.
In figure 5 the transition from the mortar solution to the bi-component one is presented.

3508
Figure 5. Mortar solution at technical offer stage and bi-component one applied during TBM advance.

3.4 Back-up design (G1 as multi-purpose gantry)

[CP] => a minimum free radial gap from 0,5 to 0,8 m between the segmental lining and the
back-up geometry in order to install a steel rib support such as HEB type to limit the risk of
convergence. For the same reasons, belt conveyor structures and pipes brackets had to be
manufactured to permit a minimum free radial gap of 30 cm from the intrados of the con-
crete rings.
[TC] => intrados diameter of the segmental lining equal to 9900 mm, back-up size inside
the equivalent diameter of 9000 mm (radially resulting in 30 cm for reinforcements and 15 cm
for back-up steering) and free available gap, behind TBM passage, equal to 30 cm where steel
ribs HEB 240 or 180 were foreseen to be installed by special devices on the TBM:
– Inside the shield: segment erector with an hydraulic clamping tool (bolted on the vacuum
plate)
– At 35÷40 m behind the shield: exploiting the gantry 1 with a dedicated steel rib erector.
In figure 6 is shown the TBM equipment to put in place the proposed steel reinforcement.

3.5 Exceptional mode

[CP] => the transfer from normal conditions to exceptional ones, causing a locking of the
TBM, could occur if a massive coal layer is visible at the face, if a squeezing and high frac-
tured soil is encountered or if a collapse of huge blocs at the face or directly on the shield
happened. In this case the Client have to formalize it permitting to the site to start all prepara-
tory works to pass in the TBM exceptional mode.
[TC] => besides all devices previously detailed (steel rib erector, coring/bolting ring. . .) the
TBM was even conceived to void the building of the invert element directly above the second-
last ring (note: normally the invert element is placed above the ring N-13, so approx. 21 m
behind the tail skin). In order to perform this operation a series of preliminary steps have to be
completed:
– Dismantling of the segment feeder
– Installing of the extension beam for the invert hoist
– Installing of the twin crutching cylinders (necessary to permit the cantilevered translation
of the invert hoist)
– Installing, alternatively with the ring building, of the clamping tool for steel rib elements

Figure 6. Clamping tool on the vacuum plate on the erector and steel rib erector on the TBM gantry 1.

3509
Figure 7. Invert element moved forward by the special hoist and segments storage above the inverts.

The invert hoist normally is in parking position and exclusively in the exceptional mode is used.
During exceptional conditions, from 8 to 12 working hours are averagely needed for a com-
plete cycle of 1,5 m long. In figure 7 the TBM is configured in exceptional mode.

4 CHRONOLOGY OF ENCOUTERED EVENTS DURING EXCAVATION

Four main relevant standstills, from the beginning of the TBM excavation at the end of
August 2016, have influenced the theoretical scheduled planning of the TBM advance in
SMP4 part 2:
1. Crossing of the 1st geological accident: from December 6, 2016 to May 12, 2017 between
rings 197 and 279, so approx. 5 months of standstill where 123 m were been bored
2. 1st cutting wheel exceptional refurbishment: from November 4, 2017 to January 26, 2018 at
the ring 1390
3. Crossing of the 2nd geological accident: from April 17, 2018 to May 23, 2018 between rings
2207 and 2266, so approx. 1 month of standstill where 88 m were been bored
4. 2nd cutting wheel exceptional refurbishment: from October 11, 2018 to November 12, 2018
1 and 3 were been defined as singular events [S] => S1 et S3, while 2 and 4 were evaluated
as a consequence of the continuous events [C] => C2 and C4.

4.1 Singular events [S]: faults with high coal fraction content and water in-let
S1 was a 10-15 m long fault marking a sudden transition to a crushed zone of unstructured
rock. It consists of black and charcoal shales with an abrupt orientation changing of the strati-
fication from sub-horizontal to vertical. Immediately, S1 involved significant water in-lets,
huge over-excavations and the cutting wheel locking. Between rings 202 and 212, totally
27800 ton were mucked out in 11 rings. So, a quantity/ring equal to 2530 ton was averagely
extracted with a pick of 5150 ton on ring 207. Several tests of cutting wheel rotation reached
the max. values of the unlocking torque 35 MNm.
S3 was a 3 m long fault characterized by schistose micaceous sandstone with an estimated coal
content of 15 %. This zone was high fractured without cohesion and with a large presence of fine
material. No changing in the orientation of the stratification was observed: sub-vertical. Between
rings 2207 and 2211, totally 16100 ton were extracted in 5 rings. So, a quantity/ring equal to 3220
ton was averagely extracted with a pick of 5155 ton on ring 2210.

4.2 Continuous events [C]: mixed face and fractured rock mass
Up to today, all along the TBM bored alignment, a series of constant phenomena were
observed:
– Systematic void between the face and the cutter head variable between 50 cm and 3,5 m
– Heterogeneity of the face characterized by an high fractured level causing the excavation of
precut rock blocks and the over-stressed of cutter head buckets
Two unscheduled and extraordinary interventions were performed to face the abnormal
damages highlighted: C2 and C4.

3510
Figure 8. 3D view of precut blocks at the face bottom and typical conditions of a mixed/blocky rock
mass.

5 IMPROVING INTERVENTIONS APPLIED FOR S OR C EVENTS

5.1 Injections: grout mixture and bi-component resin

Different types of face treatments were performed during the standstills for S1 and S3:
– Partial filling of the cutter head arms with bi-component expansive resin to limit the material
in-flow (coupled with polystyrene blocks to reduce the resin consumption). Approx. 8600 kg
in 2 phases.
– 1 m long fiber glass bars drilled through disc cages and injected with bi-component expan-
sive resin. Approx. 5200 kg.
– 8 m long casing pipes, both inclined and horizontal, equipped with an injection packer to
create a proof resin barrier at the face of the cutting wheel. Approx. 10800 kg.
– 12 m long casing pipes, both inclined and horizontal, equipped with an injection packer to
improve the cohesion of fault ground by the injection of a bi-component chemical liquid
binder – injections performed with a high pressure pump. Approx. 39500 kg.
– 3-stages areolar grouting through probing pipes through casing pipes and using an high
pressure pump. Approx. 24 m3 in 3 days of a mix formed by water, bentonite, cement, plas-
ticizer and retarding agent.

5.2 Complementary activities in S1 and S3

During the standstills S1 and S3 a series of particular operations were accomplished:
– Installation of 22 steel rib HEB 180 and 30 steel rib HEB 240 between rings 187 and 212
– Installation of 15 steel rib HEB 180 and 11 steel rib HEB 240 between rings 2207 and 2220
– Realization of self-drilling bolts by pneumatic hammers to reinforce the segmental lining
close to the shield
– Realization of multiple expansive resin barriers surrounding the last rings to completely
blocked the segmental lining through bi-component injection
– Production/injection at the face and at the extrados of the shield of water based lubrication
liquids with a natural polymer, with bentonite and with a foaming agent

5.3 Continuous updating of the cutting wheel opening ratio

In general, to definitely cope with fine material in-flows and face instability, the opening ratio
of the cutting wheel was reduced from 8 to 4,5 %. As first action during S1, original scrapers
rows plugged with thin plates were replaced with additional plates on each arm, new grill bars
and peripheral buckets were backward reinforced by steel triangular gussets.

5.4 Cutting wheel closing with special bolted steel plates

Up to S3, the cutting wheel normal closing procedure was made with the filling of the arms
with expansive resin and polystyrene blocks. From this moment, together with the new design
of buckets (see chapter 5.5) the supplier Palmieri proposed the possibility to replace the

3511
 Table 2. History of the cutting wheel open ration changing.
Event Opening ratio Comment

Start 8,0 %
S1 5,0 % 4 new plates/arm on 16 arms
4,5 % 5th plate/arm added on 8 arms
C2 5,0 % 4 plates/arm on 16 arms (gouging of the 5th plate/arm)
S3 ≪ 5,0 % As in C2 but 8 on 16 arms completely closed with see figure 10
C4 6,5 % As in S3 but with gouging of the 3rd and 4th plate on 4 arms

Figure 9. From left to right cutting wheel closing with opening ration from 8 to 5 %.

Figure 10. Closing of the arms with special bolted steel plates.

bucket tools with a bolted closing plates. This permitted to avoid the uncertainty of the com-
pressive trend of the injected expansive resin that probably it was the main cause of the pro-
gressive augmentation of the torque (due to the muck collection at the end of the arms).

5.5 Cutting wheel adjustment and refurbishment

The standstill for C2 resulted to prevent additional damages to the cutting wheel structure. Fol-
lowing approx. 2 km of boring, at ring 1390, the original peripheral hardox resistant plates were
highly worn out up to 15 mm. Besides, the bucket original design was seriously compromised.
In order to find a definitive solution for C2, supplementary more resistant wear plates were
welding outside the diameter 7,2 m and buckets (supports and tools) on all arms were replaced
with tough ones changing the tool-support bolting way and fixation system.
These exceptional maintenance operations were preceeded by the construction of a welding
chamber sizing 6 x 5 x 1,35 m (see figure 12). This one was excavated by the use of emulsion
explosive cartridges drilling the face with pneumatic hammers and coring portable machines.
The access to the face was possible through progressive rotations of the cutting wheel.
After C2, from the TBM restart, regular ultrasonic measurements of the on-going wear of the
new installed plates were carried out. In September 2018, the TBM overcame 5000 m of advance
and a sudden worsening of the wear status of the backward circumference of the cutting wheel,
as even the wear of the facing plates at the openings of each arm were noticed (see figure 13).
These facts involved another standstill (C4) to create again a smaller welding chamber
sizing 4 x 2,5 x 1 m. In the following main technical actions performed are briefly listed:
– Opening of the cutting wheel from 5 to 6,5 % as briefly described in table 2
– Reinforcement of the peripheral face of the cutting wheel with heavy welded studs
– Reinforcement of the peripheral edge of the cutting wheel with supplementary wear resistant
plates.

3512
Figure 11a. 3D face plan of the cutting wheel refurbished, wear status of the cutting wheel external part
and new bucket design applied.

Figure 11b. 3D face plan of the cutting wheel refurbished, wear status of the cutting wheel external part
and new bucket design applied.

Figure 12. As-built of the welding chamber and personnel at work for cutter head wear reconstruction.

Figure 13. Observed wear at ring 3350 and application example of heavy studs on a cutter head.

6 TBM PERFORMANCE

In figure 14 the TBM daily performance is shown. T1, T2, T3 and T4 are four time period
where TBM continuously excavates, while S1, C2, S3 and C4 are the standstills previously
described where the TBM was stopped either for a singular geological event or for extraordin-
ary maintenance operations. In table 3 is pointed out the real TBM productions and, as even
in figure 14, is evident the benefits of all technical improvements applied resulting in a regular
raising up of productions.
After S3, following the TBM restart, a series of innovative technologies were tested to verify
the authenticity of employed methods such as optical televiewing of exploratory probes in
advance. Besides, a precautionary stop decisional process was implemented and shared
between shift bosses, pilots, TBM engineer and Client. This last one permitted to define a

3513
Figure 14. TBM daily performance from September 2016 to September 2018 with detail of standstills.

Table 3. Detail of the daily and monthly TBM production in correspondent advance time periods.
Daily production [m] Monthly production [m]

Advance time period Average Maximum Average Maximum

T1 5,4 12,8 100,0 190,6

T2 10,0 22,5 238,7 406,6
T3 15,5 29,4 307,4 540,0
T4 14,5 27,7 391,6 477,1

series of thresholds in TBM parameters in order to stop the excavation and start a new
exploratory phase at the face.

7 CONCLUSIONS

Meanwhile this writing, the TBM is boring at its full capacity with an average advance speed
of 16 m/day. Moreover, the daily advance record was performed on January 12, 2019 with
24 m bored. The breakthrough in the disassembly cavern of La Praz is scheduled at the end of
July 2019 in order to start all dismantling operation of TBM and its equipment. At the end of
January 2019, 2600 m have to be still bored with the high risk to encounter high water in-
flows, hard and abrasive rock mass (as even local faults and at least one geological contact).
This project is a clear example of how much relevantly the uncertainty of geological know-
ledge could influence the advance of a tunnel boring machine.

ACKNOWLEDGEMENTS

Authors would like to acknowledge the Client TELT, the site supervision Egis/Alpina and all
construction companies involved in this project: Spie Batignolles Genie Civil, Eiffage Genie
Civil, Ghella, CMC and Cogeis. Special thanks have to be addressed to the TBM’s manufac-
turer NFM Technologies and to the belt conveyor’s constructor ROWA for their technical
support on-field. Furthermore, other suppliers have even played a key role on site assuring a
constant on-field technical assistance: Techni-Metal-System, VMT, Mapei, Palmieri Group,
Clariant, Ecocem, Sema Ventube, Lorenzetto Loris Srl, Siap, Buzzi Unicem and Condat.

3514
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

Experimental setup for studying tunnels in squeezing

ground conditions

K. Arora, M. Gutierrez & A. Hedayat

Civil and Environmental Engineering, Colorado School of Mines, Golden, USA

ABSTRACT: Squeezing ground conditions in tunnels are often associated with rock min-
eralogy, strength, ductility/brittleness, excavation sequence, and magnitude of in situ stres-
ses. Numerous methodologies and empirical correlations have been proposed in the past to
determine the level of ground squeezing conditions in tunnels. Most of the correlations are
problem-specific and limited in scope. In this work, a fundamental study of tunnel squeezing
is carried out using an experimental approach to simulate tunnel boring machine (TBM)
excavation in squeezing ground conditions. The experimental setup employs a cubical speci-
men of a soft rock/soil/synthetic material with each dimensions of 30 cm long. The specimen
is subjected to a true triaxial state of stress with different magnitudes of principal stresses
and stress levels corresponding to realistic in situ conditions. A miniature TBM is used to
excavate a tunnel into the host rock (specimen) while the rock is subjected to true-triaxial state
of stress. Embedded extensometers and strain gages glued on the surface of the tunnel liner are
used to monitor tunnel response during construction. This paper presents the details of the
experimental setup.

1 INTRODUCTION

The problem of squeezing involves time-dependent large deformation in tunnels, which may
or may not terminate during construction and arises due to high in-situ stress around the
tunnel and problematic geological and geotechnical properties (Barla 1995). Squeezing
ground has intrigued engineers over the years in completing underground construction and
resulted in the loss of time and money. Wiesmann (1912) first studied the effect of squeezing
and since then it has been studied by many researchers including Singh et al. (1992), Aydan et al.
(1996), Dube (1993), Barton & Grimnstad (1994), Goel et al. (1995), Yassaghi & Salari-Rad
(2005), Gutierrez & Xia (2009), Khanlari et al. (2012), Dwivedi et al. (2012), and Shamsoddin
Saeed & Maarefvand (2014).
The squeezing phenomenon in tunnels is still poorly understood (Kovari (1998), and
Barla (2000)). Squeezing is associated with high overburden (H), rock mass quality (Q),
the uniaxial compressive strength of intact rock (σc) and rock mass (σcm), competency
ratio (σc/γH) and tangential stress-strain response around the tunnel. Several definitions
have been proposed based on combination of the abovementioned properties (Terzaghi
(1946), O’Rourke (1984), Singh (1988), Aydan et al. (1996), Gioda & Cividini (1996),
Kovari (1998), Barla (2000)).
To study squeezing problem, it is very important to get a better understanding of the trend
in stresses and deformation that will develop around tunnels prior to the excavation. Over the
decade’s tunnel engineers have been dependent on the empirical methods with limited field
data (Schmidt, 1974; Attwell, 1978; O’ReiIIy & New, 1982; Mair et al. 1993). Various empir-
ical and semi-empirical developed over the years are problem-specific and often contradicts
with one other.

3515
 Physical models to study two and three-dimensional behavior of the ground in response to
tunneling were proposed by many researchers. A comprehensive review of such techniques for
tunneling in the soft ground is provided by Meguid et al. (2008).
Figure 1 illustrates different physical models developed all over the world to study tunneling
on soft ground and Table 1 lists all the advantages and disadvantages of the developed
models.
In this study, a scaled physical model is proposed to study the problem of squeezing in tun-
nels. Experiments are carried out on a cubical specimen of a soft rock/soil/synthetic material
with dimension of 30 cm on each side. The specimen is subjected to compressive true-triaxial

Figure 1. The various physical model developed to study tunneling in soft ground.

Table1. Advantage and disadvantages of various physical models developed to study tunneling in soft
ground conditions (modified from Meguid et al. 2008).
Method Applications/advantages Limitations

Trapdoor • 2D and 3D tests can be performed • An approximate estimate of stresses

under 1g and centrifuge. and deformation.
• Not the actual tunnel process.
Rigid tube with the • Can study failure mechanism and • No estimate of surface settlement
flexible face face stability under 1g and centri- behind the face.
fuge condition.
Pressurized airbags • Studies tunnel stability and induced • Can be used for unlined tunnels and
ground motion around tunnels in does not simulate tunnel advance.
2D as well in 3D under 1g and cen-
trifuge conditions.
Polystyrene foam and • Simulated tunnel advance process • Doesn’t give accurate results in
organic solvent under centrifuge conditions. underwater conditions.
Soil augering • Tunnels advance process in 1g • Can be used for cohesive soils only.
condition. • Not mechanized for a centrifuge.
Miniature TBM • Simulated tunneling under centri- • Only up to 25g gravitational acceler-
fuge condition. ation can be applied.
Mechanically adjust- • Easy to operate 2D tunnel excava- • It is manually controlled under the
able tunnel diameter tion process. 1g condition for 2D models only.

3516
stress state with σ1>σ2>σ3. A miniature tunnel boring machine (TBM) is designed to simulate
excavation similar to real in-situ tunneling. Monitoring is done using acoustic emission (AE),
and strain gages that are installed in the TBM and embedded in the cubical specimen. The
correlation developed from the experimental results may contribute significantly to better
understanding of the tunnel squeezing in rocks. The uniqueness of this experiment is high-
lighted by the fact that tunnel is excavated in a specimen loaded in all three directions and
with magnitudes load in the order of real field stresses.

2 EXPERIMENTAL SETUP

The objective of this experiments is to study the squeezing behavior of soft rock in
response to tunnel excavation under true-triaxial stress state. Figure 2 shows schematic-
ally the experimental setup, which includes the true-triaxial cell, miniature tunnel boring
machine (TBM), servo-controlled pumps and 115V constant speed AC motor for driving
TBM, synthetic soft rock specimen (mudstone) and data acquisition system for monitor-
ing deformations around the tunnel.
The mudstone specimen is loaded in true-triaxial stress state. The miniature tunnel boring
machine is mounted on the top lid of the true-triaxial cell. The top lid of the cell has a 76 mm
diameter circular opening which provides access to the rock surface for tunnel excavation.
Required TBM thrust and torque are provided by servo-controlled pumps and electric motor,
respectively. For continuous data acquisition, the cell is equipped with acoustic emission (AE)
sensors and strain gages embedded in the specimen. All the important aspects and working of
this experimental setup are discussed in the following sections.

2.1 True-triaxial cell

This experimental incorporates the use of true-triaxial cell developed by Frash et al. (2015) to
study the enhanced geothermal system (EGS) at laboratory scale. The apparatus is capable of
applying three independently controlled principal stresses up to 13 MPa to a 30x30x30 cm3
rock specimen.
The apparatus was designed with a mixed flexible bladder (flat jack) and passive confine-
ment system, shown in Figure 3. Each principal stress is applied via one active flat jack per

Figure 2. Schematic diagram of the proposed experimental setup to study TBM excavation in squeezing
ground conditions.

3517
Figure 3. True-triaxial cell at Colorado School of Mines (modified from Frash et al. (2015)).

principal axis when using the typical configuration for the apparatus. Specimen faces directly
loaded by the flat jacks and the opposing reaction faces supported by the frame are hereby
referred to as active and passive faces, respectively. The steel top lid was furnished with
a 63 mm diameter port to pass electrical sensor wires and hydraulic tubing for internal
sensors and jacks.
The reaction ring of the central body was constructed from A36 structural steel with the
yield strength of 250 MPa. The lids were constructed from A514 steel with the yield strength
of 700 MPa to reduce thickness requirements following stress design criteria. Sufficient lid
thickness was provided to permit drilling multiple non-intersecting 10 mm holes through the
lid while maintaining a safety factor greater than 2.0. The lid of the cell was effectively con-
sidered to be a sacrificial component. This lid is modified to conduct the TBM excavation in
squeezing ground conditions.
Active face stresses are provided by flat jacks (350 mm diameter circular Freyssinet®) and
an assembly of two 300 mm diameter round steel platens and one 300 mm square steel
platen. Each flat jack is pressurized via an independent hand pump with active digital pres-
sure monitoring. Using separate pumps bypasses pressure control issues which occur in
single pump systems with manifold valves. The square platen’s inward edges were beveled to
mitigate binding with adjacent platens. A 25 mm thickness was specified for the square
platens referencing elastic stress-deflection analysis and limit yield criterion for stress trans-
mission to the specimen corners. The square platen also provided a protective housing for
AE sensors. This design decision improved AE data quality by ensuring good face-to-face
sensor contact with the specimen, reducing noise transmission through the sensor housing
and reducing assembly time. A typical alternative AE sensor installation method in similar
true-triaxial devices involves cutting shallow holes into the specimen with consequentially
increased sensor alignment difficulty, increased assembly time, decreased stress uniformity
and likely reduced AE measurement quality (Frash et al., 2015).

2.2 Miniature tunnel boring machine (TBM)

The miniature TBM, shown in Figure 2, is an integral part of the experimental setup. The
miniature TBM is designed keeping in mind different thrust and torque requirement at a dif-
ferent level of field stress. Figure 4 presents the miniature TBM designed and fabricated at the
Colorado School of Mines. One of the important parts of this miniature TBM is the 100 kN
and 200 mm stroke cylindrical hydraulic jack. The jack is controlled by a pair of D-series Tele-
dyne ISCO pumps. The pump maintains the continuous constant pressure (CCP) at the jack
and hence, constant thrust at the face of the TBM.

3518
Figure 4. Miniature tunnel boring machine (TBM) designed and fabricated at Colorado School of
Mines.

The plunger of hydraulic jack is connected to the rotatory shaft of the miniature TBM
through a thrust bearing which allows relative rotation between the rotatory shaft and plunger
of the jack. The rotatory shaft is mainly a 300 mm long spur gear with pitch (number of teeths
per 25mm pitch diameter) is 16, pitch diameter of 38 mm and the pressure angle of 20°. The
shaft (long spur) is connected to a 50 mm button type drill bit which provides the drag action
to the rock surface similar to the soft ground TBM.
The required torque is provided by bevel and planetary gear assembly which is driven by a
115V single phase alternating current (AC) constant speed (rpm) AC motor. The main drive
gear is the pair of bevel gear having pitch 16, pitch diameter of 38 mm and the pressure angle
of 20°. A pair of bevel gear converts the rotation along the horizontal axis into the rotation
along the vertical axis.
The vertical bevel gear is axially connected to a spur gear having pitch 16, pitch diameter of
50 mm and the pressure angle of 20°. This spur gear is coupled with the shaft (long spur) of
the miniature TBM and provides the required torque to the cutter head at a constant rpm.
The longspur is also coupled with two more spur gears (pitch 16, pitch diameter of 50 mm and
the pressure angle 20°) which prevents the lateral deformation of the shaft of the miniature
TBM (See Figure 4). Therefore, the required thrust and torque for driving this miniature
TBM are provided by a pair of the servo-controlled pump at CCP and constant speed AC
motor. The muck produced due to the tunnel excavation will be clear off from the excavation
face using compressed air at regular intervals.
The whole assembly is supported by a reaction frame designed in such a way that the verti-
cal deflection of miniature TBM at the maximum thrust level will be less than 1 mm. This
reaction frame is mounted on the top of the lid of the true-triaxial cell and shown in the sche-
matic diagram of the experimental setup (See Figure 2).
The tunnel advancement rate is measured by a pair of servo-controlled pumps which is continu-
ously monitored. The torque applied by the cutter head is measured by the continuously

3519
monitored power output (as power if torque times rpm). Hence, this design of miniature TBM
provides all the flexibility in terms of the application of thrust and torque, and at the same time
continuously monitors all the essential operating parameters which are monitored in the field as
well.

2.3 Synthetic mudstone specimen

A synthetic mudstone specimen is prepared in the laboratory using the methodology proposed
by Johnston & Choi (1986). They prepared a synthetic soft rock to eliminate experimental
scatter and high variable performance for laboratory model studies. They found the model
material to be homogenous and isotropic with physical and mechanical properties well aligned
to that of natural mudstone.
In this work, artificial mudstone is prepared in the laboratory by mixing type I/II cement
(include cement provider name here), clay which is chemically hydrous aluminum silicate and
water in the correct proportion. After performing various trials with different mix propor-
tions, it was found that upon mixing equal proportions of cement, clay, and water with some
superplasticizer, a highly workable, consistent and homogenous mix is obtained. The correct
proportion of superplasticizer depends on the type and grade of the superplasticizer used. In
this case, superplasticizer used is MasterGlenium 7920 supplied by BASF. Table 2 shows the
quantity of ingredients required for preparing 1 m3 of mudstone.
The mix is poured in cylindrical and cubical molds. A cylindrical specimen of 51 mm diameter
and 102 mm length were prepared to conduct uniaxial compression tests (UCT) and conventional
triaxial tests at a confining pressure ranging from 1 MPa to 10 MPa (which is also the range of
confining stress in the true-triaxial test).
Some preliminary compressive strength tests are performed on the five cylindrical specimens
of mudstone. Average unconfined compressive strength (UCS) is observed as 6 MPa with a
standard deviation of 1 MPa.
A cubical specimen of 30 x 30 x 30 cm3 is prepared to perform laboratory scaled simulation
of TBM excavation in mudstone at stresses equivalent to the stress in the field. Based on the
deformation monitoring around the excavation, squeezing behavior of the mudstone will be
studied.

2.4 Instrumentation and monitoring

The data acquisition system (DAQ) used to monitor and control the laboratory performed
TBM excavation in squeezing ground conditions record all the essential operating parameters
as is standard for the field application while also implementing some additional elements to
take advantage of improving accessibility that an experimental setup allows. The following
section presents the details of instrumentations and monitoring in the proposed experimental
setup.

2.4.1 Instrumentation for monitoring TBM output

At the very basic level, as already discussed, the thrust of the TBM will be controlled by the
two syringe pumps under continuous constant pressure (CCP) mode. This mean TBM will

Table 2. Mix proportions for one cubic meters of synthetic

mudstone.
Ingredient (units) Quantity for 1m3 of mudstone

Cement (kg) 592

Clay (kg) 592
Water (kg) 592
Super plasticizer (ml) 24

3520
apply constant thrust at the face of the excavation. The advance rate of the TBM can be back-
calculated flow rate (Q) and time (t) data of the servo-controlled pumps. The torque provided
by the TBM will be continuously monitored by a separate DAQ which will record the power
output from the AC motor running at constant revolutions per minutes (rpm).

2.4.2 Acoustic emission (AE) to monitor damage

AE events will be monitored using Physical Acoustic Corporation (PAC) AE monitoring
system with six WSα sensor and three PCI-2 cards mounted in the Micro-II chassis. Six AE
sensors are used because in case we do not get high quality data from two or three sensor,
there should be high quality data from the minimum number of sensors required for geophys-
ical characterization.
As shown in Figure 5, six AE sensors are typically placed on test specimen surfaces and
attached directly to the specimen faces using a thin layer of vacuum gel for coupling, to attain
direct AE measurement with minimal reflection, surface interference, sensor orientation error,
or attenuation effects. The use of six sensors also enables the application of moment-tensor
analysis to classify recorded AE events according to location and failure mode, which is either
tensile dominated, shear dominated, or mixed-mode.

2.4.3 Embedded strain gauges for monitoring strain around the excavation
To monitor the strains around the excavation, various strain gauges will be embedded in a
cubical specimen of mudstone. Due to the very high water content of the synthetic mudstone,
there will be improper adhesion between strain gages and mudstone. Hence, strain gauges
can’t be directly embedded in the synthetic mudstone. A multiple point borehole extensometer
(MPBEx) will be prepared using a thin flexible material like Teflon and an array of strain
gauges as shown in Figure 6.
Since the thin Teflon sheet is more flexible than the mudstone, the embedded strain gauges
will show the deformation of the more competent member i.e. mudstone. However, this meth-
odology should be validated by performing a conventional UCT on a cylindrical specimen of

Figure 5. Six AE sensor position for true-triaxial testing on a cubical specimen.

3521
Figure 6. An array of strain gages (MPBEx) embedded in the synthetic mudstone cubical specimen.

synthetic mudstone with embedded axial strain gauges and measuring axial deformation using
a linear voltage differential transducer (LVDT). The axial strain measurements from the
embedded strain gauge and LVDT should be comparable.

2.4.4 Instrumentation in the tunnel liner

After excavating a two inches diameter tunnel, a thin flexible cylindrical liner will be installed
in the excavated tunnel. Elastic properties (Young’s modulus and Poisson’s ratio) of this liner
material will be predetermined by conducting appropriate tests. Four strain gauges, equally
spaced in the longitudinal direction, will be glued on the inner surface of the liner. The thin
annulus gap between the liner and excavated tunnel will be filled with quick set epoxy with
known elastic properties. Using the hoop stress and continuity equation at the tunnel and
liner interface, tunnel convergence will be continuously monitored.

3 ANALYSIS AND SYNTHESIS OF THE EXPERIMENTAL RESULTS

Figure 7 shows the flowchart for the post-processing of the test results. Data from the con-
tinuous monitoring of the miniature TBM, embedded strain gauges (MPBEx), tunnel liner
strain gauges and AE sensors will be closely monitored. Continuous monitoring of the mini-
ature TBM will give the cutter head thrust, advance rate, speed (rpm) and torque with time.
In this setup the cutter head thrust, and rpm will be constant throughout the experiment.

Figure 7. Flowchart for the analysis of the experimental results.

3522
 Monitoring of embedded strain gauges (MPBEx) will provide the strain around the exca-
vated boundary during and after the excavation at multiple points. On other hand, the strain
gauges installed in the liner will provide the tunnel wall convergence at multiple points after
the excavation.
Post-processing the data from the continuous monitoring using six AE sensors, installed at
the surface of the cubical specimen, will allow for the classification of recorded AE events
according to location and failure mode.
The deformations and AE events will be continuously monitored even after the TBM exca-
vation till the time no significant change is observed. Results from each test obtained during
the excavation stage and the characterization of the samples after testing will be carefully ana-
lyzed and synthesized. New methods to predict tunnel squeezing will be formulated or existing
ones will be validated and improved.
After the completion of each test, the loaded samples will be sliced into several sections
along the tunnel longitudinal axis. Each section will then be imaged and analyzed in terms of
deformations, failure and plastic zone formation around the excavation.

4 CONCLUSIONS

A novel experimental setup is proposed to simulate the TBM excavation in squeezing ground
condition. This experimental setup is capable of monitoring tunnel advance and tunnel stabil-
ity at a field stress level. Stress level up to 13 MPa stress can be independently applied in each
of the three direction (which is approximately equal to 600 m of overburden pressure). The
setup can simulate lined as well as unlined tunnels and will provide a fair estimate of stresses
and deformation around the tunnel. This experimental setup also allows to take advantages of
some additional features such as slicing of the specimen into thin sections to study the extent
of the plastic zone around the tunnel.

ACKNOWLEDGEMENTS

The authors wish to gratefully acknowledge the support from the University Transportation
Center for Underground Transportation Infrastructure (UTC-UTI) at the Colorado School
of Mines for funding this research under Grant No. 69A3551747118 from the U.S. Depart-
ment of Transportation (DOT).

REFERENCES

Adachi, T., Tamura, T., Kimura, K. & Nishimura, T., 1995. Axial symmetric trap door tests on sand and
cohesion soil, In: Proceedings of the 30th Japan National Conference on Geotechnical Engineering:
1973–1976.
Atkinson, J.H., Brown, E.T. & Potts, M., 1975. Collapse of shallow unlined tunnels in dense sand. Tun-
nels and Tunnelling 3, 81–87.
Attwell, P.B., 1978. Ground movements caused by tunneling in soil. Proceedings of the Large Ground
Movements and Structures Conference: 812–948. Cardiff, Pentish Press, London.
Aydan O, Akagi T. & Kawamoto T., 1996. The squeezing potential of rock around tunnels: theory and
prediction with examples taken from Japan, Rock Mechanics and Rock Engineering 29:125–143.
Barla G., 1995. Squeezing rocks in tunnels. ISRM News J 2(3/4):44–49
Barla G., 2000. Tunnelling under squeezing rock conditions. Department of structural and geotechnical
engineering, Politecnico.
Barton N. & Grimstad E., 1994. The Q-system following twenty years of application in NMT support
selection 12(6):428–436. Felsbau.
Chambon, P., Corte, J.F. & Garnier, J., 1991. Face stability of shallow tunnels in granular soils. Proceed-
ings of an International Conference on Centrifuge: 99–105. A.A. Balkema, Rotterdam.
Dube A.K., 1993. Squeezing under high stress conditions, assessment and prevention of failure phenom-
ena in rock engineering. In: Pasamehmetoglu AK et al ( eds ) Mine.

3523
Dwivedi R.D., Singh M., Viladkar M.N. & Goel R.K., 2013. Prediction of tunnel deformation in squeez-
ing grounds. Engineering Geology, 161:55–64.
Frash, L. P., Gutierrez, M., Hampton, J., & Hood, J., 2015. Laboratory simulation of binary and triple
well EGS in large granite blocks using AE events for drilling guidance. Geothermics, 55:1–15.
Gioda G. & Cividini A. 1996. Numerical methods for the analysis of tunnel performance in squeezing
rocks. Rock Mechanics and Rock Engineering, 29(4):171–193.
Goel R.K., Jethwa J.L. & Paithakan A.G., 1995. Tunneling through the young Himalayas a case history
of the Maneri–Uttarkashi power tunnel. Engineering Geology: 39(1–2):31–44.
Goodman, R.E., 1972. Geological investigations to evaluate stability, In: Geotechnical Practice for Sta-
bility in Open Pit Mining. Proceedings of the Second International Conference on Stability in Open Pit
Mining: 125–132. Vancouver, Canada.
Gutierrez, M., & Xia, C. C., 2009. Squeezing potential of tunnels in clays and clay shales from normal-
ized undrained shear strength, unconfined compressive strength and seismic velocity.
Hoek, E., 1971. Rock Engineering (Inaugural Lecture). Imperial College, University of London.
Johnston, I.W. and Choi, S.K., 1986. A synthetic soft rock for laboratory model studies. Geotechnique,
36(2): 251–263.
Kamata, H. & Masimo, H., 2003. Centrifuge model test of tunnel face reinforcement by bolting, Tunnel-
ling and Underground Space Technology, 18 (2):205.
Khanlari G., Meybodi R.G. & Mokhtari E., 2012. Engineering geological study of the second part of
water supply Karaj to Tehran tunnel with emphasis on squeezing problems. Engineering Geology,
145:9–17.
Kovári, K., 1998. Tunneling in Squeezing Rock (Tunnelbau in druckhaftem Gebirge). Tunnel, 5:12–31.
Lee, Y.J. and Yoo, C.S., 2006. Behaviour of a bored tunnel adjacent to a line of loaded piles. Tunnelling
and Underground Space Technology, 21:3–4.
Love, J.P., 1984, Model testing of geogrid in unpaved roads. D.Phil. Thesis. Oxford University, UK.
Mair, R.J., Taylor, R.N. & Bracegirdle, A., 1993. Subsurface settlement profiIes above tunnels in clays.
Geotechnique, 43 (2):315–320.
Meguid, M. A., Saada, O., Nunes, M. A., & Mattar, J., 2008. Physical modeling of tunnels in soft
ground: a review. Tunnelling and Underground Space Technology, 23(2):185–198.
Nomoto, T., Imamura, S., Hagiwara, T., Kusakabe, O., & Fujii, N., 1999. Shield tunnel construction in
centrifuge. Journal of Geotechnical and Geoenvironmental Engineering, 125 (4):289–300.
O’ReiIIy, M.P. & New, B.M., 1982. Settlements above tunnels in the UK – their magnitude and predic-
tion. In: Proceedings of Tunnelling’82: 173–181. IMM, London.
Park, S.H., Adachi, T., Kimura, M. & Kishida, K., 1999. Trap door test using aluminum blocks. In: Pro-
ceedings of the 29th Symposium of Rock Mechanics: 106–111. J.S.C.E.
Schmidt, B., 1974. Prediction of settlements due to tunneling in soil: three case histories. In: Proceedings
of the Second Rapid Excavation and Tunneling Conference, 2:1179–1199. San Francisco, USA.
Shamsoddin Saeed M. & Maarefvand P., 2014. Engineering geological study of NWCT Tunnel in Iran
with emphasis on squeezing problems. Indian Geotechnical Journal, 44(3):357–369.
Sharma, J.S., Bolton, M.D. & Boyle, R.E., 2001. A new technique for simulation of tunnel excavation in
a centrifuge. Geotechnical Testing Journal, 24(4): 343–349.
Singh B., Jethwa J.L. & Dube A.R., 1992. Correlation between observed support pressure and rock mass
quality. Tunneling and Underground Space Technology,7(1):59–74.
Terzaghi, K., 1936. Stress distribution in dry and in saturated sand above a yielding trap-door, In: Pro-
ceedings of the International Conference on Soil Mechanics, 1:307–311.Harvard University. Press,
Cambridge.
Terzaghi, K., 1943, Theoretical Soil Mechanics. Wiley, New York.
Tezaghi, K., 1946. Rock Tunneling with Steel Supports (RV Proctor and TL White, eds.). Youngstown,
Ohio: Comercial Shearing Co.
Whyte R.J.A., 1973. Study of progressive hanging wall caving at Chambishi copper mine in Zambia
using the base friction model concept. M.Sc. Thesis. Imperial College, University of London.
Wiesmann E., 1912. Mountain pressure. Switz J Struct, 60: 7(in German).
Yassaghi A., Salari-Rad H., 2005. Squeezing rock conditions at an igneous contact zone in the Taloun
tunnels, Tehran–Shomal freeway, Iran: a case study. International Journal of Rock Mechanics and
Mining Sciences, 42:95–108.

3524
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

Brenner Base Tunnel – challenges of gripper TBM application

for the 15 km long exploratory tunnel Ahrental in challenging
rock mass

K. Bäppler & M. Flora

Herrenknecht AG, Schwanau, Germany

ABSTRACT: The Brenner Base Tunnel will be 64 kilometers long and will be the center-
piece of the Scandinavian-Mediterranean TEN-T Corridor from Helsinki to Valetta (Malta).
It will comprise two single-track rail tunnels with a service and drainage gallery in between.
The primary focus in this paper is on the construction of the 15km long exploratory tunnel
Ahrental using a well-adapted TBM concept. Tunnelling started from Austrian Ahrental
towards the Italian border in challenging terrain using a 7.93m-diameter Gripper TBM. The
geology along the section is characterized as extremely challenging with quartz phyllite and
shale and high overburden of up to 1300m. Moreover, numerous fault zones were predicted
along the drive with expected loose to friable rock mass. The paper addresses in particular the
special TBM design and project experiences with the operation of a Gripper TBM in challen-
ging rock mass.

1 INTRODUCTION

The Gotthard Base Tunnel, which was commissioned in June 2017 as the longest railway tunnel
in the world, will cede this title in the near future to another epoch-making structure in the Alps,
the Brenner Base Tunnel. The Brenner Base Tunnel is an important link between Munich and
Verona and will contribute to sustainable mobility.
Together with the existing Innsbruck bypass in Austria, the Brenner Base Tunnel has a total
length of 64 kilometers. This will make the joint venture between Italy and Austria seven kilo-
meters longer than the Gotthard Base Tunnel.
The Brenner Base Tunnel (BBT) crosses through the Alps below the Brenner Pass. It has
some similarities with the Gotthard, which is partly attributable to the fact that the owner of the
Brenner Base Tunnel (BBT) has been working with experts from Alptransit Gotthard for many
years. Like the Gotthard Base Tunnel, the Brenner also consists of two single-track parallel
tubes and much of the technology from the Gotthard has been adopted. The experience gained
during construction of the Gotthard has been taken into consideration in the safety and rescue
concept for the BBT, for things such as the emergency stops. Connecting galleries between the
tubes are planned every third of a kilometer. These cross passages serve as escape routes in
emergency situations and comply with the highest safety standards in tunnel construction.
A special feature of the Brenner Base Tunnel is the exploratory tunnel that is being bored
over its entire length about twelve meters below and in between the main tunnel tubes, which
will also serve as a service and drainage gallery during operation. The tunnelling work on the
exploratory tunnel will provide information about the nature of the rock mass and thereby min-
imize construction costs and time.
This publication focuses on the drive for the 15 kilometer long Ahrental exploratory tunnel.
Since late September 2015, a Gripper TBM with a diameter of 7.93 meters has been boring the
tunnel from Austrian Ahrental toward the Italian border. The paper discusses the technical

3525
considerations and approaches for tunnelling in the predicted loose to friable rock mass with
squeezing rock zones and summarizes the previous experience of the Gripper TBM deployment.

2 PROJECT OVERVIEW AND REQUIREMENTS FOR THE TBM

Part of the construction lot Tulfes-Pfons envisages the construction of the 15km long explora-
tory tunnel section using TBM tunnelling between the Ahrental hub and the municipality of
Pfons with the aim of developing an optimal tunnelling, material management and safety con-
cept for the construction phase of the main tunnel. According to the technical terms of the
contract, the use of an open hard rock TBM was specified for the continuous excavation of
the tunnel. One of the challenges, together with the extremely demanding predicted geological
conditions, which will be discussed in more detail, was the assembly of the TBM in a cavern
3.5 kilometers deep inside the mountain.
The TBM advance started in the assembly and start-up cavern at km 6+922 and extends to km
22+000 (transition to tunnelling of the exploratory tunnel Wolf). An extension of the drive is
being considered based on the advance rates so far. The excavated radius of 3.95m takes into
account deformation tolerances of up to 150mm, a shotcrete outer shell of up to 300mm and a
possible (subsequent) inner shell of 250mm. Overburden between 420m in the Navistal area up to
a maximum of 1300m in the Schröflkogel area are found along the tunnelling route. The section
of the BBT thus has a comparatively low overburden height compared to the Gotthard Base
Tunnel. The expected geological conditions are extremely demanding for TBM tunnelling. Com-
pared to the Gotthard, relatively soft rocks are encountered here, with predominantly formations
of Innsbruck quartz phyllite and Graubünden shale. Numerous fault zones are predicted, which
make up about 19% of the total distance of the drive. The soft bedrock (shale and phyllite) is thus
characterized by a high degree of difficulty in some sections compared to the Gotthard.
For the characterization of the expected rock mass conditions along the Ahrental section, a
subdivision into rock mass behavior types took place. Fundamentally these indicate loose to
friable rock mass, shear failure or even squeezing rock behavior.
The following minimum requirements for the TBM were thus defined for the TBM design:
– Radial rock deformations of 150mm with a nominal diameter of 7.9m
– Handling of fault zones and radial displacements of up to 400mm
– Water runoff of up to 10l/s/10m
– Radial rock pressure of 500kN/m2
– Probe drilling for strata exploration from the working area A1
The tunnel boring machine used for the construction lot must meet very high and complex
requirements. In addition to technical performance and safety for the personnel, economic
aspects also play a prominent role.
On excavating the cavity in the partially squeezing rock zones, without appropriate counter-
measures major, long-lasting deformations of the cavity can develop. The phenomenon occurs
primarily in rock types of low strength and high deformability, which are also mainly to be
expected along the section. In the following section, the machine concept and technical consider-
ations and approaches for drives in squeezing rock mass with an open hard rock machine are
explained in more detail and previous experience summarized.

3 TBM LAYOUT AND PROJECT SPECIFIC DESIGN FEATURES

For the exploratory tunnel, Herrenknecht supplied a 200 meter long and 1800 tonne Gripper
TBM with an excavation diameter of 7.93 meters.
In keeping with the predicted geology, the machine was designed for areas of squeezing rock.
Because of the squeezing rock mass behavior, the design of the machine took into account the
reduction of the excavation diameter (geometric influence) following excavation and possible
increasing rock loads in preventing this deformation (load influence).

3526
 The cutterhead is equipped with 46 19” discs and, for transport reasons, consists of 3 parts,
which were welded together on the jobsite. By means of 3 overcutting tools, a maximum over-
cut of 100mm in radius can be achieved (maximum excavation diameter is 8130mm). The pos-
sibility of a variable excavation diameter is a suitable measure to provide more leeway for
rock mass deformations and thus pass through zones of squeezing rock mass.
Open TBMs must have a flexible front shield including an adjustable invert shoe, whose
kinematic range covers the envisaged variance of the excavation diameter. The same applies
to the gripper unit and the rear support as well as to possible relations to the excavation diam-
eter in the back-up area, such as walking legs. In contrast to shield machines, in which the
entire back-up is based on the constant internal diameter of the segments, in open machines
the remaining excavation diameter remains relevant for all following operations. Predicted
variances upward (expansion) and downward (rock mass deformation) must be taken into
account in the initial design. Devices for the installation of primary rock support directly
behind the dust shield must be designed for the smallest and largest diameters possible. This is
of particular importance for mechanized arch setting equipment.
Thanks to the adjustability of the design of the gripper shield or the front shield on all sides
(see the following illustration), the machine axis can be kept concentric to the excavation axis
even with an increased diameter.
The gripper shield can be extended to suit the maximum excavation diameter. With regard
to the defined minimum requirements for the TBM, the gripper shield is designed for a radial
load of 500 kN/m2.
With a time-dependent deformation behavior of the rock mass with a possible accompanying
reduction of the excavated diameter, in addition to the overcut the length of the shield skin is
paramount. Compared to a shielded TBM the Gripper TBM is equipped with a flexible gripper
shield and has a short front shield length. For the Gripper TBM used here, the front shield
including fingers has a length of 4420mm. The kinematics of the front shield segments (see
figure) allow variation within the structurally possible radial strokes, including a reduction of
the diameter when a predetermined rock pressure is exceeded.
The predictions of the ground conditions likely to be encountered along the exploratory
tunnel were classified as difficult in places, making the possibilities of early rock support
immediately behind the cutterhead shield very important. The challenge for the machine
manufacturer is to optimize the operator safety, degree of mechanization and flexibility with
regard to the support methods used under the extremely harsh environmental conditions pre-
vailing in this area of the machine.

Figure 1. Flexible gripper shield with adjustable roof shield (green), side wings (yellow) and invert shoe
(blue).

3527
 Immediately behind the cutterhead shield in the L1 area, two independently moving
drills are installed for rock support. The maximum drilling depth per stroke is 3,000mm
and by extending the drill rods, longer drilling depths are possible. Also installed in the
L1 area is an erector with which the steel ring arches are assembled and installed, and a
spraying manipulator with which shotcrete can be applied in the area behind the gripper
shield.
The back-up unit consists of 10 trailers and a bridge construction. In back-up 1 is the rock
support for the L2 area. It consists of an anchor drilling system with two drill carriages and a
spraying manipulator with attached operating platform. On a ring carrier behind the shotcrete
manipulator, the drill (setting angle variable between 6–9°) for probing and injection drills is
mounted, which can be moved radially 120° in the crown area.
The findings from the driving of the exploratory tunnel, in particular with regard to the
encountered nature of the rock mass, are to be used for an optimal tunnelling and safety con-
cept for the TBM tunnelling of the main tunnel tubes. To help with the geological/geotech-
nical interpretation of the tunnel face, five of the 19 inch disc cutters were equipped with the
newly developed DCLM (Disc Cutter Load Monitoring) system. The DCLM system measures
the cutting load acting on the disc cutters, which allows conclusions to be drawn about the
condition of the tunnel face and thus an optimization of the drive parameters. On the moni-
tored tracks the disc cutter load monitoring system allows detection of changing geological
conditions such as fractures, stratifications, etc. in real time. The image for visualizing the
DCLM data in the control cabin has been supplemented with an additional display for the
proportion of underload and overload. As a result, the machine operator can recognize critical
disc cutter overloads during the advance in real time as well as unstable conditions at the
tunnel face on the basis of the disc cutter underload. In addition, due to the configuration of
certain limit values, the triggering of warning messages as well as further processing via the
machine’s PLC are easy to implement.
Another five disc cutters are equipped with the DCRM (Disc Cutter Rotation Monitoring)
system, which monitors the rotational movement and temperature of the disc cutters, allowing
the optimization of tool maintenance intervals.
Additionally, a camera system provides photos of the tunnel face to the control cabin moni-
tors. The cameras are mounted in five positions across the cutterhead at a maximum radial
distance apart of 1m. During tunnelling advance the cameras are protected by a cover and
when the machine is stopped for maintenance the covers can be opened to begin taking photo-
graphs of the tunnel face.
The camera system allows the tunnel face to be photographed during every maintenance
shift and the individual images are processed into a 2D image of the tunnel face to assist in the
detection of potential tunnel face instabilities such as loose rock falls in the crown area.

Figure 2. DCLM, visualization of the load effect at the tunnel face according to color scaling in kN.

3528
Figure 3. Photo-optically recorded image of the tunnel face with five different camera positions on the
cutterhead.

4 PROJECT EXPERIENCE TO DATE AND LESSONS LEARNED

After transporting the tunnel boring machine to the jobsite, the components were transported
down an access gallery with a 12% gradient to an assembly cavern 3.5km deep inside the
mountain where the TBM was assembled. From there, the 15 kilometer advance began in Sep-
tember 2015 from Austrian Ahrental toward the Italian border. After excavating the explora-
tory tunnel, the Gripper TBM will be pulled back through the excavated tunnel.
Along the first meters of the tunnel the open TBM had to master very complex ground con-
ditions. The challenges the TBM had to face were sometimes extremely difficult rock forma-
tions with highly squeezing and swelling rock conditions with extensive invert heave and the
encounter of fault zones with sometimes large-volume rock collapses, which were also encoun-
tered several times in the course of tunnelling. Tunnel advance rates in these conditions are
largely determined by a high degree of rock support work required in the L1 and L2 area.
Between the cutterhead and the installation area A1 are approx. 5.5m of tunnelling distance
or lead time, which can be used as an “exploration element” for the L1 area to be secured.
Due to the multiple collapses that continued to characterize tunnelling even after the first
1000 meters of tunnel, modifications were carried out in the TBM’s rock support areas in
order to minimize performance losses due to increased rock support requirements.
Rock mass deformations, breakouts and collapses required additional support measures in
the L1 area directly behind the gripper shield, including shotcrete for initial rock support and
360° TH arches, which slowed the daily performance of the TBM advance. To date (as of June
2018), seven fault zones were encountered in which a reduction in performance - depending on
the thickness of the fault zone - reduced tunnelling speeds to 1-3 m/day due to the high level of
rock support required. In these difficult geological zones, in some cases with converging rock
mass the available clearance in the L1 area between the bedrock geology and the spray nozzle
for applying the shotcrete for initial roof support in the L1 area was sometimes greatly reduced.
In addition, the shotcrete equipment in the L2 area was used much more extensively and
required greater availability with reductions in cleaning and maintenance time windows,
which in turn had a direct effect on the daily performance. In order to optimize tunnelling
performance in these geologically extremely complex areas, a total modification of the shot-
crete unit in the L2 area was carried out. The nozzle was designed to allow a larger, adjustable
area between the nozzle and the tunnel wall. Also, rebound collection was completely separ-
ated from the system in order to be able to dispose of the rebound during spraying operations
and perform cleaning work in parallel with the spraying.

3529
Figure 4. Deformed rock support due to prevailing convergences.

In the course of the advance, further system improvements were carried out. The TH arches
were replaced with U-profiles and the ring beam erector was greatly modified for arch installa-
tion, which resulted in higher performance coupled with greater wear protection. Further opti-
mization of the rock support measures meant that shotcrete in the L1 area was only applied
by hand and only for initial support. The logistics were also optimized. The supply of the
tunnel boring machine with invert segments, support material, and consumables was carried
out with multi-service vehicles (MSV), which - despite the steep slope - transport the material
from the surface through the lateral access tunnels directly into the back-up area.
A consideration of the advance rates results in a very satisfactory balance. In solid rock
daily top performances of 61m and monthly performances of 825m were achieved, while regu-
lar weekly performances of around 200m were standard over many months.

5 CONCLUSION

With the construction of the Brenner Base Tunnel, another epoch-making project is being realized
in the Alps. The 15 kilometer long Ahrental exploratory tunnel is part of this major project and is
being driven through extremely complex geology. Soft rocks with converging rock mass, unstable
tunnel face conditions with accompanying breakouts and collapses at the tunnel face required and
continue to require increased rock support efforts and technology adapted to the complex condi-
tions. The use of an open Gripper TBM was based on proven technology. Gripper TBMs have
already proven themselves in numerous complex projects in the Alps, including the Lötschberg
and the Gotthard tunnels. New in the evolutionary stage of this technology are systems that help
to optimize tool maintenance intervals, such as the DCRM system and the DCLM system that
makes it possible to draw conclusions about the nature of the tunnel face, thus allowing a
kind of tunnel face scanning. Together with the newly developed camera system, in future
real-time preliminary exploration will be possible and thus an optimization of the tunnelling
parameters along with improvements in terms of performance, safety and quality. The know-
ledge gained from driving the tunnel is contributing to the development of an optimal tunnel-
ling and safety concept for the construction of the main tunnel.

3530
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

Risk management for the Brenner Base Tunnel

K. Bergmeister
Brenner Base Tunnel BBT SE, Innsbruck, Austria

ABSTRACT: Risk management, along with cost management, is a very important tool in
carrying out tunnel and infrastructure projects. Proper risk and opportunities evaluation are
especially important in large infrastructure projects. This publication proposes a novel
method for risk structuring, based on the models available in literature and including
unknown and unidentifiable risks, also known as “black swans”. In infrastructure projects, we
may also encounter so-called unidentifiable risks and risks which, from a scientific point of
view, cannot be directly inferred. The basis of the work is a mathematical relation linking
extent to probability of occurrence: It also includes the development of measures to reduce the
extent of the damage from unknown risks in large infrastructure projects.

1 INTRODUCTION

1.1 Risk types

Large infrastructure projects typically involve long planning and construction phases and a
high degree of complexity. According to Flyberg (2009), ex ante underestimations of costs and
risks for large infrastructure projects have as their consequences both cost overruns and bene-
fit shortfalls. Large infrastructural projects have a long planning and approval period and
complex project interfaces.
Besides known and identifiable risks, which can be analyzed by some deterministic or
probabilistic approach, there are also risks that are not only not quantifiable but unidentifiable
as well. For this reason, in Austria a specific guideline (ÖGG-Richtlinie) for the estimation
of risk according to levels of difficulty has been worked out in 2005 and 2016. This risk esti-
mation is based on experiential values and includes a large area of possible risks. In complex
infrastructure projects, we may also encounter so-called unidentifiable risks and risks
which, from a scientific point of view, cannot be directly inferred. Such events, were called
by Taleb (2009) “The impact of the highly improbable” or “black swans”. In the following
paper possible management strategies in order to reduce this risk will be presented. From a
scientific point of view therefore at least 3 different levels of risk types and assessment
methods can be identified.

1.2 Description of the Brenner Base Tunnel

The Brenner Base Tunnel is the central part of the Scan-Med corridor which runs for approx.
9,000 km from Helsinki (Finland) to Valletta (Malta). The Brenner Base Tunnel (BBT) will
connect Tulfes/Innsbruck, Austria, with Fortezza, Italy. The BBT consists of two main tubes
with one railway track each, i.e. trains will travel through them in one-way traffic. In between
the two main tunnels, 12 meters below, runs an exploratory tunnel. This exploratory tunnel is
being excavated for geological and hydro-geological purposes. The results of this preliminary
research will provide detailed insight, and thus it helps to minimize construction costs and
risks when realizing the main tunnel tubes. In Innsbruck the BBT will be connected with the

3531
Table 1. Risk types and their methodology.
Risk type Methodology
Identifiable and quantifiable risks Probabilistic method supported by a Monte Carlo Simulation
Recognizable but not quantifiable risks Experience based approach (e.g ÖGG 2016 Austria)
Unforeseen big events (black swan) Phenomenological approach [Bergmeister, 2013]

Figure 1. Scan-Med corridor with the Brenner Base Tunnel.

existing underground bypass. Together with the bypass of Innsbruck, the Brenner Base
Tunnel will be the longest underground railway line in the world, with a total length of 64 km.
Already since 2010 a sophisticated risk assessment procedure is carried out for the Brenner
Base Tunnel project. In Italy on the basis of a national law (D.L. 50/2016), for unforeseen
events or risks a maximum of 8% of the total cost of the project can be assumed. In Austria
since 2005 a specific guideline describes for every infrastructural project a methodology to
forecast the identifiable and quantifiable risks as well as the recognizable but not quantifiable
risks.
For the Brenner Base Tunnel as a transnational project, the probability based approach for
the identifiable and quantifiable risks has been developed and applied (Bergmeister, 2015).
The modelling of risks and opportunities for the Brenner Base Tunnel project is carried peri-
odically, at least one yearly.

3532
2 THE PROBABILISTIC BASED APPROACH FOR THE IDENTIFIABLE AND
QUANTIFIABLE RISKS

This assessment of individual opportunities and risks considers the following aspects:
– Risk type,
– Measures for risk prevention,
– Costs for the measures for risk prevention, where the respective maximum, minimum and
expected values and the probability of occurrence are all defined.
– Time-related impacts once the opportunity or risk comes to pass, where the respective max-
imum, minimum and expected values and the probability of occurrence are all defined.
– Uncertainty of the assessment.

2.1 Areas of risk

The assessment of the individual opportunities and risks for the Brenner Base Tunnel takes
place in workshops with experts and the project manager of the specific construction lot,
taking into account, at least, the following topics aimed at identifying and assessing the indi-
vidual opportunities and risks. More than 400 independent risks and approx. 20 opportunities
have been identified. In the following table 2 the major topics of risks and opportunities are
presented.

2.2 Calculation of risk provision

After assessing each individual opportunity and each individual risk, the probability of occur-
rence “P” is calculated or estimated based on experience. The possible impact “I” and/or the
damage that may be the result of a risk occurring is assessed in terms of monetary impact, to

Table 2. Risk and opportunities.

Macro topics Some specific issues
Risks
Geology, Hydrogeology, Geomecha- Uncertainties, poor rock, gas pockets. . .
nics, Tunnel construction
Natural occurrences, Disposal and Meteorological events, settlements. . .
Construction sites
Archaeological risks
Construction site logistics Material supply. . .
Spoil Asbestos. . .
Mechanical risks Loss of machinery. . .
Tender procedures Legal problems
Contractual risks, claims
Authorizations Changes..
Labour safety Additional prescriptions. . .
Environment Noise pollution. . .
Opportunities
Reutilization of excavation material Aggregates, sale..
Changes of geology, geomechanical Improved rock stability
behavior
Improved performance of TBM Open gripper TBM: excavation time > 20%
refereed to total time
Shielded TBM: excavation time > 25% refereed
to total time (Ruepp, 2018)
Reduced water inlet Stretches of dry rock
Reduced vibration transmission Changed rock behavior. . .
Changes in the logistic chain Concrete manufacturing on site

3533
determine the maximum, minimum and expected values. This gives a monetary amount for
risk provision “R”.

Ri ¼ Pi x Ii ð1Þ

The time-related impacts of a risk are also assessed with a maximum, a minimum and an
expected value. For the probability distributions, a three-point estimation was used (figure 2),
based on expert judgement [3].
The three calculated or estimated monetary or time-related values give a statistic distribu-
tion. The various steps of the calculation are shown in the flow chart of figure 3.
At least the mean total risk cost Rm and a characteristic (in the sense of statistical confi-
dence) total risk cost Rc need to be estimated (see figure 4). The latter may be viewed as a
“safe” value, similar to Value-at-Risk (VaR), widely used in portfolio management. If R
denotes the total risk cost and α (typically 0.05) the desired confidence level, then VaR is
defined [1] by:

P ½R  VaR ¼ α ð2Þ

Any interdependencies among individual opportunities or risks can be assessed with a cor-
relation matrix. All the risks were considered to have a variable economic impact and none of
them was assumed as an “event” or “shock” with a particular probability of occurrence. The
dependence was quantified and identified as being positive.
Four scenarios were analysed with regard to the dependence magnitude, namely the cases 0,
0.25, 0.50, 0.75 for Kendall’s tau τ, representing independence, weak, moderate and strong
dependence.

Figure 2. Risk assessment.

Table 3. Correlation matrix – example: R1 depend

with 0.25 on R2, R1 is independent of R3 and R1
depend with 0.5 on Ri.
R1 R2 R3 Ri
R1 1 0,25 0 0,5
R2 0,25 1 0,75 0
R3 0 0,75 1 0,5
Ri 0,5 0 0,5 1

3534
Figure 3. Flow chart of the risk analysis.

By simulating the individual risks with the Monte-Carlo method, statistic distribution
curves can be calculated for each technical field or for each construction lot. Considering the
assumed probability of occurrence, the distribution of the individual risks were simulated and
aggregated with the help of a Monte-Carlo simulation. In the Monte-Carlo simulation, random
samples are taken of all opportunities and risks to be taken into consideration. The basis for
these samples is the distribution curve. For the Brenner Base Tunnel the beta distribution of the
monetary or time-related impacts of the risks or the dependency from other risks have been
used. On the basis of various numerical calculations the beta distribution in figure 4 was identi-
fied as most appropriate by Tamparopoulos (2012) in his PhD-thesis according.
The various fractile values of the risk costs, Value-at-Risk(%) (VaR) can be determined and
calculated. Based on these curves and in accordance with the desired degree of coverage
(VaR), the scope of the monetary or time-related risks can be calculated or directly read with
a Lorenz curve. This sum curve is a graphical representation of the risk provision expressed in
a monetary value (Euros) depending on the probability of risk acceptance. With a Lorenz
curve we can see that even a small percentage increase in risk cover (less risk acceptance) can
cause significantly higher risk costs. For the Brenner Base Tunnel, a VaR of 50% was assumed
and agreed between Italy and Austria. For other infrastructural projects or underground
infrastructures a VaR of 75% is widely used. It should be noted that in all cases, the mean
(expected) value of the total cost is invariant, since the expectation operator preserves the
sums of random variables.

3535
Figure 4. The beta distribution of the monetary or time-related impact of risks (taken from Tamparo-
poulos, 2012).

Figure 5. shows the computation of the beat-curve for the estimated risk costs.

Figure 6. shows the distribution function of the risk provision for identifiable and quantifiable risks,
simulated with the Monte-Carlo method.

3536
Figure 7. shows the Lorenz curve and/or the sum curve of the risk provision for identifiable and quanti-
fiable risks, simulated with the Monte-Carlo method.

Figure 8. shows the Lorenz curve and/or the sum curve of the risk for additional time needed, simulated
with the Monte-Carlo method.

2.3 Calculation of the VaR value for the Brenner Base Tunnel
The total risk provision for the Brenner Base Tunnel with a 50% fractile (VaR50) was calcu-
lated as 697.5 mio Euro per 01/01/2017 and, with a 90% fractile (VaR90) as 839.3 mio Euro.
The total amount calculated for the opportunities was after a 5-year optimization period
(on the basis of the opportunities carried out in the risk management) ca. 2,6 mio Euro.

3 RECOGNIZABLE BUT NOT QUANTIFIABLE RISKS – THE AUSTRIAN

GUIDELINE FOR RISK ASSESSMENT

In Austria, in Switzerland (Ehrbar et al. 2003) and in Germany exists since 2017 a specific
guideline for a holistic assessment of all risks (quantifiable and not quantifiable risks).
The first Austrian ÖGG directive of 2005 and the actual guideline 2016 “Costing for trans-
port infrastructure projects” gives estimates of the required risk provision on the basis of

3537
Figure 9. Excerpt from the ÖGG Directive (2016) for the calculation of total risk provision.

many years of experiential data. The economic risk evaluation considers risks in both the pro-
moter’s and the project owner’s sphere of risk. To calculate the costs for risk provision, certain
assumptions are made based on project phase, complexity and construction ground impact.
– Project phase: Project development, preliminary design, final design, approval, tender,
realization
– Complexity: The project is categorized as difficult.
– Ground impact: Tunnels, cut&cover structures like the Brenner Base Tunnel are standing
for a high ground impact
For the actual situation, depending on the respective status of realization and on the esti-
mate of the complexity of the project Brenner Base Tunnel the risk provision for all risks
(quantifiable and not quantifiable risks) amount € 1.215 mio per 01/01/2017.

4 UNKNOWN, EXTREME OPPORTUNITIES AND RISKS

Due to their complexity and long completion horizon, mega-projects are far more likely to be
impacted by unexpected events than projects with a shorter completion time. Currently, no
quantitative models are to be found in literature that calculate risk provision for such extreme

3538
events (giga-risks), but such a model would make sense in this case. There are no experiential
or theoretical models to estimate these unknown, extreme events. In Bergmeister (2013) drew
up a suggested approach to how such extreme events could be structured for mega-projects
and, again as an approach, identified and evaluated.
In modelling unknown, extreme opportunities and risks, several independent or inter-
dependent (mutually causal) scenarios can be developed on the basis of the phenomenological
Delphi scenario method. These include, for example, mudslides caused by natural events or
floods due to high water or interruptions in financing resulting from political crises. The phenom-
ena in the early phases of an event are observed and possible processes are shown, which requires
the consideration of possible catalysts (phenomena that suddenly accelerate a process). The scen-
arios can be developed by experienced experts, supported by computer models. The question is,
of course, which scenarios hold the greatest risk or the greatest opportunity. Economic science
solves this problem using so-called operational research. Interesting goals are defined as partial
goals and their importance is summarized in a function using weighting factors. This simplifies
the problem and solves the target function. For the constant monitoring of certain phenomena or
scenarios, so-called key indicators must be set. These are important indications that an event is
imminent and require the adoption of corresponding measures.
For example, as regards the Brenner Base Tunnel, air quality conditions on disposal sites,
possible deformations of the disposal site shape and water resources are the subject of con-
stant monitoring.

5 CONCLUSION

It is important for all infrastructure projects and especially for large projects such as the Brenner
Base Tunnel to carry out risk management and periodically update the assessment of opportun-
ities and risks. For the Brenner Base Tunnel, the identifiable and quantifiable opportunities and
risks are determined with a probabilistic method. In addition to this, risk provision is also
assessed for recognizable but not quantifiable risks using the ÖGG Directive. Risks that are a
decisive factor for costs and timing must be prevented or at least reduced by specific measures.

REFERENCES

Bergmeister, K. (2013): Holistisches Chancen- und Risikenmanagement von Megaprojekten. PhD-tesis.

University of Bratislava-Vienna.
Bergmeister, K. (2015): Anwendung des holistischen Chancen-Risiken-Managements beim Brenner Basi-
stunnel. In: Tunnelbau Kompendium der Tunnelbautechnologie. Deutsche Gesellschaft für Geotechnik.
Wiley Online Library
Ehrbar, H.; Kellenberger, J.: Risk Management during the Construction of the Gotthard Base Tunnel.
Proc. Int. Symp. GeoTechnical Measurements and Modelling (GTMM), Karlsruhe, 2003
Flyvberg, B.; Bruzelius, N.; Rothengatter, W. (2006): Megaprojects and Risk. An Anatomy of Ambition.
Cambridge University Press. ISBN 0 521 00946 4
ÖGG-Richtlinie (2005, 2016): Kostenermittlung für Projekte der Verkehrsinfrastruktur unter Berücksich-
tigung relevanter Projektrisiken. Salzburg
Ruepp, A. (2018). Evaluation of excavation methods. Master thesis. BOKU University Vienna
Taleb, N.N. (2009). The black swan: The impact of the highly improbable. London.
Tamparopoulos, A. E. (2012). Cost estimation of large construction projects with dependent risks a
study on the Brenner Base Tunnel. Ph.D. – thesis BOKU University Vienna
Tamparopoulos, E.A., Spyridis, P., Bergmeister, K. (2011). Small failure probabilities and copula func-
tions: Preliminary studies on structural reliability analysis. In C. Bérenguer, Grall A., and C. Soares,
editors, Advances in Safety, Reliability and Risk Management – Proceedings of the European Safety
and Reliability Conference, ESREL 2011, pages 1115–1120, Troyes, France.

3539
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

The second Gotthard tunnel tube

G. Biaggio & V. Kumpusch

Swiss Federal Roads Office (FEDRO), Bern, Switzerland

ABSTRACT: Featuring a length of 16.9 km, the planned second tunnel tube through the
Gotthard will prove to be yet another outstanding feat in the Gotthard region. This contribu-
tion provides a short overview of the project, explains the investigation of the fault zones, out-
lines the excavation concept for piercing these zones and addresses both material management
and the reuse of the excavation material.

1 INTRODUCTION

Located on the north-south axis of the A2 motorway, the Gotthard Tunnel connects the can-
tons of Ticino and Uri between Airolo and Göschenen. The existing motorway tunnel was
opened in 1980, with forecasts based on ongoing checks indicating that it will be due for reno-
vation in around 15 to 20 years. As part of the ‘Gotthard conservation concept’ study that
began in 2009, efforts were made to identify and investigate on different feasible options for
conservation. These included the prospect of constructing a second tunnel and subsequently
renovating the first tunnel, as well as the possibility of enforcing complete closures lasting sev-
eral years in order to enable the renovation of the existing tunnel. This latter option would
have required the diversion of traffic via the pass and/or rail loading of the vehicles. On 27
June 2012, the Swiss Federal Council decided in favour of the construction of a second tunnel
tube with subsequent renovation of the existing tube. The project was approved by the citizens,
the Council of States and the National Council in a referendum on 28 February 2016.
The chosen solution significantly increases the level of safety in the Gotthard Tunnel and
ensures that the most important north-south connection will remain open during the renova-
tion of the existing tunnel tube.
When the project is completed, both tubes will feature single-lane operation with one
standard lane and one service lane in each direction (Figure 1). This ensures compliance
with the constitutional clauses relating to Alpine protection, which also preclude any
attempts to increase the capacity of transit roads in the Alpine regions.
While the first tube was being constructed, there was already a project concept for a second
tunnel tube. It was planned to subsequently expand the service and infrastructure tunnel that
runs parallel to the existing tunnel to enable the establishment of a second tube. The current
project deviates from this concept, especially as a result of the easier construction methods
featuring TBM excavation that are available nowadays.

2 PROJECT DESCRITPION

The planned second tunnel tube through the Gotthard has a total length of 16,866 m. It runs
at a standard clearance of 40 m from the service and infrastructure tunnel located east of the
existing Gotthard Tunnel. The layout of the line runs largely parallel to the existing tube
(Figure 2). The clearance is reduced in the portal areas in order to let the axis of the second
tube run along the axis of the service and infrastructure tunnel. In Airolo and Göschenen, the

3540
Figure 1. Overview of the overall system at the completed Gotthard Road Tunnel.

Figure 2. Overview of the second Gotthard road tunnel.

existing portal structures – which were designed for the inclusion of a second tube already
while the first tube was being constructed – and cut-and-cover tunnels are used as tunnel
portals. To ensure full use of the service and infrastructure tunnel once the work is completed,
the portal areas will be repositioned in advance.
The tunnel tubes are going to be connected at regular intervals by 62 walk-through cross-
passages and six vehicle-accessible cross-passages. A majority of the new cross-passages will
be arranged in the extension of the existing cross-passages to the service and infrastructure
tunnel.
Because of the slab track and the intermediate ceiling, the standard profile (Figure 3) will be
divided into three main areas: the driving area (clearance envelope for one standard lane and
one service lane with walkways on both sides according to the first tube), the intake and
exhaust duct above the intermediate ceiling and two service ducts beneath the slab track (with
one service duct intended for the routing of a 380 kV line). The tunnel will be given a double-
lining construction (excavation support and inner lining) and an umbrella sealing. The drain-
age water will be collected and guided to the portals via water lines located in the invert while
the water from the road surface will be directed into the wastewater line via slotted channels
and siphon shafts.

3541
Figure 3. Standard profile of the second Gotthard road tunnel.

Ventilation poses a major challenge when it comes to long road tunnels. When it comes to
the second road tube through the Gotthard, a ventilation system will be installed which
enables one-way as well as two-way traffic in case of maintenance. The ventilation systems of
the two tunnel tubes are independent of one another, with three new underground ventilation
control centres and two portal control centres being created for the second tube. Only those
ventilation shafts that are already present in the existing underground ventilation control
centres and provide a sufficient ventilation cross-section will be used for both tubes. These
ventilation shafts will be used both for exhaust air and the intake of fresh air for the traffic
area. In case of an emergency, the tunnel will be divided into five ventilation sections. The
fumes are to be drawn out of the driving area via exhaust flaps into the intermediate ceiling
and blown out through the ventilation shaft by the respective control centre. Jet fans will also
be used to control the longitudinal flow in the driving area.
The accompanying measures will also involve the rearrangement of the Airolo motorway
connection, including tasks such as the creation of the Galleria di Airolo and corresponding
terrain modelling. This cut-and-cover twin-tunnel will be 1086 m long and take the form of a
gallery in certain sections on the valley side, making a significant contribution to the eco-
logical upgrading of the valley floor.

3 GEOLOGY AND HYDROGEOLOGY

3.1 Geology
Thanks to the excavation of the adjacent first tunnel tube the geology of the second Gotthard
road tunnel to be driven through is well known (Figure 4). Approaching from the north, the

3542
Figure 4. Geological longitudinal profile.

tunnel crosses the Aar Massif that consists mainly of granites and paragneiss. It then passes
through the Urseren zone, which is formed of meta-sediments from the Mesozoic and
Permo-Carboniferous periods. The third formation is the Gotthard Massif, largely compris-
ing paragneiss, granite gneiss and granite. The Nufenen zone composed of meta-sediments
(from the Mesozoic period) is reached at the Airolo portal. The portal areas themselves fea-
ture loose rock sections of varying lengths.
Along the axis of the planned second Gotthard road tube it is expected to encounter two
problem zones: these will subsequently be referred to as the ‘northern fault zone’ (Mesozoic/
Permo-Carboniferous) and the ‘southern fault zone’ (Guspis zone).
Located in the Urseren zone at the transition between the Mesozoic and Permo-Carbonifer-
ous sections, the northern fault zone is around 270 m long and largely consists of clay and
sericite schists. Limestone, gypsum and dolomite are also expected in this zone. The rock in
the northern fault zone shows moderate to very high levels of fault gouging and demonstrates
both friable and squeezing rock behaviour.
The southern fault zone is located in the Guspis zone of the Gotthard Massif. Cover-
ing a distance of nearly 300 m, it contains a variety of more weathered and fault-
gouged and also more compact and firm biotite gneiss and schists, hornblende schists
and individual amphibolites. Dyke rocks are also expected to be frequent. As with
the northern fault zone, friable and squeezing conditions are also anticipated in this
section.

3.2 Hydrogeology
The crystalline rocks show average to good permeability, especially along the joint planes. Old
crystalline gneiss and schists have a somewhat lower level of permeability than the crystalline
rocks. The Alpine metamorphic sedimentary rock located above this, which originates from
the Mesozoic and Permo-Carboniferous periods, also features low permeability and indeed
acts largely as an impermeable layer.
Despite the high underground water level and increased permeability in the area of the fault
zones and the fissured rock, along rock borders and in hydrothermally disintegrated rock, the
overall expectations based on the measurements performed in the existing system point to the
presence of moderate amounts of water.
Areas bearing more water are particularly clustered in the southern zones near to the
portal. The amount of underground water at the portal is forecast to be 30 – 40 l/s in the
north and 95 – 110 l/s in the south.

3543
4 EXCAVATION CONCEPT

Given the length of the tunnel, the planned excavation concept involves one TBM drive each
approaching from the north and the south portal. Both drives are to proceed at an incline and
will feature an excavation diameter of 11.8 m.
The loose rock sections in the portal areas will undergo conventional excavation in advance.
Afterwards a further amount of rock (400 m in the north, 750 m in the south) will be exca-
vated by blasting drivage. This is due to the very cramped conditions in the portal areas,
which will affect the assembly and launching of the TBMs, as well as additional issues relating
to the construction programme and risk-related considerations.
In order to reduce the risks and optimise the overall construction programme, the northern
and southern fault zones (totaling approx. 570 m) will be conventionally excavated and
secured in advance. Once this has taken place, the two TBMs will be pushed through these
areas. The fault zones will be reached via separate access tunnels approaching from the north
(approx. 4.4 km) and the south (approx. 5 km), which will be created at an early stage. These
tunnels will be excavated using Gripper TBMs with an excavation diameter of 6.0 m.

5 PIERCING OF GEOLOGICAL FAULT ZONES

5.1 Experience gained from the excavation of the first Gotthard road tunnel
During the construction of the first Gotthard tube, it was likewise necessary to pass through
the two fault zones featuring friable to squeezing rock behaviour.
In the northern fault zone, there were considerable difficulties relating to the excavation of
the service and infrastructure tunnel. There were rockfalls that generated so called chimneys,
incidents of water ingress that left the rock showing signs of softening and generally low sta-
bility times of a few hours. Grouting did not lead to any noticeable improvement in the prop-
erties of the rock and the accruing seepage water was only partially displaced.
Based on the experience made while building the service and infrastructure tunnel, the main
tunnel was excavated using the so called ‘German’ method (core method) in the area of the
northern fault zone. This involved the sidewall headings and the crown area being excavated
first and after completing the tunnel walls the core was removed.
In addition to the loose rock-like conditions, it was also necessary to deal with squeezing
rock behaviour as average convergences of approx. 0.4 m were measured. It was possible to
establish a link between the occurrence of the convergences and the level of permeability: the
deformations occurred early on in the case of high permeability, while deformations were only
detected after a couple of hours or days in areas where permeability was low.
Squeezing conditions were also experienced along a nearly 400 m section in the area of the
southern fault zone, both in the service and infrastructure tunnel and in the main tube.
In light of the experience gained with the excavation of the service and infrastructure
tunnel in the Guspis zone, an exploratory heading was driven in advance on the axis of
the first tube. Considerable amounts of convergence occurred, especially in foliated sec-
tions of the paragneiss, and this could no longer be overcome by means of a simple steel
support. For this reason, and also as a result of what happened with the north drive in the
Mesozoic/Permo-Carboniferous section, the decision was taken to also excavate the main
tube via an intermediate access point using the core method. A high level of rock pressure
was detected in the process, with the resulting convergences amounting to as much as 1.5 m
in the side walls (Figure 5).

5.2 Investigation of the fault zones

Starting from the existing service and infrastructure tunnel, a total of eight sub-horizontal
exploratory probe holes each measuring approx. 70 - 75 m were created across both fault
zones in 2016. This was done in order to investigate the geological conditions and to extract

3544
Figure 5. Squeezing rock during construction of the Gotthard Road Tunnel.

rock samples for laboratory testing. Three exploratory boreholes were drilled in the Jurassic
limestones and clay schists of the Mesozoic layer in the Urseren zone and five boreholes were
generated in the Guspis zone. Pressure gauges were installed at two boreholes and dilatometer
tests were performed in five boreholes.
In addition, the laboratories of the École polytechnique fédérale de Lausanne (EPFL) and
ETH Zurich were commissioned to perform the following rock mechanics tests on rock
samples:
– uniaxial compressive strength tests
– Tri-axis tests (with and without consideration of the pore water pressure)
– Split tensile strength test
– Direct shearing tests
– Cerchar abrasivity tests
These were used to determine the strength and deformation properties of the fault-
gouged rocks.

5.3 Excavation and support concept

Based on the forecast geology and the experience gained from the first Gotthard tube, poten-
tial danger patterns were defined for the fault zones and a subsoil model divided into homoge-
neous areas was created.
Given the poor-quality properties of the rock, it was decided that both fault zones would be
excavated by means of conventional excavation.

3545
 In the attempt of achieving ring closure as quickly as possible, the planned excavation con-
cept for the fault zones involves full-face excavation. The excavation cross-section (Da = 14–
14.5 m) features room for securing, tolerance for movement on the part of the TBM and space
for convergences (0.65 m respectively 0.5 m) based on the deformation calculations (see
Section 5.4).
Grouted anchor bolts will be deployed in the crown area as an auxiliary construction meas-
ure in order to ensure working safety. Steel fibre shotcrete will also be applied to the profile
and the face, with face bolts also being installed in the latter (110, l = 8 m, every 12 m). In
order to allow the deformations to occur in a controlled manner, the plan calls for a yielding
lining (as was used in the Tavetsch intermediate massif of the Gotthard Base Tunnel). This
will involve the installation of mesh, steel ring beams with sliding connections (each featuring
two interlocking, moveable TH-beams of type TH 44) and a radial bolting system (self-drilling
anchors, l = 8–12 m). The ring beams are covered with shotcrete, in which inlays are imple-
mented to keep specific areas open allowing a more controlled deformation (Figure 6). Only
once most of the forecast deformations have occurred, the resistance will then be increased –
firstly by completely shotcreting the open inlay areas and thus fixing the TH-beams in place,
and secondly by ensuring that the excavation support reaches its target thickness of 50 cm
through the application of mesh-reinforced shotcrete (Figure 7).

Figure 6. Yielding lining in area L1.

Figure 7. Yielding lining in area L2.

3546
Figure 8. Characteristic curve for rock according to characteristic values of rock.

5.4 Deformation calculations and remaining uncertainties

The convergences in the fault zones were calculated using the characteristic curve procedure.
Additionally a FE model was created for the southern fault zone. The calculations were also
used to investigate the influence of the wall effect on short fault zones, thereby accounting for
the beneficial influence this has on the behaviour of a thin fault zone in the neighbouring
unaffected area. Together with the characteristic curve of the excavation, this makes it pos-
sible to determine the best time for the installation of the TH-beams and for the closing of the
sliding connections so as to enable optimum loading of the excavation with only minor
deformations.
Despite the prior knowledge relating to the first Gotthard Road Tunnel tube and good-
quality investigation of the rock, the anisotropy of the rock and the wide range of rock char-
acteristics mean that the calculations remain subject to some uncertainties (Figure 8).
In addition, the consideration given to the pore water pressure plays a major role as it sig-
nificantly influences the short-term and long-term development of the convergences. There is
also still an outcrop gap at the northern border of the southern fault zone.
These issues will have to undergo more in-depth analysis in the course of subsequent project
planning.

6 MATERIAL MANAGEMENT

6.1 Reuse concept

The project has set the following objectives in relation to material management:
– The excavated material must be utilised as much as technically possible.
– The environmental impact of material management must be kept to a minimum.
– The material management concept must be economically acceptable.
– Material management must not be the ultimate performance-defining factor.
The excavation material is primarily generated from the main drives. Additional excavation
material is also amassed from auxiliary structures, access tunnels and caverns. The geology,
which has been documented in detail and was further investigated via the exploratory boreholes

3547
Figure 9. Reuse of excavated material.

drilled in 2016, enables positive forecasts to be made in relation to effective recycling. A total of
nearly 6.5 million tons of excavation material will be accumulated (Figure 9), half at the north
portal and half at the south portal.
It will be possible to use around 1.6 million tons as aggregates for concrete once this has
been processed. Around 4.9 million tons of uncontaminated material that is not suitable for
processing or is rejected during material processing will be recycled in the course of filling pro-
cedures or renaturation. Lightly contaminated silt and lightly polluted excavation and dis-
mantling material cannot be used for renaturation or terrain modelling and will be disposed in
compliance with the law.

6.2 Material logistics

Due to the cramped conditions in the vicinity of the portal and the exposed location with
respect to natural hazards (especially at the portal area in Göschenen), the material manage-
ment installations will largely be located in the Airolo area.
The excavation material will be transported from the face to the triage points located out-
side of the portals by means of a conveyor belt. The material that is accumulated in the north
drive and is suitable for processing will be taken by train through the Gotthard Tunnel from
the Göschenen portal to Airolo. Once there, both this material and the material from the
south drive will be transported via conveyor belt to the processing plants in Stalvedro.
The processed aggregates will then be transported from Stalvedro back to the silos of the
concrete plants via conveyor belt and by train (in the case of Göschenen). With the exception
of the preliminary lots, it will be possible to cover the full amount of gravel and aggregate
required for the concrete through the processing of the A material.

6.3 Filling and renaturation

A certain amount of the excess non-polluted excavation material (B material) will be deposited
in the immediate surroundings of Airolo. The rearrangement of the Airolo connection features
plans for various adjustments to the terrain that are associated with relevant material require-
ments. These also include the Galleria di Airolo, around 1000 m of which will be covered with
excavation material – thereby significantly upgrading the valley floor (Figure 10).
A further approx. 2.8 million tons of excavation material will be used in the canton of Uri
for the renaturation of the shallow water zones of the Reuss delta and will therefore serve as
embankment protection for the Reuss plain (Figure 11). At the beginning of the 20th century,
the point at which the Reuss flowed in Lake Uri consisted of an expansive shallow water zone
featuring multiple islands. Following the canalisation of the Reuss and the removal of gravel,
which started at the turn of the 20th century, the river landscape ceased to develop in a natural
manner. The shorelines moved 200 to 300 m inland in certain areas depending on the water

3548
Figure 10. Visualisation of covering of Galleria di Airolo.

Figure 11. Visualization of renaturation of shallow water zone in Lake Uri.

level, with around 24 ha of reed meadows and fields falling victim to this human intervention.
Between 2000 and 2007, a total of 3.3 million tons of excavation material from other tunnel
projects was used for the purpose of renaturation in seven stages. The excavation material
from the tunnel projects relating to the new Axenstrasse and the second Gotthard road tunnel
will make it possible to complete this renaturation process. ‘Lakefill III’ is a separate project
being run by the canton of Uri that was granted approval back in summer 2017.

7 CONCLUSIONS

The building contractor is convinced that the excellent collaboration among all participants
has enabled the development of a convincing project. Following public disclosure, it will be
possible to quickly turn this project into a reality. Once the construction period of around
eight years has been completed, the second Gotthard road tunnel will be opened. This is esti-
mated to take place in 2028, depending on the duration of the approval process. The commis-
sioning of the second Gotthard road tunnel will be immediately followed by the renovation of
the existing first tunnel, which will take around three years. After around eleven years of

3549
construction work a modern and safe overall system featuring two tunnels – each containing
one lane – can be put into operation.

REFERENCES

IG Gottardo Due. 31/08/2017. Construction project. Secondo tubo San Gottardo, Bozza Rossa
Swiss Federal Roads Office (FEDRO). 25/10/2017. Die zweite Gotthard-Strassenröhre. Infodossier zur
Genehmigung des Generellen Projekts durch den Bundesrat.
Cf. https://www.astra.admin.ch/astra/de/home/themen/nationalstrassen/sanierung-gotthard.html
Wild, K. & Amberg, J. 13/10/2017. Presentation ‘Problematik Durchörterung Störzonen’, STSym autumn
workshop

3550
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

Dubai Strategic Sewerage Tunnel – challenges to infrastructure at

unprecedented depths in the region

D. Brancato, A. Ayoubian, M. Joye & G. Monks

Parsons Corporation

F.A. Al Awadhi & J. Tharamapalan

Dubai Municipality, Dubai, UAE

ABSTRACT: In the rapidly developing Emirate of Dubai, the Dubai Strategic Sewerage
Tunnel (DSST) Project will upgrade the existing sewage infrastructure from a pumped system
to a gravity-based system. The project will include almost 250 km of link sewer, approximately
82 km of tunnels with inside diameters ranging from 3.5 to 6.5 meters, and two terminal pump
station shafts with diameters of 43 meters and 63 meters. The general subsurface conditions
are surficial sands underlain by cemented calcareous sand and weak rock, with groundwater
near ground surface. Preliminary design was performed for the largest diameter pump station
shaft final liner and invert slab, and tunnel liner. Additional analyses include a cylindrical
buckling and uplift evaluation of the final shaft structure under computed loads.

1 PROJECT OVERVIEW

Dubai Municipality (DM) has embarked on a major program to expand and improve their exist-
ing sewerage system by converting from a pumped system to a deep gravity tunnel system. This
Dubai Strategic Sewerage Tunnels (DSST) program will provide the infrastructure to capture and
treat flow for the rapidly developing Emirate of Dubai. The program will be expanded in future
to suit land use changes. The project will optimize operations for DM by eliminating more
than 100 subsidiary and main pump stations and localized Sewage Treatment Plants (STP)
and diverting and centralizing sewage flows to the Al Warsan and Jebel Ali sewage treatment
plants (WSTP and JSTP). The DSST project has two main catchments shown in Figure 1. The
Deira catchment terminates at WSTP, and the Bur Dubai catchment flows to JSTP, which are
called herein DSST-DR and DSST-BD, respectively.

1.1 Existing Sewerage System

The existing system includes both subsidiary and main pump stations, with the main pump
stations transferring flows by force main directly to the STPs. There are twelve main pump
stations in the existing sewerage system. Each of these main pump stations receive flows from
subsidiary pump stations and in many instances, from direct gravity sewers. There are over
100 subsidiary pump stations owned and operated by DM. Flows from these subsidiary pump
stations are transferred either directly via pumps or by intermediate gravity networks to the
main pump stations.

1.2 Proposed Sewerage System

The DSST-DR catchment will include approximately 15 km of deep sewer tunnel (3.5 m diam-
eter) ranging in depth from 35 to 80m below ground surface with a 790 mega liter per day

3551
Figure 1. Catchment areas for the Dubai Strategic Sewerage Tunnels for the Al Warsan and Jebel Ali
Treatment Plants.

(MLD) ultimate capacity terminal pump station and 55 km of link sewer. The DSST-BD
catchment has approximately 60 km of tunnels (ranging in diameter of 3.5 to 6.5 m) and
varies from 22 to 108m in depth (Figure 2). DSST-BD has 196 km of link sewers and a 4,270
MLD ultimate capacity terminal pump station. The link sewers will divert existing sewerage
networks (i.e. subsidiary and main pumping stations and localized STPs) to the tunnels. The
diameters for the terminal pump station shafts are 43 m for the WSTP shaft and 63 m for the
JSTP shaft based on preliminary concepts developed for the stations.
In addition to the deep tunnels and the terminal pump station shafts, other ancillary shafts
will be constructed. The shaft types include tunnel boring machine (TBM) launch or construc-
tion shafts (CS), TBM reception shafts (RS), vortex drop structure shafts (VDS), and main-
tenance shafts (MS). The elevation view of Figure 2, the primary segment of the DSST-BD
tunnel, shows the locations of 27 shafts: five (5) construction shafts, one (1) receiving shaft,
fourteen (14) maintenance shafts, and seven (7) vortex drop structure shafts. The vortex drop
structure shafts intercept and divert link sewer flows to the main tunnel. Flows intercepted by
construction shafts from link sewers or other tunnels can be accommodated by internal struc-
tures built within the relatively large diameter shafts. The maintenance shafts provide manhole
access to the tunnel and are spaced approximately every 2 to 3 km.

Figure 2. Schematic of the DSST-BD vertical alignment, showing locations of incoming flows.

3552
Figure 3. Terminal Pump Station – Preliminary Concept.

Not including the two terminal pump station shafts, there will be more than 40 intermediate
shafts for the DSST project. These shafts are proposed with minimum excavation diameters of
10 m for the maintenance shafts, 12 m for the reception and vortex drop shafts, and 20 m for the
construction shafts. Permanent diameters will range from 2 to 4 m for the construction,
reception, and maintenance shafts, and between 5 to 10 m for the vortex drop shafts.
The single pump station concept will require terminal pump station shafts to be very large
and very deep. Figure 3 shows a section view of a terminal pump station. The alignment was
reviewed to identify locations where an intermediate pump station could be located to reduce
the depth and size of the pump stations. The evaluation did not indicate a cost-benefit to the
intermediate pump station concept. Benefits with respect to depth were offset by land avail-
ability and operational considerations. Therefore, the preliminary design is based on the single
Terminal Pump Station (TPS) concept. The preliminary pump station designs include wet and
dry wells with single-stage vertical mixed flow volute pumps.

2 SUBSURFACE CONDITIONS

The Dubai area is covered with a middle to upper tertiary group of rocks, underlying the
recent soil/desert sand or beach sand of varying thickness of 0.5–17m. The coastal sands are
largely Holocene, and near the coast consist of mostly calcium carbonate (CaCO3) derived
from carbonate sediments, seashells, and coral reefs.
Underlying the Holocene sands is Pleistocene Ghayathi formation. The Ghayathi Forma-
tion consists of light brown to reddish brown, poorly cemented calcareous sand, white Calcar-
enite, calcium carbonate rich sandstone with generally distinct cross bedding. These
Calcarenite Sandstones are generally up to 25m thick.
Underlying the Pleistocene Sandstone Conglomerates is the Mio-Pilocene Barzaman forma-
tion. The Barzaman formation consists of Pliocene and Miocene buffwhite/creamy white Silt-
stone, Calcisiltite, and light reddish brown Conglomerate. The thickness of the Barzaman
formation is up to 60m.
Underlying the Mio-pliocene Barzaman formation is the Miocene Fars formation. The Fars
formation consists of reddish brown/greenish grey Siltstone, Mudstone, Claystone, and color-
less to white Gypsum.
Based on the finalized tunnel alignments and by considering the available geotechnical data,
the majority of the main tunnel alignment will cross the Barzaman formation. Typical rock core
photographs of Barzaman Siltstone/Calcisiltite (left) and Conglomerate (right) are shown in
Figure 4. However, the shallower sections of the Deira tunnel north alignment will encounter
the Ghayathi formation. A typical rock core photograph of Ghayathi Sandstone is shown in

3553
Figure 4. Typical Rock Core Photograph of Barzaman Siltstone/Calcisiltite (left) and Conglomerate (right).

Figure 5. Typical Rock Core Photograph of Ghayathi Sandstone.

Figure 5 The tunneling operations, along the deepest sections of Bur Dubai tunnels, at the area
of Jebel Ali treatment plant, will cross the top layers of Fars formation (mudstone units).
An extensive subsurface investigation for the design of the project was conducted along the
tunnel and link sewers alignment. The field investigation included geotechnical borings with
downhole seismic velocity measurements, in-situ Packer permeability tests, falling head perme-
ability tests, and installation of groundwater monitoring wells. A wide ranges of laboratory
tests were conducted on soil and rock samples collected during subsurface investigation. Soil
tests included Atterberg Limit tests, moisture content, hydrometer, and chemical tests. Tests
on rock cores included uniaxial compressive strength test with measurement of Young’s
Modulus, point load index, indirect tensile strength test (Brazillian), total hardness, slake dur-
ability test, axial swelling test, Cerchar Abrasivity Index (CAI), petrographic analysis descrip-
tion, petrographic point count analysis, and X-Ray diffraction clay joints.

3 DESIGN CHALLENGES

One of the primary goals of the project is to develop sewer schemes to decommission all pump
stations to the new sewerage system. The two terminal pump stations are very large and require
managing risk associated with the unique conditions for the shafts. The goals of the sewerage
tunnel investment are best satisfied when the system includes the deep shafts and tunnels, but
must also satisfy the 100-year design life. This requires a focus on developing a durable design.

4 PERMANENT STRUCTURE DESIGN

The preliminary design for the segmental liner of the sewerage tunnel and the Jebel Ali pump
station shaft at the terminus of the Bur Dubai alignment was performed. The tunnel outlet at
the Jebel Ali shaft has an invert depth of 110 meters below ground surface, an internal diam-
eter of 6.5 meters, with a liner thickness of 350 mm. The shaft itself has a final depth of about

3554
118 meters, a proposed internal diameter of 63 meters, a final liner thickness of 2.5 meters,
and an invert slab varying in thickness of 3.5 to 6.5 meters. Both the shaft liner and shaft
invert slab are to be cast-in-place reinforced concrete, and the tunnel liner is proposed as pre-
cast concrete segments with steel fibre reinforcement.
The evaluated condition is a long-term case where all anticipated hydrostatic loads, earth
loads, and in the case of the shaft liner, dead loads due to the weight of the liner, are fully
acting. The tunnel is expected to reach the permanent design condition relatively soon after
construction. The shaft, however, is assumed to reach long term condition when the dia-
phragm wall is completely inundated and the temporary shear keys along the intrados of the
diaphragm, necessary for construction, have deteriorated. In the case of the shaft, intermedi-
ate conditions should be considered in the final design.
The finite element program, PLAXIS 2D, was used to perform the analysis, though the
unique shaft conditions, including the extreme depth and large diameter, warranted independ-
ent consideration of buckling and buoyancy resistance.

5 SHAFT DESIGN

For long term loading, the final shaft liner must be designed to take the full hydrostatic water
pressure, its self-weight, and any forces applied at the connection of the final liner to the
invert slab. A diaphragm wall will be used for excavation support and will also resist the
ground pressures in long term as well as the forces transferred from the invert slab due to
uplift. These loads can be approximately calculated with an axisymmetric model shown in
Figure 6. Separate closed-form analyses are performed to check cylindrical buckling and
buoyancy of the structure.

5.1 Finite Element Model

The shaft construction begins with the installation of a diaphragm wall around the shaft per-
imeter to support the ground during excavation. After construction of diaphragm wall,

Figure 6. Elevation View of Proposed Jebel Ali Shaft and Corresponding Plaxis 2D analysis.

3555
sequential excavation within the shaft is performed along with dewatering of the shaft
followed by installation of the final liner using a top-down construction method. To
temporarily support the permanent liner, shear keys will be constructed at the interface
between the final liner and the diaphragm wall. Waterproofing is proposed to be placed
between the diaphragm wall and the final liner on the inside face of the diaphragm wall.
The following phases were used in the PLAXIS finite element analysis to simulate the
construction sequence:
1. Initial Phase: compute the initial ground stresses based on the ground properties and water
table depth
2. Install Diaphragm Wall with 5 meter embedment depth below the proposed invert slab
3. Excavate the shaft and dewater by wells placed outside of the shaft
4. Install Invert Slab
5. Cease dewatering: invert slab and diaphragm wall support both earth and water pressure
6. Install Final Liner
7. Allow water pressure to act on the final liner only (this stage of the analysis represents a
long term condition)
The final liner force results are presented in Figure 7 and show, despite the high radial force
and assuming a 20 cm eccentricity, the pin connection between final liner and invert slab
induces large bending moment along the vertical axis of the shaft walls. This is because the
invert slab acts as a restraint against the inward elastic movement of the shaft wall. These high
stresses for the deep shaft will require special consideration for the connection design and
reinforcement at the toe of the liner.
The Jebel Ali Shaft invert slab must resist a water head of approximately 125 meters, or
about 1200 kPa. Instead of designing the invert slab as a flexural member, the invert slab
can be designed to allow the uplift pressure to arch to the diaphragm wall. This phenomenon
was modeled in the axisymmetric finite element model. To ensure no tension could develop
in the invert slab when the outside dewatering was ceased, the concrete was modeled as a
Mohr-Coulomb material with zero tensile strength. As Figure 8 shows, the principle stresses
are compression and arch from the bottom of the invert slab at the center to the top corner,
and the minor stresses are showing no tension has developed in the topside of the invert
slab. By enabling the arching to develop, the required reinforcement can be significantly
reduced.

Figure 7. Jebel Ali Shaft Liner Force Results from the PLAXIS 2D Axisymmetric Model.

3556
Figure 8. Principle stress in Jebel Ali Shaft invert slab (compression is negative).

5.2 Elastic Buckling and Buoyancy

For cylinders with thin walls, defined as having a ratio of diameter to wall thickness (D/t
ratio) greater than 20, a collapse pressure can be calculated. The Jebel Ali Shaft can be con-
sidered as a thin walled cylinder as it has a D/t ratio of 63m/2.5m = 25.2. A conservative esti-
mate of the collapse pressure would be for an infinitely long cylinder with a Poisson’s ratio of
0, represented by equation 1 (Sturm, 1936).

2 t3
Pcr ¼ E 3 ð1Þ
ð1 
Þ D
2

At the deepest point, the shaft liner will support about 110 m of water, or about 1.08 MPa.
Using equation 1, the critical pressure is 4.75 MPa. The factor of safety for the Jebel Ali Shaft
against buckling is calculated to be 4.41.
For buoyancy, it is assumed the water table is at the ground surface, and the uplift resisting
forces include the weights of the shaft elements, and the soil friction along the entire depth of
the diaphragm wall. The uplift force acts on the bottom of the diaphragm wall and the
bottom of the invert slab. Using only shaft weight and soil friction (neglecting adhesion), an
adequate factor of safety against uplift was calculated for the Jebel Ali Shaft.

6 TUNNEL DESIGN

The tunnel liner will consist of one-pass, bolted, gasketed pre-cast reinforced concrete segmen-
tal liner. During detailed design phase, the analysis of the segmental lining cross-section will
consider soil-structure interaction and constructions sequences of the tunnel will be carried
out. The tunnel segments must also be designed to resist manufacturing loads, handling, stack-
ing and TBM jacking forces.
For the preliminary tunnel liner design, hand calculations along with finite element analysis
using PLAXIS 2D plane strain models were performed. The model consists of four soil layers
and a plate and interface element to define the tunnel liner. The construction sequence was

3557
Figure 9. PLAXIS 2D model for the 6.5 m diameter tunnel.

simulated by allowing for some level of in-situ stress relaxation before the liner was activated
in the model. The representative model domain is shown in Figure 9.

7 CONCLUSIONS

The rapidly growing Emirate of Dubai is undertaking the DSST project to transform the exist-
ing sewerage system from a pumped system to a fully gravity-based system. The project will
replace more than 100 main and subsidiary pump stations with two very deep pump stations
constructed at the Al Warsan and Jebel Ali treatment plants. The proposed gravity system
will have nearly 250 km of new link sewer, 75 km of deep tunnels (diameter between 3.5 m
and 6.5 m), and more than 40 shafts. The design forces on these structures may be quite high,
and analysis methodologies must be developed to ensure all temporary and long-term condi-
tions are considered. Some of the most challenging aspects of the project include accounting
for the risks associated with the very large terminal pump stations and ensuring the system is
operational for more than 100 years.

REFERENCES

Sturm, R.G. 1936. A Study of the Collapsing Pressure of Thin-Walled Cylinders. University of Illinois:
Urbana, IL

3558
 Tunnels and Underground Cities: Engineering and Innovation meet Archaeology,
Architecture and Art, Volume 7: Long and deep tunnels – Peila, Viggiani & Celestino (Eds)
© 2020 Taylor & Francis Group, London, ISBN 978-0-367-46872-9

The hugest and more complex belt conveyor system in the longest
tunnel under construction in the world: Brenner Base Tunnel

M. Bringiotti
GeoTunnel S.r.l., Genoa, Italy

S. Portner & V. Grasso

Marti Technik AG, Moosseedorf, Switzerland

E.R. Vitale
Ghella S.p.A., Roma, Italy

ABSTRACT: Marti has been working on the Italian side of the Brenner Base Tunnel project
since the beginning, starting not only with the bridge which overtakes the Highway, the Isarco
river, a National Road, and the Railway line, but also with the supply of the long belt for the
Aica-Mules pilot bore. These structures have been designed for those preliminary tunnels as
well as for an extensive use in the ongoing project, which is related to the excavation of the
main Brenner lines on the Italian side. The operative job sites which the system involves are
Mules 1, Genauen 2, Unterplattern and Hinterrigger. These are linked to the already existing
plants The system is able to handle contemporarily excavated material produced by n. 3 TBMs
and various D&B tunnels (ca. 2.100 t/h capacity) and also brings crushed material to the batch-
ing plant installed in the logistic knot cavern, in an Industry 4.0 full integrated optic.

1 INTRODUCTION

Since the beginning of 2017 the BTC consortium, formed by Companies Astaldi, Ghella, PAC,
Cogeis and Oberosler, has been involved in the construction of the longest underground railway
link in the world: the Brenner Base Tunnel, which forms the central part of the Munich-Verona
railway corridor.
The whole project consists of a straight railway tunnel, which reaches a length of about
55 km and connects Fortezza (Italy) to Innsbruck (Austria); next to Innsbruck, the tunnel will
interconnect with the existing railway bypass and will therefore reach a total extension of
about 64 km.
The tunnel configuration includes two main single-track tubes, which run parallel with a 70m
span between each other through most of the track, and linked every 333 m by cross passages
(Figure 1).
Between the two main tunnels and driven 12 meters below, an exploratory tunnel will be
excavated first. Its main purpose during the construction phase is to provide detailed informa-
tion about the rock mass Furthermore, its location allows important logistic support during
the construction of the main tunnel, for transportation of excavated material as well as that of
construction material. During the operations, it will be essentially used for the drainage of the
main tunnel.
The excavation process is divided into 2 blocks; the first one will be bored by n. 3 TBMs (n. 2
for the main tunnels and n. 1 for the exploratory tunnel, toward North). The second one will be
bored with traditional method, including mainly drill & blast in the competent material and spe-
cial drilling techniques in the faulty zones, toward South and in some areas toward North.

3559
Figure 1. Brenner Basis Tunnel scheme.

The Company CIPA S.p.A. is currently handling most of the tunnel excavations activities
in the lot named “Mules 2–3” with the French drilling partner Robodrill SA These mainly
consist of:
– Excavation and lining of the ADIT Tunnel at the Trens Emergency Stop and the Central
Tunnel, with a total length of approx. 4,500 m;
– Excavation and lining of the Exploratory Tunnel by traditional method, with a total length
of approx. 830 m;
– Excavation and lining of the Main Tunnel toward South, East tube and West tube in single
track section, with a total length of approx. 7,320 m;
– Excavation and lining of the Main Tunnel toward South East tube and West tube in
double track section, with a total length of approx. 2,590 m;
– Excavation and lining of 19 Connecting Side Tunnels linking the two main tubes, with a
total length of approx. 900 m.
All the material excavated in traditional method by the subcontractors Cipa, Europea92
and LSI, runs from the central cavern up to the surface, and sized by crusher, by means of
MT belt conveyors. The material is also divided in 3 different geological classes and moved
according to needs in 2 different depony areas.

2 GEOLOGICAL CONDITIONS

As anticipated, the Brenner Base Tunnel is the high-speed rail link between Italy and Austria
and therefore establishes a connection with North East Europe. It consists in a system of tun-
nels, which include two single-track tunnels, a service/exploratory tunnel that runs 12 m below
them and mostly parallel to the two main tunnels, bypasses between the two main tunnels
placed every 333 m, and 3 emergency stop stations located roughly 20 km apart from each
other. The bypasses and the emergency stops are the heart of the safety system for the oper-
ational phase of this line.
Average Overburden is, between 900 and 1.000 m, with the highest one about 1.800 m at
the border between Italy and Austria.
The excavation will be driven through various geological formations forming the eastern
Alpine Area. Most of these are metamorphic rocks, consisting of Phyllites (22%), Schist
(Carbonate Schist and Phyllite Schist, 41%) and Gneiss of various origin (14%). In addition,
there are important amounts of plutonic rock (Brixen Granite and Tonalite, 14%) and rocks
with various degrees of metamorphism, such as marble (9%).
Among the tectonic structures in Italy, we find the above mentioned Periadriatic Fault. As
mentioned, the unknown characteristics of the rock masses along this stretch determined the

3560
Figure 2. Geological distribution.

necessity of excavating the exploratory tunnel long before the main tunnels, in order to allow
the identification of lithological sequences of the various types of rock mass within this heavily
tectonized area, as well as a detailed study of their responses to excavation. These analysis were
subsequently used to adjust the consolidation and support measures both for the exploratory
and for the main tunnels. The excavation of the exploratory tunnel inside the Periadriatic Faults
allowed to determine the actual sequence of lithologies within this area.
A geological description is necessary in order to understand the chosen excavation process.,
This is divided in mechanized tunnelling for the good and competent rock types along the
exploratory and the main tunnels toward North, D&B for a limited distance toward South (2–
3 km) and in some minor short tunnels as well as several bypasses, and traditional method in the
faulty areas. It is also important to understand difficulties related to the excavation in the hard
granite as well the related problems linked to the abnormal wear and tear due to high abrasivity.
A summary of the encountered rock sequences is illustrated with Figure 2.

3 THE IDEA

The conveyor belt system allows a straightforward and efficient mean of transportation for
material going both in and out from the underground rock crusher, and in and out from the
underground concrete batching plants. It evolves while the project is carrying out, and is grad-
ually implemented in accordance to the work’s progress.
The conveyor system has been rationalized during the tender phase maintaining its high
potential, flexibility and capacity. The differentiation of the two belts in Aica (belt 1 and belt
2) allow each single band to be allocated with a certain type of material; therefore belt 1 is for
material A (good material for concrete) and belt 2 is for material B+C (semi-good and not
concrete-accepted material). The differentiated configuration avoids alternating different
material types with temporary stocking buffer on the same conveyor. However, the system
provides the possibility to switch the material types in Aica according to necessities. This kind
of realization allows an easier management of the whole system.
Such works may be identified in three major phases. This report is aimed to describe the
choices taken within the Construction Project and to list them congruently to the time-related
phases.
Without taking into consideration the specific calculations regarding the Executive Work
Program’s Space-Time Diagram, the 4 major phases may be described as follows:

3561
– Phase 1: excavation of the Exploratory Tunnel (CE) with the conventional excavation
method and the TBM assembly chamber; realization of the excavation of the first part of the
North Line Tunnels (GLN) by traditional system and excavation of the South Line Tunnels
always by traditional method (done by Cipa S.p.A. with its Drilling Partner Robodrill SA);
– Phase 2: continuation of the activities following the previous phase; realization of “in
cavern” assembly areas for the two TBMs, which will excavate the Line Tunnels (toward
North), excavation prosecution of the South Line Tunnels with drill & blast system (GLS)
and realization of the mechanized excavation of the CE Northwards.
– Phase 3: prosecution of the precedent activities, excavation of the North Line by mechan-
ized tunneling (GLN) and subsequent finishing job site works.
– Phase 4: main TBM tunnels secondary final lining. This also represents a complex activity,
which will take more than 1 year. A part of this job will be anticipated during TBM excava-
tion, deviating logistic and traffic from one tube to the parallel one.
The phases are integrated and shown in the complex lay out indicated in Figure 3.
All belts generally are connected to the five excavation fronts with the Logistic Knot (N). The
system is flexible, but actually in the current phase belts are directly used only for the North
bound tunnels (3 TBMs). South bound (conventional excavation) the material is transported by
dump truck up to the logistic Knot, where after resizing (by the crushing plant housed under-
ground within the cavern) it is transferred and transported by belt conveyor system.
After that, from Knot N:
– Material type A is carried towards the jobsites Mules and Genauen 2 (depony area reached
by a 180m long bridge belt conveyor – 300 ton/h – which crosses the A22 highway, the
Isarco river, the National road S.S. 12 and the existing railway line, Figure 4), it is stocked
and/or crushed outside and/or transported back to the underground batching plant, in
order to provide DB tunnels’ primary and secondary lining.

Figure 3. Belt lay out with cavern zoom (center left).

Figure 4. Highway A22 transfer belt to Genauen 2 intermediate waste dump.

3562
– Material type A/B+C is transported to Hinterrigger jobsite and is used to produce the pre-
cast elements and as pea-gravel for the TBM’s back filling. Concrete segments and pea
gravel will be then transported back into the tunnel with a dedicated rolling stock through
the AICA Tunnel.

4 BELT CONVEYOR SYSTEM DESCRIPTION

Up next is a description of the whole belt conveyor system taking into consideration the three
main project’s phases thus in order to provide detailed information about each specific instal-
lation. In particular it may be detailed by material class handled by each system (specifically
type A in case material is high quality, and types B+C when material quality is poorest, mater-
ial type A may be used to produce “working concrete” and sprayed concrete), its origin and
destination, its potential and capacity in ton/h

4.1 Phase 1
During Phase 1, the system includes:
– Existing belt no. 1, located at Mules – Aica stretch, with a 500 ton/h capacity.
– Belt no. 2, which has been be implemented on the same stretch and juxtaposed to belt
no. 1, thus resulting in a total capacity of 2.100 ton/h, having a capacity of 1600 ton/h.
Figure 5 shows the existing tunnel section and the related clearances of the two belts, as
well as the rail system with the booster section where the two runaways become one.
The new conveyor belt system planned for the Aica Tunnel has been designed in order to
allow transportation of material toward Hinterrigger in accordance the tunnel’s geometries.
These geometries grant a tunnel alignment with a reduced curve radius along the initial part
of the tunnel. During the tender phase, changes in the tunnel geometries were explicitly forbid-
den, disqualification being the penalty.
This has been quite an important technical constraint; having the already existing Aica tunnel
a very restricted section, contractually forbidden to provide suitable enlargements (niches) in
order to host the necessary booster stations, it hasn’t been possible to apply standard solutions.
MT and BTC designed compact sized boosters in order to fit in the limited spaces available,
leaving a sufficient train runaway.
Therefore, the two belts (no. 1 and no. 2) start at the Exploratory Tunnel chamber, arrive
at the Aica – Unterplattner jobsite and continue on the outside. Belt no. 2 is designed to trans-
port the incoming material from all excavation fronts (Main Line Tunnels, Exploratory
Tunnel, Trens Emergency Stop, FdE Access Tunnel and new Logistic Knot) from the cham-
ber’s Exploratory Tunnel toward Aica.
Already existing belt no. 3 connects the Unterplattner jobsite with Hinterrigger, and has a
capacity of 500 ton/h. Transition between belt no. 1 and belt no. 3 occurs through an existing
transfer chute located in the Unterplattner jobsite.
Belt no. 4 is built in juxtaposition to belt no. 3; its capacity is 1.600 ton/h. Belt no. 4 will be
of the same type as belt no. 2 is therefore an extension. The new belt is located on surface

Figure 5. Belts assembly in the Aica-Mules exploratory tunnel with the booster assembly details.

3563
from the Aica portal to the entrance of the Unterplattner tunnel, a part being on an elevated
structure and the other sitting on the ground.
The Figure 6 shows the belt conveyor portion standing in front of the Aica-Mules Explora-
tory Tunnel, together with its general layout out applied in the Unterplattner area.
Figure 7 shows the general layout out applied in the Hinterrigger area with the various dis-
tribution points.
During the excavation of the remaining tunnel stretches excavated with conventional
method in the Exploratory Tunnel up until the assembly chamber of the TBM, the transporta-
tion of the muck is carried by dump trucks until the TBM dismantling chamber. From the
existing chamber, material will be loaded on belt conveyors no. 1 and no. 2. Within said
tunnel, later belt no. 7 will be assembled.
Belt no. 8 runs through the whole Mules Adit tunnel toward Mules, as mentioned, with a
capacity of 600 ton/h. Just a note on CE material; muck coming from the excavation activities
within the Exploratory Tunnel will be transported to Hinterrigger by means of conveyor belt
system. The plant is able to separate CE material from A and B+C, but it will be moved only
to Hinterrigger

4.2 Aggregate handling within the Logistic Knot

All belts coming from the GL tunnel discharge material onto the 3 conveyor belts available at
the Logistic Knot; which are:
– One for material A towards Mules
– One for material A towards Hinterrigger
– One for material B+C towards Hinterrigger
The two belts towards Hinterrigger merge in order to transport material into the shaft that
connects GL and GE tunnels.
The related material will go down from the upper level, and will be unloaded either on belt
no. 1 or on belt no. 2, according to its classification and quantity.
Figure 8 (top right) are a close-up representation of belt no. 10 (the picture may be also
referred to belt no. 9), which unload on a lower level through the vertical shaft (on the left).
Belt no. 11 connects belt no. 13 with belt no. 8b. related to the transport of material type A,
this belt allows to carry material outwards, thanks to a belt plow switch, supplying the Mules
jobsite with good quality material thus in order to mix fresh concrete. Belt capacity is about
1.600 ton/h.

Figure 6. Unterplattern general lay out.

Figure 7. Hinterrigger general lay out.

3564
Figure 8. General logistic knot plan view lay out and details.

Figure 8 shows the general layout out of the logistic knot.

As previously reported, in the caverns, transferring devices are provided for each belt; this
is the so-called Switch system. It consists of a belt plow system, transferring material on differ-
ent belts according to classification. One belt is dedicated to material type A and one to mater-
ial type B+C. These cross the tunnel connecting the two chambers, reaching the connecting pit
area down to the Exploratory Tunnel chamber (km 10+4).
Belt no. 12 connects the East bound Tunnel with the Logistic Knot. In this phase the mater-
ial can be handled in the Movable Crusher located in the GLE assembling chamber. This
material converges in the Logistic Knot and heads to jobsites Hinterrigger or Mules. It comes
from the GLE, the GLO and the CE excavations. The belt’s capacity is 1.000 ton/h.
Belt no. 13 connects the West bound Tunnel with the Logistic Knot and is implemented
during this phase. Its capacity is 1.000 ton/h.
Belt no. 23 allows the handled material (crushed and screened in Mules) to return through
the Mules Adit. This material is used to produce concrete with the batching plant available in
the Logistic Chamber. This belt’s capacity is 350 ton/h. Figure 9 shows the related
installation.
Belt no. 24 carries material handled by the Movable Crushing Plant to be transferred on
belt no. 12. The material then reaches the Logistic Knot. The belt’s capacity is 200 ton/h;
during this phase, it receives the excavated material of the GLE, GLO and CE. From there,

Figure 9. Batching plant installed in cavern.

3565
material reaches the connecting shaft, goes down to the chamber in the Exploratory Tunnel
(km 10+4), and then transported to Hinterrigger (classes A and B + C) or in Mules (class A).
Belt no. 25; as previously mentioned, handles the outgoing material from Mules Adit to the
surface. The exiting belt no. 25 reaches the south part of the jobsite area, next to the first transfer
chute. Through a specific Switch, from there, the muck can either be discharged on ground, on a
temporary stock pile, or it can befurtherly transported by the secondary belts (i.e. branch 1 and
branch 0) to a second transfer chute toward Genauen 2. Belt no. 25 has a capacity of 600 ton/h.
The peculiarity of this system is the 2 level belt; in which the upper part is able to move
material from the cavern, through Mules Adit to the surface, and the lower part handles
crushed and sized aggregates (from the surface close to Mules Adit) to the underground batch-
ing plant. It has been specially designed in order to handle the various and different demands,
in terms of aggregate size and mucking capacity).
Belt no. 26 handles material to Belt no. 23, allowing material handling back to the under-
ground batching plant. The belt’s capacity is about 350 ton/h. It handles the crushed material
resized in the crushing plant by means of vibrating dosing transfer chutes.
Figure 10 shows the general plant layout in the Mules site.
Belts no. 27 and no. 28 are in the Hinterrigger jobsite, down from Belt no. 4. They are
meant to handle material types B+C; Belt no. 28 is a slewing conveyor, suitably designed for
stock piling. The capacity of both is 1.600 ton/h each.
Belt no. 29 handles material from Belt no. 3 and serves, as the mentioned, for type A mater-
ial stock piling. Belt no. 29 has a maximum capacity of 500 ton/h.

4.3 Phase 2
All of the mentioned belts are installed during Phase 2.
The excavation of the Exploratory Tunnel with conventional method up to the TBM dis-
mantling chamber is conducted during phase 1. Transportation of muck is preferably handled
with conveyor belts when the supply time is favorable. In case of delay, the material is handled
by means of dump truck. During said phase Belt no. 7 handles the mechanized excavation
material to the Logistic Knot, from where it is transported to Hinterrigger as material type A
or B+C. Belt no. 7 has a capacity of 500 ton/h.
All the material excavated by D&B is moved to the underground crushing plant, and then
transported via belt conveyor on surface (Hinterrigger or Mules).

4.4 Phase 3
In main phase 3 the previously installed belts are continuously extended in accordance to the
mechanized excavation advancement rate. together with the belts for the mechanized excava-
tion towards North.
Belt no. 7 is extended consequently to the advancement of the mechanized excavation in the
Exploratory Tunnel. The extension belt station is located within the TBM launch chamber
(km 1+29). From there, the excavation material is carried to the chamber at km 10+47 with a
fixed belt. Muck is transported from the TBM belts in the GLN to 100mc transfer chutes in
order to manage potential minor stops without interrupting the excavation activities as well as
to level off the flow of material coming from the node.

Figure 10. General plan overview site Mules with external crushing plant and extraction pits.

3566