CIRIA C671
London 2009
Tunnels:
inspection, assessment
and maintenance
L McKibbins
Mott MacDonald Ltd
R Elmer
Golder Associates (UK) Ltd
K Roberts
Atkins
Classic House, 174–180 Old Street, London EC1V 9BP
TEL: +44 (0)20 7549 3300 FAX: +44 (0)20 7253 0523
EMAIL: enquiries@ciria.org WEBSITE: www.ciria.org
Tunnels: inspection, assessment and maintenance
McKibbins, L, Elmer, R, Roberts, K
CIRIA
CIRIA C671
© CIRIA 2009
RP712
ISBN: 978-086017-671-8
British Library Cataloguing in Publication Data
A catalogue record is available for this book from the British Library.
Keywords
Transport infrastructure, facilities management, health and safety, knowledge
management, materials, materials technology, regulation, site management, sustainable
construction, whole-life costing
Reader interest
Classification
Asset management, civil
infrastructure, tunnel
condition appraisal,
inspection, maintenance
and repair
AVAILABILITY Unrestricted
CONTENT
Advice/guidance document
STATUS
Committee-guided
USER
Asset owners, managers, designers,
contractors, tunnel and civil engineers
Published by CIRIA, Classic house, 174-180 Old Street, London, EC1V 9BP
This publication is designed to provide accurate and authoritative information on the subject matter covered. It is
sold and/or distributed with the understanding that neither the authors nor the publisher is thereby engaged in
rendering a specific legal or any other professional service. While every effort has been made to ensure the accuracy
and completeness of the publication, no warranty or fitness is provided or implied, and the authors and publisher
shall have neither liability nor responsibility to any person or entity with respect to any loss or damage arising from
its use.
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means,
including photocopying and recording, without the written permission of the copyright-holder, application for
which should be addressed to the publisher. Such written permission must also be obtained before any part of this
publication is stored in a retrieval system of any nature.
If you would like to reproduce any of the figures, text or technical information from this or any other CIRIA
publication for use in other documents or publications, please contact the Publishing Department for more details
on copyright terms and charges at: publishing@ciria.org or tel: 020 7549 3300.
ii
Summary
This guide provides infrastructure owners, consulting engineers, contractors and
maintenance managers with guidance on the management, condition appraisal,
maintenance and repair of the structural elements of existing infrastructure tunnels,
focusing primarily on older infrastructure and certain tunnel types. It is based on a
detailed review of published literature and infrastructure owners’ procedures,
consultation with experts and practitioners within the field, and case studies illustrating a
wide variety of tunnel maintenance issues, repairs and incidences.
The purpose of the guide is to:
present current good practice
provide a guide for routine management
recommend assessment, maintenance and repair strategies to give best value for
money
help knowledge sharing.
Tunnels remain a vital part of the transport and services infrastructure in the UK and
other countries. However, they are facing many challenges associated with their extended
period in service, changing requirements and the continuing subsurface development of
modern cities. To ensure the continued efficient use of these assets in the future it is
necessary to manage and maintain their structural elements carefully, with due regard to,
and an adequate understanding of, their special characteristics and needs. In several
important ways these are distinct from those of more modern structures, and the effective
stewardship of older infrastructure tunnels requires some specialist knowledge and a
particular approach. This guide provides information and guidance to assist those
responsible for this task in achieving their aims.
The book is divided as follows:
Chapter 1: introduction and general background information on the document, including
advice on how and where to find help.
Chapter 2: an overview of tunnel construction history, techniques and materials,
behaviour and performance, which is intended to be particularly useful to readers with
less experience in this type of structure.
Chapter 3: advice on tunnel management, statutory obligations, health and safety and
environmental considerations, and strategies for condition assessment and maintenance
planning.
Chapter 4: condition appraisal, including information and guidance on carrying out
inspections, investigations, monitoring and structural assessment of tunnels.
Chapter 5: the selection and enforcement of structural maintenance and repair
techniques.
Chapter 6: advice on dealing with water ingress in tunnels where this is problematic.
CIRIA C671 • Tunnels 2009
iii
Chapter 7: summary of recommendations for good practice, and discussion of future
research and development needs.
Appendixes A1 to A7 give more detailed information to support Chapters 1 to 7.
Appendix A1 includes case studies illustrating the practical nature of the issues discussed
in Chapters 1 to 7.
iv
Acknowledgments
Authors
Leo D McKibbins BSc (Hons) MSc CEng MIMMM FGS
Leo is a principal engineer with the Special Services unit of Mott MacDonald Ltd
consulting engineers, currently working on the delivery of unified specifications for the
Crossrail project. Originally trained as a geologist at University College London he
entered the field of engineering after an MSc in Geomaterials at Queen Mary College
(University of London). He has over 12 years experience in the investigation, condition
assessment and design of remedial measures for a wide variety of civil engineering
structures including tunnels. He has particular expertise in the specification, assessment
and remediation of construction materials and dealing with the causes and effects of water
ingress into tunnels.
Richard Elmer BSc (Hons) MSc CEng MIMMM MCSM
Richard is a senior geotechnical specialist at Golder Associates (UK) Ltd. He is a
geotechnical advisor with 20 years experience in investigation, assessment, design and
construction supervision of underground works in rock and soil. His expertise includes
rock mass characterisation, design of ground support measures and management of
geotechnical assets, particularly relating to rail infrastructure. Educated at Southampton
University (BSc Geology) and Camborne School of Mines (MSc Mining Geology), Richard
has worked on tunnelling projects worldwide including those in Australia, China, Turkey,
Malaysia and Europe.
Kevin Roberts BSc (Hons) CEng MIMMM
Kevin is a principal engineer for Atkins. He is currently seconded onto the Crossrail
project within the Arup-Atkins Framework Design Consultancy for Crossrail Ltd. Before
this he was seconded into BCV & SSL Metronet working on London Underground PPP
contract for civil works as the deep tube tunnels inspection and assessment delivery
manager. In this role he advised the civils maintenance teams responsible for the LU
tunnels on issues of tunnel maintenance. He has close to 25 years experience in ground
investigation, assessment, design and construction supervision of earth structures and
tunnels in rock and soil. Educated at Surrey University (BSc Civil Engineering), Kevin has
worked on a variety of geotechnical and tunnelling projects in the UK, Africa, North
America, Hong Kong, China and Europe, including long-term overseas placements.
CIRIA C671 • Tunnels 2009
v
Project Steering Group
Following CIRIA’s usual practice, the research project was guided by a steering group,
which included:
Brian Bell (chairman)
Network Rail
Simon Brightwell
Aperio Ltd
Geoff Edgell (Prof)*
CERAM
Robert Ford
Highways Agency
Peter Harris*
Donaldson Associates
Tim Hughes (Prof)
Cardiff University
Gerald Kerr
Health and Safety Executive
Jack Knight
Scott Wilson (formerly with Charles Haswell & Ptnrs)
Donald Lamont (Dr)
Channel Tunnel Safety Authority
John Lane
RSSB (Rail Safety and Standards Board)
Andrew Lawrence
Arup
Jim Moriarty
London Underground
Ganga Prakhya (Dr)
Sir Robert McAlpine
Chris Reynard
British Waterways
Tony Salmon
London Underground
Tim Simpson
Atkins and Metronet
Colin Sims
Network Rail
Len Smith
Transport for London
Brian Thomas
Transport for London
Peter Wright*
Tube Lines
* Corresponding members
Funders
RSSB (Rail Safety and Standards Board)
Health and Safety Executive
London Underground
Tube Lines
Metronet
Network Rail
Scottish Water
Research contractor
John Perry (Dr) (project director) Mott MacDonald
vi
Leo McKibbins (lead author)
Mott MacDonald
Richard Elmer (author)
Golders (formerly with Mott MacDonald)
Kevin Roberts (author)
Atkins
The case studies and adaptations of technical papers were written and contributed by the
authorship team and members of the Project Steering Group, together with:
Robert Hills
Donaldson Associates
David Jarvis
Owen Williams Railways
Chris Levy
Mott MacDonald
Chris W Rees
May Gurney
Martin Roach
Metronet
Danny Swannell
Owen Williams Railways
Ian Wilson
Network Rail
Appendix A7 worked examples by:
Giuseppe Simonelli
Mott MacDonald
CIRIA Project managers
Project managed and directed by Chris Chiverrell. The project proposal was developed by
Dr Andrew Pitchford and Natalia Brodie-Greer (née Brodie-Hubbard).
In memoriam
This publication is dedicated to the memory of Jack Knight who tragically died before its
completion. As a member of the project steering group he gave valuable input to many
chapters, constructive comments on initial drafts and willingly gave to the project team a
download of his considerable experience and knowledge of all things “tunnelling” for the
benefit of the final publication and its readers. He will be sadly missed by the tunnelling
community and those fortunate enough to have worked with him.
CIRIA C671 • Tunnels 2009
vii
Contents
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .iii
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .v
Boxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xiv
Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xiv
Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xx
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxii
Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxix
1
2
Introduction and background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
1.1
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
1.2
Purpose and scope of work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
1.3
Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
1.4
Issues dealt with in this guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
1.5
How to use this guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
Construction and behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6
2.1
Tunnel construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6
2.1.1
History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6
2.1.2
Construction method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8
2.1.2.1 Cut-and-cover tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
2.1.2.2 Bored tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
2.2
2.3
2.1.3
Excavation methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
2.1.4
Stress redistribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17
2.1.5
Ground failure mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19
2.1.6
Temporary support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20
2.1.7
Primary and secondary linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21
Construction shafts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
2.2.1
Shaft construction techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
2.2.2
Shaft eyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
2.2.3
Closed shafts (blind shafts) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
Masonry linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
2.3.1
Lining profile, thickness and quality . . . . . . . . . . . . . . . . . . . . . . . . .27
2.3.2
Lining construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28
2.3.3
Inverts and drainage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
2.3.4
Brickwork bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
2.3.5
Construction joints in brickwork . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
2.3.6
Masonry materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
2.3.6.1 Mortar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31
2.3.6.2 Stone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34
2.3.6.3 Brick . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34
2.3.7
viii
Structural behaviour of masonry . . . . . . . . . . . . . . . . . . . . . . . . . . . .36
2.4
Metal linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37
2.5
2.4.1
Cast iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37
2.4.2
Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41
Pre-cast concrete linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41
2.5.1
Lining forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41
2.5.1.1 Bolted pre-cast concrete lining . . . . . . . . . . . . . . . . . . . . . . .42
2.5.1.2 Expanded concrete linings . . . . . . . . . . . . . . . . . . . . . . . . . .43
2.5.2
2.6
Casting methods and reinforcement . . . . . . . . . . . . . . . . . . . . . . . . .43
Tunnel performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .44
2.6.1
Structural deterioration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .44
2.6.2
Materials deterioration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47
2.6.2.1 Masonry linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47
2.6.2.2 Metal linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49
2.6.2.3 Deterioration of concrete linings . . . . . . . . . . . . . . . . . . . . .51
2.6.2.4 Deterioration of unlined tunnel support . . . . . . . . . . . . . . .55
2.6.3
Effect of fire on tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55
2.6.3.1 The influence of structural form . . . . . . . . . . . . . . . . . . . . .56
2.6.3.2 Concrete and masonry linings . . . . . . . . . . . . . . . . . . . . . . .56
2.6.3.3 Metallic linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58
2.7
Shaft performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .60
2.7.1
3
Effect at ground level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61
Tunnel asset management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .63
3.1
The need for tunnel management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .63
3.2
Special requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .63
3.3
Loss of performance and its consequences . . . . . . . . . . . . . . . . . . . . . . . . . . .64
3.4.1
Appraisal of current condition, performance and serviceability . . . .66
3.4.2
Maintenance strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69
3.4.2.1 Planned maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69
3.4.2.2 Reactive maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70
3.4.3
Maintenance planning and prioritisation . . . . . . . . . . . . . . . . . . . . .70
3.4.3.1 Assessment of tunnel criticality . . . . . . . . . . . . . . . . . . . . . . .70
3.4.3.2 Effect of maintenance strategy on tunnel performance . . .71
3.4.3.3 Effect of maintenance strategy on inspection intervals . . . .72
3.4.3.4 Optimising planned maintenance strategies . . . . . . . . . . . .72
3.4.3.5 Deferral of maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . .73
3.4.3.6 Minimising disruption from tunnel maintenance . . . . . . . .73
3.5
3.6
Tunnel management procedures and tools . . . . . . . . . . . . . . . . . . . . . . . . . . .73
3.5.1
Tunnel information requirements . . . . . . . . . . . . . . . . . . . . . . . . . . .73
3.5.2
Tunnel management systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75
3.5.3
Tunnel identification and referencing systems . . . . . . . . . . . . . . . . .75
3.5.4
Managing risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76
3.5.5
Whole-life asset costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77
Health and safety and environmental management . . . . . . . . . . . . . . . . . . . .78
3.6.1
CIRIA C671 • Tunnels 2009
Health and safety management . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78
ix
3.6.2
Competence and training of staff . . . . . . . . . . . . . . . . . . . . . . . . . . . .80
3.6.3
Heritage conservation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
3.6.4
Environmental conservation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
3.6.4.1 Conservation bodies and environmental legislation . . . . . .81
3.6.4.2 Wildlife conservation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83
3.6.4.3 Managing environmental impact . . . . . . . . . . . . . . . . . . . . .83
3.7
Tunnel operational safety and fire risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87
3.8
Management of tunnel shafts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
3.9
4
3.8.1
Shaft identification and location . . . . . . . . . . . . . . . . . . . . . . . . . . . .90
3.8.2
Maintaining shafts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90
3.8.3
Development of land above shafts . . . . . . . . . . . . . . . . . . . . . . . . . . .91
Management of closed and disused tunnels . . . . . . . . . . . . . . . . . . . . . . . . . .91
Condition appraisal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .93
4.1
Types and sources of information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .93
4.2
Desk studies and existing information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .94
4.3
Visual inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95
4.4
4.5
4.6
4.7
4.8
4.9
4.3.1
Advantages and limitations of visual inspection . . . . . . . . . . . . . . . .95
4.3.2
Types of visual inspection and inspection intervals . . . . . . . . . . . . . .96
4.3.3
Competence of inspection staff . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99
4.3.4
Visual inspection procedures and techniques . . . . . . . . . . . . . . . . . .99
4.3.5
Optimising inspection procedures and results . . . . . . . . . . . . . . . .100
Tunnel investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102
4.4.1
Objectives of tunnel investigation . . . . . . . . . . . . . . . . . . . . . . . . . .102
4.4.2
Investigation strategy and reliability of results . . . . . . . . . . . . . . . .102
4.4.3
Techniques for tunnel investigation . . . . . . . . . . . . . . . . . . . . . . . . .104
4.4.4
Selection of investigation techniques . . . . . . . . . . . . . . . . . . . . . . . .105
4.4.5
Optimising tunnel investigations and results . . . . . . . . . . . . . . . . .107
Tunnel monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109
4.5.1
Objectives of tunnel monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . .109
4.5.2
Monitoring instrumentation and techniques . . . . . . . . . . . . . . . . . .110
4.5.3
Selection and design of monitoring systems . . . . . . . . . . . . . . . . . .110
Preparing for inspections and investigations . . . . . . . . . . . . . . . . . . . . . . . .112
4.6.1
Risk assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .112
4.6.2
Access, programming and timing . . . . . . . . . . . . . . . . . . . . . . . . . .113
Location and inspection of tunnel shafts . . . . . . . . . . . . . . . . . . . . . . . . . . . .114
4.7.1
Detection and location of unknown hidden shafts . . . . . . . . . . . . .114
4.7.2
Shaft inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114
Interpretation of inspection and investigation data . . . . . . . . . . . . . . . . . . .116
4.8.1
The importance of good interpretation . . . . . . . . . . . . . . . . . . . . . .116
4.8.2
Considerations for interpretation . . . . . . . . . . . . . . . . . . . . . . . . . .116
Structural assessment of tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .122
4.9.1
Assessment in principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123
4.9.1.1 Qualitative assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . .124
4.9.1.2 Analytical assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125
x
4.9.1.3 Cast iron and steel linings . . . . . . . . . . . . . . . . . . . . . . . . .127
4.9.2
Multi-level assessment procedure . . . . . . . . . . . . . . . . . . . . . . . . . .129
4.9.3
Structural defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132
4.10 Reporting on and interpreting asset condition . . . . . . . . . . . . . . . . . . . . . . .133
5
4.10.1
Reporting inspection and investigation results . . . . . . . . . . . . . . . .133
4.10.2
Initial evaluation and identification of sensitive structures . . . . . . .134
4.10.3
Interpretation of results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .135
4.10.4
Condition ratings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .136
Selecting and carrying out works on tunnels and shafts . . . . . . . . . . . . . . . . . . . . . .138
5.1
Selection, planning and preparation for works . . . . . . . . . . . . . . . . . . . . . . .138
5.1.1
Planning and programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .138
5.1.2
Managing risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139
5.1.3
Selection of techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .140
5.1.4
Method statements and risk assessments . . . . . . . . . . . . . . . . . . . . .141
5.1.5
Completion of works and beyond . . . . . . . . . . . . . . . . . . . . . . . . . .142
5.2
Tunnel repair measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .142
5.3
Routine (preventative) maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144
5.3.1
Tunnel cleaning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145
5.3.2
Drainage maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .148
5.3.3
Management and removal of vegetation . . . . . . . . . . . . . . . . . . . . .149
5.3.4
Repointing of masonry-lined tunnels . . . . . . . . . . . . . . . . . . . . . . .149
5.3.5
Application of protective coatings . . . . . . . . . . . . . . . . . . . . . . . . . .151
5.3.5.1 Metal tunnel linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .151
5.3.5.2 Concrete, brick and masonry linings . . . . . . . . . . . . . . . . .152
5.4
Remedial repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .152
5.4.1
Masonry linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .153
5.4.1.1 Patch repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .156
5.4.1.2 Crack repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .159
5.4.1.3 Ring separation repair . . . . . . . . . . . . . . . . . . . . . . . . . . . .160
5.4.2
Metal tunnel linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161
5.4.2.1 Cast iron lining repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . .162
5.4.2.2 Wrought iron and steel repairs . . . . . . . . . . . . . . . . . . . . .166
5.4.2.3 Alternative repair solutions . . . . . . . . . . . . . . . . . . . . . . . .167
5.4.3
Concrete tunnel linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .168
5.4.3.1 Concrete repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169
5.4.3.2 Other types of treatment and repair . . . . . . . . . . . . . . . . .173
5.5
Strengthening and structural improvement . . . . . . . . . . . . . . . . . . . . . . . . .175
5.5.1
Replacement and strengthening existing tunnel linings . . . . . . . . .175
5.5.1.1 Replacement of tunnel lining . . . . . . . . . . . . . . . . . . . . . . .177
5.5.1.2 Tunnel strengthening . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178
5.5.1.3 Replacement of structural elements . . . . . . . . . . . . . . . . . .180
5.5.2
Underpinning of masonry-lined tunnels . . . . . . . . . . . . . . . . . . . . .183
5.5.2.1 Continuous strip foundations . . . . . . . . . . . . . . . . . . . . . . .185
5.5.2.2 Piling methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186
CIRIA C671 • Tunnels 2009
xi
5.6
5.5.3
Invert repair (strengthening/replacement) . . . . . . . . . . . . . . . . . . .188
5.5.4
Rock stabilisation: unlined tunnels . . . . . . . . . . . . . . . . . . . . . . . . .189
Treatment of tunnel shafts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .194
5.6.1
Access for working . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .195
5.6.2
Shaft lining maintenance, repair and decommissioning . . . . . . . . .196
5.6.2.1 Deteriorating cross-members . . . . . . . . . . . . . . . . . . . . . . .196
5.6.2.2 Water ingress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .196
5.6.2.3 Shaft lining stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .197
5.6.2.4 Relining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .197
6
5.6.3
Filling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .197
5.6.4
Grouting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .198
5.6.5
Capping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .199
Water ingress and control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201
6.1
General considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201
6.2
Passive measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .204
6.3
6.2.1
Drip trays (including guttering and down pipes) . . . . . . . . . . . . . .205
6.2.2
Secondary lining systems (or drainage membrane) . . . . . . . . . . . .205
6.2.3
Weep holes (and pipes) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .207
6.2.4
Channelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .207
Active measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .207
6.3.1
Caulking, bolt holes, grummets and grout holes . . . . . . . . . . . . . .208
6.3.1.1 Caulking joints in segmental linings . . . . . . . . . . . . . . . . .208
6.3.1.2 Sealing bolt holes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211
6.3.1.3 Sealing grout holes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211
6.3.2
6.4
Surface sealing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211
Grouting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .212
6.4.1
Grouting technique selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .212
6.4.1.1 Cementitious grouts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .215
6.4.1.2 Chemical (resin) grouts . . . . . . . . . . . . . . . . . . . . . . . . . . .216
6.4.2
Grouting masonry-lined tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . .219
6.4.2.1 Grouting procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . .221
6.5
7
6.4.3
Metal or pre-cast concrete segmental lined tunnels . . . . . . . . . . . .224
6.4.4
Concrete-lined tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .225
6.4.5
Void grouting behind linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .225
Alternative measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .226
6.5.1
Groundwater lowering (dewatering using well-points) . . . . . . . . . .226
6.5.2
Electro-osmosis (dewatering) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .227
Recommendations and future needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .229
7.1
Recommendations for good practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .229
7.2
Areas requiring further research and future needs . . . . . . . . . . . . . . . . . . .231
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .233
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .246
Regulations and standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .247
xii
Regulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .247
Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .247
Appendices
A1
Case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .249
Case study 1
Remedial treatments to Folkestone Rail tunnels . . . . . . . . . . . . . . . . . .250
Case study 2
Investigation and treatment of ground instability and water ingress at
Blackheath tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .267
Case study 3
Strengthening of Brunel’s Thames Tunnel . . . . . . . . . . . . . . . . . . . . . .281
Case study 4
Standedge North Railway Tunnel: investigations and design of major
remedial works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .288
Case study 5
Geophysical surveying to identify hidden shafts . . . . . . . . . . . . . . . . . .304
Case study 6
Relining of Blisworth Tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .306
Case study 7
Leak sealing and rehabilitation of Sewer Tunnels . . . . . . . . . . . . . . . . .310
Case study 8
Management of a disused and deteriorated rail tunnel . . . . . . . . . . . . .316
Case study 9
Reconstruction of an underground line tunnel at Old Street . . . . . . . .321
Case study 10
Inspection and maintenance of a raw water tunnel . . . . . . . . . . . . . . . .328
Case study 11
Investigation and construction joint mapping of Haymarket Tunnels . .335
Case study 12
Relining of Sugar Loaf Tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .338
Case study 13
Structural monitoring strategy for the Channel Tunnel . . . . . . . . . . . .340
Case study 14
Invert reconstruction and other structural repairs to Netherton
Canal Tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .346
Case study 15
Piling adjacent to deep and near-surface tunnels in London . . . . . . . .350
Case study 16
Predicting and monitoring the effects of adjacent construction on
masonry-lined tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .354
Case study 17
A feasibility-based risk matrix for option selection . . . . . . . . . . . . . . . . .359
Case study 18
Tunnel fires, collapses and other serious incidents . . . . . . . . . . . . . . . .369
A2
Sources of existing information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .385
A2.1 Sources of historical information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .385
A2.2 Sources of geological and hydrogeological information . . . . . . . . . . . . . . . .387
A2.3 Aerial photographs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .389
A2.4 Utilities and services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .389
A2.5 Walkover survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .389
A3
Visual inspection procedures and observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .393
A3.1 Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .393
A3.2 Observation and recording . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .393
A4
Inspection, investigation and monitoring techniques . . . . . . . . . . . . . . . . . . . . . . . . .403
A4.1 Inspection, mapping and simple on-site tests . . . . . . . . . . . . . . . . . . . . . . . .404
A4.2 Sampling and testing techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .410
A4.3 Specialist non-destructive investigation techniques . . . . . . . . . . . . . . . . . . .419
A4.4 Techniques for monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .429
A5
Detection and location of hidden shafts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .433
A5.1 Multi-phase approach to shaft location . . . . . . . . . . . . . . . . . . . . . . . . . . . . .433
A6
Investigation and assessment of unlined tunnels and shafts . . . . . . . . . . . . . . . . . .442
A6.1 Desk study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .442
CIRIA C671 • Tunnels 2009
xiii
A6.2 Reconnaissance visit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .442
A6.3 Detailed survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .442
A6.4 Scan line mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .443
A6.5 Rock mass mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .444
A6.6 Rock mass classification systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .445
A7
Guidance on structural assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .447
A7.1 Limit state assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .447
A7.2 Assessment principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .448
A7.3 Worked examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .452
Worked example 1: Cast iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .453
Worked example 2: Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .461
Worked example 3: Masonry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .469
Boxes
Box 3.1
Dealing with bats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84
Box 5.1
Assessing the nature of a crack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .160
Figures
xiv
Figure 2.1
Typical tunnel profiles for UK railways . . . . . . . . . . . . . . . . . . . . . . . . . . .9
Figure 2.2
Typical section through a C&C railway tunnel . . . . . . . . . . . . . . . . . . . . .11
Figure 2.3
Typical cross-section of a bored railway tunnel . . . . . . . . . . . . . . . . . . . .11
Figure 2.4
Typical excavation sequence for canal tunnels . . . . . . . . . . . . . . . . . . . . .13
Figure 2.5
Typical cross-sections and dimensions of narrow and wide canal tunnels . .13
Figure 2.6
Typical annular infill for lined tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . .14
Figure 2.7
Hand-excavation and spoil removal from the top-heading of a tunnel,
using the English method of construction popular in the 19th century,
showing temporary timber supports. A completed bottom heading is
also visible . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
Figure 2.8
Stress redistribution around a circular tunnel . . . . . . . . . . . . . . . . . . . . .18
Figure 2.9
Stress concentrations around a non-circular opening . . . . . . . . . . . . . . .19
Figure 2.10
Temporary support formwork and replacement with a multi-ring
brickwork lining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20
Figure 2.11
Typical railway tunnel shaft construction details . . . . . . . . . . . . . . . . . . .23
Figure 2.12
Temporary support detail at shaft eye . . . . . . . . . . . . . . . . . . . . . . . . . . .24
Figure 2.13
Typical rail tunnel shaft eye construction details . . . . . . . . . . . . . . . . . . .24
Figure 2.14
Possible states of construction and ventilation shafts . . . . . . . . . . . . . . . .25
Figure 2.15
Examples of open and closed shafts in a brick-lined tunnel. The closed
shaft, on the right, has been capped off just above the eye so is easily
visible, but this is often not the case . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
Figure 2.16
Rail tunnel (Clifton Hall tunnel) with multi-ring masonry lining and
structural invert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27
Figure 2.17
An example of one method of brickwork bonding for masonry arches . .30
Figure 2.18
A construction joint picked out by its shadow using low-angle
lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31
Figure 2.19
Two views of construction joints: clearly visible joint where dog-toothing
is absent (a) and joint is more difficult to spot, but is marked by subtle
irregularity of brickwork and slightly wider vertically aligned joints (b) . . . .31
Figure 2.20
Original drawings from Rotherhithe Tunnel (1908) with bolted grey
cast iron sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .38
Figure 2.21
Typical construction and joint details for a London Underground
bolted grey iron lining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39
Figure 2.22
Bolted cast iron lining with water seepage at joint . . . . . . . . . . . . . . . . . .40
Figure 2.23
Example of a 3D FE model of a cast iron lining incorporating a vertical
crack in the sidewall, shown in white . . . . . . . . . . . . . . . . . . . . . . . . . . . .41
Figure 2.24
Typical bolted pre-cast concrete lining . . . . . . . . . . . . . . . . . . . . . . . . . . .42
Figure 2.25
Section through Potters Bar tunnel expanded pre-cast concrete lining . .43
Figure 2.26
Typical forms of lining deformation in brick-lined tunnels . . . . . . . . . . .47
Figure 2.27
Deep spalling of soft red brick near to a tunnel portal caused by
freeze-thaw damage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49
Figure 2.28
Collapse of part of masonry lining at the waterline in a canal tunnel
due to a combination of deteriorative mechanisms (moisture saturation
and leaching, salt weathering and freeze-thaw) . . . . . . . . . . . . . . . . . . . .49
Figure 2.29
Corroded cast iron lining in Aldwych shaft . . . . . . . . . . . . . . . . . . . . . . .51
Figure 2.30
Acid attack of tunnel lining at Bond Street, London Underground . . . .52
Figure 2.31
Concrete spalling from segmental lining sections . . . . . . . . . . . . . . . . . .54
Figure 2.32
Gasket deterioration of circle joints and around key block in concrete
segmentally lined tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .54
Figure 2.33
Simplified method for determining the zone of influence of tunnels (a)
and shafts only (b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61
Figure 3.1
Information required for an assessment of tunnel serviceability . . . . . . .67
Figure 3.2
Outline process for assessing and maintaining serviceability of tunnels . . .68
Figure 3.3
Relationship of serviceable life, performance and maintenance
interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71
Figure 3.4
Proprietary bat brick artificial roost and suggested locations for
installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84
Figure 3.5
Results of collapse of material into an incompletely filled shaft (1909) . . .89
Figure 4.1
Two views down a tunnel shaft. Water ingress and the presence of
shaft furniture can obstruct inspection and other work in shafts and
should be taken into account when planning access . . . . . . . . . . . . . . .115
Figure 4.2
Diagram illustrating the application of the limit analysis method to a
masonry tunnel lining in principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . .127
Figure 4.3
Assessment of cast iron linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .128
Figure 4.4
Exploiting the symmetry conditions to avoid boundary condition
problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .130
Figure 5.1
Thick accumulation of soot on a rail tunnel crown . . . . . . . . . . . . . . . .146
Figure 5.2
Guttering and downpipe system that has been installed to channel
water ingress from a tunnel wall into the invert drain, but has not
been maintained so that it is no longer effective . . . . . . . . . . . . . . . . . .149
Figure 5.3
Several visibly distinct phases of patch repair to an old rail tunnel
lining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .156
Figure 5.4
Typical patch repair to two courses of brickwork (a) with pinning
detail (b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .157
Figure 5.5
Carrying out patch repairs using temporary supports . . . . . . . . . . . . .158
Figure 5.6
Installation of stitching bars along a crack . . . . . . . . . . . . . . . . . . . . . . .159
Figure 5.7
Brick lining pinning for grouting ring separation . . . . . . . . . . . . . . . . .161
Figure 5.8
Typical example of a damaged circle joint flange of a bolted cast iron
lining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162
CIRIA C671 • Tunnels 2009
xv
xvi
Figure 5.9
Typical example of flange strapping in cast iron lined tunnels . . . . . . .162
Figure 5.10
Typical example of plate repair to cast iron tunnel segment pan . . . . .163
Figure 5.11
Metal stitching process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .165
Figure 5.12
Example of metal stitching of cast iron tunnel lining . . . . . . . . . . . . . . .165
Figure 5.13
Strengthening repair of buckled steel section lintel used in an
opening of a cast iron lined tunnel due to structural defect . . . . . . . . .167
Figure 5.14
Typical patch repair of pre-cast concrete tunnel lining . . . . . . . . . . . . .171
Figure 5.15
Example of cracking in pre-cast expanded concrete tunnel lining . . . .172
Figure 5.16
Use of ribs and sprayed concrete to provide a secondary lining . . . . . .179
Figure 5.17
Underpinning a tunnel portal structure by piling to prevent structural
movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .187
Figure 5.18
Details of a replacement invert (a) details of an overslab invert (b) . . . .189
Figure 5.19
Support of unlined tunnels using rock bolts . . . . . . . . . . . . . . . . . . . . .191
Figure 5.20
Examples of different types of rockbolts . . . . . . . . . . . . . . . . . . . . . . . . .194
Figure 5.21
Grouted plug remedial measure for deteriorating shaft lining . . . . . . .199
Figure 5.22
Potential failure mechanism of a shaft cap located at rock head level . .199
Figure 6.1
Drip trays to control water ingress before (a) and after installation (b) . .206
Figure 6.2
Reconstruction of the lining of the Mersey Tunnel (a) exposed painted
cast iron and stainless steel support members for panels during
installation works (b) partially completed section of secondary lining
with panels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .207
Figure 6.3
Seepage from circumferential joints in a pre-cast concrete lined
tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .210
Figure 6.4
Typical seepage from cast iron segmental lining grout hole . . . . . . . . .211
Figure 6.5
Summary of grouting techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .213
Figure 6.6
Grouting operation in progress in a pre-cast concrete segmental
lined tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .219
Figure 6.7
Series of longstanding point leaks from the lower part of the arch in a
brick lined tunnel, made clear by the thick deposits of carbonate that
have built up on the brickwork . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .220
Figure 6.8
Section through a multi-ring brick arch illustrating the positioning of
the access holes relative to the structure . . . . . . . . . . . . . . . . . . . . . . . . .222
Figure 6.9
Elevation showing a typical access hole pattern . . . . . . . . . . . . . . . . . . .222
Figure 6.10
Section through cracked masonry arch showing typical grout access
hole layout (note structural thickness and type of grout used to
determine the access hole centres) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .223
Figure 6.11
Elevation and structural drawing of an access hole pattern for sealing
a joint in brickwork . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .223
Figure 6.12
Closely spaced access holes to deal with point leaks . . . . . . . . . . . . . . . .224
Figure A1.1
Abbotscliffe tunnel portal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .250
Figure A1.2
Geological section and corresponding view of west portal . . . . . . . . . .251
Figure A1.3
Fracturing in the masonry lining wall in Abbotscliffe tunnel . . . . . . . . .251
Figure A1.4
Window panel through lining exposes chalk at extrados . . . . . . . . . . .253
Figure A1.5
Trial pit through ballast to expose footings . . . . . . . . . . . . . . . . . . . . . .253
Figure A1.6
Lydden Spout, February 2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .255
Figure A1.7
Remedial treatment at Lydden Spout . . . . . . . . . . . . . . . . . . . . . . . . . . .255
Figure A1.8
Drilling for rock dowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .257
Figure A1.9
Rotating cutter head removing brick . . . . . . . . . . . . . . . . . . . . . . . . . . .258
Figure A1.10
Excavated wall panel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .258
Figure A1.11
Applying sprayed concrete to crown . . . . . . . . . . . . . . . . . . . . . . . . . . .259
Figure A1.12
Typical array of monitoring instrumentation . . . . . . . . . . . . . . . . . . . . .260
Figure A1.13
View of Martello tunnel portal (a) and details of its lining profile (b) . . .262
Figure A1.14
Example condition matrices. Inspection June 1964 (a) and Inspection
February 2002 (b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .262
Figure A1.15
Re-lining in Martello Tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .264
Figure A1.16
Framing to patch repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .264
Figure A1.17
Brick excavation in Martello . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .265
Figure A1.18
Martello wall panel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .265
Figure A1.19
Water streams from the base of one of the plastic sheets used to deflect
its flow down the tunnel wall rather than spouting into the running
area of the tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .267
Figure A1.20
Site investigation resulted in some additional subsidence at the ground
surface, affecting an area of about 1 m², which subsided by around 300
mm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .270
Figure A1.21
Idealised cross-section through tunnel at location of water ingress
showing inferred ground conditions and water pathway between
perched water table and tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .271
Figure A1.22
Design for the geogrid capping layer . . . . . . . . . . . . . . . . . . . . . . . . . . .275
Figure A1.23
Construction of the capping layer using geotextile and
engineering fill . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .275
Figure A1.24
Finite element structural modelling results for tunnel lining subjected
to full ground loading, hydrostatic water and grout pressures during
injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .277
Figure A1.25
Section of the 3D laser-scanning survey results showing one side of
the tunnel intrados folded flat as a 2D image . . . . . . . . . . . . . . . . . . . . .277
Figure A1.26
Design of the grout injection scheme, showing amber zone (1.5 m to
3 m offset from tunnel extrados) where strict controls on drilling and
injection were adopted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .279
Figure A1.27
The original tunnel after removal of services, track and ballast . . . . . .282
Figure A1.28
Completed lining, including architectural features . . . . . . . . . . . . . . . .282
Figure A1.29
Invert construction in progress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .284
Figure A1.30
Temporary propping to tunnel during cross-passage reconstruction . .284
Figure A1.31
Fixing waterproofing membrane to new tunnel lining . . . . . . . . . . . . .285
Figure A1.32
Cross-section through the tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . .289
Figure A1.33
Repair histories of tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .293
Figure A1.34
Pre-cast concrete cess trough and CHS pile . . . . . . . . . . . . . . . . . . . . . .298
Figure A1.35
Plan of remedial works showing piles . . . . . . . . . . . . . . . . . . . . . . . . . . .299
Figure A1.36
Pre-cast block in the six foot with ballast retention box . . . . . . . . . . . . .300
Figure A1.37
Invert construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .301
Figure A1.38
Relining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .302
Figure A1.39
Colwall Old Tunnel: concrete shaft cap exposed after targeting by
geophysical survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .305
Figure A1.40
Tunnel intrados marked out in 1 m squares to allow condition mapping –
this area exhibits some spalling of brickwork at the crown . . . . . . . . . . . .307
Figure A1.41
Patch repairs underway supported off steel centering . . . . . . . . . . . . . .308
Figure A1.42
Completed patch repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .308
Figure A1.43
Construction of concrete segmental lining within the tunnel shield . . . . . .
Figure A1.44
Grouting behind the tunnel lining to stabilise and help to waterproof it . .309
CIRIA C671 • Tunnels 2009
xvii
xviii
Figure A1.45
Partly-constructed two-ring oval profile sewer with timber heading,
Piccadilly Circus c1928 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .312
Figure A1.46
Deterioration of lower part of arch corresponding with typical location
of water inflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .312
Figure A1.47
Cross-section showing the relative location of the two tunnels (a) and
a 3D representation of sand lens (b) . . . . . . . . . . . . . . . . . . . . . . . . . . . .322
Figure A1.48
Settlement profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .327
Figure A1.49
General condition of lining and silt deposits in first leg of tunnel 1986
inspection. Note the absence of any significant biological growth . . . . .329
Figure A1.50
View of intake shaft access with ladders that were, in the absence of
contrary information, assumed to be unsafe so that alternative safe
access methods were required (a) and entry to dewatering shaft using
a safety winch and tripod (b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .331
Figure A1.51
Image from the 2005 inspection showing 3.66 m dia. tunnel with
persistent old longitudinal cracks in cast in situ concrete lining at
crown and shoulder positions made visible by the use of low-angle
lighting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .332
Figure A1.52
Severe spalling to original brick lining . . . . . . . . . . . . . . . . . . . . . . . . . .336
Figure A1.53
Haymarket south tunnel – GPR with joint mapping . . . . . . . . . . . . . . .337
Figure A1.54
Sugar Loaf Tunnel after relining with sprayed concrete . . . . . . . . . . . .339
Figure A1.55
Instrumentation box for piezometric and vibrating wire strain gauges
(VWSG), installed in cross-passage for permanent access . . . . . . . . . . .343
Figure A1.56
Convergence measurement on the upper part of the tunnel by invar
line, with the help of the hydraulic access platform . . . . . . . . . . . . . . . .343
Figure A1.57
Digital plotter with data-logger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .345
Figure A1.58
Section through the original canal tunnel . . . . . . . . . . . . . . . . . . . . . . .346
Figure A1.59
Details of replacement invert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .348
Figure A1.60
Schematic diagram of instrumentation and survey methods . . . . . . . . .349
Figure A1.61
Graph of tunnel convergence during the construction process . . . . . . .349
Figure A1.62
Deep tunnels: piling and pile-cap construction . . . . . . . . . . . . . . . . . . .351
Figure A1.63
Deep tunnel – influence of piling from the face of the tunnel . . . . . . . .351
Figure A1.64
Subsurface tunnel (seven ring masonry) – piling and pile cap
construction close to the tunnel walls . . . . . . . . . . . . . . . . . . . . . . . . . . .352
Figure A1.65
Subsurface tunnels (five ring masonry) – propping of tunnel during
construction of a pile cap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .352
Figure A1.66
Pile cap construction in steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .353
Figure A1.67
Typical sections showing the proposed development near tunnels . . . .355
Figure A1.68
FE model of the tunnels and soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .355
Figure A1.69
Typical construction operations, excavation on south side (a) and
excavation between tunnels (b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .356
Figure A1.70
Effect of unsymmetrical excavation vs predicted in situ stresses . . . . . .356
Figure A1.71
Capacity factor vs. imperfection (a) and construction of transfer beams
over tunnels (b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .357
Figure A1.72
Monitoring stations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .357
Figure A1.73
Plan and section of the tunnel after the accident showing timber
bulkheads and backfill . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .372
Figure A1.74
Scraper being used to level off the ingressed sand . . . . . . . . . . . . . . . . .372
Figure A1.75
Superheated, fuel-rich gases combust as oxygen becomes available
at the top of the shafts – the plumes of fire reached 50 m above
ground level and caused closure of the local A-road . . . . . . . . . . . . . . .374
Figure A1.76
Investigators stand amid the twisted wreckage of one of the wagons . .375
Figure A1.77
Rubble fills the tunnel below the collapsed area of the tunnel crown,
revealing a void behind the lining with construction timbers still
in place . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .375
Figure A1.78
Postulated subsidence mechanism showing clay plug failing in
undrained shear and relative locations of the tunnel and the Cornwallis
building . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .380
Figure A1.79
Possible collapse mechanism of tunnel . . . . . . . . . . . . . . . . . . . . . . . . . .381
Figure A3.1
Extensive whitish surface encrustations of carbonate-minerals (typically
calcite) on a masonry tunnel lining – these have been leached out of
the mortar by water seepage and gradually deposited on the lining
surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .399
Figure A3.2
Typical appearance of surface-wet masonry (wetness index of 2 or 3 in
accordance with the classification given in Table A3.1) near to a tunnel
portal that is gradually spalling due to freeze/thaw cycling . . . . . . . . . .400
Figure A4.1
Photography combined with oblique backlighting can be a very
useful aid to recording areas of surface-wetness on tunnel linings,
because these are highly reflective . . . . . . . . . . . . . . . . . . . . . . . . . . . . .405
Figure A4.2
Results of combined joint mapping and ground penetrating radar
(GPR) survey to identify construction features and defects within a
masonry-lined rail tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .407
Figure A4.3
Taking a 100 mm core through a brickwork tunnel lining. The lightweight
rig is bolted to the wall. Progress can be slow in hard masonry materials
and in the investigation of rail tunnels use is often made of heavier and
more powerful coring equipment mounted on track trolleys . . . . . . . . . .411
Figure A4.4
Compressive strength testing of a 300 mm diameter concrete core
while simultaneously measuring strain . . . . . . . . . . . . . . . . . . . . . . . . . .413
Figure A4.5
Trial-pit through ballast at base of tunnel sidewall to prove invert
depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .415
Figure A4.6
Intrusive investigation through metallic segmental tunnel lining . . . . .415
Figure A4.7
Flat jack developed by Cardiff University . . . . . . . . . . . . . . . . . . . . . . .419
Figure A4.8
Resistivity survey traverse line laid out over a tunnel with suspected
hidden shafts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .424
Figure A4.9
Radar survey from a track-mounted cradle, with the aerial, mounted
on the end of a telescopic arm, swept over the intrados at the location
of a possible hidden shaft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .426
Figure A4.10
Surface preparation (a) and ultrasonic testing (b) of cast iron segmental
lining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .428
Figure A5.1
Chimney clearly marks the location of a shaft at ground level – not
all are easily located . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .435
Figure A5.2
Results of a ground resistivity survey traverse, taken along the crown
of a masonry tunnel lining that clearly identify a potential location of a
hidden shaft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .436
Figure A5.3
Results of a ground penetrating radar (GPR) survey traverse along the
crown of a masonry tunnel lining that identify a potential shaft location
of a hidden shaft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .437
Figure A5.4
Drill sequence to locate obscured shaft when position has been
reasonably well-established . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .439
Figure A5.5
Dynamic probing at possible hidden shaft location . . . . . . . . . . . . . . . .439
Figure A5.6
Iso-surface interpolated from dynamic probing to locate backfilled
shaft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .440
CIRIA C671 • Tunnels 2009
xix
Figure A5.7
Aerial photograph from which tunnel alignment and possible location
of construction shafts can be discerned based on topography. Ideally
such photographs can be viewed as stereo pairs . . . . . . . . . . . . . . . . . .441
Figure A6.1
Example scan line discontinuity log . . . . . . . . . . . . . . . . . . . . . . . . . . . .444
Figure A6.2
Example discontinuity map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .445
Tables
xx
Table 1.1
Where to find information and guidance on specific topics . . . . . . . . . . .5
Table 2.1
Timeline of tunnel development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7
Table 2.2
Classification of ground conditions in the 19th century . . . . . . . . . . . . .12
Table 2.3
Change in construction methods over time . . . . . . . . . . . . . . . . . . . . . . .12
Table 2.4
Degree of disturbance due to excavation method . . . . . . . . . . . . . . . . . .17
Table 2.5
Ground failure mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19
Table 2.6
Mortar mixes and compressive strengths used in the UK, and
corresponding strengths of masonry using different bricks . . . . . . . . . .33
Table 2.7
Comparison of typical strength and density values of some common
UK building stones with other construction materials . . . . . . . . . . . . . . .34
Table 2.8
Properties of some old bricks used in bridge and tunnel construction . . .35
Table 2.9
Statistical analysis of properties of brick samples from old railway
structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36
Table 2.10
Example results of finite element modelling of cracked cast iron lining,
as shown in Figure 2.23. This suggests that the presence of cracks has
only minor influence on tunnel deformation . . . . . . . . . . . . . . . . . . . . . .41
Table 2.11
Summary of causes of masonry deterioration . . . . . . . . . . . . . . . . . . . . .48
Table 2.12
Summary of causes of metal deterioration . . . . . . . . . . . . . . . . . . . . . . . .50
Table 3.1
Direct and consequential cost of tunnel incidents . . . . . . . . . . . . . . . . . .65
Table 3.2
Examples of hazards and risk mitigation measures for tunnels . . . . . . . .77
Table 4.1
Current tunnel structure inspection requirements of the main UK
infrastructure owners: Network Rail (NR), Highways Agency (HA),
British Waterways (BW) and London Underground (LU) . . . . . . . . . . .97
Table 4.2
Recommended methods for direct investigation of tunnel parameters . .106
Table 4.3
Interpretation of common inspection and investigation observations . .117
Table 4.4
Closed form solution for analysis of tunnel lining . . . . . . . . . . . . . . . . .131
Table 5.1
Repair techniques for tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .143
Table 5.2
Summary of typical defects of brick and masonry tunnel linings and
possible remedial solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .154
Table 5.3
Principles and available methods for prevention and repair of
deterioration to structural concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . .169
Table 5.4
Concrete repair methods and materials . . . . . . . . . . . . . . . . . . . . . . . . .171
Table 5.5
Summary of tunnel lining replacement and strengthening techniques . .181
Table 5.6
Summary of different rockbolt types . . . . . . . . . . . . . . . . . . . . . . . . . . .192
Table 6.1
Summary of passive and active water ingress control measures . . . . . .203
Table 6.2
Grouting techniques with relevant ground types . . . . . . . . . . . . . . . . . .214
Table 6.3
Joint aperture range for various cement grouts . . . . . . . . . . . . . . . . . . .216
Table A1.1
Options for water control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .263
Table A1.2
North tunnel lining details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .290
Table A1.3
Geological succession within Standedge tunnels . . . . . . . . . . . . . . . . . .291
Table A1.4
Fault locations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .292
Table A1.5
Summary of rock mass quality assessment . . . . . . . . . . . . . . . . . . . . . . .294
Table A1.6
Summary of range of adopted Hoek-Brown Constants . . . . . . . . . . . . .295
Table A1.7
Summary of deformation modulus correlations . . . . . . . . . . . . . . . . . .297
Table A1.8
Predicted and observed movements during excavation on south . . . . .358
Table A1.9
Feasibility matrix for initial assessment of options for box tunnel
relining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .364
Table A1.10
Detailed matrix for further assessment of shortlisted options . . . . . . . .367
Table A2.1
Primary sources of infrastructure-specific sources of tunnel
information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .385
Table A2.2
Sources of historical information – contact details . . . . . . . . . . . . . . . . .391
Table A2.3
Geological and other sources of information – contact details . . . . . . . .392
Table A3.1
Example of descriptive wetness index system . . . . . . . . . . . . . . . . . . . .400
Table A4.1
General, specialist, testing and monitoring techniques for tunnel
investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .403
Table A4.2
Techniques used in the investigation of ground around tunnels . . . . .416
Table A4.3
Usefulness of engineering geophysical methods for geotechnical
investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .421
Table A5.1
Survey techniques for the identification and location of hidden tunnel
shafts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .436
Table A7.1
Partial factors of safety for actions recommended by Eurocodes and
British Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .450
CIRIA C671 • Tunnels 2009
xxi
Glossary
xxii
Action (F)
An action is a force (load) applied to a structure (direct action)
or a deformation caused, for example, by temperature changes,
moisture variation, uneven settlements.
Adit
A horizontal or sloping passage between a tunnel and the
ground surface or an adjacent underground structure.
Appraisal
Includes the range of activities that can be involved in the
evaluation of a tunnel’s condition and performance, ie the
gathering of existing data, inspection, investigation and
structural assessment.
Arch ring
The load bearing part of an arch containing one or more
overlapping rings or leaves of masonry.
Ashlar
Masonry consisting of blocks of stone square dressed to given
dimensions and laid in courses with thin joints.
Aquiclude
Soil or rock forming a stratum, group of strata or part of a
stratum of very low permeability, which acts as a barrier to
groundwater flow.
Aquifer
Soil or rock that forms a stratum, group of strata or part of a
stratum that is water bearing.
Assessment
Here used specifically to imply the evaluation of a tunnel’s
structural capacity and performance, typically by one of several
analytical methods and commonly using proprietary software
applications. Assessment can be carried out as part of a more
wide-ranging appraisal of a tunnel’s condition and
performance.
Backing or Backfill
Material used to fill an excavation or give support behind a
structure.
Batch
Quantity of material (here commonly grout or mortar) mixed at
one time.
Bedding plane
The plane of stratification in sedimentary rock, which may also
be present in building stone produced from it.
Bed joint
Horizontal joint in masonry.
Blind shaft
A temporary shaft that has been covered, sealed or capped in
such a way as to render the position of the shaft discernible.
Bond
An arrangement of masonry units where the vertical joints (end
joints) of one course do not coincide with those immediately
above and below. The bond type refers to the relative
arrangement of construction units in masonry, eg the presence
and combination of units laid as headers and stretchers. The
most common bond in tunnel sidewalls is English bond, whereas
arches are often constructed in stretcher bond (indicating no
connection between rings).
Cap
A structural slab placed over a shaft, capable of supporting the
weight of any ground above it and any superimposed load.
Cast iron
An iron-carbon alloy produced in a blast furnace containing up
to four per cent carbon.
Cementitious grout
A grout containing cement and water as major ingredients.
Chemical grouts
Any grouting material characterised as a pure solution with no
particles (other than impurities) in suspension.
Compaction grouting
A grouting method similar to displacement grouting. Grout
generally does not enter the soil pores but remains in a
homogeneous mass that gives controlled displacement to
compact loose soils.
Centring
Temporary structure on which an arch is supported during
construction, normally made from timbers.
Competent person
A person who, by reason of theoretical and practical training,
actual experience or both, is competent to perform the task or
function or assume the responsibility in question and is
authorized to perform such a task or function.
Condition appraisal
See Appraisal.
Conservation
Work carried out with the aim of maintaining or restoring the
features of a tunnel that are important to its character, in
particular the visible parts of its structure.
Crown
The highest point of the internal curved surface of a tunnel
cross-section.
Cut-and-cover (C&C)
A method of tunnel construction in which the tunnel structure
is built in an open excavation and covered by fill.
Deepwell
A groundwater extraction well of sufficient dimension to accept
a submersible pump.
Discharge
The flow rate pumped out by a groundwater control system.
Drawdown
The amount of lowering of the water table in an unconfined
aquifer or of the piezometric level in a confined aquifer by a
groundwater control system.
Effect (E)
An effect (or action effect) on structural members, (eg internal
force, moment, stress, strain) or on the whole structure (eg
deflection, rotation). For ultimate limit states, an effect is a
quantity associated with the actions and with the structure to be
analysed, that can be directly compared with the resistance of
the structure or part of it. For serviceability limit states, effects
can be displacements, crack opening or other quantities relevant
to functioning of the structure (crack opening for instance can
be important if the structure is intended to be watertight).
Electro-osmosis
A groundwater control method used in very low permeability
soils where an electric potential difference is applied to the
ground to induce groundwater flow.
English bond
A strong method of building walls by laying bricks together in
alternating courses of headers and stretchers. The most
common bond in tunnel sidewalls.
Essential maintenance
Rehabilitation works required to address specific inadequacies
in function and performance, eg reinstatement of deteriorated,
damaged or failed elements essential to serviceability.
CIRIA C671 • Tunnels 2009
xxiii
xxiv
Explosive spalling
The rapid loss of the surface layers of concrete during a fire,
particularly in high strength concrete (HSC).
Extrados
The outer (convex) curve of an arch or circular/semicircular
element, which in the case of a tunnel lining or shaft may be,
but is not always, in direct contact with the adjacent ground.
Fill
Material used to occupy a void.
Fit for purpose
A single performance criterion against which a tunnel may be
judged, indicating that it meets the full range of performance
criteria set by the asset owner, for example, in terms of its safety,
functionality and maintainability.
Garland
A type of drain, formed within an excavation at the level of an
impervious stratum that underlies permeable strata, to intercept
water that would otherwise flow into the excavation.
Grommet
Material used to bung a hole, typically a grout plug used in a
tunnel lining.
Groundwater
Water contained within, and flowing through, the pores and
fabric of soils and fissures in rock.
Grout
A liquid material injected into a soil or rock formation that gels,
stiffens or sets with time and thereby changes the physical
characteristics of the formation.
Grouting
The injection of grout material under pressure into void spaces
either in naturally occurring substances such as soils or fissured
rocks, or in artificial cavities such as those found in porous
masonry or behind tunnel linings.
Haunch
The lower section of an arch ring towards its springing.
Header
A brick laid with its longest dimension normal to the face of a
single ring or skin of brickwork, interconnecting adjacent rings/
skins.
Heading
A tunnel with a small cross-section.
Heave (base)
Lifting of the floor of an excavation or structure, usually caused
by unrelieved pore water pressure or from high stresses in
natural invert materials that behave in a brittle-plastic manner.
Heavy ground
Ground where excavated faces need support relatively quickly.
Hidden shaft
A temporary shaft that has been buried, covered, sealed or
capped in such a way as to render the position of the shaft
indiscernible.
High strength
(HSC)
A term used to describe concrete that can attain a high concrete
strength, typically 50 MPa or above. HSC potentially exhibits
good durability characteristics but can suffer from explosive
spalling in fires.
Hinge
A more or less local situation at which, due to a tensile crack, the
structure can rotate as if it were an articulation.
Historic tunnel
One that has some recognised historical value, through rarity or
in terms of social, cultural or engineering heritage, and is
subject to statutory protection, eg through listed building or
Scheduled Ancient Monument status. Normally applies to visible
parts only (ie portals) but exceptionally can include the tunnel.
Hydrogeology
The study of the interrelationship of the geology of soils and
rock within groundwater.
Intrados
The inner (concave) curve of a circular or semicircular element.
In the case of a tunnel lining or shaft, defines its internal space.
Intervention
An action carried out to rectify or arrest continuing
deterioration and/or loss of performance of a tunnel through its
protection, maintenance, repair or enhancement.
Invert (tunnel invert)
The bottom surface of a tunnel.
Lime mortar
Pure lime (also known as fat or non-hydraulic lime) is produced
from pure limestone and relies upon gradual reaction with
atmospheric carbon dioxide (carbonation) to harden and
develop strength. Pure limes produce a mortar that is typically
weaker and more porous and permeable than impure limes
with a degree of hydraulic (water-dependent) set or those to
which Portland cement has been added (a process known as
gauging).
Lined tunnel
A tunnel in rock or soil where a lining is provided.
Lining
Permanent or temporary cover to the rock or soil surface at the
wall of an excavation for a tunnel, shaft or adit.
Loss of fines
The movement of clay, silt or sand sized particles out of the
ground towards a sump or well or through a tunnel lining.
Maintenance
All the operations necessary to maintain a tunnel in a
serviceable condition until the end of its life, comprising routine
maintenance and essential maintenance.
Masonry
The work of a mason, strictly referring to work in stone,
however commonly used to refer generally to work in either
brick or building stone, as it is here.
Metal linings
For the purpose of this guide, it includes grey and spheroidal
graphite (SGI) cast iron and steel (including stainless steel)
linings.
Mortar
A mixture of lime and/or cement, sand and water used to bind
bricks and masonry in construction, or a highly viscous,
particulate grout.
Open shaft
An unfilled shaft visibly detectable from both the top and
bottom.
Packer
A device inserted into a drillhole through which an injection
pipe passes. Usually an expandable device activated
mechanically, hydraulically or pneumatically.
Pattern bolting
Installation of rock bolts on a regular pattern and/or at
equidistant centres, ie on a square grid or square grid with one
in the centre of each square.
Permeability
A measure of the ease with which water can flow through the
pores of soil or rock (also known as coefficient of permeability,
hydraulic conductivity).
Permeation grouting
A grouting process for replacing water in voids between soil
grains or particles with grout fluid at a low injection pressure
without disturbing the natural structure of the ground.
CIRIA C671 • Tunnels 2009
xxv
Planned maintenance
Maintenance that is premeditated to keep the tunnel in a fully
serviceable condition rather than reactive in response to
inadequate performance. It can be subdivided into two types:
1
2
xxvi
Periodic (carried out regularly at predetermined intervals).
Condition-based (carried out in response to a perceived or
anticipated loss of performance).
Pore pressure
The interstitial pressure of water within a mass of soil, rock, or
concrete.
Portal (tunnel portal)
An entrance or a structure that forms an entrance to a tunnel.
Pozzolan
A cement additive comprising silica in reactive form, which can
impart hydraulic set. It can be either naturally occurring (eg
volcanic ash) or artificially produced (eg brick dust or
pulverised fuel ash, PFA).
Primary lining
Structural lining of a tunnel/adit/shaft.
Reactive maintenance
Maintenance that is carried out in response to inadequate
performance.
Recharge well
Replenishment of groundwater artificially via wells to reduce
drawdowns external to a groundwater control system or as a
means of disposing of the discharge.
Refuge
An area where the tunnel cross-section is locally widened to
provide shelter to staff from traffic using the tunnel.
Ring
Either a single layer or leaf of a brickwork lining consisting of
stretcher bricks, which may or may not be bonded to adjacent
leafs using headers, or an assembly of segments, one segment
wide, which forms a complete section of a lined tunnel or shaft.
Ring separation
Loss of bonding between adjacent rings of brickwork (not
necessarily an open gap).
Rise
Vertical height of arch from springing level to the crown of the
intrados.
Rehabilitation
Work that involves bringing features of a deteriorated tunnel
back into a satisfactorily functional state.
Resistance (R)
Resistance is the capacity of a member or component, or a crosssection of a member or component of a structure, to withstand
actions without mechanical failure.
Rock
Relatively hard naturally occurring part of the Earth’s crust that
has not been broken down into loose material that can be
readily excavated by hand.
Routine maintenance
Routine work carried out with the aim of preventing or
controlling deterioration, including inspection and monitoring
activities, and general housekeeping and minor repairs.
Rubble masonry
Term describing many different types of masonry, the main
types being random rubble (stone as it comes from the quarry)
either coursed or un-coursed, and squared rubble, either
coursed or un-coursed.
Secondary lining
Extra lining to the primary lining for improvement or
enhancement of performance or for decoration.
Segment
An arc-shaped preformed component that forms part of the
lining of a tunnel, shaft or adit.
Segmental arch
Arch whose intrados comprises a segment of a circle smaller
than a semicircle.
Semicircular arch
Arch with an intrados with a semi-circle profile, ie 180º, so that
the rise is half the span.
Service life (or
serviceable life)
The period of duty after which replacement or major renewal/
refurbishment, rather than continued use, is anticipated to be
justifiable on an economic or operational basis.
Set
The condition reached by a cement paste or grout when it has
lost plasticity to an arbitrary degree through hydration.
Shaft
Vertical or steeply inclined excavation, usually of limited crosssection in relation to its depth.
Shaft eye
The intersection of a shaft with a tunnel.
Shallow arch
Arch where the rise is smaller than half the span.
Shield
A mobile structure, commonly cylindrical, used to support the
ground at the tunnel face ahead of the tunnel lining.
Sidewall
The vertical or near-vertical internal surfaces of a tunnel,
forming a curved or straight plane that defines its sides. In an
arched tunnel the sidewalls extend up to the springings where
they support the arch. In earlier tunnels the sidewalls were
constructed to be near vertical, but later were often curved to
provide a more structurally efficient ovoid cross-section.
Soffit
The underside of an element.
Soil
Mineral material that results from the weathering of rock.
Soldier
Masonry unit laid with its longest dimension upright and
parallel with the face of the wall, ie bedded on a face having
smaller dimensions.
Spalling
Flaking and loss of material (either rock, stone, brick or
concrete) from an exposed surface normally caused by frost, salt
action or mechanical action, or, in the case of reinforced
concrete, also by corrosion of embedded metallic reinforcement.
Springing
Point, line or plane from which an arch or vault springs, located
at the junction between the supporting sidewalls/abutment and
an arch or vault.
Stretcher
A masonry unit laid with its longest dimension horizontal and
parallel to the face of the wall.
Stretcher bond
A masonry bond in which bricks are laid in courses with
overlapping joints with their longest dimension parallel, so that
all bricks are laid as stretchers. This bond is commonly used in
masonry arches, where it indicates that there is no structural
connection between brick rings.
Tunnel
An enclosed underground structure, horizontal or sloping, that
has been constructed by some means (eg cut-and-cover, boring,
jacking) to provide access for something (eg vehicles, utilities).
CIRIA C671 • Tunnels 2009
xxvii
xxviii
Tunnel engineer
A person responsible for the technical and engineering
processes of tunnel management, eg carrying out or making
decisions regarding condition assessment, serviceability,
performance restrictions and requirements for maintenance,
repair and alteration.
Unlined tunnel
Usually in rock where structural support is provided by rock
engineering methods (eg rock-bolting) and so a lining is
structurally unnecessary and not provided.
Wellpoint
Small diameter shallow well normally installed at close centres
by jetting techniques.
Wrought iron
A material produced by hammering and rolling billets of iron.
Zone of influence
The 3D volume of ground adjacent to a tunnel, including any
shafts or adits, which may be affected by its construction and its
later life, eg through structural instability or collapse, and
particularly the area of land surface above the tunnel that lies
within this zone.
Abbreviations
ALARP
As low as reasonably practicable
BR
British Rail (now Network Rail)
BW
British Waterways
CDM
The Construction (Design and Management) Regulations 2007
CP
Cathodic protection
CCP
Current cathodic protection
C&C
Cut-and-cover
E
An effect (or action effect) on structural members or the whole structure
ESR
Excavation support ratio
F
Force (load) applied to a structure
FE
Finite element (method of structural analysis)
GIS
Geographical Information System
GRC
Glass reinforced cement
GRP
Glass reinforced plastic
GPS
Global positioning system
HA
Highways Agency
HGV
Heavy goods vehicle
HSC
High strength concrete
ITA
International Tunnelling Association
LCA
Life cycle assessment
LU
London Underground
MCI
Migrating corrosion inhibitor
NR
Network Rail
PFA
Pulverised fuel ash
PHEW
Panel for Historical Engineering Works
PIARC
World Road Association
PPE
Personal protective equipment
PRC
Plastic reinforced concrete
Q
Rock mass quality
QRA
Quantitative risk assessment
RMR
Rock mass rating
SCMI
Structures condition marking index (system used by Network Rail)
SCOSS
Standing Committee on Structural Safety
SGI
Spheroidal graphite iron
SNCOs
Statutory nature conservation organisations
SPA
Special protection area
SAC
Special area for conservation
SACP
Sacrificial anode cathodic protection
SBR
Styrene-butadiene rubber
SNCO
Statutory Nature Conservation Organisation
CIRIA C671 • Tunnels 2009
xxix
xxx
SSSI
Site of special scientific interest
TMS
Tunnel management system
TBM
Tunnel boring machine
TMS
Tunnel management system
UIC
International Union of Railways
WLC
Whole-life costing
1
Introduction and background
1.1
BACKGROUND
Humans have been excavating and constructing tunnels for thousands of years. In
Neolithic times (2000–4000 BC) tunnel excavations were made for purposes including
shelter, burial, defence and mineral extraction. Later, civilisations (particularly the
Babylonians, Egyptians, Greeks and Romans) constructed, and occasionally lined, tunnels
for the transport of clean water and sewage waste. Technology remained basic and rates of
progress very slow for many centuries up to the invention of gunpowder and rudimentary
drilling machines in the 17th century. In the United Kingdom the Industrial Revolution
led to the rapid expansion of the canal system in the 18th century and then, in the 19th
century, the sewerage and rail systems, driving further advances in tunnelling technology.
This resulted in the construction of hundreds of miles of tunnels, many of which remain
an integral part of the UK’s transport and distribution infrastructure today. Over time, the
range of lining materials available has evolved from timber, masonry (stone and brick),
iron, reinforced concrete, to steel and sprayed concrete. Modern tunnelling technology,
including increased mechanisation and more sophisticated design and construction
techniques, has enabled successful construction of tunnels in increasingly technically
challenging environments.
The physical nature of the UK’s current and in-use infrastructure tunnels reflects a
complex mix of past needs, technologies, available materials and mechanical means, and
spans the period from those constructed in the 18th and 19th centuries, which have
already exceeded the normal life expectancy of modern structures, to recent 20th and
21st centuries tunnels. There are very few cases where it would be practical, economically
justifiable or socially acceptable to substantially replace these ageing assets. The capacity of
UK infrastructure and transport tunnels is largely fixed and finite with relatively little new
build, so the key challenges for infrastructure owners are to maintain assets efficiently and
to provide optimum availability and throughput, which requires careful management of
the existing tunnel stock. Gradual and progressive deterioration in service can be
overcome by a considered, coherent and planned maintenance, repair and upgrade
strategy. A rigorous asset management approach to tunnel assessment, maintenance and
repair is increasingly necessary to deliver optimum asset performance.
This guide promotes good practice in all aspects of tunnel assessment, maintenance and
repair, combining current thinking and technology, and providing coherent guidance.
Included in this guide is a selection of case studies from recent tunnel works, which give
real examples that will be useful for those responsible for tunnels.
1.2
PURPOSE AND SCOPE OF WORK
This publication provides guidelines for the management, appraisal, maintenance and
repair of tunnels, and advice on issues such as conservation, health and safety, and the
environment.
The purpose of the guide is to:
present good practice
CIRIA C671 • Tunnels 2009
1
provide a guide for routine management
recommend assessment, maintenance and repair strategies to give best value for
money
help knowledge sharing
identify gaps in knowledge.
This publication is principally concerned with the civil engineering aspects of tunnels with
a large enough section to allow routine man access, ie with an internal diameter of at least
1.75 m. However much of the information here may apply to tunnels of smaller diameter.
The types of tunnel linings under consideration are defined as primary support systems
by the British Tunnelling Society Tunnel lining design guide (BTS and ICE, 2004), as they
bear directly onto the ground. They may also constitute a permanent support system in
the case of one-pass lining types (where there is no extra lining). Due to the great diversity
of tunnel types, ages and construction methods, it has been necessary to limit the scope of
this guide primarily to those tunnels (and their shafts, adits and drainage) constructed in
the first half of the 20th century and before. These are:
bored and cut-and-cover tunnels
unlined tunnels
masonry (brick and stone) lined tunnels
metal lined tunnels (grey and spheroidal graphite cast iron and steel)
pre-cast segmental concrete lined tunnels.
Certain types of tunnel have been excluded, although some of the information included
here may still be relevant to them, and certain aspects specifically relating to them are
dealt with in passing:
in situ concrete lined tunnels
sprayed concrete lined tunnels
immersed tube tunnels
jacked tunnels (eg pipe jacked tunnels used in trenchless technology, jacked box
linings etc).
Also excluded is:
tunnel equipment and associated infrastructure (eg pumping systems, electrical and
communication systems, trackform and highway pavements) other than highlighting
situations where this is directly affected by or integral to a tunnel’s structural
performance.
This publication provides guidance on the asset management of tunnels. It is not intended
as a design guide for tunnel assessment or remedial works, although these areas are
discussed.
2
1.3
APPLICATION
This guide is intended for:
clients who are infrastructure owners
those responsible for the management and care of tunnel assets
engineers responsible for assessing, maintaining and repairing tunnels.
The main UK tunnel owners are railway authorities, highway authorities, navigable
waterway authorities, local authorities and statutory service providers.
1.4
ISSUES DEALT WITH IN THIS GUIDE
Topics of particular importance in the management of tunnels include:
the need to consider many tunnels as having an indefinite service life ie their longterm closure and complete replacement/reconstruction is unlikely to be feasible at any
time in the foreseeable future because they form indispensable elements of vital
infrastructure
the need to investigate and evaluate the existing structure, its performance and
materials, taking into account issues such as complex structural behaviour and
interaction with adjacent ground, lack of design to modern codes, the presence of
defects and the original variability and in-service deterioration of materials
consideration of changes in external factors ie urban development increasing ground
loading, changes in water table, increased live loading on shallow tunnels etc, and
possible change in use of the structure from that originally designed for
the necessity of regular maintenance to ensure continued performance and
serviceability while minimising unnecessary repair expenditure, closures and traffic
restrictions.
consideration of the effectiveness of repairs and alterations, and their likely influence
on the long-term performance, maintenance and whole-life cost of the structure
the significant influence of tunnel performance on the performance and efficiency of
the infrastructure as a whole, and the resulting high impact of restrictions in use and
tunnel closure
the particular difficulties associated with carrying out work in tunnel environments,
necessitating particular care in selection, design and planning so as to minimise
disruption to normal tunnel operation
the particular access, safety and environmental issues, and their associated
requirements and management implications, in managing and maintaining tunnels.
CIRIA C671 • Tunnels 2009
3
1.5
HOW TO USE THIS GUIDE
This guide is divided into seven sections, plus supporting appendices, each including
advice and guidance on particular aspects of tunnels.
Chapter 1 Introduction
General background information, scope and limitations, how to use this guide.
Chapter 2 Construction and behaviour
Basic principles of tunnels, their history, construction and materials, behaviour and performance.
Chapter 3 Asset management
Tunnel asset management, strategies and systems for condition appraisal and maintenance, and health
and safety and environmental considerations.
Chapter 4 Condition appraisal
Methodologies for tunnel inspection, investigation, monitoring and structural assessment.
Chapter 5 Selecting and carrying out works
Maintenance and repair techniques, selection and execution of works.
Chapter 6 Water ingress and control
Methods for the reduction and control of water ingress in tunnels where this causes problems.
Chapter 7 Summary of recommendations and future needs
Overall summary of recommendations, discussion of future research and development needs followed
by a list of references.
4
Table 1.1
Where to find information and guidance on specific topics
General topic
Specifically
Where to find
guidance
History of tunnels
Excavation and construction (general)
Lining construction and materials
2.1.1
2.1.2
Masonry – 2.3
Metals – 2.4
Concrete – 2.5
The structural behaviour of tunnels
and causes and signs of loss of
performance and deterioration
Structural elements
Tunnel performance
Structural damage and deterioration
Materials deterioration
Effects of fire
Approach to structural assessment
2.1.2
2.6
2.6.1
2.6.2
2.6.3
4.9
Tunnel shafts and adits
History and construction
Management aspects
Performance and behaviour
Potential effect at ground surface
Location, inspection and investigation
Carrying out works
Finding hidden shafts
2.1, 2.2
3.2
2.7
2.7.1
4.7
5.6
Appendix A5
Ensuring tunnel serviceability
through a maintenance and repair
programme
Tunnel management (general)
Maintenance planning and strategies
Management concepts and tools
Decommissioning and managing closed tunnels
1.1
3.4
3.5
3.9
Investigation and evaluation of
existing tunnel structure
Condition assessment (general)
Finding and using existing information
Sources of existing information
Principles of visual inspection
Preparation for visual inspection
Inspection procedures and observations
Tunnel investigation
Monitoring
Investigation and monitoring techniques
Structural assessment of lined tunnels
Structural assessment techniques in detail
Influence of defects and deterioration
Investigation and assessment of unlined tunnels
Interpretation and reporting
4
4.2
Appendix A2
4.3
4.6
Appendix A3
4.4
4.5
Appendix A4
4.9
Appendix A7
4.9.3
Appendix A6
4.10
The selection, design and execution
of maintenance and repair methods
Maintenance and repair (general)
Selection, planning and preparation
Carrying out routine maintenance
Information on repairs and remedial techniques
Structural lining replacement and strengthening
Rock stabilisation in unlined tunnels
Treatment and filling of tunnel shafts
Dealing with water ingress
5, 6
5.1
5.3
5.4
5.5.1
5.5.4
5.6
6
Access, safety and environmental
issues
Planning works and controlling risk
Access requirements
Health and safety management
Ensuring operational safety and fire safety
Competence and training of staff
Environmental and ecological issues
5.1
5.6.1
3.6.1
3.7
3.6.2, 4.3.3
3.6.4
Understanding the history of
tunnels, how they were built and the
materials used
CIRIA C671 • Tunnels 2009
5
2
Construction and behaviour
2.1
TUNNEL CONSTRUCTION
2.1.1
History
Originally excavated for mining near-surface natural resources, for shelter, burial
chambers or as part of defensive structures, the use of tunnels dates back many centuries
but for infrastructure purposes was not widespread until the Industrial Revolution of the
18th and 19th centuries in Europe. The development of different excavation and lining
techniques through time influences the form of tunnel deterioration present today.
The first unlined tunnels in the UK were constructed by Stone Age man around 2000 BC
to mine flints from chalk deposits in East Anglia. These tunnels were excavated through
weak rock using hand tools such as deer antlers. Early tunnel excavation through harder
rock relied upon the use of hammers and wedges or by fire quenching, where the tunnel
face was heated by fire and suddenly cooled by cold water causing the rock to shatter.
The earliest recorded infrastructure tunnel was a brick lined structure passing beneath the
Euphrates River in ancient Babylon. This was a cut-and-cover tunnel that relied on the
diversion of the river to allow construction of a brick arch structure in a trench, and later
backfilling before returning the river to its original course.
The Greeks and Romans constructed unlined or masonry-lined tunnels both for water
supply conduits and for highways. However, during the Middle Ages tunnelling was
predominantly restricted to military use and it was not until the 1600s that the use of
tunnels for infrastructure was renewed.
The advent of gunpowder and rudimentary drilling machines in the 17th century allowed
much faster excavation progress to be made, fed by the demands of a developing canal
system that reached a peak in the late 18th century. Tunnelling was further improved in
the mid 19th century with the development of dynamite and pneumatic drills to meet the
requirements of the expanding rail system. Tunnelling through soft ground during the
1800s was improved through the development of the tunnelling shield, invented by Marc
Brunel and further modified by Greathead (1895). The shield allowed ground to be
excavated while protecting the miners as a lining to support the tunnel was installed
behind an extension of the shield. Compressed air was first used at this time to control
water inflow. Later water control measures included ground freezing and grouting.
Over time, the lining material for tunnels has developed from timber, through brick and
masonry, and cast iron to reinforced concrete and steel. Each lining type has particular
characteristics that influence current condition, which will be explored in Chapters 3, 4
and 5 of this guide.
Recent tunnelling developments have included the use of more sophisticated explosives
and drilling methods in hard rock tunnelling coupled with sprayed concrete linings, while
increased mechanisation and development of tunnel boring machines (TBMs) has helped
both soft ground and hard rock excavation. Improvement in the understanding of ground
response through the development of soil and rock mechanics, including the founding of
elastic/plastic theory, has helped tunnel design progress throughout the 20th century.
6
A timeline for infrastructure tunnel developments is given in Table 2.1.
Table 2.1
Timeline of tunnel development
Ancient tunnels
Neolithic
Excavation for shelter and mineral extraction
c2160 BC
1 km long brick lined pedestrian tunnel beneath the Euphrates, Babylon
c700 BC
200 m water supply tunnel, Jerusalem
c500 BC
1 km long water supply tunnel through limestone, Samos
c200 BC
Construction of lined qanats for water supply, Middle East and China
36 BC
First road tunnel – on Naples to Pozzuoli Roman route
Greek/Roman times
Various water supply and drainage tunnels
Middle Ages
Military tunnels
Industrial Revolution
c17th
Development of navigational tunnels
1679
Gunpowder first used in infrastructure tunnel construction – Languedoc, France
1760-1830
Expansion of the canal system in the UK, involving the construction of tunnels typically with
structural brickwork arch linings
1811
Longest and deepest UK canal tunnel (Standedge) completed, Huddersfield, UK
1823
First UK road tunnel, constructed beneath Reigate Castle
1826-1829
First railway tunnel – Liverpool to Manchester, UK
1826-1900
Expansion of the rail system in the UK, involving the construction of many new tunnels. These
are typically with structural brickwork arch linings or unlined in areas of hard rock
1841
Completion of Marc Brunel’s Thames Tunnel using the first tunnelling shield
1858
Final UK canal tunnel completed – Netherton Tunnel
1863
Metropolitan Line (cut-and-cover) opens between Paddington and Farringdon Street, London
1864
Dynamite invented, which together with development of pneumatic rock drills allowed faster
excavation
1869
First cast iron segmentally lined tunnel excavated using the first circular shield, Tower Hill,
London
1871
First Alpine tunnel completed, Frejus, France
1879
First use of compressed air to balance water pressure in soft ground tunnelling, Hudson River
Tunnel, USA
1880
First tunnel beneath English Channel attempted
1890
First deep tube tunnel opens between King William Street and Stockwell, London
1897
First UK sub-aqueous road tunnel, Blackwall, London
1904
First part of New York Subway opens, USA
1900+
Reinforced concrete and steel supersede brickwork and iron as the engineering materials of
choice for new tunnels and prompt the use of new structural forms
CIRIA C671 • Tunnels 2009
7
Table 2.1
Timeline of tunnel development (contd)
Modern developments
2.1.2
20th century
Development of mechanised tunnelling methods
1936
First bolted reinforced concrete segments introduced
1940s
onwards
Improvement of ground exploration techniques and understanding of ground response through
development of soil and rock mechanics
1946
Rock mechanics first applied to steel arch support design (Terzaghi)
1950s
Introduction of rock bolts and sprayed concrete linings
1962
Coining of the term New Austrian Tunnelling Method (NATM) for tunnel support systems that allow a
high degree of convergence in initial linings to establish equilibrium before installation of a final lining
1970s
Development of circular segmental tunnel lining design
1970s
Development of TBM’s to include earth pressure balance and slurry types
Late 20th
century
Innovations such as fibre reinforcement, ground freezing, laser/infra-red survey techniques for
construction and assessment, settlement prediction techniques, finite element and difference
analysis, asset register management and fire protection design
1994
Channel Tunnel opens between UK and France
1999
Jubilee Line extension opens, London
Construction method
It is important to understand the method used to construct a tunnel as this is likely to be
significant when considering its performance or the possible modes of deterioration or
failure. For a thorough description of tunnel excavation and construction technology and
methods in the 19th century refer to Simms (1844) and Gripper (1879). A summary is
provided here.
The method of construction adopted was controlled principally by ground conditions
(geology and groundwater) but also influenced by location, length, contemporary
technology and economics. The excavation method and intended use will also have
influenced tunnel geometry. Typical profiles for lined UK rail tunnels are illustrated in
Figure 2.1. In most cases, unlined tunnels have been excavated to optimise the strength
and stability given by the geological structure so will be varied throughout the length of
the tunnel and may be square or even triangular in cross-section.
A large proportion of the UK’s tunnel infrastructure was built in the second half of the
19th century when the decision on whether to construct a tunnel or a cutting was typically
based on a consideration of the relative economic factors involved. For example, Gripper
(1879) stated: “when a cutting attains 70 feet in depth, it is generally advisable to
introduce a length of tunnel”. However, in certain situations tunnels were constructed at
shallow depths because of other factors such as lack of a suitable site for depositing spoil,
difficulties in obtaining suitable skilled workforce or necessary materials, or the influence
of the landowner.
Two construction techniques were used to build most rail tunnels:
8
1
Cut-and-cover (C&C).
2
Boring.
Shallow service tunnels for water and sewage systems, and the shallower parts of urban
metro systems, were frequently constructed by cut-and-cover methods, whereas canal and
rail tunnels were usually bored. However, the construction of any tunnel was rarely
uniform from one end to the other. The engineer reacted whenever necessary by
modifying construction methods, changing lining thickness, introducing drainage
channels or weep holes to manage the inflow of water around the lining, or using other
measures to enhance the short-term integrity of the structure. Many tunnels are hybrids,
with most of their length bored but portions at each portal constructed using cut-andcover techniques.
Figure 2.1
a
Rectangular
b
Circular
c
Straight sidewall, vaulted roof
d
As c, but with segmental/unlined
sidewalls
e
Battered sidewall, vaulted roof
f
Semi-elliptical
Typical tunnel profiles for UK railways (Railtrack, 1996)
CIRIA C671 • Tunnels 2009
9
g
Elliptical cut-and-cover
h
Parabolic
i
Segmental (oval)
j
Shallow Gothic arch
k
Tall Gothic arch
Figure 2.1
Typical tunnel profiles for UK railways (Railtrack, 1996) (contd)
2.1.2.1
Cut-and-cover tunnels
The choice of C&C or boring was dependent on ground conditions, depth of overburden,
proximity of existing buildings and whether the land above the tunnel could be disturbed.
C&C typically involved excavating a large trench with the tunnel built as a box inside it.
Once complete, the top of the tunnel was covered with the excavated material (with
varying degrees of compaction) and the surface returned to use. For the London
underground railways, this would usually be a street. A typical section through a C&C
tunnel is given in Figure 2.2, although wall thickness would vary along a tunnel due to
variations in ground conditions.
10
Figure 2.2
Typical section through a C&C railway tunnel (Railtrack, 1996)
In particularly poor ground or adjacent to sensitive buildings, a concrete saddle was
sometimes pre-cast and buried between two lines of driven piles to allow excavation
beneath (see Figure 2.1g). Where ground conditions were favourable, such as in rock,
sidewalls could be left unlined. Loading on a C&C tunnel crown can be high as the
ground above is not self-supporting, and usually compounded by the added component of
live load.
2.1.2.2
Bored tunnels
As the depth of overburden increased, or where disturbance at the ground surface was not
permissible, tunnel boring techniques were used. A typical cross-section of a bored rail
tunnel, with the principal components annotated, is given in Figure 2.3.
Figure 2.3
Typical cross-section of a bored railway tunnel (Railtrack, 1996)
The boring method usually consisted of sinking shafts or driving adits from the ground
surface at various points along the line of the tunnel and excavating out laterally from the
CIRIA C671 • Tunnels 2009
11
base until the tunnel was completed. This allowed simultaneous excavation on several
faces to speed construction. The construction shafts were either left open for ventilation or
often sealed and backfilled with spoil to create blind or hidden shafts (Section 2.2).
A very basic classification scheme was adopted by the Victorian engineers to characterise
the ground, as summarised in Table 2.2. This was used both in cost estimation and
support design.
Table 2.2
Classification of ground conditions in the 19th century
Term
Description
Heavy ground
Ground that was not self-supporting, where excessive pressures were imposed on the excavation,
requiring considerable temporary and permanent support – typically deep tunnels.
Light ground
Ground that was self-supporting and required little or no temporary support during the
construction of the lining – typically shallow tunnels.
Gripper (1879) suggested that heavy ground could be expected at depths under 50 ft
unless tunnelling in a strong geological formation. However, this classification scheme
failed to take account of the huge range of possible ground conditions, and particularly
the effect of groundwater. This has led to a legacy of problems that continues to this day,
including many early tunnel failures some of which are described in the case studies in
Appendix A1. Table 2.3 provides a summary of the change in construction methods
adopted over time for Victorian UK rail tunnels (NR, 2004a).
Table 2.3
Change in construction methods over time
Year
Construction method
1830
Predominantly shaft sinking followed by horizontal excavation from the base. Shaft diameter was
kept to a minimum although many shafts were sunk.
1850
A combination of pneumatic rock drills, blasting and hand labour was used. Fewer shafts required
and the diameter of some of the shafts was increased.
1886
Compressed air machines used in the UK with air locks and shields. The number of shafts sunk
reduced and the diameter increased.
The age of the tunnel gives an indication of the likelihood and number of shafts, and the
probability of encountering problems that are associated with shafts. Over time, improved
excavation techniques were accompanied by improved design through greater
understanding of stress redistribution around underground openings. Tunnel profiles
developed from vertical sidewall, vaulted crowns and flat inverts to parabolic or elliptical
shapes (see Figure 2.1). Tunnels may have been constructed either with or without an
invert. Early tunnels generally bore onto the underlying ground through footings. The
lack of a structural invert frequently leads to problems, particularly where the drainage is
blocked.
In canal tunnels tunnelling typically involved the excavation of a top heading between
shafts, with the roof supported by timbers in light ground. Once complete, the topheading provided a haulage route for spoil as the bench (second heading) was excavated
to the base level of the invert, as shown in Figure 2.4. A masonry lining would then be
constructed from the invert upwards. Canal tunnels were generally constructed to allow
the passage of one seven foot narrow boat at a time, but some longer tunnels on important
canals, for example, Blisworth tunnel on the Grand Union Canal (then the Grand
Junction Canal), were built wider to allow narrow boats to pass one another within the
tunnel. Typical cross-sections of wide and narrow canal tunnels are given in Figure 2.5.
12
Figure 2.4
Typical excavation sequence for canal tunnels (courtesy Jack Knight)
Figure 2.5
Typical cross-sections and dimensions of narrow and wide canal tunnels
(courtesy Jack Knight)
Around the time of the Industrial Revolution many bored rail tunnels were constructed
using the English method (see Figure 2.7) whereby an initial bottom haulage heading was
advanced ahead of the main tunnel. An upper heading was then constructed and the
crown supported. The top heading was first enlarged to the full tunnel width, followed by
the lower part to complete the full face. The lining was constructed from invert up,
supported by timber props. The gap between the outside of the lining and the ground was
CIRIA C671 • Tunnels 2009
13
then often filled by one of the methods shown in Figure 2.6. Where temporary support
was provided by timber crown or drawing bars (wooden logs indicated in Figure 2.6d),
complete infill was difficult as timbers were drawn forward leaving an unfilled void. The
amount of overbreak requiring to be filled was controlled by ground conditions and care
was taken with excavation. Typically, drill and blast tunnelling lead to greater overbreak
and larger voids behind the lining. Lining types are discussed in detail in Sections 2.3 to
2.5. Other methods adopted included the German system in which multiple box headings
were advanced before completion of the top arch and central section, and the timberintensive Austrian system that started with a robustly constructed central heading followed
by a crown heading and full face excavation (Muir Wood, 2000).
In light ground, where no lining was installed, rockfalls were common, exacerbated by
groundwater inflow loosening rock blocks surrounding the tunnel.
During the early industrial period, engineers often had very little or no information on
ground conditions and generally used a reactive approach to any problems encountered.
Some problems that affect these tunnels today are associated with this reactive approach,
including:
temporary timber supports built into the tunnel lining if removal would have caused
failure during construction (with time these weaken and rot with the potential to
create local instability)
large annular voids due to overbreak
variable construction methods, with combinations of cut-and-cover and bored
techniques to reduce costs or reflecting weaker ground at or near the tunnel portals
poor packing of material to fill voids.
In light ground (typically rock) some tunnels were left unlined but were provided with a
structural portal to safeguard against weathering and unstable ground around the portal
area and to give the tunnel an aesthetically pleasing appearance from the outside.
a
b
Figure 2.6
c
14
d
Typical annular infill
for lined tunnels
(Railtrack, 1996)
2.1.3
Excavation methods
The means by which ground was removed as the tunnel advanced has a significant
bearing on the current condition of the tunnel. Three excavation methods were used:
1
By hand (for soil or weak rock).
2
Drill and blast (for rock).
3
Mechanical excavation (for soil or rock).
Hand excavation
Hand excavation was widely used in the past for digging tunnels through soft ground or
weak rock. Early tunnels were dug using hand-held tools. More modern tunnels were
excavated using pneumatic clay spades. However, recent health and safety legislation
regarding hand-arm vibration problems (commonly known as white finger) has restricted
the use of this technique in recent years.
Figure 2.7
Hand-excavation and spoil removal from the top-heading of a tunnel, using the
English method of construction popular in the 19th century, showing temporary
timber supports. A completed bottom heading is also visible
Of all the excavation techniques, hand excavation causes the least disturbance to the
ground that remains in situ, and overbreak is minimised.
Drill and blast
Drill and blast has been in use for tunnel excavation through rock since the mid-1800s. It
is a cyclic process consisting of:
1
Drilling the face.
2
Charging the drill holes.
3
Firing the round.
4
Ventilating the excavation.
5
Scaling.
CIRIA C671 • Tunnels 2009
15
6
Spoil removal.
7
Supporting the rock walls.
Blasting causes damage to the surrounding rock mass through several mechanisms (Hoek
and Brown, 1980). When an explosive contained in a borehole is detonated, the high
pressure gases generated impact the borehole walls, causing a high pressure wave to be
propagated outwards into the rock. The zone immediately surrounding the blast-hole is
crushed where the compressive strength of the rock is exceeded by the pressure wave.
Beyond this zone, there is a region of radial cracking, formed where the tangential stress
component of the stress field exceeds the tensile strength of the rock. The length of these
radial cracks dictates the extent of the disturbed zone caused by blasting and prescribes
the likely extent of problems in unlined tunnels today.
Records rarely detail the precise method of blasting for old tunnels, so the extent of
disturbance around a tunnel may require direct investigation.
Mechanical excavation
Mechanical methods of excavation involve the cutting of ground by discs, picks or drilling
bits. These tools are mounted on a variety of apparatus, including tunnel boring machines
(TBMs), roadheaders and mobile miners.
A tunnel boring machine cuts a circular tunnel by rotating a circular cutting head against
the ground. The cutting head is fitted with picks to cut soft ground or disc cutters to break
hard rock. The TBM moves forward by extending thrust rams, which may operate either
against the tunnel lining, the tunnel invert or a reaction ring that is secured to the rock by
thrust pads. Spoil is gathered up by arms or buckets on the cutting head that load a
conveyor belt running through the TBM. The conveyor belt can load rail cars, dump
trucks or another haulage conveyor. Variations of TBM include earth pressure balance
machines in which the face is supported by compressed air and slurry machines that mix
spoil with mud to assist removal.
Various researches have concluded that tunnels excavated by TBM show no structural
instability due to excavation disturbance, although rock damage may occur where shear
zones are crossed. One study has suggested that damage by drill and blast is five times that
caused by TBM excavation (Mott MacDonald, 1992).
The roadheader, first developed in the mid-20th century, is a tracked chassis on which is
mounted a pivoted cutting boom. The boom mounts a cutting head fitted with picks, discs
or tricones, which may rotate about, or transversely to, the axis of the boom. Spoil is
collected at the front of the roadheader and conveyed to the rear of the machine. A
roadheader can cut a variety of excavation profiles.
Recent experience of excavation by roadheader in chalk marls (A20 road tunnels) has
shown that the disturbed zone generally extends to two to three joint spacings from the
excavation wall. This is produced by the rock blocks at the face of the excavation being
jostled by the action of the picks.
Also falling within the definition of mechanical excavation is the use of a mechanical
excavator under the protection of a shield. This method will also have caused little
disturbance to the surrounding ground.
Excavation disturbance
Table 2.4 provides a summary of the likely degree of disturbance due to the excavation
method in different ground conditions.
16
Table 2.4
Degree of disturbance due to excavation method
Ground conditions
Hard rock, eg
limestone, granite
Weak rock, eg chalk
Soft ground, eg
London Clay
Era
Method
Excavation disturbance
Ancient
Hand tools, fire quenching
Low
Industrial Revolution
Drill and blast
Moderate to high, depending on
blasting method
Modern
Drill and blast, TBMs
Low to moderate for drill and
blast, low for TBMs
Ancient
Hand tools
Low to moderate depending on
rock structure
Industrial Revolution
Hand tools
Low to moderate depending on
rock structure
Modern
Mechanical excavation, eg road
headers/TBMs
Low
Ancient
Hand
Low
Industrial Revolution
Hand, shield
Low
Modern
Mechanical excavator, TBMs,
cut-and-cover
Low
As a tunnel is excavated, the ground responds due to stress redistribution. The majority of
movement happens shortly after excavation, but movement continues over a period of
years. The behaviour of the ground depends on the ground conditions and the in situ
stress conditions. The principal processes that can cause disturbance to the ground
surrounding an excavation are:
disturbance associated with the excavation technique
disturbance associated with stress redistribution
time-dependent degradation.
Each of these may influence the current condition of ageing infrastructure tunnels.
2.1.4
Stress redistribution
Stresses exist in undisturbed ground that result from the weight of the overlying strata
and its geological history. This stress field is disturbed by the creation of a tunnel. In soft
ground, the stresses try to close the opening and are resisted by lining the tunnel. In rock,
stresses high enough to exceed the strength of the rock may be induced, which could
cause failure unless support is installed. Determination of the stress field around an ageing
tunnel may be required to optimise the design solution for remedial measures where the
problem is stress related.
A typical pattern of stress redistribution around a circular tunnel is given in Figure 2.8,
indicating concentration of stresses close to the tunnel opening. For non-circular
openings, stresses will be concentrated at corners (see Figure 2.9). These locations
typically exhibit the worst deterioration in tunnel lining or loosening of rock surrounding
unlined tunnels.
Stress redistribution can also be influenced by horizontal components of in situ stress that
were present before the tunnel was constructed, skewing the stress distribution pattern. In
extreme cases, this can lead to rock bursting in sidewalls but this is unlikely to be seen in
infrastructure tunnels in the UK.
CIRIA C671 • Tunnels 2009
17
Unlined tunnels can also suffer from the effects of long-term time dependant stress
change as exhibited in some water, sewerage and road tunnels (McQueen, 2005) or
through the action of swelling clays.
The stress field around an existing tunnel can be modelled using various computer-based
numerical methods including finite difference or finite element techniques, or boundary
element methods. These can be carried out in both two and three dimensions.
Figure 2.8
18
Stress redistribution around a circular tunnel (Hoek and Brown 1980)
Figure 2.9
Stress concentrations around a non-circular opening (Hoek and Brown 1980)
2.1.5
Ground failure mechanisms
Once an opening is created, the ground may fail by one or a combination of the
mechanisms summarised in Table 2.5.
Table 2.5
Ground failure mechanisms
Ground/structural
conditions
Predominant consideration
Failure mechanism
Construction
Maintenance
Soft ground
Squeezing and flowing ground, short stand up time.
Soil
Effective shear strength insufficient
Invert failure in softer
ground
Lack of confinement and water ingress. Poor drainage
leading to softening
Blocky jointed rock
Gravity falls of blocks from roof and sidewalls,
controlled by geometry of excavation in relation to
discontinuities in rock mass (Hoek and Brown 1980)
Massive rock
Few stability problems (where stresses surrounding
excavation < approx 1/5 unconfined compressive
strength of intact rock)
Pillar failure
Excessive loading of rock pillars, eg between two
adjacent running tunnels
Joint infill deterioration
Progressive failure due to water ingress washing out.
Discontinuity infilling reducing inter-block strength
leading to loosening
Weathering
Degradation of rock fabric due to chemical or
mechanical action of environmental influences such as
temperature, water or wind
Stress change
Long-term stress changes leading to rock or lining
degradation or failure
External factors
Extra loading from new construction of adjacent
tunnels, building over the tunnel, piling, seismic events,
terrorist attack, traffic loading on shallow tunnels
CIRIA C671 • Tunnels 2009
19
To prevent such ground failure and reduce maintenance, support is installed during
construction or may be enhanced later if conditions change. The type of support falls into
three categories:
1
Temporary support – installed at the tunnel face to prevent immediate collapse.
2
Primary support – to provide long-term stability to the tunnel.
3
Secondary or functional support – installed if the final use of the tunnel requires a
particular surface.
The following sections describe the various continuous support types typically installed in
infrastructure tunnels.
2.1.6
Temporary support
Early unlined tunnels were supported locally using timber props and beams. More recent
unlined tunnels use methods that provide both temporary and primary (long-term)
support:
spot rock bolts to hold individually identified rock blocks or wedges
pattern bolting
bolting with rockfall protection mesh
sprayed concrete.
In soft ground extensive temporary timber propping was used to maintain the opening
before a final lining was installed. The form of the timber support was complex and varied
(Gripper, 1879). A typical layout is given in Figure 2.10.
Figure 2.10
20
Temporary support formwork and replacement with a multi-ring brickwork lining
2.1.7
Primary and secondary linings
Where long-term stand-up time (the length of time an underground opening will stand
unsupported after excavation) is insufficient, primary and sometimes secondary tunnel
linings are installed. These have several functions:
structural support of the adjacent ground
protection of the internal space of the tunnel from water ingress and falling debris
protection of the adjacent ground as a result of deterioration from the effects of
exposure to the air, the passage of vehicles, water flow etc
to provide a regular intrados profile that defines a consistent internal space in the
tunnel and can be used to attach cables and other services
in rock tunnels secondary linings were used for a variety of purposes including
protection, improvements in aesthetics, lighting and ease of cleaning, ventilation
efficiency.
In unlined tunnels, the primary structural support is provided by methods such as rockbolting where a lining is not required for other purposes.
Ground types in which tunnel linings were used include:
self-supporting ground where the lining was conceived as non-structural and not
carrying any load at the time of construction, although loading may have occurred
later
broken ground requiring a structural lining capable of resisting ground pressure
soft cohesive material, such as clay, requiring linings to support significant earth
pressures
non-cohesive material, such as gravel or sand, which caused the greatest difficulty in
tunnelling and required strong fully structural linings to support full overburden
pressure
wet ground in situations where water ingress could be a potential problem.
By the middle of the 19th century, tunnels in sound rock were considered simple work
(Gripper, 1879) but much greater difficulties were faced when tunnelling in unstable
ground, requiring temporary support and the construction of a structural lining. The
influence of ground conditions and disturbance of overlying ground was recognised, and
an important distinction was made between tunnelling in light or heavy ground. The
easier conditions were encountered in light ground, where the depth of the tunnel crown
below the ground surface was relatively great and the ground self-supporting, so that it
did not slump back to rest upon the temporary supports and lining as it was constructed.
In contrast heavy ground was typically encountered where the tunnel crown was relatively
shallow and disturbance from the tunnelling work caused the full load to come upon the
temporary supports (“with a depth from the surface to the tunnel top of 50 feet (15 m) or
less, heavy ground may be looked for, unless the geological formation is a strong one”,
Gripper, 1879). In heavy ground tunnel construction required special measures including
greater temporary support from frequent large timbers (used up to 2 feet and 6 inches
(0.76 m) in diameter) and a thickened lining (exceptionally up to 4 feet (1.2 m) thickness
of brick masonry). Also, the influence of geological features such as faults, joints and joint
orientation, joint fillings and swelling clays were understood and known to present special
difficulties for tunnelling.
CIRIA C671 • Tunnels 2009
21
The conditions at the tunnel extrados can vary significantly. This is related to the type of
ground through which the tunnel was driven and to the construction methods employed.
The possible scenarios are:
voids left between the lining extrados and the ground. The depth of these voids
generally increases towards the crown
voids between the extrados and the ground fully or partially filled with rubble infill,
fallen rock or timber left during construction. In some brick lined tunnels there can
also be brickwork piers (sleeper walls) between the extrados and the ground, as shown
in Figure 2.4
solid contact between the ground and the lining. This situation could be achieved in
the original construction, by positioning the lining directly against the ground or by
tight backfilling of any voids. The situation can be reached as a result of movements of
the ground after the construction of the lining.
Three principal primary lining types are identified in this guide, based on their material
type:
1
Masonry linings (including brick and stone).
2
Metal linings (including cast iron and steel).
3
Pre-cast concrete linings.
These are discussed in Sections 2.3, 2.4 and 2.5 respectively.
2.2
CONSTRUCTION SHAFTS
In the 19th century, hand-excavation of tunnels required the construction of several
vertical shafts at intervals along the tunnel’s horizontal alignment (and sometimes also
adits at other orientations) to allow excavation to proceed from many working faces, and
for disposal of spoil and ventilation. Once complete, some shafts were left open for
ventilation, while others were infilled and/or capped off to avoid the burden of
maintaining them. However, all shafts both open and closed (whether backfilled or not)
present a maintenance liability today, as described in Section 2.7.
2.2.1
Shaft construction techniques
Early shafts were sunk using a method known as steining in which an oak or elm
hardwood diaphragm ring was used as a base for the construction of the shaft lining, so
that the ring and the lining sank as the soil was excavated from beneath. Any voids behind
the lining were packed as work progressed. The steining method sometimes incorporated
a type of shaft shield called a barrel, this ensured safe excavation without the threat of the
shaft walls collapsing inwards. The sides of these barrels were designed to allow it to sink
gradually. Plumb-bobs were used to ensure the shaft maintained its verticality.
Another popular method was to construct the shafts in lengths (see Figure 2.11). A depth
of shaft was excavated and a diaphragm oak or elm ring placed at the bottom. A
brickwork shaft lining was then constructed. Once the length was complete, the next
length was excavated leaving a bench of earth beneath the diaphragm ring. Sections of the
earth bench were removed and props inserted to support the shaft lining above. The
remainder of the bench was then removed and another diaphragm ring positioned. The
next section of lining was then built between the props, which were either removed or
built into the shaft lining if they were too difficult to remove. The lengths of the shaft were
constructed with the lining keyed into the surrounding ground, and the annular void
22
packed tightly with dry earth. In an attempt to waterproof a shaft, clay plugs were
installed through water bearing horizons or a clay backing was rammed behind the lining.
It was considered important to obtain a bond between the shaft and the ground to lessen
the weight on the tunnel lining through to the shaft eye.
Some construction shafts were relatively short in depth and were excavated without
permanent linings, but with the sides of the shaft shuttered with lagging boards and
waling to safeguard against debris falling.
2.2.2
Shaft eyes
Shaft eyes were constructed using brickwork, concrete, cast iron or stonework with
complex temporary support, as shown in Figure 2.12, and with the final appearance as
shown in Figure 2.13. The load passing through the shaft eye into the tunnel lining is
difficult to assess. The skin friction afforded by the shaft lining and the extent to which the
shaft is keyed into the earth is very difficult to calculate, so if available, contemporary
records can provide valuable information. Ingress of water through the shaft and/or the
absence of a drainage system serve to reduce skin friction and material integrity. The shaft
eye is under vertical load from the shaft and also circumferential load from the tunnel
arch ring, resulting in a wedging action that serves to stiffen the eye. The shaft eye may be
subject to shear failure.
Figure 2.11
Typical railway tunnel shaft construction details (Railtrack, 1996)
CIRIA C671 • Tunnels 2009
23
Figure 2.12
Temporary support detail at shaft eye
Figure 2.13
Typical rail tunnel shaft eye construction details (Railtrack, 1996)
2.2.3
Closed shafts (blind shafts)
Open tunnel shafts will always remain a maintenance liability and should be examined in
the same way as any other component of the tunnel. Some shafts required only for
construction were closed when the tunnel was completed, although such construction
shafts were often left open to provide extra ventilation. As such, they rarely appear on
design drawings. With the passing of the steam age, the need to ventilate tunnels has been
reduced and, in a bid to reduce maintenance costs and to stabilise shafts, some of these
have now been filled in. The possible current states of shafts are illustrated in Figure 2.14.
24
Type:
Figure 2.14
1
2
3
4
5
Possible states of construction and ventilation shafts (Network Rail, 2004b)
1
Open shafts (see Figure 2.14, Type 1) are those that were left open for ventilation
purposes.
2
Blind shafts (see Figure 2.14, Types 2 and 3) are temporary construction shafts that
have been sealed or capped so that they are discernible from within the tunnel, at the
ground surface or both.
3
Hidden shafts (see Figure 2.14, Types 4 and 5) are temporary construction shafts that
have been sealed or capped so that their location is not visually discernible.
Loosely packed spoil or brick rubble was used to fill redundant shafts following the
completion of the tunnels. In more recent times a variety of materials have been used that
stabilise a shaft, but do not impose excessive load on the tunnel roof at the base of the
shaft. The latter is strengthened before the filling operation. Reinforced concrete saddles
have been used for this purpose. Further details of shaft capping and filling techniques
are included in Section 5.6.
Figure 2.15
Examples of open and closed shafts in a brick-lined tunnel. The closed shaft, on the right, has
been capped off just above the eye so is easily visible, but this is often not the case
CIRIA C671 • Tunnels 2009
25
Shafts were often part-filled with an arched support some distance from the top of the
shaft or, wherever the engineer judged adequate support could be gained, such as at the
level where the rock head was encountered. Determining the status of a shaft is difficult if
adequate records were not kept or are no longer available. If a shaft is infilled completely,
the tunnel lining directly beneath the shaft should be examined very closely for signs of
distress, especially following a prolonged spell of wet weather, which may have the effect of
increasing the overburden pressure on the lining. Some shaft caps were constructed using
timber decking, and the condition of old timber structures varies significantly.
Knowledge of closed shafts associated with a tunnel is extremely important but finding
them can prove challenging, even with contemporaneous construction records. Once
infilled, a shaft was rarely recorded by the engineer. Examples of amended construction
drawings are also rare. Moreover, few edits were made detailing the (then) new status of
filled shafts and over the years many of the records that were amended were stored
improperly with many being lost altogether. The records that do exist, such as
construction drawings and site notes, indicate the whereabouts of the shafts and how they
have been infilled. The method engineers often used to monitor progress was to write on
the construction drawings, along the line of the tunnel, dates when excavation reached a
certain point so that on completion of the tunnel there was a series of dates plotting its
progress from start to finish.
Multiple headings were driven to speed up the excavation and reduce long construction
periods. Sometimes additional shafts were sunk, which were not recorded on construction
drawings. Progress dates written on the construction drawings often give the appearance
that headings were being excavated away from a point underground without a shaft being
sunk for access. Other indications of the presence of a hidden shaft are marker stones or
mounds of rubble on the ground above the tunnel. Circular wet patches on the crown of
the tunnel are not uncommon and may give an indication to the presence of a blind shaft.
However, there are certain patterns and features pointing to blind shafts. In capping the
shaft and supporting the shaft eye the support timbering, called sills, were often propped
off the tunnel lining at haunch level. After the sills were removed, filling in and sealing the
holes left by the timber beams proved to be difficult. Careful scrutiny of the lining at such
locations may show a row of small wet patches where the sill beams were located.
Furthermore, the bond of the brickwork is often irregular where the sill holes were
plugged up. The lining thicknesses beneath capped and open shafts will usually be thicker
than elsewhere within the tunnel, but only over a short length of tunnel.
Techniques for locating hidden shafts are addressed in Appendix A5.
2.3
MASONRY LININGS
Masonry, both brick and stone, was the predominant material used in the construction of
tunnel linings for the rapid development of the canal, rail and sewerage systems in the
19th century. The majority of these tunnel linings were of brick. However stone masonry
was often used regionally where stone supplies were readily available (typically in Scotland,
parts of Wales and the south-west of England), and some linings are hybrids of stone and
brick masonry. Wherever possible the material excavated during tunnel construction was
used to produce the lining material on-site. Clay spoil was used to produce bricks while
sands and possibly lime were used for the production of mortar. Where bands of rock
were encountered they could be cut to shape and used to line other parts of the tunnel,
reducing the requirement for more labour-intensive brick and mortar production.
Linings constructed from stone masonry were typically of ashlar, consisting of a single ring
of relatively large and heavy, regularly cut and shaped stone blocks with thin mortar
26
joints. Their size and weight made individual blocks difficult to handle and place,
particularly in the higher parts of the tunnel arch, and so hybrid tunnels, with thick stone
sidewalls supporting brickwork arches, were sometimes constructed.
In most parts of the UK the lack of easily-won stone and the relative availability of the raw
materials for brick-making meant that brickwork was the material of choice for lining
construction. Brickwork also had the advantage of comprising small, easily transported
and handled regular units that were suited to the construction of arches in confined
spaces and did not have to be cut to shape according to their position in the arch.
However, to achieve the necessary lining thickness typically required construction of
several skins of brick in sidewalls and arches with multiple rings, and a relatively large
supply of mortar because the proportion of mortar in brickwork is generally much higher
than in good quality ashlar stonework.
2.3.1
Lining profile, thickness and quality
Masonry linings are typically built as arches. As with masonry arch bridges, the lack of
structural knowledge or understanding of soil or rock mechanics resulted in an empirical
approach to masonry arch tunnel design. A variety of cross-sectional profiles were
adopted, but earlier tunnels tended to have straight, vertical sidewalls supporting an arch.
This profile is typically found in early canal tunnels (see Figure 2.5) and rail tunnels where
it was a more efficient profile to accommodate trains without unnecessary excavation.
Later construction frequently adopted curved sidewalls to achieve a more structurally
efficient ovoid transverse profile. In rail tunnels a horseshoe shape (see Figure 2.16) was
frequently adopted, whereas in sewerage systems a circular or egg shaped profile was
common. Lining thickness was determined by simple rules of thumb, dependent on the
type of ground (light ground being self-supporting and heavy ground not) and past
experience based on trial and error.
Figure 2.16
Rail tunnel (Clifton Hall tunnel) with multi-ring masonry lining and structural invert
CIRIA C671 • Tunnels 2009
27
Figure 2.16 shows an original drawing of a 19th century rail tunnel with multi-ring
masonry lining and structural invert supporting a central drain. Note the bonding pattern
between the rings, the all header bond in the arch and invert brickwork, and the brick
piers in the backfilled void above the crown extrados connecting the arch with the
ground. Also note that the sidewalls, footings, invert and most of the arch use Lias mortar
(a natural cement) whereas the crown has been constructed using cement (an early
Portland-type), which would have been stronger and less tolerant of movement, but more
expensive so used sparingly. In old multi-ring tunnel linings the presence and nature of
connections between arch rings is highly variable, sometimes even within the same tunnel,
and may have a significant influence on structural behaviour. The pattern of bonding
shown in Figure 2.16 is unusual (compared with a more commonly used pattern shown in
Figure 2.17) and may have been a contributory factor in the catastrophic collapse of this
tunnel in 1953 (see Case study A1.18, Section 18.2).
Brickwork linings in heavy ground might comprise six rings of brick, although they could
be significantly thicker where difficult conditions were encountered. Fewer rings were
used for construction in better ground, with linings of four rings in thickness being
common. Non-structural linings could be constructed as thin as two rings. The excavation
methods employed resulted in some irregularity in the ground profile and, just as it was
not easy to avoid overbreak, underbreak was also common and would be accommodated
by a local reduction in lining thickness. The possible ring thickness of a given tunnel can
be assessed using contemporary engineering guides such as Molesworth’s Pocket book of
engineering formulae (Molesworth, 1862).
2.3.2
Lining construction
After excavation, masonry linings were built up from foundation level using wooden
falsework (centerings) to form the arch. Excavation and construction proceeded in lengths
that were often about four yards (3.7 m) in good ground, although they could be less than
2 yards (1.8 m) in poor ground. In each length the arch was built so that the extrados
profile was below the thick longitudinal timber bars supporting the roof of the excavation,
so that between the lining extrados and the excavated ground profile there was a
relatively large space containing timbers and empty voids. At the completion of each
length, these longitudinal bars were drawn forward, sliding them forward to support the
heading for the next length of construction. Because it was understood that voids between
the arch extrados and the ground were undesirable (“it should be an invariable rule never
to leave a vacuity behind the work”, Simms, 1844), the void space was packed solid with
clay, broken bricks or other material. Another common practice consisted of building
sleeper walls (small masonry pillars) off the extrados to support the ground on the lining
(visible in Figure 2.16).
Figures 2.10 and 2.16 show common profiles through a masonry-lined tunnel with the
space between the lining crown and the ground filled with brickwork piers and rubble
packing. The large longitudinal timbers often used could, despite the efforts of tunnellers,
sometimes become wedged in place and could not be withdrawn so had to be left in
position (eventually rotting to form voids with loss of support to the ground). It was not
uncommon for improperly supervised workers to leave the area between the lining
extrados and the ground unfilled or not packed tight.
Considering the particularly difficult conditions of construction, workmanship of masonry
tunnel linings is often surprisingly good. However, instances of poor workmanship and
variations in construction are known to occur. This can mean irregular thickness and/or
shape, inadequate brick mortar bond or complete lack of mortar and temporary elements
such as timbers left to rot within the lining. Behind the visible intrados there was always
the temptation for poorly supervised contractors and labourers to cut corners, for instance
28
by laying bricks dry (without mortar) or even by constructing fewer brick rings than
required. Many of these features have the potential to affect the tunnel’s performance and
structural capacity but are not directly visible at the intrados, making them difficult to
identify.
2.3.3
Inverts and drainage
An example of a tunnel with a structural masonry invert and drainage culvert is shown in
Figure 2.16. The presence (or absence) of an invert is an important influence on the
behaviour of the masonry-lined tunnel. While in light ground tunnel linings typically
rested on stepped footings, tunnels in heavy ground were typically constructed with a
structural invert to stiffen the lining and provide restraint to settlement and movement of
the sidewalls and arch. This also provided a sound ground surface on which to place
ballast or construct a roadway, dependent on the tunnel’s intended use. In road or rail
tunnels, which it was necessary to keep dry, central culverts were often included for
drainage to one of the portals. Unfortunately, maintenance has often been neglected and
drains have frequently become silted up or been damaged by other maintenance
operations such as tamping of ballast. Flooding is a problem in some rail tunnels because
of the adverse effect on tunnel equipment and the softening of the trackbed leading to
pumping and deflection of rail level and alignment. Furthermore, excessive water can
lead to softening of the invert and loss of support to the tunnel side walls resulting in
structural defects in the lining.
2.3.4
Brickwork bonds
Several types of masonry bonds are possible, particularly in multi-ring arches where it is
necessary to deal with the difference in radius of individual brick rings. This requires the
bedding joints between adjacent rings to be offset, or kept parallel with varied joint
thickness. Examples of different bonding methods for multi-ring arches are shown in
Figures 2.16 and 2.17.
Typically tunnels consist of English bond throughout (known as single ring work) or
English bond up to springing level only with running bond (all bricks laid as stretchers,
bonded by headers where made possible by the coincidence of courses in adjacent rings)
in the arch above. The advantage of using running bond in the arch is that the mortar
joints remain regular and relatively thin throughout the arch barrel’s thickness. If English
bond is used the mortar joints need to become thicker toward the extrados, resulting in a
high proportion of mortar to brick in the rear rings of a lining, which may be six rings or
more thick. Because mortar (and particularly the lime-based mortar used in many tunnels
in the 19th century) is weaker and less resistant to deterioration than brick, it was
recognised by some engineers, that such construction might be less durable, particularly
where the ground was wet. They also recognised that the lower modulus of the mortarrich outer rings might allow them to yield more under load, resulting in stress being
concentrated in the stiffer intrados ring. However, some engineers still advocated the use
of English bond because they considered regular bonding between brick rings was
necessary to ensure that the arch acted in a structurally composite manner. Frequently the
method of masonry arch bonding is similar for tunnels constructed for the same owner,
due to particular policies being adopted.
CIRIA C671 • Tunnels 2009
29
Note
Many other methods have been used and often headers have been added wherever the brick courses
in adjacent rings have coincided, or they may be completely absent so that there is no connection
between rings in the arch
Figure 2.17
An example of one method of brickwork bonding for masonry arches
(courtesy Jack Knight)
2.3.5
Construction joints in brickwork
Each constructed length of lining was keyed into the alternating offset toothings of the
courses in the preceding length. Because slight settlement frequently occurred between
the construction of one length and the next, or due to slight differences in the level of the
initial course of brick, joints between lengths were often slightly irregular. They can often
be seen at the intrados as slight distortions in the bedding planes and irregularities in the
mortar thickness, forming vaguely discernible linear features regularly spaced along the
tunnel circumference (see Figures 2.18 and 2.19). Where difficulties were encountered in
tunnelling, such as an area of weak and unstable ground or water ingress, the construction
lengths were often reduced to two yards or even less. The lining might be thickened to
provide more structural stability, and/or an invert added for rigidity and resist
convergence of the tunnel walls. Joint spacings, particularly changes in spacing, can
provide important clues to the hidden structure of the tunnel, changes in ground
conditions behind the lining and even the location of hidden shafts.
More detail on the significance of construction joints and their use in interpreting ground
conditions and other features of a tunnel’s construction are given in Section A4.1.4.
2.3.6
Masonry materials
The response of masonry to loads is influenced by the way in which the materials have
been used in its construction, their original physical characteristics and any later changes,
including deterioration. Differences can also be expected between masonry elements with
different joint widths (proportions of mortar) or incorporating mortars of different
strength and compressibility.
30
Figure 2.18
A construction joint picked out by its shadow using low-angle lighting
a
b
Figure 2.19
Two views of construction joints: clearly visible joint where dog-toothing is
absent (a) and joint is more difficult to spot, but is marked by subtle irregularity
of brickwork and slightly wider vertically aligned joints (b)
2.3.6.1
Mortar
The functions of mortar in masonry are:
to provide an even contact surface between the masonry units (brick or stone) and to
promote even load transfer between them, avoiding excessive local stresses that might
otherwise develop at points of contact
to physically bind the masonry units together as part of the masonry fabric and
allowing it to function as a composite material, ie by influencing its important physical
characteristics such as compressive strength and modulus of elasticity (the use of weak
and flexible lime-based mortars conferred upon masonry arches some degree of
articulation and allowed them to respond plastically to stresses)
CIRIA C671 • Tunnels 2009
31
to provide a preferential pathway for the movement of moisture through a masonry
structure, allowing it to breathe, and to act as a sacrificial component where
deterioration would be concentrated, rather than in the masonry units themselves.
In stone masonry of well-cut ashlar, blocks typically rest directly on one another or on a
very thin bed of mortar, which was probably used as much as a lubricant during
construction as for any other reason. In such construction, the behaviour of the masonry
is principally dependent on the properties of the stone. Where stone was less well-dressed,
thicker mortar beds were required to provide a uniform bearing surface, and the mortar
becomes an important influence on masonry behaviour. Likewise, mortar characteristics
become more significant to the structural behaviour of the masonry where it is present in
high proportions, eg in rubble masonry and mortared rubble fills. Similarly, in brickwork,
strength and other aspects of structural behaviour will be influenced to some degree by
the physical characteristics of mortar, depending on the thickness of mortar beds.
In thin mortar beds between well-dressed stone, mortar is likely to be in a state of triaxial
compression, where its strength is less significant. Conversely, in thicker and less even
beds, such as those between uncoursed stone, mortar is more likely to be subject to nonuniform stresses and there is a greater potential for compression, with the potential for
load transfer directly between masonry units.
Lime mortars
Traditionally, masonry mortars were produced using lime cements, the lime (calcium
carbonate, CaCO3) typically being derived from natural limestone, including chalk. The
characteristics of the lime mortars were dependent on the nature of the raw materials used,
the presence of any impurities or additions, and the process of production, particularly the
firing conditions in the kiln. Lime cements can exhibit different properties:
pure limes (also known as fat limes) are produced from pure limestone or similar
materials. They are non-hydraulic cements that harden slowly by reaction with
atmospheric carbon dioxide only, known as carbonation or air-setting
hydraulic limes are those which do not rely entirely on reaction with atmospheric
carbon dioxide to set, but also have some element of hydraulicity, ie set by chemical
reaction with water. This can result from the inclusion of clay impurities in the
limestone raw material or from the direct addition of hydraulic material to it, which
affect a setting action when mixed with water.
Mortars based on pure and hydraulic limes can exhibit significant variability in their
characteristics, principally related to their degree of hydraulicity. At one extreme, a very
pure lime that relies entirely on air-setting produces a mortar that strengthens and cures
very gradually over long periods, but remains relatively weak and plastic, typically with a
crumbly texture like a dry crumbly biscuit. At the other extreme, a mortar based on a
strongly hydraulic lime sets and achieves a higher strength much more rapidly in the
presence of adequate moisture.
Hydraulic limes, natural cement and Roman cement
Many of the lime mortars used in the past have been produced from limestone with some
degree of clay impurity, which naturally imparted a degree of hydraulicity to the mortar
without any further processing. As the degree of hydraulicity of the lime increases, its
characteristics become less like those of pure lime and more similar to those of a Portlandtype cement, ie exhibiting more rapid set, greater strength and brittleness and lower
permeability. Traditionally, hydraulic limes were classified as eminently hydraulic,
moderately hydraulic and feebly hydraulic. Natural cements could be produced by
burning certain clayey-limestone at moderate temperatures to produce something that
32
had some of the properties of Portland-type cements. They were sometimes called and
marketed (misleadingly) as Roman cement and could be used to produce a relatively
strong, dense and impermeable mortar, commonly used for construction and repair in the
19th century (notably including Brunel’s Thames Tunnel). The ability of natural cement
to set hard quickly in wet conditions and resist the action of water (when compared to
more weakly hydraulic limes) meant that it was particularly in demand for the
construction of tunnels. However its relatively high cost compared with lime mortar
meant that in brickwork linings it was sometimes specified for use in the extrados rings
only, or to provide cement covering to the rear of the extrados (Simms, 1844).
Modern cements and gauging of mortar
Another method of introducing an element of hydraulicity to a lime mortar is by adding
hydraulic material to the mix. Since the late 19th century this has meant the addition of
Portland-type cements, which rely entirely on chemical reaction with water to set, so are
fully hydraulic cements. These were included in mortar mixes in an attempt to overcome
the potential disadvantages of traditional lime mortars, ie that their weakness did not suit
them to use in the newly developing thinner-walled masonry designs, and that they can
take a very long time to cure and harden, especially in wet conditions when susceptible to
damage from water and frost. However, mortars based only on cement and sand at an
optimal ratio of 1:3 are too harsh and difficult to work with, and produce a mortar that is
too strong and inflexible for most applications, so the cement was instead added to lime
and sand mixes to impart hydraulicity to the set by a process know as gauging (the
proportioning of materials by volume). Such cement:lime:sand mixes are the most
commonly used today for general purpose masonry.
Before the 20th century and the prevalence of modern cements, lime mortars were
commonly gauged with other hydraulic materials known as pozzolans. The most popular
of these were types of volcanic ash imported from the continent (for example, Trass or
Tarras from Belgium). However brick dust was more cheaply and readily available and
could provide an element of hydraulicity to mortar mixes. Brick dust was commonly used
and has been found in the mortar of some 19th century tunnels.
Table 2.6 gives mix proportions for mortars and indicative ranges for their compressive
strengths, and also values for the compressive strength of the masonry produced using
different brick strengths (assuming standard brick dimensions).
Table 2.6
Mortar mixes and compressive strengths used in the UK, and corresponding strengths of
masonry using different bricks (from Sowden, 1990)
Type of mortar (by volume)
Mortar
designationa
Cement:
lime:sand
Masonry
cement:
sand
(i)
1:0–0.25:3
–
(ii)
1:0.5:4.5
1:2.5–3.5
(iii)
1:1:5–6
(iv)
Mortar
strength
range
(N/mm²)
Cement:
sand+
plasticiser
Brick strength (N/mm²)
7
20
35
50b
70c
Characteristic compressive strength
of brickwork (N/mm²)
11–16
3.5
7.5
11
15
19
1:3–4
4.5–6.5
3.5
6.5
9.5
12
15
1:4-5
1:5–6
2.5–3.6
3.5
6
8.5
11
13
1:2:8–9
1:5.5-6.5
1:7–8
1–1.5
3
5
7
9
11
(v)
1:3:10–12
1:6.5-7
1:8
0.5–1
2
4
6
7.5
8.5
(vi)
0:1:2–3d
–
–
0.5–1
2
4
6
7.5
8.5
(vii)
0:1:2–3e
–
–
0.5–1
2
3
3.5
4.5
5
Notes:
a
b
From BS 5628–1 (BSI, 1992)
Class B engineering brick
CIRIA C671 • Tunnels 2009
c
d
e
Class A engineering brick
Hydraulic lime
Pure lime
33
Both English Heritage <http://www.english-heritage.org.uk> and The Scottish Lime
Centre <http://www.scotlime.org> can provide information and support on the use of
lime cements. Sources of lime materials are listed in the Teutonico (1997). Further
information on lime cements, aggregates and mortars is available in Ashurst and Ashurst
(1988), Ashurst (1997), Allen et al (2003) and Ellis (2002).
2.3.6.2
Stone
Stone is a term applied to construction material quarried from a natural rock resource. It
is one of the oldest building materials known to man, and since the earliest times of
civilisation has been the preferred material for the construction of permanent and
important buildings. As well as its aesthetic appeal, the most notable feature of stone is its
potential for exceptional durability. However, the stone used in tunnel linings was often a
by-product of excavation rather than a material selected specifically for its durability.
Some existing tunnels have problems with stone deterioration either because of original
poor selection and use, subsequent implementation of inappropriate repairs, or because
deterioration has been hastened by the harsh conditions of the tunnel environment.
A wide range of rock types have been used as building stone (see Table 2.7) but in the UK
the most commonly used were the sedimentary rocks limestone and sandstone, and, in
some areas (particularly the north and west of England, Scotland and Northern Ireland),
igneous rocks such as granite. A description of the range of rock and stone types available
and their geological and engineering characteristics is beyond the scope of this document,
which is limited to a brief discussion of some of the principal factors affecting their
principal characteristics, selection and performance in service. For detailed information
see Smith (1999).
Table 2.7
Comparison of typical strength and density values of some common UK building stones
with other construction materials (after Geological Society, 1999)
Typical compressive strength
(kN/m²)
Typical density (kg/m³)
Stokeground Bath – base bed
22.5
2126
Stokeground Bath – top bed
13.8
1988
Portland Roach
52
2100
Portland Whitbed
36
2200
Welsh Blue Pennant
158
2630–2850
Clipsham limestone
32
1826
Woodkirk Yorkstone
54
2400
100–350
2500–3200
Concrete (typical)
40
2240
Bricks (typical commons)
20
1800
Bricks (engineering class A)
70
Up to 2800
Masonry unit material
Granites
2.3.6.3
Brick
Clay bricks are produced by firing natural clay at high temperatures until the clay
minerals melt and fuse to form a combination of vitreous and new mineral phases. The
composition and characteristics of the fired brick depend on the original composition of
34
the clay, and the temperature and duration of the firing process. Brick colour depends on
the raw clay materials used in their manufacture, and can be influenced by the addition of
other minerals and pigments.
Traditionally, clay known to be suitable for brickmaking was dug from the ground and
weathered for some time to dry it, before being mixed and hand-thrown into individual
moulds. The earliest firings were done by heaping the bricks and fuel together and
covering with turf, but simple kilns followed – a single clamp of a brick arch covered with
turf being one of the earliest, followed by round brick kilns. The enormous demand for
bricks by the middle of the 19th century led to the development of the first brickmaking
machines, and kiln design was constantly being improved to increase efficiency.
The enormous quantity of bricks required for the construction of masonry tunnel linings
of any significant length or thickness meant that brickmaking was usually carried out onsite, wherever possible using the raw materials excavated from the tunnel, and where
necessary improved by blending it with more suitable material from the locality. When
introduced during the brickmaking process, the raw materials can be diverse in
composition and condition, so the quality of the bricks within a single structure,
particularly large structures like tunnels, can vary. For instance, bricks that were fired in
the centre of the clamp were subject to burning and baking at high temperatures, tending
to be better quality. In contrast, bricks from the outer part of the clamp were often poorly
fired, weaker and less durable. The fired bricks were graded according to their quality so
that they could be used appropriately, the best being reserved for facing work at the
tunnel intrados. If they were not re-fired, poorer quality bricks were frequently used in
rings behind the facing or as random rubble or fill.
Beyond their original variability, the process of ageing and deterioration of bricks in the
frequently harsh environment of tunnels is another factor that has influenced the current
condition and physical characteristics of their masonry.
Table 2.8
Properties of some old bricks used in bridge and tunnel construction (Railtrack, 1996)
Source
Description
Manufacture
date
Compressive
strength (kN/m²)
Elastic modulus
(kN/mm²)
Sugar Loaf Tunnel (north
of Llandovery)
Large regular handmade
red brick
1850–1900
34.1
6.2
Alfreton Tunnel
Large regular handmade
blue brick
1850–1890
55.2
9.1
Watford Tunnel
Long thin distorted
handmade red brick
1840–1850
22.9
3.9
Bridge 251 near
Grantham
Small rough handmade
red brick
1852
26.3
8.7
Bridge 31 near
Windermere
Small rough handmade
red brick
1840–1850
19.6
1.5
Bridge 49 near Wing
Regular handmade red
brick
1850–1900
28.8
6.8
Bridge 65 on Preston to
Lancaster Line
Large handmade red
brick
1850–1910
25.6
9.4
Harringworth Viaduct
Small rough handmade
blue brick
1840–1850
17.9
12.9
Note:
The cross-sectional area for a small brick is 210 mm² and for a large brick 250–300 mm²
CIRIA C671 • Tunnels 2009
35
Table 2.8 shows the results of tests carried out to discern the principal physical properties
of bricks from a variety of railway structures (bridges and tunnels) over the period
1840–1910, showing the range of variability in compressive strength and particularly in
the elastic modulus.
Research conducted by British Rail (Temple and Kennedy, 1989) involved an extensive
testing programme to determine the compressive strength and elastic properties of
brickwork from masonry structures of different ages (mostly between 1840 and 1910)
from across the UK, and applied statistical techniques to draw general conclusions that
could be used for tunnel assessments. The results, which are summarised in Table 2.9,
illustrate the considerable difference in strength and modulus of old blue bricks
(engineering bricks) and old red or yellow bricks (probably various non-engineering class
bricks eg stock bricks and gault bricks) used in these structures. It was also noted that
individual bricks from the same sample often showed considerable variability in their
physical characteristics. It should be emphasised that the results given in the table are
typical values only and that there are many different types of brick with differing
properties to those quoted.
Table 2.9
Statistical analysis of properties of brick samples from old railway structures (Temple and
Kennedy, 1989)
Characteristic strength
(N/mm²) a
Modulus of elasticity
(kN/mm²)
Poisson’s Ratio
Red and yellow bricks
16.5
5.2
0.11
Blue bricks
70b
15.6
0.16
Notes
a
b
Value exceeded by 90 per cent of bricks tested in a large sample.
Value is suggested, based on typical results from a small sample size.
Further information on the historic production of bricks and their characteristics is
discussed in Hammond (1981).
2.3.7
Structural behaviour of masonry
Masonry is unable to resist tension stresses or bending. So in masonry structures, loads are
resisted only by compressive axial stresses. Masonry structures are geometrical elements
that resist actions only when they can include, within their geometry, a thrust line in
equilibrium with the external loads. In general, from a structural point of view, of the
three conditions any structure has to verify – strength, stiffness and stability – in masonry
structures stability (static equilibrium) is the most relevant, although serviceability
requirements should also be satisfied.
As a result of their inability to resist bending forces, masonry structures under loading will
deform and crack unless they can resist those loads through a path of compression
internal forces. In consequence, cracking is quite common in masonry structures and
should not be automatically associated with structural distress. Also, the durability of
masonry is not as severely affected by cracking as, for example, reinforced concrete, and
in many cases the plasticity of most historical lime mortars will allow those cracks to be
gradually sealed in a process known as autogenous healing, involving the precipitation of
lime dissolved in pore-water.
As a composite material, the stress state of masonry, even under simple loading conditions
is quite complex. As a result of this, under compression, masonry fails by developing
indirect tension cracks in the units, parallel to the direction of load.
36
For further discussion, see Hendry (1998).
A detailed consideration of the structural behaviour of masonry is beyond the scope of this
document. For further information and guidance, refer to McKibbins et al (2006) and for
the derivation of masonry properties for structural assessment, to Hendry (1990). For line
of thrust clarification, reference should be made to Heyman (1982).
2.4
METAL LININGS
2.4.1
Cast iron
Cast iron has been used since the end of the 18th century for permanent linings to shafts
(in 1795 grey iron was used as tubbings in circles for a shaft lining at Walker Colliery on
Tyneside) but it was not until 1869 that it was first used as a permanent lining in a tunnel –
for Tower Subway under the Thames, which is still in use today.
Cast iron segments were assembled into a ring under the protection of a shield, which
temporarily supported the ground as tunnelling advanced. As the shield advanced, the
ground closed around the lining putting the ring into compression, fully supporting the
tunnel. In stiff clays where stand-up time allowed, the ring could be built directly against
the ground rather than within the shield. The joints between segments were sometimes
caulked, typically using lead or a fibrous caulking material.
Cast iron was mainly used from the late 19th century for medium to large diameter road
and rail tunnels constructed in soft ground. The first deep tunnels to use a bolted cast
iron lining were the City and South London Railway opened in 1890 (see Greathead,
1895 for a contemporary account) and the Waterloo and City Railway (1898). Up to the
1940s, cast iron was specified for all deep tunnels for London Underground. With the
demands for raw materials for re-armament ahead of WWII, the subsequent increase in
cost of bolted cast iron, coupled with technical advances in other lining forms, led to the
introduction of concrete linings. However, escalators, station tunnels, concourses etc and
underground railways in water bearing ground still used bolted cast iron.
The result is that the majority of London Underground’s 300 km of deep level tunnels are
cast iron lined. They were constructed with 11 feet 8.25 inches (3.6 m) or 12 feet 6 inches
(3.8 m) running tunnels, and 21 feet 2.5 inches (6.4 m) platform tunnels. The running
tunnel linings typically consist of six segments plus a key at the crown, with three bolt
holes in each 18 inches (457 mm) long radial flange and eight or nine bolt-hole in the
circumferential flanges. While the radial bolts are usually in place, a recent assessment
indicated that up to half the circumferential bolts tend to be missing in some areas (Tube
Lines 2005, unpublished).
Two types of cast iron have been used in tunnel lining segments: grey iron (with free flake
graphite) and spheroidal graphite iron. Grey iron was widely used in soft ground tunnels
until spheroidal iron was developed in the 1960s, and which, with its higher tensile
strength, allowed thinner, wider sections to be designed. However, grey iron continued to
be used in some stiff cohesive ground tunnels, such as on London Underground’s Victoria
line and even the Jubilee line (Fleet line) constructed in the 1970s. All circular road
tunnels under rivers used grey cast iron tunnel linings up to the mid-1960s. Examples
include the Blackwall road tunnel under River Thames (1892–1897), the Mersey
Queensway Tunnel under River Mersey (1925–1934) and the Rotherhithe Tunnel under
the Thames (1904–1908, see Figure 2.20). A contemporary account of the construction of
the Blackwall Tunnel is provided in Hay and Fitzmaurice (1897).
CIRIA C671 • Tunnels 2009
37
Figure 2.20
Original drawings from Rotherhithe Tunnel (1908) with bolted grey cast iron sections
Early linings were generally cast in low quality grey iron with a new design for each
tunnel. Two types were typically used: heavy lining for water-bearing ground and light
section for London Clay. From the 1930s a grading system was introduced, a higher grade
is indicative of increased tensile strength (for further information on iron grading, see BS
1452:1990). Grade 10 or 12 iron (minimum tensile strengths 150 N/mm² and 180 N/mm²
respectively) was generally specified although occasionally higher grades were adopted.
Typical drawing details for a bolted grey iron lining from a London Underground tunnel
are given in Figure 2.21.
38
Figure 2.21
Typical construction and joint details for a London Underground bolted grey iron lining
Expanded (articulated) grey cast iron linings were developed between 1949 and the 1950s
– these were unbolted. A short experimental length was driven in 1958 followed by a 1.9
km length for the LTE Victoria Line in 1960–1961. The lining was designed to be
interchangeable with bolted grey cast iron linings and had six segments per ring and with
a flange depth half the thickness of the bolted lining. The small flange gave a relatively
narrow width for locating the shoes of the shield rams, which made these linings
susceptible to construction and handling damage. There was a saving in weight and cost,
but the main advantage was speed of lining erection and the rate of tunnel advance.
Spheroidal graphite (SGI) bolted cast iron linings have been used in the UK only since the
late 1960s. First experimental use of this lining was a pilot tunnel constructed in June
1968 for an enlargement of a crossover tunnel on the Victoria line extension to Brixton.
By the early 1970s, use of SGI was not economical for small to medium sized tunnels
(below 5 m to 6 m) compared with bolted grey cast iron. However, for larger tunnels the
saving in weight offset the large increase in cost of the material over the grey cast iron
linings. SGI linings continue to be used, most recently on parts of the Jubilee Line
extension.
Behaviour of cast iron linings
A recent survey of LU cast iron lined tunnels (unpublished, Tube Lines, 2005) found that
typical visible defects included:
corrosion and rust effects
cracks, fractures and broken flanges
delamination of the lining segments.
water, silt or sand ingress (Figure 2.22)
missing jointing/caulking material
movement/displacement of segments
open joints
loose and missing bolts/grommets including grout plugs.
CIRIA C671 • Tunnels 2009
39
However, the frequency of such defects is low and the effect on the overall integrity of the
lining is minor.
Figure 2.22
Bolted cast iron lining with water seepage at joint (courtesy Tube Lines)
Research is in progress on the current behaviour of cast iron tunnel linings found in the
London Underground system. London Underground Engineering Standard E3322
allows a basic analysis using the linear elastic continuum model of Muir Wood (1975), as
modified by Curtis (1976). As this is a conservative solution, structural modelling has been
undertaken using 3D finite element programs that analyse stress distribution under
loading and model the behaviour of joints and cracks.
The analysis has shown the following (unpublished, Tube Lines, 2005):
the ground load acting on the back of the lining causes a convex deflection to occur to
the inside of the pan, in turn causing the circle flanges to splay outwards slightly (this
is possible because of the deep caulking groove on the circle flanges, see Figure 2.21)
the stresses between joints are transmitted through the solid contact between pan to
pan and circle flange to circle flange. The radial flanges do not transmit compression,
and are slightly in tension relative to each other, because of the effect of circle flange
distortion causing a similar consistent distortion of the radial flange
the convex deflection of the pan has the effect of attracting tangential stresses to the
centre of the pan. Stresses in the pan at the connection with the circle flanges are
much lower than in the centre
there is evidence from the model that at 40 m depth the radial joints open a little.
Tangential tensile stresses at the crown and invert in the circle flanges also appear to
exceed the allowable values locally, as stresses are concentrated at these points.
Vertical deflection at the crown is about 13 mm (0.68 per cent).
maximum tangential compressive stresses can be seen around axis level, inside the
tunnel.
An example of finite element modelling of a cracked lining is shown in Figure 2.23, with
the results given in Table 2.10.
40
Table 2.10
Example results of finite element modelling of cracked cast iron lining, as shown in Figure
2.23. This suggests that the presence of cracks has only minor influence on tunnel
deformation
Deflections
Model
Stresses
Crown/invert
(mm)
Axis (mm)
Maximum
(tension) (MPa)
Minimum (max
compression) (MPa)
Uncracked
-13.3
8.2
140
-114
Vertical crack
-11.1
14.4
118
-154
Horizontal crack
-13.4
8.4
111
-104
Figure 2.23
Example of a 3D FE model of a cast iron lining incorporating
a vertical crack in the sidewall, shown in white (courtesy Tube lines)
2.4.2
Steel
Steel linings have rarely been used in infrastructure tunnels in the UK due to their high
cost. Typically, they have been used for short lengths of tunnel through particularly
adverse ground conditions or for complex openings or transitional sections. Bolted steel
tunnel linings were used for the Dungeness Power Station cooling water tunnels and a few
other projects in the USA and Europe. Expanded steel tunnel linings were used at Oxford
Circus and King’s Cross stations where tunnelling was close to other structures.
2.5
PRE-CAST CONCRETE LININGS
2.5.1
Lining forms
First used in the US in East Boston in 1892, pre-cast concrete linings were introduced in
the UK in 1903 but were not extensively used until the 1930s. The first standard lining
designs were available from the late 1940s.
Four types of pre-cast concrete lining have been used:
1
Bolted (or dowelled) – suitable for most ground conditions.
2
Expanded flexible – principally used in small diameter tunnels through London Clay.
CIRIA C671 • Tunnels 2009
41
3
Smoothbore grouted – first introduced in 1903 but only available as standard lining in
late 1950s, generally used in soft ground or weak rock.
4
Expanded grouted – used in all ground conditions in modern tunnels.
To prevent water ingress, each type can be used with backfill grouting around the lining
annulus and with or without gaskets between segments. The principal lining forms are
described in the following sections.
2.5.1.1
Bolted pre-cast concrete lining
This type of lining was similar in form to the bolted cast iron linings used extensively on
the London Underground system and could be used interchangeably with them. It was
first used during the construction of extensions to London Underground’s Central line in
1937 as a result of a shortage of raw materials for cast iron. Its main features included
concrete stiffeners to help take shield ram forces and a reduction of the number of bolts
around the circumferential joint from 52 (cast iron) to 31.
While used in air raid shelters beneath London in WWII, the use of bolted pre-cast linings
increased considerably after the war, especially in sewer tunnel construction. Standard
lining designs were available for use in all ground conditions. Where a smooth internal
bore was required, as for a sewer or water tunnel, an internal or secondary lining was
used. These were originally brick or cast in situ concrete or a combination.
In the UK, standard bolted rings were generally 2 feet (0.61 m) wide although widths up
to 2.6 feet (0.76 m) have been used. A typical bolted pre-cast concrete lining arrangement
is shown in Figure 2.24.
Figure 2.24
Typical bolted pre-cast concrete lining
(courtesy TRRL)
42
2.5.1.2
Expanded concrete linings
In the 1940s, it was recognised that bolting of the longitudinal joint of the pre-cast
concrete linings took little bending moment and that bolting of the circumferential joints
was only required to aid erection or, through water bearing ground, to ensure water
tightness. So linings were developed in which the segments were joined only by guides
such as dowels or grooves on the joint surfaces. The lining was expanded against the
ground by either driving a wedge shaped segment into the ring or jacking the ring tight
and backfilling the resulting gap. The use of an expanded lining requires an accurate
tunnel profile although back grouting can be carried out to ensure continuous
ground/lining contact.
The first expanded concrete lining used in the UK was the Don-Seg lining for the
experimental tunnel for the Metropolitan Water Board Thames-Lee Valley scheme
(1950–1951). Expanded linings were first used on medium sized tunnels on the London
Underground experimental length of the running tunnel for the Victoria line in 1961.
Later designs for the Victoria line included the Halcrow lining and the Mott Hay and
Anderson lining (1963). The lining type was also used on the British Rail tunnel at Potters
Bar, in 1955. Details of the lining for this latter project are shown in Figure 2.25.
Figure 2.25
Section through Potters Bar tunnel expanded pre-cast concrete lining (after Muir Wood, 2000)
2.5.2
Casting methods and reinforcement
The pre-cast concrete segments are cast in moulds, and variations in quality of these can
have an impact on the tolerance of the segment dimensions. Out of tolerance segments
lead to imperfect rings and stress concentrations can lead to spalling where there is
uneven contact between segments.
Steel reinforcement may be included in concrete segments to:
increase section resistance to tensile and bending stresses during handling and
erection
withstand permanent ground loading.
Poor reinforcement placement in the moulds resulting in insufficient cover depth of
concrete can lead to the spalling defects often seen in such segments today.
CIRIA C671 • Tunnels 2009
43
2.6
TUNNEL PERFORMANCE
This section considers the causes of loss of tunnel performance. Tunnel performance is
affected by the condition and performance of the structure and the chosen construction
materials. The following sections describe typical issues found in ageing tunnels, divided
into structural and material problems. Particular problems associated with shafts and the
effects of fire on tunnels are dealt with separately in Sections 2.7 and 2.6.3.
2.6.1
Structural deterioration
A failure of a structure can:
lead to injury or loss of life
disrupt traffic flow, which may have detrimental economic effects
cause frustration for users and associated parties/neighbours
damage service and infrastructure furniture that may be housed within the structure,
or cause damage to adjacent services and neighbouring properties
necessitate costly and disruptive remedial/replacement works.
The following factors contribute towards structural instability:
lack of invert or its inadequacy to resist heave or swelling of the tunnel floor
ill-advised alteration works, eg lowering invert, making additional openings,
inadequate repairs, badly specified repair work using unsympathetic techniques or
incompatible materials
abrasion and scour in canal and water tunnels
impact from vehicular traffic
damage through accident, vandalism or terrorism
defects that were built-in at time of construction (for example, cracking of pre-cast
segments due to handling damage of tunnel lining components or from over thrust
from tunnel shields, degradation of gaskets due to poor material choice)
changes in ground loading (for example, by increased loading due to development
above the tunnel or change of in situ stress regime, eg through cliff regression,
unloading of near surface tunnels due to excavations at surface, influence of mining
related ground movement) or changes in internal loading
change in the function of the tunnel leading to changes in environment and internal
loadings
reduction in effective thickness of structural elements, eg tunnel lining from
weathering, corrosion, spalling or erosion due to the flow of water
concentrated loading at joints
A variety of groundwater issues including:
44
settlement induced by softening beneath sidewall footings as a result of excess of water
(naturally or as a result of drainage failure)
increase in loading due to swelling of clays and marls
attack by aggressive groundwater
loss of shear strength on rock joints and bedding planes, which can result in load
being transferred onto the lining in blocky ground
reduction of lining confinement through dissolution of limestone or outwash of fines
causing voids
where there is no invert, saturation of the tunnel floor resulting in inadequate
resistance to possible inward movement of side walls
the introduction of more efficient drainage or drying of the lining as part of a
remedial solution resulting in changes in water flow inducing surface settlement and
possible structural problems at the surface.
Structural failure may develop if defects or degradation of tunnel components are not
addressed, as in the case of water ingress through joints or cracks in the lining leading to
excessive corrosion and ultimate failure. Cracking of a lining may be present as the result
of non-structural defects such as those due to careless handling of lining components
during construction, but may also be a precursor to structural failure, eg due to altered
ground loading.
One key visible sign of structural distress is lining distortion (Railtrack, 1996) of which
there have been many examples. Distortions or bulges can reduce the clearance in a
tunnel considerably and in extreme cases can lead to either inadequate clearances for
traffic, or a localised collapse.
The main cause of distortion results from pressure from the surrounding ground acting
on the lining. Ground movements can be caused by numerous factors and can manifest
themselves in many different ways. Behavioural characteristics of ground vary
tremendously, according to soil or rock type. Rock movements can occur due to
movement along discontinuities such as faults, shear zones and joints in any plane or
orientation. In granular soils, the shear strength of the ground is low. Tunnels excavated
through this type of ground are susceptible to considerable loading. Also, there is the
possibility that load distribution will change due to the porous nature of the soil and the
through flow of water causing erosion and undermining of the lining. Where there is a
gap between lining and ground, increased loading on the lining may be caused by
accumulation of debris from ground movements or by point loads from rock movement.
Clays and fine grained soils are also likely to load tunnel linings. Furthermore, some clays
or soft rocks such as shale and mud rocks, when exposed to air and water, or a reduction
in confining pressure, are susceptible to expansion causing pressure to be exerted on the
lining extrados.
The effect of ground pressure is worsened by some tunnel construction details, such as
poor packing of the lining at the time of construction resulting in a lack of support. This
packing was intended to spread the load evenly across the lining. However, if poorly
executed, or not carried out at all, it causes the pressure to be applied as point loading.
An inconsistent tunnel profile can indicate bulging, which may be a response to changes in
ground loading or structural weakening (see Figure 2.26). Not all bulges are problematic
though - some will have existed for some time, possibly dating back to construction and
the earliest years of the tunnel’s life. During construction, the removal of the formwork
could have resulted in movement of the lining due to the process of loading the lining and
also due to the lining not having achieved design strength. A result of this movement is an
undulating lining surface, which has the appearance of one that is bulged. A lining that
deformed 100 years ago can be a threat to the structural integrity of the tunnel, but rarely
is. Live distortions occur alongside other features such as cracks or open joints, loose or
spalled brickwork.
CIRIA C671 • Tunnels 2009
45
Ground conditions can change due to many factors such as an excess of groundwater or a
geological fault. The structural condition of a lining, in terms of its stress distribution, will
also change as a result. These changes manifest themselves in terms of crushing, cracking,
heaving, bulging and shearing. Such problems are worsened by poor quality lining
material or a lining that has been poorly constructed.
Cracking occurs due to tensile bending, shear and tension within the lining. An example
of tensile bending would be the cracks that appear in the middle of a bulged area. Shear
cracks are a result of differential loading or settlement and appear as a lip or step in the
lining. Tension cracks usually appear in the horizontal plane due to settlement.
Crack types and patterns help to identify the mode of failure of a tunnel lining. Vertical
pressure can result in cracks appearing in the crown, sometimes coupled with crushing
(compression) at the mid-haunch level (see Figure 2.26a). Lateral pressure results in
tension cracks in the sidewalls and haunches with crushing in the crown (see Figures 2.26b
and 2.26c). Details of the annular support conditions help to establish the significance of
the deformation and lining damage. Figure 2.26d shows a local deformation with both
faults arising due to lateral earth pressure. Figure 2.26e shows a total profile distortion.
Where annular support conditions are favourable, the lining can retain its overall shape.
Poor annular support may result in serious profile distortion and where ground pressures
are very large, serious deformation is unavoidable although lining deterioration can be
mitigated by improving annular support.
Where tunnels have been constructed through expansive soils, such as clays, and no invert
has been provided, the soil may expand upwards into the tunnel lifting the formation
(Figure 2.26f). Where an invert has been constructed, rising formation level is indicative of
invert failure. As well as swelling of the underlying strata, other possible causes are
excessive water pressure, mining subsidence and excessive lateral forces on the side walls.
In tunnels built through soft ground where adequate foundations or an invert have not
been provided, the tunnel lining is susceptible to subsidence. The formation level will
often not change but will rise in relation to the sinking tunnel arch resulting in reduced
clearances. In rail tunnels, this can cause particular problems with overhead
electrification.
Such deformations could lead to structural failure if the change in condition is sudden
and no advanced warning given. So it is important to investigate the principal cause of any
defect that becomes apparent through any form of inspection, be it cursory or as part of
an inspection regime that may be implemented as part of long-term maintenance
programme.
46
a
b
c
d
e
f
Figure 2.26
Typical forms of lining deformation in brick-lined tunnels (Railtrack, 1996)
2.6.2
Materials deterioration
Tunnel performance can be influenced by changes in the properties of its structural
materials, ie weathering and corrosion due to external or internal tunnel conditions such
as water ingress or chemical attack. This section considers the principal factors that can
cause such deterioration.
2.6.2.1
Masonry linings
The deterioration of stone, brick and mortar is a complex and wide-ranging topic, and
can only be briefly summarised here. It is worth considering that the majority of
deterioration is related either directly or indirectly to the presence of water and the
chemical contaminants it often contains. This highlights the importance of taking
measures to keep masonry dry, and where this is not possible to allow it to dry and drain
freely. Further information is provided in CIRIA C656 (McKibbins et al, 2006) and
Sowden (1990).
Contributory mechanisms for deterioration of masonry in all types of structures are
summarised in Table 2.11.
CIRIA C671 • Tunnels 2009
47
Table 2.11
Summary of causes of masonry deterioration
Deterioration mechanism
48
Consequences
Freeze-thaw cycling
Where masonry is persistently wet and exposed to repeated freeze-thaw cycles,
this can cause spalling of masonry units and mortar loss from joints. It is most
likely in masonry at or near to portals or open shafts, which are likely to be
subject to a greater number of freeze-thaw cycles (see Figure 2.27).
Physical salt weathering
Transport and precipitation of salts can cause softening, crumbling, flaking,
blistering and laminar spalling of mortar and masonry units.
Sulfate attack
This is generally an expansive reaction between sulfates (present in
groundwater, soil and rock) and components of the cement matrix of mortar
causing its deterioration into a flaky, crumbly non-structural material. Sulfate
attack may also affect bricks and some types of stone with similar results.
Leaching and corrosive attack
The mortar’s calcium hydroxide and calcium carbonate components are
particularly vulnerable to attack by acidic water, and their loss creates
secondary porosity that can weaken materials and in turn aggravates the
effects of other agents like freeze-thaw. In anaerobic conditions, particularly in
tunnels carrying sewage, corrosive hydrogen sulphide may be produced.
Leaching may result in staining and whitish deposits on masonry surfaces.
Biological attack
Tree roots can cause serious damage to the structural fabric of the tunnel even
tens of metres below the ground surface. Other plants can disrupt masonry at
portals. Smaller organisms that may be found in damp areas of the tunnel
fabric can cause deterioration by increasing porosity and facilitating leaching,
and by other mechanisms. The microbial anaerobic conditions can lead to low
pH resulting in attack of grout, concrete and metal.
Repair with unsympathetic materials
The use of overly-hard mortar can lead to masonry units losing their faces and
edges. The use of overly-hard masonry units in repairs can damage adjacent
original fabric. Use of impermeable materials can increase saturation and
redirect moisture into other components or parts of the structure, accelerating
their deterioration. Corrosion of ferrous elements can cause spalling of
adjacent masonry.
Expansion and contraction (thermal,
and wetting and drying cycles)
This can result in internal fracture of the units and spalling, and loss of mortar
from the joints.
Moisture saturation
Units are vulnerable to environmental agents that cause deterioration. The
nature and extent of the saturation is a function of the type and amount of
porosity. Movement of moisture can result in washout of fines from particulate
materials, eg from the ground behind the lining, causing weakening and
instability.
Ground movements
The development of additional stress or change in stress distribution due to
ground movement can lead to cracking or loosening of masonry units, which in
extreme cases can lead to loss of structural integrity of the lining.
Cyclic loading and fatigue effects
Cyclic loading such as from repeated passage of vehicles or trains principally
affects the invert of tunnel structures, unless they are near-surface. There is
little information available on the effects of this action and fatigue exhibited by
masonry linings, but research carried out by Cardiff University of Cardiff
(Roberts et al, 2006) indicates that this does occur and could potentially be of
structural significance.
Figure 2.27
Deep spalling of soft red brick near to a tunnel portal caused by freeze-thaw damage
Figure 2.28
Collapse of part of masonry lining at the waterline in a canal tunnel due to a combination of
deteriorative mechanisms (moisture saturation and leaching, salt weathering and freeze-thaw)
2.6.2.2
Metal linings
Cast iron may corrode, although from experience gained on the London Underground
system in general there is little evidence of deterioration of cast iron linings, as evidenced
by linings removed from various locations and by coring carried out as part of tunnel
assessment projects. This is partly because of the natural corrosion resistance of cast iron
and partly because of the benign conditions inside the tunnels and the grout-protected
environment against the clay on the extrados. One notable exception is at a location south
of Old Street on the Northern line (City branch), which is described in detail in Case study
A1.9).
CIRIA C671 • Tunnels 2009
49
Where deterioration occurs, it may take many forms, as summarised in Table 2.12.
Table 2.12
Summary of causes of metal deterioration
Deterioration mechanism
Details
Graphitisation corrosion
This is a form of corrosion peculiar to cast iron. The corrodible elements of its
microstructure are leached out of the surface leaving a soft, spongy, skeleton of
graphite and other corrosion resistant constituents. There may be no visible sign of
this type of corrosion having occurred but the element will be weakened.
Cracking
This may be due to impact damage, overloading, presence of casting defects, thermal
shock, etc or at weld repairs. Cracking could also indicate possible problems resulting
from exposure to fire or changed ground loading.
Stray current corrosion
Holing may occur when current flows through paths other than the intended circuits
resulting in protection where the current enters the metal structure and potentially
high rates of local metal loss where it leaves.
Microbially induced corrosion
Bacterial activity can adversely affect the structure of the lining material, both
anaerobic, as can occur in compacted clay soils (typically by the action of sulfate
reducing bacteria) and/or in aerobic conditions (typically by sulphur oxidising bacteria).
On cast iron, graphitisation occurs, the iron being converted to its sulfide, leaving a
matrix of low mechanical strength.
Corrosion of metals is an electrochemical process with an anode reaction where metal is
oxidised, corresponding to an equal cathode reaction where typically oxygen is reduced.
In aerobic conditions trivalent iron is precipitated as iron hydroxide, which binds with soil
particles to form a crust on the metal surface. In well-aerated oxygen-rich soil the initial
rate of corrosion is high but slows as the iron hydroxide crust forms and limits oxygen
supply, although microbially-induced corrosion can continue due to the action of sulfatereducing bacteria. In the anaerobic conditions of waterlogged soil, the corrosion rate is
initially lower due to reduced oxygen supply but the rate is not reduced by formation of
corrosion products.
Factors affecting the susceptibility of metal linings to corrosion include:
Soil aggressivity
This is controlled by soil porosity, drainage and ground water constituents. The likely rate
of corrosion may be assessed by measuring:
soil resistivity, which is indicative of the soil’s moisture content and soluble salt
concentration
redox potential, which is indicative (together with sulfate levels) of the soil’s
susceptibility to support anaerobic bacterial corrosion
dissolved salts (eg soluble sulfate and chloride) in the groundwater
pH of soil and groundwater
the type and concentration of any aggressive contaminant in the surrounding soil or
groundwater.
Of the range of soil types peat is potentially the most corrosive soil followed by clay. The
acidity of peat comes from the degradation of the organic material that produces humic
acids. In such conditions hydrogen evolution can replace oxygen reduction as the cathode
reaction, with dissolved metal ions forming complexes with the humic acid. Soils with the
potential to cause severe corrosion problems are those with good electric conductivity,
such as clay. Good conductivity allows the anodic process on a small spot to correspond to
a cathodic process on a large area, causing a rapid and concentrated attack. Measuring the
soil resistivity is a way to estimate the corrosivity of the soil. Waterlogged soils are
potentially more corrosive than dry soils.
50
Macro-corrosion cells
Typically these occur when a metallic element runs through two different soil types
creating differential oxygen conditions. For example, if one part of a tunnel lining is in
contact with well aerated soil and another with poorly aerated soil, oxygen reduced at the
well-aerated area can cause corrosion of the part in the poorly aerated soil.
Calcium carbonate protection
Precipitation of a natural calcium carbonate coating or scale from the groundwater on the
surface of a metallic lining can protect against corrosion by the groundwater. The
tendency for a water to be scaling or not can be determined by water analysis and
calculation of its Langellier index.
Internal tunnel conditions
Water leakage, high humidity, poor ventilation, presence of acidic gases (eg sulfur dioxide)
and any aggressive deposits on the tunnel surfaces exacerbate deterioration on the
intrados of the tunnel. Surface deposits can build-up through mechanisms such as
fluctuations in ground water level, which continuously replenish aggressive agents. The
presence of a coating either organic (eg application of bituminous paint layer) or inorganic
(eg grout layer in contact with the outer surface of the lining) will also affect deterioration.
Figure 2.29 shows an example of corrosion due to water flowing down the inside of a shaft
while the results of sulphuric acid attack resulting from microbial action are shown in
Figure 2.30.
Figure 2.29
Corroded cast iron lining in Aldwych shaft (courtesy Tube Lines)
2.6.2.3
Deterioration of concrete linings
The most common deterioration problems found in pre-cast concrete lined tunnels are:
Cracking and spalling
Cracking and/or spalling (as shown in Figure 2.31) is frequently the result of construction
damage due to either poor segment casting or installation. It can also be the result of
changed loading conditions and deterioration in service (see the following sub-sections on
Reinforcement corrosion and Tunnel fires).
CIRIA C671 • Tunnels 2009
51
Figure 2.30
Acid attack of tunnel lining at Bond Street, London Underground (courtesy Tube Lines)
Reinforcement corrosion
In the alkaline environment of freshly-cast concrete, steel reinforcement remains in a
passive state and is protected from corrosion. However, where chloride ions are present or
the alkalinity of concrete is reduced by carbonation, depassivation may occur and, where
adequate moisture and oxygen are available, corrosion can proceed. Corrosion of the
reinforcing steel produces hydrous ferrous oxides of greater volume than the original
steel, generating expansive forces resulting in cracking and spalling of the concrete cover,
steel bar wastage and loss of bond between the steel and concrete.
Carbonation involves the reaction of atmospheric carbon dioxide with phases of the
cement matrix, and progressively penetrates the concrete from its outer surfaces, reducing
its alkalinity. It does not occur in very wet or very dry conditions, and in good quality
dense concrete its progress is slow and it may take many decades to reach reinforcement
where it is protected by an adequate depth of cover. However in conditions with the right
level of humidity where concrete is highly permeable (eg weak, porous or honeycombed),
or where the cover depth to reinforcement is low, it can also cause problems much sooner.
Chloride ions can either be cast into the concrete (normally through contamination of
ingredients such as aggregate or mix-water. In older concrete (pre-1976), chloride was
intentionally but misguidedly added to accelerate hardening) or they can penetrate from
outside (for example, in saline water or water contaminated with de-icing salts). As with
carbonation, the ingress of chlorides into the concrete from an external source is
progressive and depending on conditions it may take many decades for reinforcement to
become depassivated in good, dense impermeable concrete with adequate depth of cover
to the reinforcement. At a similar level of concentration, chlorides that were cast into the
concrete at the time of its production are typically less deleterious than those that have
ingressed from external sources, because a certain proportion of cast-in chloride ions
become chemically bound-in to the concrete and are effectively inert. However, if the
concrete becomes carbonated these bound-in chloride ions are released and can
contribute to depassivation and corrosion of the reinforcement.
52
Sulfate attack and acid attack
Sulfate attack occurs where there is either an external source of sulfate and water or
where a sulfate bearing aggregate has been used in the concrete. Sulfates occur naturally
in groundwater, soils and rocks. In the UK, the most common source is groundwater in
gypsum-bearing soils, and clays and mudstones such as the Oxford Clay and the Mercia
Mudstone. Sulfates typically affect hardened concrete by reacting with the calcium
aluminate hydrates present in the cement to form either gypsum (hydrated calcium
sulfate) or ettringite (a hydrated calcium sulfoaluminate). These are expansive reactions,
leading to disruption and softening of the cement matrix. In severe cases disintegration of
concrete can occur through the full depth of a section. Another, rarer, form of sulfate
attack involves the formation of thaumasite. Sulfate attack can be exacerbated in presence
of acids, which also attack concrete leading to a softening and disintegration of the cement
paste.
The literature relating to sulfate attack is complex, and for detailed information on the
causes and mechanism of deterioration and guidance on assessing the risk to and avoiding
problems with new construction see BRE Special Digest 1 (BRE, 2005).
Freezing and thawing
As with masonry (see Section 2.6.2.1) concrete in a saturated or near-saturated state is
susceptible to damage through cyclic freezing and thawing. This leads to loss of strength
and cohesion, cracking and spalling that can reduce the effective structural thickness of
segments and reduce concrete cover, increasing the susceptibility of reinforcement to
corrosion through carbonation or chloride ingress (see above). Because exposure to cyclic
fluctuations in temperature is necessary, freeze-thaw damage is typically confined to
concrete that is at or near to portals and shafts and is unlikely to be a problem in deep
tunnels that are insulated by the ground.
Tunnel fires
Concrete typically exhibits good resistance to damage from fire, but fires confined in
tunnels (in particular hydrocarbon fires) can generate exceptionally high temperatures
that are sustained over long periods, and high temperature differentials across sections,
which can result in severe cracking and spalling in pre-cast concrete linings. The effect of
tunnel fires is discussed in detail in Section 2.6.3.
Gasket failure
It is common for segmental pre-cast concrete linings to have gaskets between the segments
as part of the waterproofing system. If the waterproofing is ineffective, the resulting
leakage can lead to corrosion and damage to the lining (Figure 2.32). Over the last 20
years the gaskets used in tunnelling have developed from early, relatively ineffective,
bitumastic strips to much more effective co-extruded EPDM rubber and hydrophilic
gaskets. In order for any type of gasket to be effective, the lining should be built and
maintained such that there is tight control over steps, lips and gaps. Excessive steps, lips
and gaps will adversely affect the performance of the gasket (Shirlaw et al, 2006).
CIRIA C671 • Tunnels 2009
53
Figure 2.31
Concrete spalling from segmental lining sections
Figure 2.32
Gasket deterioration of circle joints and around key block
in concrete segmentally lined tunnel (courtesy Golder Associates)
Although these are perhaps the most common deteriorative processes affecting concrete
tunnel segments, other forms of deterioration such as physical salt weathering, leaching of
cement components by the passage of water and alkali-silica reaction are possible. The
diagnosis of the true causes and assessment of the consequences of concrete deterioration
is best carried out by a competent materials specialist, but for more information refer to
the Concrete Society (2000).
54
2.6.2.4
Deterioration of unlined tunnel support
The deterioration of unlined tunnels in terms of ground movement or failure is discussed
in Section 2.1.5. This section describes modes of failure in support regimes that are used
to reinforce unlined tunnels.
Rock bolts/dowels
The principal mode of deterioration for rock bolts and dowels is corrosion (see Appendix
J of BS 8001 (BSI, 1989) for a full description of corrosion mechanisms). Corrosion is
usually evident from the condition of the head of the bolt or dowel, but may occur along
the length of the bar. Corrosion can result from the absence or poor initial installation of a
corrosion protection system, degradation or cracking of the grout or resin surrounding
the bar or disruption to the protection system caused by ground movements.
There can also be deterioration of face plates and loosening of locking nuts over time,
which can have the effect of untensioning bolts and reducing the effectiveness of dowels.
The integrity of installed bolts or dowels can be assessed non-destructively using a recently
developed percussive technique (Starkey et al, 2001). Other systems using ultrasonic and
radio frequency methods have also been developed (HSE, 2003a). Where pattern bolting
has been installed, it may be appropriate for pull out testing to be carried out to
determine the effectiveness of the support. However, careful analysis of the support and
tunnel condition is required before this is undertaken.
Rockfall protection mesh
Corrosion of plastic coated wire mesh is common, particularly if the mesh is pinned to the
rock face where the plastic protection is easily damaged, usually during installation. Very
old mesh is frequently seen to fall away from rock bolt locations due to corrosion. These
areas can preferentially corrode due to water using the bolts as a pathway or due to bimetallic action between the face plates and the mesh.
The integrity of rockfall protection mesh can also be compromised by an excessive weight
of material that has fallen from the tunnel wall to be restrained by the mesh.
Sprayed concrete
Sprayed concrete deterioration usually takes the form of delamination from the surface to
which it has been applied. It can result from poor original application, poor concrete
mixing, or installation of inadequate drainage. The lack of drainage leads to the build-up
of water pressure behind the sprayed concrete, forcing it away from the tunnel wall.
2.6.3
Effect of fire on tunnels
Tunnel fires can be quite different in character to those occurring in above-ground
structures, potentially generating higher temperatures and sustaining them over a much
longer period of time. Generally, but not always, tunnel fires are vehicle fires, which have
some important characteristics that are not shared by most building fires, in particular:
fuel tanks, which can rupture or explode causing a very rapid increase in fire growth/
extent/severity
cargoes, which provide a high fuel density and result in larger fires with greater
volume of smoke
passenger density, which can be very high, with major implications for rescue from
confined spaces.
CIRIA C671 • Tunnels 2009
55
Experiences of serious fires in modern tunnels suggest that temperatures at the lining
normally average 600 to 700°C, but can reach 1300°C or more locally. The severity of
combustion depends on the nature and quantity of the available fuel (hydrocarbon fires
pose a particular hazard) and the ventilation. Because there is nowhere for the fire plume
to escape upwards, heat is retained close to the fire and radiation reflected by the smoke
layer and the material surrounding the tunnel is many times greater than for a fire an
unconfined space.
Recent experimental studies (Ingason and Lonnermark, 2005) suggest that the heat
release of a HGV fire in a tunnel (carrying a normal cargo, rather than a hazardous one)
may be between 100–200 MW, which is significantly greater than the values of 50 MW or
less that were typically used in the design of existing infrastructure tunnels. The Channel
Tunnel fire of 1996 reached temperatures of 1000°C (Kirkland, 2002), and a fire involving
fuel tankers in Summit Tunnel in 1984, described in Case study A1.18, burned for several
days reaching temperatures of over 1500°C. So fire can place exceptional demands on the
structural materials present in tunnels.
The effect of fire on tunnel structural elements and materials is considered in the
following sections.
Operational and safety aspects of tunnel fires and other hazards are considered in
Section 3.7.
2.6.3.1
The influence of structural form
The guide by BTS and ICE (2004) considers two main types of structural members:
1
Flexural members (eg members of rectangular tunnels).
2
Compression members (eg those of circular tunnels).
It comment on how the structural form of tunnels affects their ability to resist fire:
in non-circular tunnels or tunnels with non-uniform cross-section formed from
reinforced concrete, the principal load condition is controlled by considerations of
bending. Spalling of the soffit and loss of reinforcement in that zone will significantly
reduce the capacity of the section
in circular tunnels, the principal load condition is hoop compression. In concrete and
masonry linings, the reduction of capacity may only be governed by the amount and
rate of spalling. In circular sections, reinforcement in concrete often provides only
secondary structural support, so its loss may not have the same significance as for
non-circular tunnels.
As a result of a fire the movement of the tunnel lining, its stiffness, effective section and
interaction with the ground may change and these factors should be considered in its
design and any post-fire assessments.
2.6.3.2
Concrete and masonry linings
Concrete and masonry are non-homogenous materials whose thermal conductivity is low,
typically 50 times less than structural steel. This means that they heat up slowly, are less
severely affected by relatively short-duration fires, and even in longer duration fires the
depth to which damagingly high temperatures penetrate may be quite limited. So tunnel
56
linings with adequately thick sections often perform well in fires and areas requiring
repair may be low. They are non-combustible and do not emit toxic fumes on heating.
Although resistance of concrete to fire damage is typically considered good, it is generally
inferior to that of masonry constructed with brick and other burned clay products as they
have already been exposed to high temperatures during manufacture and so are relatively
stable in fire endurance tests. Historically, masonry walls have demonstrated excellent fire
resistance, provided that the foundations and supporting structures remain stable. Past
experience indicates that little damage may be caused and structural integrity maintained
even during very prolonged and severe hydrocarbon-fuelled tunnel fires (for example, as
described in Case study A1.18). The performance of stone masonry in fires is not generally
as good as brickwork, but will depend on the type of stone.
The principal detrimental effects of exposure to fire in these materials are:
loss of effective structural section through spalling and delamination
irreversible loss of strength (particularly in concrete)
in some situations, thermal warping and buckling (of masonry).
The low thermal conductivity of these materials means that in thick sections when only
one side of the section is exposed to heat, as is the case in tunnel fires, temperature
gradients across the section can be large. This can lead to high internal stresses and loss of
effective structural section through spalling and, in multi-ring masonry linings, cracking
and delamination between rings. This can have a detrimental effect on the stability of the
tunnel lining.
Concrete
Spalling is particularly a problem for high-strength concretes (HSC) with compressive
strengths of 55 MPa or more, where explosive spalling can result in the rapid loss of the
surface layers of the concrete during a fire, increasing the rate of transmission of heat to
the core concrete and the reinforcement.
Spalling is attributed to the build-up of water and air pore pressure during heating. HSC
is believed to be more susceptible to this pressure build-up than normal strength concrete
because of its low permeability (Kodur and Sultan, 1998, Lie and Woolerton, 1988). The
extremely high water vapour pressure, generated during exposure to fire, cannot escape
because of the high density (and low permeability) of HSC. This pressure often reaches
the saturation vapour pressure, which at 300°C is about 8 MPa. Such internal pressures
are often too high to be resisted by the HSC, which has a tensile strength of about 5 MPa
(Kodur, 1999).
The Channel Tunnel fire in 1996 caused severe damage to tunnel rings because of the
spalling of concrete that completely destroyed some areas of the 450 mm thick concrete
lining, exposing the chalky soil behind. It resulted in injuries to eight people, closure of
the tunnel for six months and an economic loss approaching £1m per day (Ulm et al,
1999). The severity of the spalling was attributed to the high strength of the concrete.
Typically, concrete begins to suffer irreversible loss in strength once heated to
temperatures in excess of 300°C, depending on its composition and nature. Concrete with
siliceous aggregate is more affected than other types. This strength loss is accompanied by
an even greater loss in Young’s Modulus, although this is believed to be at least in part
recovered over time. For practical purposes, 600°C can be considered as the limiting
temperature for structural integrity of concrete made with Portland cement (Neville,
CIRIA C671 • Tunnels 2009
57
1995). Sudden temperature changes, such as those that might be caused by rapid
quenching of a fire by water, can lead to greater reductions in strength.
In reinforced concrete, as the reinforcing steel approaches a temperature of 600°C it loses
about 50 per cent of its yield strength and becomes susceptible to buckling and distortion.
This is reversible on gradual cooling. Heating to 800°C may result in a permanent
reduction in yield strength of between 30 per cent (for cold-worked bars) and five per cent
(for hot-rolled bars). However, even at temperatures below this, if a rapid loss of
temperature associated with sudden quenching occurs there may be a permanent loss of
ductility that can severely reduce the load carrying capabilities of reinforced elements
(NCSCCMI, 1994). Prestressing steels experience a permanent loss of strength at lower
temperatures than for reinforcing bars, affecting cold-drawn and heat-treated steels at
about 300°C and 400°C respectively. Additionally, spalling of prestressed concrete that
exposes steel strand can indicate a loss of prestress, resulting in reduced capacity that
should be properly investigated before any repairs are carried out. Further information
on the performance of concrete exposed to fire, and advice on the assessment and
reinstatement of fire-damaged concrete, is given in the Concrete Society TR33 (Concrete
Society, 1990).
Masonry (brickwork and stone)
By comparison, masonry units of clay brickwork show little strength loss when heated to
temperatures up to 1000°C, although the mortar is affected at lower temperatures, similar
to those of concrete, resulting in a loss of bond strength between brick and mortar.
However, results from over 200 full-scale fire tests carried out in Australia indicated that
concentrically loaded masonry walls do not suffer sufficient strength loss at elevated
temperatures to fail in compression (Gnanakrishnan and Lawther, 1990), rather they tend
to fail through excessive deflection caused by buckling under high differential thermal
gradients, predominantly affecting elements with thin sections. The effect of fire on
stonework is not so predictable because the properties of the many different types of stone
vary considerably, but some generalisations can be made. At high temperatures
(600–800°C) the strength of most stones is seriously affected and if thermal shock occurs
the stone can disintegrate. At lower temperatures (200°C–300°C) damage is usually
restricted to colour changes, for example, the reddening of iron-containing stones
(Chakrabarti et al, 1996).
2.6.3.3
Metallic linings
Metallic materials behave very differently in fires to masonry and concrete, and depend
on the type of metal, its production and the construction form. In new construction,
additional protection is commonly specified for steel structural elements to meet fire
resistance specifications. The guide by BTS and ICE (2004) argues that fire protection is
not needed except where there is a risk of a high-temperature (generally hydrocarbon)
fire. Where protection is necessary it can be difficult to find an acceptably economic
solution, but the use of intumescent paint or an internal lining of polypropylene fibre
reinforced concrete can be effective. A discussion of protection systems is beyond the scope
of this document.
Steel tunnel linings
Steel begins to lose strength at temperatures above 300°C and reduces in strength at a
steady rate until about 800°C. The residual strength is significantly decreased at this
temperature and is about 50 per cent of its room temperature strength. Beyond 800°C the
reduction in residual strength is more gradual until the melting temperature at about
58
1500°C. The capacity of steel to accept high levels of strain increases significantly at higher
temperatures (Lawson and Newman, 1990).
Strength reduction factors for steel with increasing temperature are given in BS 5950-8
(BSI, 2003a) and research has shown that this approach is valid for older mild steel
provided that the yield stress adopted is appropriate to the steel being assessed (Bussell,
1997).
Also the response of steel in a fire depends on the rate of heating due to a creep
component of the deformation at temperatures above 450°C. The phenomenon of creep
results in an increase of deformation (strain) with time, even if the temperature and
applied stress remain unchanged. High temperature creep is dependent on the stress
level and heating rate. The occurrence of creep indicates that the stress and the
temperature history should be taken into account when estimating the strength and
deformation behaviour of steel structures in fire.
Hot-rolled structural steel will regain virtually all of its strength when it cools back to
ambient temperature from 600°C, but exposure to higher temperatures will result in a
reduction in strength on returning to ambient temperature. The extent of the strength
reduction depends on the grade of steel (Lawson and Newman, 1990).
Cast iron tunnel linings
London Underground cast iron deep tube tunnel linings are either composed of grey cast
iron, spheroidal graphite iron or flexible grey cast iron, all with differing properties. The
melting temperature of cast irons is in the range 1150 to 1300°C, at which point these
materials are fully molten.
The strength of cast irons is retained up to temperatures of at least 400°C, but there is
historical evidence of catastrophic failure at temperatures below this as a result of cracking
(Bussell, 1997). Cracking could either occur during the fire (and may lead to immediate
structural collapse), or when applying cold water to put out the fire leading to explosive
shattering of the iron and progressive structural collapse. Fire testing carried out by the
Greater London Council Scientific Services Branch in 1984, found previously loaded cast
iron beams cracked when the unloaded beam was hosed down. It is thought that the
cracking is a result of significant locked-in thermal stresses in the previously compressed
area of the beam (Bussell, 1997). In buildings sprinkler systems would normally be
activated before cast iron elements reached temperatures at which this cracking
phenomenon occurred, but the authors are not aware of any infrastructure tunnels in the
UK now that have such systems.
The thermal expansion of cast iron is lower than that of steel at 1 × 10-5°C-1 but it is
unable to undergo distortion in a fire due to the material’s weakness in tension (Lawson
and Newman, 1990). It is often the thermal expansion properties of cast iron that result in
its failure in a fire where it is in contact with other materials. Due to its brittle nature it is
vulnerable to the distortion of the structure around it. For example, if cast iron elements
are attached to steel elements, expansion of the steel during heating from a fire could lead
to tensile bending failure of the cast iron elements.
The residual strength of cast iron is similar to steel up to 600°C, but at temperatures above
this the reduction in strength is more significant than that of steel.
CIRIA C671 • Tunnels 2009
59
2.7
SHAFT PERFORMANCE
As discussed in Section 2.2 (on shaft construction) shafts can be a liability in terms of their
own maintenance and their effect on the tunnel and the land above it. In certain
circumstances shafts can become unstable, potentially leading to collapse causing damage
to the tunnel and to people and property above it within their zone of influence. Figure
2.14 shows the possible states of construction shafts.
Open shafts are frequently responsible for the ingress of water into tunnels because they
can act as vertical drains for groundwater above it, particularly where they intercept
permeable strata. They also allow the circulation of cold air so that shaft linings and the
adjacent areas of tunnel often suffer from freeze-thaw damage. Their condition should be
assessed periodically through inspection (see Section 4.7) to allow maintenance and repair
as necessary.
Capped, filled and partly filled shafts may also act as drains, directing water down to the
tunnel from overlying ground. Shafts capped at their top but left open at their base
(Figure 2.14, types 2 and 3) require inspection and maintenance as for open shafts, but
present additional access difficulties. Filled shafts (Figure 2.14, type 4) do not allow
inspection other than at the tunnel intrados, but if properly filled with stable and
lightweight material should present a reduced risk of collapse. However, there are
circumstances in which shaft stability is at risk:
where shafts have been improperly filled or contain voids or unstable material, or
where capping has been constructed with timbers that rot and become unstable.
Settlement can result in voids that migrate upwards to be expressed at the ground
surface above
where friction between the shaft lining and the surrounding ground is lost, the tunnel
lining supporting the shaft at its base is subject to increased loading and may become
unstable. This could be caused by the passage of water along the lining/ground
interface or the shrinkage of the ground and/or lining in dry weather
where deterioration of the tunnel lining occurs at its structural connection with the
shaft. Its capacity to bear load is reduced, particularly where original construction
timbers may have been left in place. This may be exacerbated by the passage of water.
It is vital that the location of all shafts is known so that the risks can be assessed and, where
necessary, mitigated. However, the existence of shafts was not always recorded at the time
of construction, and where unrecorded shafts are not visible from the ground surface or
from within the tunnel, their presence may remain unknown. Such shafts present an
uncontrolled risk to the tunnel and the land above it. So it is important to locate all shafts,
including those that are hidden and blind, and apply a process of condition assessment,
maintenance and repair as for the rest of the tunnel. Network Rail has a policy of marking
the location and, where possible, the extent of all hidden shafts directly on the tunnel
lining so that they may be easily located and particular attention can be focused on these
areas during visual inspections (Network Rail, 2004b).
It is worth considering that in recent UK history the most serious tunnel collapse (Clifton
Hall), which resulted in multiple fatalities and a major public inquiry in 1953, was because
of the instability of a hidden and unknown construction shaft. Several other serious tunnel
collapses have occurred for the same reason. (the Clifton Hall collapse is described in Case
study A1.18, among other similar incidents).
60
2.7.1
Effect at ground level
Consideration of the potential effect of a tunnel at ground level requires an evaluation of
its zone of influence ie the volume of surrounding material that is potentially affected by
it, particularly by a collapse of the tunnel and any shafts or adits. It is important to assess
the zone of influence for tunnels for the purpose of risk management. An assessment of
the zone of influence is necessary for controlling liability in cases where the use of the land
above the shaft changes, increasing the overburden. The assessment can be used to
identify land and property use above the tunnel and produce a list of potentially affected
landowners, and to assess the consequences of tunnel deterioration and collapse, in
particular where hidden shafts are suspected or known to be present (Section 3.7), or
where tunnels become disused and should be maintained to control risk rather than as an
operational asset (Section 3.9).
An example of the effect of tunnel collapse at the ground surface is the collapse of a
disused rail tunnel on the Canterbury to Whitstable line in 1974, which resulted in the
collapse and demolition of a University of Kent building located above it, described in
Case study A1.18.
An initial determination of the potential zone of influence of a tunnel can be based on the
properties of the ground and the depth of cover to the tunnel, as illustrated in Figure 2.33.
Figure 2.33
Simplified method for determining the zone of influence of tunnels (a) and shafts only (b)
(after Network Rail, 2004a)
Values for tunnel diameter and depth of shafts can be taken from drawings or inspection
records where available. The geology in the region of the tunnel can typically be obtained
from drift maps, and ground characteristics, including worse case friction angles for the
materials that are present, can be taken from published data (eg Hoek and Bray, 1977).
Where variations in ground conditions occur, the worse case geological characteristics of
the possible variations can be taken and applied uniformly throughout the overburden to
give a conservative result, and ensuring that the extent of the zone is not underestimated.
In this case, the lowest value of the slope friction angle for all materials occurring should
be applied uniformly throughout the overburden as shown in Figure 2.33.
CIRIA C671 • Tunnels 2009
61
Evaluations should be made at sections taken at the location of tunnel shafts and where a
change in overburden profile is evident (from long-sections or contour lines on
topographic maps). The results may be plotted onto a map of the ground surface above
the tunnel to define the limits of the zone of influence and identify land-use and
ownership within it.
The method described provides a straightforward means of making what is likely to be a
relatively conservative estimate, but in certain situations it may be too simplistic (Network
Rail, 2004a). It could, however, be useful in providing an initial assessment for identifying
areas where risk to development is significant enough to merit a more sophisticated sitespecific assessment method with a greater degree of accuracy. The best approach for the
determination of the zone of influence should follow discussions between tunnel engineers
and geotechnical specialists. This will include a consideration of the level of risk that, for
example, depends on land-use at the ground surface above.
In the past, various rules of thumb have been proposed for assessing the safe distance
from a tunnel for further development: Price et al (1968) suggested that, for multi-storey
development, a minimum distance was equal to the depth of superficial deposits, up to a
maximum of 30 m. It has also been suggested that a safe distance can be defined by the
dimensions subtended by an angle of 45° to the ground surface from the point where the
sides of the shaft intersect rock head (NCB, 1982) or that, in situations where the
overburden is not exceptionally weak, a distance of twice the overburden thickness, up to
a maximum of 15 m depth may be used (Bell, 1975). However, a more realistic safe
distance can be assessed on a shaft-by-shaft basis by considering the local circumstances,
primarily the geological properties of the soils and rocks present, and the state of the
shaft, followed by stability analysis using classical mechanics methods (Healey and Head,
1984).
62
3
Tunnel asset management
3.1
THE NEED FOR TUNNEL MANAGEMENT
Tunnels are a vital element of the transport and services infrastructure. They have
typically performed well in service because of their construction, many for at least 100
years and some for considerably more, and often with relatively little maintenance, repair
and alteration. However, changes do occur, some sudden and others more gradual, and
without intervention tunnel condition cannot safely be considered steady state, for
example:
in some cases tunnels have been modified to accommodate a change in use from that
originally designed for (for example, canal tunnels converted to accommodate
railways), which may impose changes in loading to the tunnel invert or lining not
previously envisaged
other modifications, such as infilling of shafts, can also influence loading on the
structure
urban development over time may also have an effect, particularly on relatively
shallow utility and metro system tunnels in cities where there is increasing subsurface
construction, for example, tall buildings with deep piles, basements and other tunnels
ground movements and hydrological variations (for example, changes in land
drainage and water tables) may bring about changes in ground pressure, either
directly (for example, hydrostatic pressure) or indirectly (for example, swelling clays)
the natural processes of weathering and decay result in deterioration of tunnel
materials, typically causing loss of strength and cohesion, which may result in
redistribution of loading and reduced structural stability
where structural elements are not readily visible for inspection (for example, where
they are obscured by sheeting or a secondary lining, or where access cannot be easily
gained) deterioration can progress undetected unless special measures are taken to
monitor their condition.
Like any other structure, tunnels have a finite life before significant renewal or
replacement works are required. However, the huge resources required and the impact of
disruption associated with wholesale renewal and replacement of the tunnel infrastructure
means that wholesale removal is not a feasible option. In effect, it is necessary to consider
such assets as having an indefinite life and aim to devise management, maintenance and
repair strategies that will ensure their continued safety and serviceability well into the
future. To achieve this it is imperative that the highest standards of asset stewardship are
established and maintained, and this requires the development and application of efficient
management policies and procedures, supported by adequate resources.
3.2
SPECIAL REQUIREMENTS
There are several features and characteristics of ageing tunnels that require special
consideration in their management:
CIRIA C671 • Tunnels 2009
63
many are among the oldest elements of the transport infrastructure and have
particular maintenance and repair needs that may differ from those of modern
structures
they are often very individual in their character, behaviour and maintenance needs
typically they lack information regarding their design, construction, hidden structure
and important environmental factors such as local ground conditions
their structural behaviour and performance is complex and not as well understood as
that of modern structures, presenting difficulties for structural assessment
access is often restricted and conditions within tunnels may be poor. This may hamper
inspection, investigation, maintenance and repair work, and makes it difficult,
disruptive and expensive. Because of this there is a risk that it may be neglected or
undertaken in a sub-optimal fashion
visual inspection is limited to the intrados, and it is difficult to get reliable information
about what is going on behind this
the effectiveness of repairs and alterations and their likely influence on the long-term
performance and maintenance of the structure are not well understood
hidden features, such as tunnel shafts, may be difficult to access and inspect, or their
presence may not be known, which can be a hazard to safety.
Infrastructure tunnels represent a huge capital investment, which should be protected,
and the benefits of developing effective ways of dealing with these challenges are likely to
be considerable.
3.3
LOSS OF PERFORMANCE AND ITS CONSEQUENCES
The consequences of loss of tunnel performance are described in the following sections:
Safety in operation
Factors such as age, increased traffic loading, inadequate or poor maintenance and
deferred repairs reduce tunnel performance and may ultimately compromise operational
safety. For certain types of tunnel safety in daily operation is largely a matter of the
mechanical and electrical components that are installed and renewed on a much shorter
life cycle than the civil structure. Structural instability means that operational safety cannot
be guaranteed, public safety is jeopardised and complete closure of the tunnel may be
necessary. Other types of failure may not compromise the structural integrity of the tunnel
but can nonetheless result in serious accidents and injuries, for instance material falling
from a tunnel crown into the running area below. For certain types of tunnel
infrastructure, such as utilities, the public safety issues may be less onerous but hardly less
tolerable for other reasons. Risks to those in and over the tunnels should be considered.
See Section 2.7.1 for consideration of the potential area of influence of a tunnel and its
shafts both underground and at the ground surface.
Disruption and customer dissatisfaction
While it is impossible to avoid unplanned disruption completely, a proactive approach to
maintenance and planning repair works can significantly reduce it. Increasing demand on
services places greater pressures on infrastructure owners to ensure smooth operation,
and there may be direct penalties for failure to comply with operational performance
targets. Also to the negative user-perceptions associated with disruption and delays to
services, in certain situations there is also increased user risk.
64
Costs of accidents, failure and repair
When unplanned repair becomes necessary, significant costs are likely to be incurred.
These costs may extend beyond the direct cost of remedial works to the provision of
access, temporary restrictions, provision of alternative services, lane and line closures in
road and rail infrastructure, and reduction in revenue. Where a tunnel failure has
occurred or a structural fault has resulted in accident or injury, the need for formal
investigations may entail further loss of service with consequential loss of revenue and
possibly punitive financial penalties.
Serious incidents, such as tunnel fires, may require significant periods of closure to effect
repairs and any necessary upgrades of tunnel systems to prevent recurrence and improve
safety. Where major infrastructure tunnels are concerned, the national economic
consequences can be considerable. Table 3.1 gives three recent examples of the direct and
consequential economic costs of major tunnel fires. In each case the total national
consequential cost is significantly greater than the cost of the tunnel repairs alone. Aside
from the economic cost, the Tauern and Mont Blanc fires between them cost dozens of
lives and left many more injured.
Table 3.1
Direct and consequential cost of tunnel incidents (after Rock and Ireland, 2005)
Tunnel/incident
Cost of repairs (€)
Total cost (€)
Lives lost
Channel Tunnel fire (1996)
48.5m
253m
0
Mont Blanc Tunnel fire (1999)
189m
392m
39
Tauern Tunnel fire (1999)
8.5m
29m
12
Managing tunnel maintenance
As noted in Section 3.1, most tunnels are required to have an indefinite life: the
maintenance of a tunnel can be defined as all the operations necessary to maintain it in a
serviceable condition indefinitely, including:
condition appraisal (inspections, testing and monitoring, structural assessments)
routine maintenance (typically involving like-for-like replacement of the tunnel fabric
to maintain efficient functioning and preserve condition)
interventions (to carry out vital repairs to and modification of the structural fabric in
response to deterioration and loss of performance, or adaptations to meet new
requirements, eg for higher loadings, health and safety or control equipment)
emergency actions (eg in response to unforeseen incidents).
It is necessary for asset managers to develop effective and efficient management strategies
that maintain the tunnel fit for purpose and help to avoid the need for emergency action,
ensuring safe operation at an adequate level of service. These strategies should also align
with the long-term objectives of the infrastructure owner and meet statutory and
regulatory requirements. This can be achieved by a system of maintenance planning and
management for tunnel assets, carried out through a formalised system of procedures.
This allows the asset manager to identify the maintenance needs of the tunnel stock as a
whole and of individual structures, and to develop and justify suitable maintenance plans
to address these both system-wide and on individual tunnels. This information is included
in the asset management plan, which documents management objectives for the assets and
sets out a clear strategy for achieving them.
CIRIA C671 • Tunnels 2009
65
A suitable maintenance planning process includes several elements and stages:
undertake an investigation into the history of the tunnel, using suitable historic
sources (see Appendix A2)
compile and maintain a tunnel inventory and database
carry out periodic condition appraisal of tunnel stock
identify maintenance needs
assess and prioritise maintenance needs (value management)
develop optimal solutions for prioritised maintenance needs (value engineering)
consider resource availability and prepare work plans and schedules
programmed maintenance works
keep tunnel records updated with current information
monitor and improve the management process through continual feedback.
This approach can be used to ensure that safety, performance and business objectives are
met, to determine the resources required, and to make best use of available resources
through sustainable maintenance work plans. If such a system is properly devised, fully
initiated and adequately resourced it will provide improved asset performance and return
on maintenance investment.
For example, road tunnel maintenance for Highways Agency tunnels is to be undertaken
in accordance with HA standard BA 72 (HA, 2003), and this provides useful guidance for
other road tunnels. While the majority of continuing maintenance tasks for highway
tunnels relate to tunnel equipment, there are also sections on tunnel structure cleaning
and maintenance that are relevant to other tunnel infrastructure.
3.4.1
Appraisal of current condition, performance and serviceability
To ascertain maintenance requirements it is necessary to gather and periodically update
and evaluate information relating to tunnel performance and condition by a process of
condition appraisal. In this context, the term appraisal is used in its broader sense to
encompass all the activities undertaken to determine the adequacy that a tunnel can
perform its functions. Each of the main infrastructure owners has its own internal
procedures and systems for determining the maintenance needs of its structures,
including tunnels, but they are mostly based on a similar principle: information on
individual tunnels is obtained through regular visual inspections. This is supplemented by
more detailed inspections less frequently, and by further investigation works and/or
structural assessment where necessary. Additional information on the tunnel, its past
performance and maintenance history is also considered. In particular, increasingly
sophisticated surveying and monitoring techniques are being used to meet the evergreater demand for information on the behaviour of civil engineering structures,
including tunnels. The information obtained is used as a basis for making informed
decisions regarding safety, serviceability and performance.
The range of information used to appraise a tunnel’s serviceability is given in Figure 3.1.
66
A thorough investigation
into the history, type and
method of construction
including any remaining
temporary works
Thickness and
capacity of any
lining present
As-built details including
voids, shafts etc.
Particular attention shall
be given to locating
blind or hidden shafts
Structural features
affecting operational
safety including
clearances for traffic
Construction
materials their
strength, current
conditions and levels
of deterioration
Geology of the
surrounding material
and its influence on
the tunnel lining
Risk from current or
abandoned mineral
extraction workings
Effect of present
type of traffic on
the tunnel structure
Serviceability
Presence of water
and known
watercourses
Condition and
significant
defects
History, type
and method
of any repairs
Cover and
tunnel
dimensions
Effect on
tunnel of
other works
Presence of combustible materials
in the tunnel (including shafts, adits,
portals, and other passages) and
the vulnerability of the tunnel to fire
Figure 3.1
Information required for an assessment of tunnel serviceability (Network Rail, 2004a)
Using this information, current tunnel performance and condition is assessed against
serviceability criteria assigned by the asset owner. These criteria will include standards for
safety as well as structural and operational performance and will vary according to the
infrastructure type and owner policies and objectives. The results of the appraisal are used
to assess the overall condition of the tunnel and identify any changes or trends, to plan
routine and preventative maintenance to preserve tunnel condition, and to trigger
reactive repairs to correct unacceptable performance where a tunnel is not considered
serviceable (see Figure 3.2).
CIRIA C671 • Tunnels 2009
67
Figure 3.2
Outline process for assessing and maintaining serviceability of tunnels
The information produced in the course of condition appraisal should be collated and
recorded in a suitable format, incorporated in the asset inventory and used to highlight
any changes in tunnel condition, and determine its serviceability and level of performance
against specified performance requirements. This information forms the basis for
assessing the tunnel’s needs and determining appropriate management actions, such as:
adequacy of existing routine maintenance regime
additional routine maintenance requirements
changes in frequency of inspections
requirements for further inspections and their objectives
need for structural assessment
essential maintenance requirements
requirements for safety measures (restrictions of use, regular monitoring).
The interval between inspections is related to the importance of the tunnel and the
perceived degree of risk associated with it. Also to routine/planned inspections, certain
observations and incidents may lead to a requirement for special investigation of a tunnel’s
condition and performance (see Section 4.3.2). This might involve an increase in the
frequency of visual inspection, or carrying out specific investigations involving a variety of
testing and monitoring techniques to assess the structural condition of the tunnel, the
nature and cause of any defects, their extent and potentially the rate of deterioration. This
information can be used to evaluate the tunnel’s performance against safety and
serviceability requirements, determine the optimal management strategy and assess the
need for maintenance and remedial works.
68
3.4.2
Maintenance strategies
Where maintenance is required, this involves implementation of routine works (periodic,
often cyclic, planned maintenance tasks to repair minor defects and prevent or slow future
deterioration) and interventions (repair and rehabilitation of the structural fabric in
response to deterioration and loss of performance). Maintenance typically includes both
planned (proactive) or unplanned (reactive) activities, and depend on knowledge of
current condition, often obtained by periodic inspection and an assessment of tunnel
performance against requirements (see Figure 3.2).
When maintenance resources are limited it is sometimes the case that routine works are
neglected or given a lower priority than they deserve. This can be counterproductive in
the long-term. What began as minor maintenance issues can develop into serious
problems if not dealt with at an early stage, often with significant repercussions for tunnel
serviceability in the interim period and the eventual cost and disruption associated with
rectifying problems that were avoidable in the first place. Unless there is good justification
otherwise, asset managers should establish a proactive regime of preventative routine
maintenance for all tunnels: maintenance and repair programmes should deal with the
causes, and not just the effects, of deterioration. Advice on routine maintenance is given in
Chapter 5.
Planned and reactive maintenance strategies are discussed further in the following subsections.
3.4.2.1
Planned maintenance
Planned maintenance can be subdivided into two types:
1
Periodic maintenance is carried out regularly at predetermined intervals, the intervals
being based either on calendar time (eg quarterly, annually, biennially) or on actual
functional time in operation (eg after 1000 hrs operation). The former is more
common for a tunnel structure, whereas the latter is more typically used for tunnel
equipment. Periodic maintenance is suitable where maintenance requirements are
relatively regular and foreseeable, or where condition-based maintenance is
unfeasible. The maintenance interval is important because if it is set too high it will
result in unnecessary work and wasted resources, but if set too low tunnel
serviceability, and sometimes safety, may fall below acceptable standards. Once
sufficient experience and information has been gained, maintenance intervals may be
optimised.
2
Condition-based maintenance aims to provide maintenance as it is needed so that
intervention is always at an optimal time and resources are not wasted. It is the most
common and potentially the most suitable method for maintaining tunnel structures.
However it requires a good knowledge of current condition and an adequate
understanding of tunnel performance and deterioration to define suitable,
measurable triggers for activating maintenance interventions. Condition checking is
typically carried out by regular inspections to identify visible evidence of loss of
performance at an early stage, allowing problems to be quickly resolved before they
start to affect safety and serviceability. The principal limitation of this approach is that
it depends upon identifying and responding appropriately to detectable criteria, and
works better for more evident defects such as cracking and spalling, but less well for
those that do not show clearly visible symptoms.
CIRIA C671 • Tunnels 2009
69
3.4.2.2
Reactive maintenance
Reactive maintenance consists of carrying out corrective remedial works once loss of
performance has occurred. If there is a positive aspect to reactive maintenance, it is that
the initial maintenance intervention may be deferred. However, it is not an economical or
sustainable policy for long-term stewardship of assets such as tunnels as it has several
potential drawbacks compared with planned maintenance:
it is not possible to budget or plan for maintenance
maintenance is likely to be more disruptive and costly
it can allow deterioration to spread and affect other elements
the asset condition worsens and maintenance demand is increased in the long-term
good knowledge of current condition is particularly critical
there is a greater risk to operational efficiency and safety.
Although planned preventative maintenance can reduce the risk of the need for reactive
maintenance, it cannot be avoided altogether. Asset owners should be adequately
prepared for unforeseen malfunctions and failures of the tunnel structure and associated
equipment (that may need emergency actions).
In both planned and reactive maintenance, it is important that good records are kept of
the work that has been carried out, preferably to include records of location, type and
extent of repairs, their cost, materials used, and before and after photographic records,
sketches or dimensioned drawings. Over time, the accumulated information will be
invaluable in future maintenance assessment and planning, and can be used for
monitoring deterioration rates to provide a more realistic assessment of future
maintenance needs.
3.4.3
Maintenance planning and prioritisation
3.4.3.1
Assessment of tunnel criticality
It is often necessary to prioritise maintenance needs based on an assessment of the
criticality and sensitivity of individual infrastructure elements because resources for asset
maintenance can be limited. This requires infrastructure owners and operators to identify
those elements that are most critical to ensuring the safety and efficient operation of their
networks. Tunnels are often among these, so frequently merit a high priority for
management activities.
Criteria for prioritisation may include:
70
risk assessment
tunnel condition and assessed safety factors
degree and consequences of substandard performance and failure
importance of route
minimisation of maintenance costs
organisational policy
environmental considerations
budgetary constraints.
Operationally critical tunnels can be identified by an assessment of their location within
primary transport and distribution routes, volume of transport and possible diversion
options so as to consider the impact of loss of performance and tunnel closure on the
network infrastructure. For critical tunnels, depending on the condition, it may be
possible to justify a higher frequency of routine inspections relative to non-critical tunnels
and other structures (see Section 4.3.2), so maintenance and repair works may benefit
from being given a higher priority. This approach requires careful consideration of the
relative risks and benefits involved, but makes good sense in terms of efficient asset
management.
A risk-based approach (see Section 3.5.4) can be used to assist with identifying highcriticality structures, and with prioritisation and planning to achieve optimum use of
resources.
3.4.3.2
Effect of maintenance strategy on tunnel performance
The frequency and scope of maintenance intervention will depend on the desired level of
performance of the tunnel, as illustrated in Figure 3.3. In service, deterioration (and
possibly other factors) result in a gradual loss of tunnel performance over time. This
translates into an increasing loss of reliability and risk of failure until the full service life of
the tunnel is reached and major rehabilitation or renewal is required. In the meantime,
maintenance interventions are carried out to prolong the service life and keep tunnel
performance and reliability at acceptable levels. The frequency and scope of maintenance
influences the margin of safety between actual tunnel performance and the intervention
level where performance becomes unacceptable.
Ideally, if the rate of reliability loss (represented by the slope of the curve in Figure 3.3)
were accurately known, asset managers would be able to identify optimal timings for
maintenance intervals and interventions. However, in reality, the rate of deterioration and
The optimal maintenance strategy will be one that provides the desired
level of performance over the longest period in the most economical
way. This will vary according to the policy of the infrastructure owner,
the asset type and the specific characteristics of the actual asset.
Figure 3.3
Relationship of serviceable life, performance and maintenance interventions (from Patterson
and Perry, 1998)
CIRIA C671 • Tunnels 2009
71
its effect on reliability is difficult to evaluate for most tunnel structures and assessing the
optimal timings for maintenance and repair relies heavily on engineering judgement. This
engineering judgement is more likely to be accurate if it is supported by good quality
reliable data on current and past tunnel condition and performance.
For tunnels, which are frequently expected to have an indefinite life, and where closures
and restrictions for carrying out maintenance and repair work are particularly costly and
disruptive, there is good justification for expenditure on a programme of regular planned
maintenance. This will keep the structure at a safe margin above the intervention level
and defer the requirement for more major rehabilitation works.
3.4.3.3
Effect of maintenance strategy on inspection intervals
The optimum inspection interval for individual tunnels is likely to differ, depending on
their type, condition, deterioration and accessibility, and the consequences of hazards
occurring. A suitable balance should be maintained between tunnel condition, the level of
preventative maintenance and inspection interval. In relative terms, inspections at longer
intervals may be acceptable in tunnels with a high reliability, ie those that are known to be
in good condition and are subject to regular preventative maintenance. Tunnels that have
a low reliability, ie those that are in poor condition and are maintained in a reactive
fashion, need shorter inspection intervals. There are potential advantages of varying
inspection intervals rather than specifying fixed intervals, but the risks associated with
increasing inspection intervals need to be adequately assessed on a tunnel-by-tunnel basis
and often there is insufficient data available to do this adequately. Because of this, most
large infrastructure owners have fixed-interval inspection regimes that are adequate to
ensure the safety and serviceability of the numerical majority of their tunnels. Some have
provision for reducing inspection intervals for the minority of tunnels that are identified
as being particularly sensitive, for example, due to continuing deterioration or following
structural repairs. Inspection intervals are discussed further in Section 4.3.2.
3.4.3.4
Optimising planned maintenance strategies
As discussed in Section 3.4.2, a proactive system of maintenance should be established for
tunnel assets, based on a good understanding of a tunnel’s past history, its current
condition and its likely future requirements. This information can be used to formulate a
strategy for planned, preventative maintenance, in which tunnel condition is maintained
at or slightly above the optimum level.
Wherever possible, maintenance should treat the cause of loss of performance, as well as
its effect. This is one of the main principles of preventative maintenance. However,
putting this theory into practice can present difficulties:
72
1
The circumstances that cause deterioration may not be readily detectable (eg changes
in ground pressure or loss of materials strength) and the maintainer may become
aware of the need for maintenance only after the tunnel’s performance has been
affected. Maintenance then becomes reactive.
2
It may not always be feasible to treat causes rather than symptoms. For example, in a
long tunnel suffering from widespread water ingress, it may be more economical to
try to directly mitigate the problems the water causes rather than attempt to control
or prevent water ingress throughout the tunnel. Frequently, a successful maintenance
strategy will aim to provide preventative maintenance and treat causes of loss of
performance, but it should be prepared to carry out reactive maintenance and
treatment of symptoms where this is unavoidable or can otherwise be justified.
It is important that effective asset management relies on accurate and comprehensive
information concerning a tunnel’s environment, structure and past and current
performance. This requires skilful and conscientious research and organisation, and
management of existing and new data, which will be discussed in the following sections.
3.4.3.5
Deferral of maintenance
Where maintenance and repair is unduly deferred, this may have negative impact on the
efficiency and economy of tunnel management as well as a temporary reduction in tunnel
performance and serviceability. Deferral of essential maintenance may require interim
measures to ensure the continuing safety of the tunnel and its users, for instance
restrictions on capacity or requirements for extra monitoring and special inspections. In
certain circumstances the deferral of maintenance may be unavoidable due, for example,
to operational or budgetary constraints, or it may be justifiable in terms of perceived
benefit. In either case, the implications of maintenance deferral should be assessed,
particularly any potential impact on tunnel safety and serviceability.
3.4.3.6
Minimising disruption from tunnel maintenance
It is vital to minimise disruption to the normal service from management activities,
including carrying out condition assessments, maintenance and repairs because tunnels
can have a considerable influence on the operation of the whole infrastructure. For certain
types of infrastructure, costs associated with access and necessary disruption to service may
account for most of the total cost of such activities. Because of this the need to ensure
continuity of service and minimise disruption is often the overriding influence in the
selection of maintenance and repair schemes, and planning and programming are key
elements in the success of any works carried out. Tunnel closures need to be planned in
detail to ensure best use of the time available. Where access is at a premium, it is advisable
to co-ordinate all foreseeable inspection, investigation, repair, maintenance, renewal and
other works to the tunnel structure and associated equipment (drainage, mechanical and
electrical systems, ventilation and pumping systems etc). Where necessary, diversions of
traffic or services should be planned in advance and carefully managed.
For tunnels that depend on the function of mechanical and electrical systems (for
example, road tunnels) the optimal maintenance period of the tunnel structure may be
influenced by the maintenance period or serviceable life of these systems. For example, in
a road tunnel a 20-year frequency of structural refurbishment may be desirable, given the
18-year nominal design life of tunnel fans and lighting given by Highways Authority
standard BD53 (HA, 1995). Disruption to tunnel services can be minimised by planning
and co-ordinating maintenance and refurbishment/repair activities in this way (Rock and
Ireland, 2005).
3.5
TUNNEL MANAGEMENT PROCEDURES AND TOOLS
3.5.1
Tunnel information requirements
Comprehensive knowledge of an asset is fundamental to its effective management. Tunnel
owners should make efforts to collect and collate all existing information on their tunnel
assets, and to store this information safely in a form that can be accessed by those who
might need it - including asset managers, engineers, consultants, maintainers and repair
contractors. Tunnel inventory asset files should be established and managed as an
important element of the infrastructure. Asset files may be maintained in either hard copy
CIRIA C671 • Tunnels 2009
73
or electronic format (and preferably both) in a database where information on the
structure, its condition, maintenance needs and management plan can be stored for later
retrieval.
In tunnels built since the Construction (Design and Management) Regulations 2007, there
exists a legal requirement to produce a health and safety file and store it so as to allow
easy access and retrieval of information. In certain circumstances this may also be a
requirement for older tunnels. Compliance is typically achieved by integrating this
information as part of an asset management system. The health and safety file should
include:
a description of the work carried out
residual hazards and how they have been dealt with
main structural principles
any hazards associated with materials used
information regarding the removal or dismantling of installed plant and equipment
health and safety information about cleaning and maintenance equipment
the nature, location and marking scheme of significant services (eg fire services)
information and as-built drawings of the structure, its plant and equipment.
For older infrastructure, where there is no health and safety file, it is important that
comparable data is collated and held in a tunnel register to perform a similar function,
and also to provide additional information for optimising the efficiency of management
and maintenance of the tunnel. A tunnel register should ideally include, but not
necessarily be limited to, the following:
74
unique tunnel identifiers (name, number)
location data (map reference, road/route details, land-use within zone of influence)
owner and maintaining agent
tunnel age, type, form of construction, main structural elements and materials, length
and dimensions
construction history and special features, for example, areas where difficult tunnelling
conditions were encountered, changes in tunnel profile, lining materials or thickness,
presence or absence of an invert (either from available historical records or inferred
from more recent observation and investigation)
presence and location of shafts (known or suspected) or other features that might
present special risks or require particular management actions
local geological, hydrological and environmental data
details of tunnel use (eg traffic frequency, types and speeds)
performance data (eg capacity, any restrictions on serviceability)
details of tunnel interface and interaction with other parts of the infrastructure
access information for all parts of the tunnel with methods and details of any special
access requirements
hazard identification/risk log
copies of registers of known hazardous materials, eg asbestos registers
emergency planning information (emergency access, escape plans, contacts etc)
details of outside parties and activities that may affect the tunnel, eg piling works over
or adjacent to the tunnel.
current inspection results and history of previous inspections, investigations,
assessments and condition appraisals
history of maintenance, repairs and other works, including any health and safety files
produced from works carried out in compliance with CDM Regulations
schedules for planned inspections, maintenance and repairs
information on tunnel equipment and services
details of services either carried in or close to the tunnel with up-to-date emergency
contact numbers etc maintained (eg local water authorities to be contacted in case of a
sudden increase of water ingress to the tunnel)
any statutory designations or restrictions (eg listed status or environmental
designations)
historical records and documentation (drawings, articles etc) including details of sources
other information (eg incidents such as flooding, emergency incidents).
Compiling comprehensive tunnel data is likely to result in voluminous records. Strict
procedures should be established for the management and maintenance of this data to
ensure that the most appropriate and up-to-date information is available and identified.
Electronic information can be managed and manipulated with computer-based asset
management software, which provides an opportunity to make more effective use of
existing knowledge, but given past experience with data recording and storage formats, it
is important to take steps to guard against obsolescence and ensure that data remains
readily accessible and usable. Further information on the development of tunnel records is
given in a paper by the International Tunnelling Association (ITA, 1987).
3.5.2
Tunnel management systems
Asset knowledge on tunnels should be collected, stored, managed and retrieved
throughout their service life. As the asset experiences deterioration and local failures,
planned rehabilitations, routine maintenance, upgrades, modifications and other
associated activities need to be recorded. Databases are ideal tools for such tasks, and these
should be integrated as part of the tunnel management system (TMS). The TMS
comprises a framework that allows efficient organisation of tunnel maintenance, including
activities such as information management, condition appraisal and maintenance and
repair planning, which can be used to inform, guide and support management decisions.
A TMS stores information on individual tunnels that it can use to carry out a variety of
engineering and economic assessments. It can be a powerful tool for owners, providing
assistance on organisational policy, adhering to statutory requirements, making, recording
and justifying management decisions, determining the best use of limited resources, and
formulating and presenting business cases for obtaining funding.
It is important to appreciate that even the most sophisticated management systems rely on
the quality and integrity of available data. Inadequate or inaccurate data is likely to lead to
poor management decisions, whereas good quality data allows more effective and efficient
management of the tunnel stock. It is important that such systems are easy to update with
new information and to maintain.
3.5.3
Tunnel identification and referencing systems
It is desirable that tunnel assets and sub-assets are identified by a unique number or code,
ideally with number or code plates attached directly to the asset for ease of identification
on-site.
CIRIA C671 • Tunnels 2009
75
In many types of infrastructure, systems for longitudinal measurement (along railway
lines, pipelines, canals, major roads etc) are already in place and can be used for this
purpose, but it is also necessary to record accurately the location of features around the
circumference of the tunnel. This can be done by reference to the structural form of the
tunnel, using descriptive words such as sidewalls, haunches, soffit, crown etc. There are
benefits in using a grid system that divides the tunnel longitudinally and circumferentially
into regular sections so that features can be assigned to an individual cell on the grid. This
method is particularly suited to recording information on an electronic database. The
optimum size of the cells will vary according to the size and nature of the tunnel and the
type of activities it is to be used for. For example, in the 1980s British Waterways went
through a process of marking up the intrados of all their tunnels with a 1 m spaced grid.
Although this required a significant initial expenditure of resources, it has since proved a
valuable aid to tunnel management and in particular to tunnel inspection and
specification of repairs.
3.5.4
Managing risk
The purpose of the risk assessment process is to systematically identify significant risks,
allowing prioritisation of actions to minimise and manage them. Risk assessment
procedures can be applied by asset managers to ensure that both performance and safety
objectives are met within a business framework and that funds are justified and allocated
in response to safety and business needs.
The need for risk assessment of tunnel assets arises principally to satisfy statutory safety
obligations, as discussed in Section 3.6.1. These regulations require that hazards are
identified and assessed and that adequate levels of safety are maintained/assured. A
tolerable level of risk can be identified, above which measures are to be used to ensure
that risks are reduced to as low as is reasonably practicable (ALARP). Reasonably
practicable is interpreted in law to mean that safety measures should be undertaken unless
the cost, in terms of money, time and trouble, is grossly disproportionate to the safety
benefit. All of the principal transport infrastructure owners and operators have asset
management processes that encompass such risk assessment procedures.
Examples of safety hazards that might be considered, along with possible risk reduction
measures for existing tunnels, are given in Table 3.2.
The ALARP principle allows that safety improvements should not be pursued at any cost
and only if the cost of averting the risk is not grossly disproportionate to the risk averted.
However, relative to other elements of the infrastructure the typically high replacement
value of tunnel assets and their criticality to operations may influence what is considered
reasonable in terms of minimising risks or recovering from accidents (for example, after a
fire or partial collapse).
The Standing Committee on Structural Safety (SCOSS), an independent body that
maintains a continuing review of building and civil engineering matters affecting the
safety of structures, is a further source of information concerning the assessment and
control of risk <http://www.scoss.org.uk>.
76
Table 3.2
Examples of hazards and risk mitigation measures for tunnels
Hazard
Risk mitigation measures
Structural instability
Accidental impacts (due to road,
water or rail vehicles)
Fire (from plant or services
contained in the tunnel or
vehicles or goods transported
through the tunnel or external
events)
Adjacent construction
(boreholes, piles etc)
Flooding (direct, from failure of
sealants, raised groundwater
pressure or partial collapse of
tunnel, or indirect from entry at
portals etc)
Explosion (from internal or
external sources)
3.5.5
principal means of risk mitigation is by condition appraisal, particularly carrying
out regular visual inspections at an adequate frequency to identify signs of
structural distress
use of appropriate monitoring systems and instrumentation
adequate routine maintenance (and, where necessary, repair) of structurally
sensitive elements reduces the risk of structural instability.
likelihood of impact reduced by considering vehicle use and adequate
clearances, or providing high-visibility signage where appropriate
consequences reduced by ensuring appropriate vehicle-strike response
procedures (reporting and response system, emergency tunnel closures and
engineering assessments), by frequent inspection and possibly by
reinforcement of vulnerable elements.
main control measure is through the use of fire-retardant materials in construction
consequence reduced by using fire preparedness plans
consequence reduced by using and installing fire protection systems, training
staff and involving the fire services
likelihood of fire reduced by controlling access, vehicle use and embargos on
materials carried through the tunnel
consequences reduced by installation of dry/wet mains.
likelihood reduced by restrictions on development within areas that can
influence the tunnel
regular walkover surveys above the tunnel to identify changes in land-use and
development that might affect the tunnel
consequences reduced by analysis and monitoring of the tunnel and adjacent
ground during any permitted construction works.
likelihood reduced through adequate inspection and condition appraisals being
carried out
consequence reduced by enforcing emergency preparedness plans
consequence reduced by the installation of emergency pumps and detection
equipment.
using increased security measures
as with the risk from fire by controlling access, vehicle use and embargos on
materials carried through the tunnel.
Whole-life asset costs
In new construction, whole-life costing (WLC) provides a rational basis for decision
making, allowing comparison of a variety of alternative construction schemes and aiding
the selection of one that is most economical or appropriate to the current or expected
financial position. For new structures a suitable design life may be specified that will allow
replacement/refurbishment in a planned manner and provide a basis for making decisions
on the optimum timing and extent of maintenance works. Similar principles can
potentially be applied to the maintenance of existing structures to assist with comparison
of alternative maintenance and repair schemes. In practice it is more difficult to set up a
reliable model, particularly where structures are expected to have a long (or even
indefinite) life, such as tunnels, because the long-term requirements, the likely frequency
of expenditure and an appropriate future discount rate are difficult to estimate. Also there
is a need to consider the specific infrastructure requirements, which will tend to dominate
the maintenance costs. For example, if rail possessions or traffic management programmes
are required, they will distort the relative merits/costs of maintenance/repair methods.
Determining and including such factors with adequate weighting can present further
problems. There is a risk that whole-life cost models can become overcomplicated, but on
the other hand if they are too simplistic this may defeat the whole object of the exercise
and their results may be misleading.
CIRIA C671 • Tunnels 2009
77
The most appropriate and realistic discount rate to apply is a contentious issue, because
this has a significant influence on the results, particularly when considering long-term
assets such as tunnels. The discount rates now recommended by HM Treasury and
applied in the public sector can mitigate against tunnel maintenance activities, because
they show a worse rate of return than delaying tunnel replacement and major
refurbishment. The danger is that this could potentially lead to an unfeasibly large
requirement for tunnel replacements in the future, which may be unsustainable in terms
of demands on resources and the disruption to transport and distribution networks.
Despite these challenges whole-life costing represents a rational approach to evaluating
alternative maintenance and repair strategies and, in certain circumstances, could
potentially provide a useful framework for helping to consider these complex issues where
they relate to tunnels, so long as its limitations are recognised. Some major transport
infrastructure owners are now developing life cycle asset management tools based on
whole-life costing principles, for example, Network Rail’s STAMP (structures asset
management process), which recognises that deterioration of assets is inevitable without
intervention, and seeks to model that deterioration and the consequences of alternative
maintenance strategies available.
3.6
HEALTH AND SAFETY AND ENVIRONMENTAL
MANAGEMENT
3.6.1
Health and safety management
Working in tunnels can be dangerous and regulations and practical measures exist to
ensure the safety of those working on them and members of the public who may use them
or be affected by them. Owners and operators of tunnels have obligations to maintain
assets in a safe condition to protect employees, those not in their employment and the
environment from unreasonable or unacceptable risks so far as is reasonably practicable.
These obligations arise from statute or licence conditions, and also from responsibilities
under common law duty of care, and have a significant impact on asset management
policy and operational procedures. All major tunnel owners and operators have internal
asset management policies and procedures that provide a framework for satisfactorily
fulfilling these obligations. Health and safety and environmental management
considerations with particular relevance to tunnels are discussed in the following sections.
Risks associated with loss of tunnel performance or collapses are managed by the process
of condition assessment, maintenance and repair, which is the principal subject of this
document. Carrying out maintenance, repair, refurbishment and reconstruction works
presents a range of hazards, including:
78
1
Risks to tunnel users and members of the public. In some situations it may be
necessary to carry out work in areas where a tunnel is not entirely closed to use or
where members of the public are near and might be affected. Risks are also presented
where works might trigger structural instability or collapse, which might affect people
or other structures within the tunnel’s potential zone of influence (see Section 2.7.1).
Safe systems of work should consider the safety of the public as well as workers.
2
Inherent hazards of the environment. The confined nature of tunnels makes them
potentially hazardous worksites requiring, for example, working in or near a live road
or rail traffic environment, working over or near water, working in confined spaces,
exposure to risks from falling objects or materials, working at height, exposure to
noise and vibration and use of access equipment – with all the hazards associated with
such conditions. The poor, dark and wet conditions that may be present in some types
of tunnel can lead to increased risk from normal worksite hazards such as slipping
and tripping, and the presence of vermin or waste material may lead to biological
hazards, including leptospirosis (Weil’s Disease).
3
Exposure to hazardous materials. These may be included in the fabric of the tunnel
or services contained therein (eg asbestos), be used in carrying out works (eg chemical
treatments), or arise in the course of carrying out works (eg dust, fumes and
poisonous or flammable gases). Gas testing and ventilation are fundamental
considerations, including the risk of methane migration into tunnel voids, and the
potential for build-up of fumes from generators and other equipment to be used in
the works (Swannel, 2003). Planning for works should include an assessment of such
hazards and the need for temporary ventilation and atmospheric monitoring.
4
Use of plant and equipment. As with any other civil engineering work, there are risks
associated with the use of heavy plant, such as cranes and excavators, light plant, such
as generators, and hand-held tools, such as angle-grinders. Special access equipment
and scaffolding is often required, and this can also introduce hazards to the work.
5
Risk from fire or explosion. Works within tunnels may require the use of equipment
that could make operation potentially hazardous such as the use of cutting gear or
petrol/diesel driven plant. Associated problems could include the movement and
storage of flammable materials (including gas bottles) and their inclusion in
temporary or permanent works. Additionally, local concentrations of flammable
natural gases may be present and should be checked for in confined spaces, and the
use of intrinsically safe electrical equipment may be required. It is necessary to
identify and assess the risks from potential fire hazards and adhere to the same
policies on fire precautions as during other construction work.
6
Means of escape and emergency access. Many of the normal hazards associated with
carrying out works on surface structures are potentially more dangerous in enclosed
tunnel environments because of the limited means and routes of escape, possible
distance to points of egress, problems with communications and difficulties for access
of emergency services should accidents occur. So it is particularly important that safety
and emergency procedures are considered, adequate provisions are made and all
parties are suitably briefed before beginning any potentially hazardous activities
within a tunnel.
The potential hazards associated with any work should be identified and risks carefully
managed so far as is reasonably practicable to reduce them to an acceptable level and to
comply with statutory requirements. Industry guidance and standards are available to
assist in complying with these requirements, in particular BS 6164 (BSI, 2001c). Note that
such publications will not reflect changes made to legislation and industry practice since
their publication or most recent revision.
The comfort and welfare of staff involved in inspections, investigations, maintenance and
any other tunnel works should be considered and adequately provided for, not only to
satisfy health and safety requirements but also to assist them in carrying out their work to
a high standard and improving the quality of the results (as discussed in Section 4.3.5).
Those carrying out works in tunnels should be physically fit, properly trained to a
certified standard, suitably experienced and have a good understanding of basic tunnel
safety requirements (see Section 3.6.2 on Competence).
Those involved with planning and executing inspections should be aware of the relevant
health and safety hazards to individuals and the environment and, at a minimum, ensure
that they are dealt with in accordance with statutory requirements. Risk assessments
should be carried out to ensure that hazards are identified, risks are assessed and where
necessary measures taken to minimise risks to acceptable levels. Inspectors should always
CIRIA C671 • Tunnels 2009
79
be alert, and aware of procedures to be followed and people to be contacted in case of
emergencies. This can be assisted by the preparation of a method statement for the works,
which is a formal requirement of many of the larger infrastructure owners. If a safe system
of work that mitigates all risks cannot be generated due to a lack of information, then
sufficient investigation should be carried out to supply that information. For example, this
could mean further desk study research, a reconnaissance visit/walkthrough of the tunnel
or a preliminary investigation with its scope limited to safely obtaining the necessary
information.
Owners of infrastructure should have their own health and safety management systems to
allow them to meet legal requirements. Also to complying with their own safety
management systems, consultants and contractors should adhere to the owner’s systems
when carrying out any work on-site. When construction work is being carried out on an
operational site that is under the control of the owner, co-ordination may be necessary to
clarify who is in control of the work area. The CDM Regulations 2007 (HSE, 2007) include
requirements relating to the control of construction work.
A comprehensive description of current health and safety legislation can be found in Tyler
and Lamont (2005). However, it should be remembered that legislation is liable to change
and it is the responsibility of those involved in the management of tunnels and tunnel
works to ensure that current legislation is adhered to.
3.6.2
Competence and training of staff
BS 6164 (BSI, 2001c) includes advice on the competence of staff as follows:
“The most vital contribution to health and safety in any tunnelling operation is
through competent engineers and managers, and a competent workforce.
Competence is gained through a combination of training and experience. All
persons underground should be competent for the environment in which they are
working and for the work tasks and activities they are required to carry out.
Engineers, managers and supervisors should be competent both with respect to the
work under construction, and in the techniques of management, communications
and supervision. Evidence of competence such as the achievement of recognised
qualifications should be sought.”
It is a legal requirement for all persons at work to be given appropriate training in health
and safety related to the risks they might encounter at work. This may require specialist
training for first-aides, those operating plant and machinery or working in hazardous
environments such as confined spaces. Induction training should be given before any
person starts work underground, whether as a new employee or as a person new to a
particular project because the specific hazards of working in tunnels may vary from those
associated with working in other environments. Longstanding employees are particularly
vulnerable because the familiarity and routine that come with experience may lead to a
false sense of security and increase risk. Refresher training should be provided at suitable
intervals.
It is particularly important that people working in tunnels are physically fit to carry out
their work in this environment, and that any health problems or disabilities do not
constitute a hazard to themselves or those around them. To this end, some companies
carry out regular medical assessments of their employees and require that they be
certified as fit to work in tunnel environments.
80
For safety-critical tasks and in certain industries it may be a requirement to demonstrate
continuing competence through periodic reassessment.
3.6.3
Heritage conservation
The heritage authorities have a general duty to conserve the built heritage, and works on
tunnels or parts of tunnels with recognised historic value. Those within certain areas that
have special environmental protection may require their consultation and co-operation:.
The authorities are:
1
English Heritage.
2
Historic Scotland.
3
Northern Ireland Environment and Heritage Services.
4
Cadw (the historic environment agency of the Welsh Assembly Government).
Very few tunnels are protected as listed structures but many can be considered as being of
historic importance. However many tunnel portals and tunnel-related structures are listed
under the Planning (Listed Buildings and Conservation Areas) Act 1990 for England and
Wales (and equivalent legislation elsewhere in the UK) and are subject to statutory
controls. Tunnels may also be afforded protection under a variety of designations of the
land where they are sited, for instance as a conservation area, Site of Special Scientific
Interest (SSSI), Special Area for Conservation (SAC) or National Parks (see Section 3.6.4).
These designations highlight the need for a special approach to the management and
conservation of existing structures, and frequently indicate special statutory protection
and restrictions on any works that may affect them or the surrounding land. Works that
affect only the settings of listed or Scheduled Ancient Monument structures do not require
listed building consent or Scheduled Ancient Monument consent, but setting is a material
consideration in planning applications.
Where a tunnel’s historic value is recognised by some form of statutory designation, such
as listed building status, this is likely to have a significant effect on the options available for
maintenance and repair works, particularly those affecting the original fabric or
historically significant alterations. Such works require careful and co-operative
management. Such works should be carried out in a manner that is sensitive to the
tunnel’s important historic features, and with the advice and consent of the relevant
heritage bodies.
The Panel for Historical Engineering Works (PHEW) is an advisory body run by the ICE
that has a database of works of an historic nature including tunnels. Although not
mandatory it is recommended that their advice should be sought in any matter that relates
to the repair or alteration of what might be an historic tunnel.
3.6.4
Environmental conservation
Infrastructure owners have statutory obligations in respect of the environment and these
have to be reflected within their asset management policy. Also to these statutory
requirements there are various other reasons why it is in the interest of tunnel owners to
consider the environmental aspects of their tunnels. Asset owners are already taking steps
to satisfy their obligations in these respects by the formulation of environmental policies
and action plans, with a requirement to carry out environmental audits on infrastructure
projects.
CIRIA C671 • Tunnels 2009
81
3.6.4.1
Conservation bodies and environmental legislation
Works associated with tunnels and associated infrastructure may affect protected sites or
protected species. The nature conservation bodies (known as the Statutory Nature
Conservation Organisations or SNCOs) have responsibility for promoting the conservation
of wildlife and natural features:
1
Natural England.
2
Scottish Natural Heritage.
3
Northern Ireland Environment and Heritage Service.
4
Countryside Council for Wales.
There are various categories of sites with designations for environment and conservation
(both statutory and non-statutory) at international, national, regional or local level that
can affect tunnels. These designations afford varying levels of protection and carry with
them restrictions on the types of activities that can take place, which are likely to have a
significant influence on any works undertaken within them. They stipulate procedures
that should be followed for notifying relevant authorities and gaining permissions to
undertake any work. This should be considered from the start of a project, and may have
significant effect on the selection of works and method of working, and also on the
programming and cost of works. For example:
under the Wildlife and Countryside Act 1981, operations on Sites of Special Scientific
Interest (SSSI) must be agreed with the appropriate SNCO, and species listed as
protected must not be killed or have their habitat damaged without a licence. Special
consideration must be given to any work on such sites to minimise disruption to
habitats and employ environmentally friendly methods of working
under the Habitat Regulations, 1994, SNCOs can permanently ban operations that
they consider may damage SAC (special areas for conservation) or SPA (special
protection area) designated sites. Although appeals can be made, these must be on the
basis that the works are for imperative reasons of overriding public interest and that
no alternative solutions exist. Where appeals are granted, compensatory works are
likely to be required, ie the creation of suitable replacement habitat that should ideally
be ecologically functional before the original habitat is damaged.
Where work may visibly affect protected sites, for instance on the land above the tunnel or
at the portal areas, this may require the involvement and permission from the relevant
SNCO. Although work within tunnels may not be externally visible, it still has the
potential to cause environmental damage either directly, through pollution or the
disturbance and destruction of wildlife habitats and species, or indirectly through its
effects on the local environment (eg changes in the local hydrological regime).
The ethos underpinning EU environmental legislation is the precautionary principle ie
that prevention is better than cure and that it is important to prevent foreseeable acts of
environmental damage. The continual review of legislation to ensure that good practice is
always followed should become part of the design process. For example, the actions
required to prevent pollution during construction are usually relatively easy and cheap to
do compared with the cost of the clean-up if pollution occurs. It is particularly important
to take measures to prevent pollution where tunnels pass near watercourses or through
aquifers.
Newton et al (2004) provides a useful summary of wildlife legislation and planning guidance
relevant to the UK construction industry, and how it affects those involved with construction.
82
3.6.4.2
Wildlife conservation
Tunnels can provide habitats for a variety of flora and fauna, including bats, birds,
amphibians, reptiles, insects and small mammals. The Wildlife and Countryside Act 1981
(as amended) and the Countryside and Rights of Way Act, 2000 afford protection to
certain endangered species of wildlife, and the presence of certain species of plants and
animals can have a profound effect on routine maintenance and repair works. Seeking
ecological advice at an early stage in a project that might affect protected species or their
habitats is important in determining and mitigating the potential impacts, and may avoid
serious repercussions to progress and budget.
The major infrastructure owners typically recognise the value of wildlife on their land and
work with the SNCOs to manage protected habitats. The preservation and management
of wildlife habitats should be incorporated into the asset management plans for structures,
and reflect overall environmental management targets. To this end, the major
infrastructure owners employ environmental management staff as part of management
teams to assist with determining environmental policy and to liaise with other specialists in
measuring and achieving environmental targets.
For more detailed information about protected flora and fauna useful reference sources are:
CIRIA C587 Working with wildlife (Newton et al, 2004)
CIRIA C502 Environmental good practice on site (Coventry and Woolveridge, 1999a)
Defra <http://www.defra.gov.uk>.
All tunnel works should include consideration of the potential presence of bats and other
protected species (see Box 3.1).
3.6.4.3
Managing environmental impact
Environmental aspects to be considered in management and maintenance of tunnels
include:
consumption of limited resources (materials, energy)
air, noise and water pollution
soil and waste
discharge of water from the tunnel
safety
visual impact
land-use
flora and fauna.
Where the environment and sustainability are concerned, construction and development
are particularly sensitive issues. Environmental appraisals are now mandatory elements of
the planning and design of new transport routes, but are less commonly considered for
the maintenance and repair of existing infrastructure. Environmental legislation, coupled
with a greater understanding of the potential impact of construction activities, is leading to
the incorporation of environmental concepts and aims as a core policy of national and
local authorities. Efforts are being made to develop methodologies for assessing and
comparing the real environmental impact of alternative infrastructure management
policies. A good example of such an approach is given in Steele et al (2003) where a life
CIRIA C671 • Tunnels 2009
83
Box 3.1
Dealing with bats
Old tunnels, as well as shafts and adits, are a favourite roosting place of bats, which are strictly
protected by UK and EU law due to their rapidly declining numbers. They inhabit both rural and urban
sites and are easily disturbed by maintenance and repair works. Both English Nature and the
Countryside Council for Wales have highlighted damage and losses to bat populations associated with
the routine maintenance and demolition of old structures.
The Bat Conservation Trust summaries bat legislation in their Professional Support Series leaflets, and
as follows:
the Wildlife and Countryside Act 1981 provides protection for all bats and their roosts and requires
consultation with English Nature before carrying out activities that might harm or disturb bats
and/or roosts
the Countryside and Rights of Way Act 2000 adds the word reckless (in England and Wales) to the
offence of disturbing a bat or damaging/destroying a place a bat uses for shelter of rest (ie a bat
roost). This is important legislation because it protects bats and roosts from reckless and/ or
international disturbance/damage
under the EC Habitats Directive it is considered an offence to damage or destroy a breeding site or
resting place of any bat, or to deliberately capture, kill or disturb a bat. Most development and
maintenance works affecting bats and/or roosts require a habitats regulations licence that must be
applied for and obtained from the Department of the Environment, Food and Rural Affairs (DEFRA).
Activities such as repointing and repair of masonry may result in the disturbance of bats and loss of the
cracks and crevices necessary for roosting and hibernation, which may be illegal under the above
legislation. In some circumstances licences may be obtained from DEFRA to permit actions affecting
bats or their roosts that are normally prohibited by law, but it will be necessary to demonstrate that the
proposed works are necessary for public health or safety, or for reasons of overriding public interest.
Applicants must demonstrate that there is no satisfactory alternative and suitable mitigation measures
are likely to be required, including restrictions on the timing of works, protection of existing roosts or the
provision of alternative roosts. There is likely to be a requirement to monitor the bats and the adequacy
of the mitigation measures, and this may take considerable time. It is advisable to seek the services of
a professional environmental consultant with appropriate experience at an early stage of planning when
considering works that might affect bats or their roosts.
Where provision of alternative roosts is required, a variety of proprietary bat-boxes and other artificial
roosts are available for such uses, including bat bricks which can be included at suitable locations within
a masonry lining (Figure 3.4). When considering the use of such artificial roosts it is important that
expert advice from a bat specialist is sought to assist with their selection and location.
Figure 3.4
Proprietary bat brick artificial roost and suggested
locations for installation (courtesy Norfolk Bat Group)
Where tunnels are to be closed and portals blocked, the presence of bats is likely to require inclusion
of suitable measures for allowing their continued access and egress.
Bat mitigation guidelines (Mitchell-Jones, 2004) have been published by English Nature, and include
detailed guidance on bats, their habitats, bat surveys and acceptable mitigation plans for development
and construction, with case studies that may be useful to those who have to deal with bat-occupied
structures.
Further guidance is included in:
Bats in buildings (SNH, 2004)
Bats, development and planning in England (BCT, 2002)
Nature conservation in relation to bats (HA, 1999).
84
cycle assessment (LCA) approach has been applied to the management of a masonry arch
bridge.
A detailed consideration of the topics of managing environmental impact and
sustainability is beyond the scope of this publication, but reference should be made to
existing good practice guidance, for example:
CIRIA C502 Environmental good practice on site (Coventry and Woolveridge, 1999a) for
practical advice on environmental responsibilities when planning and executing civil
engineering works and how to fulfil them satisfactorily.
CIRIA C571 Sustainable construction procurement (Addis and Talbot, 2001) includes
advice on successful techniques and strategies for delivering construction projects that
encourage environmental responsibility.
BRE IP14/04 Environmental sustainability in bridge management (Steele, 2004) sets out a
method of considering environmental sustainability in bridge management, many of
the concepts of which could be applied to the management of tunnels.
Choice of materials
The choice of materials used for the maintenance, repair and construction can not only
affect the local environment but can contribute towards effects on the wider environment.
With the exception of some renewable sources, all energy sources and processes requiring
the use of energy release CO2 into the atmosphere. CO2 is a greenhouse gas and is
implicated in climate change, which affects the species, habitats and built environment
around us. The production and processing of new materials inevitably requires energy
and may have other environmental impacts, eg noise, pollution and land-use. Careful
consideration of the relative environmental impact of alternative materials used in
maintenance and repair can provide environmental benefits. For example, for equivalent
quantities of lime and cement mortars, lime production uses between 47 and 70 per cent
of the energy needed for cement production, with corresponding reductions in emission
of pollutants (Pritchett, 2003). Although the LCA of materials is often not straightforward.
With results that are subject to uncertainties and dependent on many assumptions, the
process provides a logical framework for helping asset managers consider alternative
strategies for achieving environmental objectives.
Waste, re-use and recycling
Waste affects the environment in several ways: loss of valuable resources, need for landfill
space, and the unnecessary production of additional materials. It may also lead to
unnecessary pollution. Wherever practicable, original materials should be re-used unless
they have already proven unsuitable or are in a state such that they are unlikely to provide
adequate performance. Where original materials are unavailable or unsuitable, used and
recycled materials that are not a part of the original structure may be considered, and
sourced locally wherever feasible, to reduce the demand for production and transport of
new materials. Where waste is unavoidable, measures should be taken to avoid pollution
and minimise its environmental impact.
Good practical guidance on waste minimisation and recycled materials can be found in:
CIRIA C513 The reclaimed and recycled materials handbook (Coventry et al, 1999b), which
summarises the opportunities for re-using and recycling materials with information
on their properties, performance, specification and use
CIRIA SP133 Waste minimisation in construction – site guide (Guthrie et al, 1997) is aimed
at construction workers to illustrate practical ways that they can help minimise waste
on site.
CIRIA C671 • Tunnels 2009
85
Prevention and control of pollution
Although a consideration of the consequences of pollution immediately bring to mind
contamination of the air, water and soil, the impact of noise pollution and other less
tangible, transient and indirect consequences of carrying out works should also be
considered.
When working on tunnels it is important to prevent materials from entering groundwater
where they can cause pollution problems, and suitable mitigation measures should be
incorporated into the working methods. This is especially important where potential
pollutants and hazardous materials are being used, or where tunnel works might affect
watercourses, aquifers and sensitive ecological sites. Activities such as injection of chemical
grouts or other materials containing potentially hazardous components into the ground
around the annulus of tunnels present a particularly high level of risk, and in the past
pollution incidents from such activities have had severe consequences for workers, the
public and the local environment (for example, see the section on the Hallandsås Tunnel
in Case study A1.18). Particular care should be exercised when using materials, such as
chemical grouts, which rely on the mixing of two or more components. In some the fully
combined and reacted end-product is environmentally innocuous, but the individual
components themselves may be highly toxic with potentially severe consequences if
unreacted material enters the environment.
Airborne pollution may also present a hazard, particularly where work may involve the
generation of dusts or gases, the use of sprays, or directly or indirectly involve the
disturbance of hazardous materials such as asbestos, requiring suitable hazard
identification and risk assessment, and may be subject to specific controls under relevant
legislation.
It is equally important that the general operation of the tunnel does not affect the
environment. In particular, discharge of ingress water from tunnel sumps and pumping
stations from transport tunnels. The otherwise clean groundwater may become
contaminated within the tunnel from pollutants arising as a result of the nature of the
traffic using the tunnel, ie contamination by hydrocarbons (oils, diesel and petrol spillages
etc) from road vehicles.
Under the Environmental Protection Act (1990) it is an offence to deliberately or
accidentally pollute controlled waters (all watercourses, lakes, lochs, coastal waters and
groundwater) and any discharges into them require consent from the relevant
environmental agency. Other waste produced on construction sites is subject to the duty of
care under the Environmental Protection Act, 1990 and may be subject to control under
the Waste Management Licensing Regulations, 1994 (separate legislation applies in
Northern Ireland).
Detailed guidance on water pollution is given in CIRIA C532 (Masters-Williams et al,
2001), which identifies potential sources of water pollution from within construction sites
and discusses effective methods of preventing its occurrence. Further guidance is given in
the Pollution Prevention Guidelines (PPGs) published by the Environment Agency for
England and Wales and equivalent agencies in Scotland and Northern Ireland, in particular:
PPG1: General guide to the prevention of pollution (EA, 2001a)
PPG5: Works in, near or liable to affect watercourses (EA, 2000)
PPG6: Working at construction and demolition sites (EA, 2001b)
PPG23: Maintenance of structures over water (EA, 2002)
These are available to download free from <http://www.environment-agency.gov.uk>.
86
3.7
TUNNEL OPERATIONAL SAFETY AND FIRE RISKS
Many of the factors that lead to accidents on normal transport routes, such as road
junctions, blind corners, level crossings and obstacles on railway tracks, are absent in
tunnels. When accidents do occur the confined nature of tunnels makes evacuation and
rescue more difficult and, in particular, a fire that would be a manageable incident
elsewhere can prove to be catastrophic. Several serious fires in tunnels over recent years
have put tunnel safety on the public agenda. This concern is directed towards both road
and rail tunnels. Recent fires in road tunnels resulted in 39 fatalities in the Mont-Blanc
Tunnel (Austria, 1999), 12 fatalities in the Tauern Tunnel (Austria, 1999), and 11 fatalities
in the St-Gotthard Tunnel (Switzerland, 2001). Most recently, in 2005, a fire in the eight
mile long Frejus tunnel linking Italy and France killed two drivers and kept the tunnel
closed for several months. In rail tunnels, fire in a funicular tunnel at Kitzeinhorn
(Austria, 2000) resulted in 155 fatalities and a metro tunnel fire in Daegu (South Korea,
2003) 198 fatalities, the result of an arson attack. Aside from the tragic loss of life, such
incidents can have serious long-term effects on the local infrastructure and reduce public
confidence in the safety of transport systems.
In response to these incidents, recent years have seen new initiatives launched at a variety
of national, European and international levels, involving bodies such as the International
Union of Railways (UIC), the World Road Association (PIARC) and the International
Tunnelling Association (ITA). These international initiatives aim to produce, and where
possible harmonise, safety regulations. In each of the tragedies mentioned, smoke was the
major killer and death tolls could have been lower with improved fire engineering,
including a better understanding of the tunnel structure’s response to fire and better
planning based on the behaviour of people involved in such incidents.
The principal safety risks in transport tunnels are:
fire
structural collapse
collision
derailment (of trains).
In tunnel fires, the most important parameters affecting the consequences are:
risk of vehicles stopping and becoming trapped in the tunnel
smoke generation, ventilation conditions and dispersal
time required for evacuation.
In 2001 the UIC published a leaflet outlining recommendations for measures to increase
safety in railway tunnels, covering the fields of infrastructure, rolling stock and operations
(UIC, 2001). The following priorities were agreed, their order reflecting a decreasing
degree of effectiveness, especially in the event of a fire:
1
Prevention.
2
Mitigation.
3
Escape.
4
Rescue.
Recommendations for each of these priorities were made, including elements such as the
inherent technical safety of rail and rolling stock systems, fire suppression systems,
CIRIA C671 • Tunnels 2009
87
improved systems for communication between train staff, operations centres and
passengers, emergency and evacuation training for staff, and the provision of escape
plans, routes and aids (such as handrails and signage showing escape routes).
A working group of the ITA has published recommendations for the protection of new
and existing road tunnels from the effects of fire (Russell, 2004). The guidelines include
minimum levels of fire-resistance to ensure structural stability is maintained for a period
of time that will allow safe evacuation, working time for fire and rescue staff, and to
prevent collapse, which could have catastrophic effects both below and above ground. For
achieving this in tunnels with concrete and metallic linings, as well as those including
elements such as anchorages and ceramic tile finishes, the guidance recommends a system
of general thermal protection, although such methods were not considered necessary for
structural linings comprising clay brick masonry or structural stonework.
The protection methods identified included:
upgrading the fire resistance of the structure
application of coatings that delay heat transfer to the structure
construction of secondary linings
installation of fire protection materials.
Although structural safety is important, smoke and asphyxiation is often the biggest
hazard in tunnel fires and it is important that any materials used are inflammable and do
not give rise to toxic gases when exposed to the extremely high temperatures that can be
generated in such incidents.
In the UK the fire brigade is normally consulted during the design of new tunnels, and
involved in devising suitable fire-fighting methods and evacuation drills for existing ones.
For existing tunnels it is frequently unfeasible to make significant changes to the structure
to improve safety, however improvements may still be made such as updating rolling stock
and taking operational measures. Multi-million euro improvement programmes have
already been initiated by many countries with older tunnels, including Austria, France,
Germany and Switzerland (Muncke and Zuber, 2004).
Detailed information concerning recommendations for tunnel safety and information
concerning tunnel fire risks, prevention and mitigation are beyond the scope of this guide
but for further information readers should refer to the following for current guidance:
the tunnel safety and tunnel fires working groups of the UIC
<http://www.uic.asso.fr>
ITA <http://www.ita-aites.org>
PIARC <http://www.piarc.org>
relevant European Community (EC) directives.
Another organisation promoting and researching tunnel fire safety is UPTUN, with a
collaborative project specifically targeted at ensuring a pan-European approach towards
the improvement of fire safety in European tunnels through the development of new
technologies and procedures <http://www.uptun.net>.
Further guidance is given in Beard and Carvel (2005) particularly Chapter 6 concerning
fire safety in concrete tunnels.
88
The effect of fire on tunnel structural elements and materials is discussed in Section 2.6.3.
3.8
MANAGEMENT OF TUNNEL SHAFTS
Shafts frequently present an increased burden of maintenance and a variety of problems
for tunnel management. They may present hazards to people and property on the ground
surface through:
accidental or intentional entry
gradual movement or sudden collapse of ground (see Figure 3.5)
presence of gas (which may be combustible or poisonous)
pollution and loss of water supplies.
They may also present risks to the serviceability and operation of the tunnel, and to any
people and vehicles using it, through:
Figure 3.5
unstable material falling into the tunnel
partial or complete collapse into the tunnel
changes in loading on the tunnel lining at the shaft eye resulting in lining instability
collection and diversion of water into the tunnel.
Results of collapse of material into an incompletely filled shaft (1909)
Where the presence and location of a shaft is known, these risks may be managed through
a programme of condition assessment, maintenance and repair as for other parts of the
tunnel, but problems occur where:
access is difficult or impossible (eg shafts are capped at both ends) so shaft condition
cannot be directly assessed
the adequacy and stability of shaft fill material is unknown.
CIRIA C671 • Tunnels 2009
89
3.8.1
Shaft identification and location
It is necessary for tunnel managers to take steps to determine the presence and location of
all shafts associated with their tunnels with a suitable degree of confidence, relative to the
potential risk presented to their operations and to the health and safety of their employees
and the public.
A typical procedure for managing the risk from suspected shafts is:
1
Assessment of likelihood of shaft presence.
2
Assessment of potential risks (including determining the zone of influence, see Section
2.7.1).
3
Determine location of shaft.
4
Determine specific level of risk.
5
Assess the requirement for risk management.
6
Take action to control any unacceptable risks to acceptable levels.
Strategies and techniques for the location of suspected shafts are discussed in detail in
Appendix A5.
3.8.2
Maintaining shafts
Although tunnel shafts can be considered an additional element or extension of the main
tunnel bore, their management and maintenance can present special challenges, in
particular gaining safe and adequate access for inspections, investigation, maintenance
and repair. Despite the associated difficulties, it is particularly important that the
maintenance of shafts is not neglected or treated as secondary to the maintenance of the
rest of the tunnel, because many of the serious incidents that have occurred in tunnels
have been associated with problems in tunnel shafts (several examples are presented in
Case study A1.18).
The safety and serviceability of any shaft plug, capping or covering must also be
considered where these are present. There are instances where fatalities have arisen as a
consequence of the unavailability of shaft access cover details.
The inspection of shafts, including access and health and safety considerations, is
considered further in Section 4.7.
Shafts and adits may require treatment to achieve one or more of several objectives:
to prevent accidental or intentional access and falling hazards
to cut subsidence or collapse of the ground surface
to control or prevent the escape of gases to the atmosphere
to control or prevent the collection and transfer of water into the tunnel
to allow development at the ground surface
to maintain the linings of unfilled shafts in a safe condition.
Guidance on these treatment methods and discussion of the particular access and health
and safety issues associated with working on shafts is provided in Section 5.6.
90
3.8.3
Development of land above shafts
If development is planned above a tunnel with shafts, the shafts will need to be made safe
or the area around the shaft left free from development. The diameter of the safety zone
for shafts with uncertain stability depends on expected size, the nature of the surrounding
ground and whether the method of filling or capping is known. Section 2.7.1 discusses the
assessment of the zone of influence of tunnels and shafts.
3.9
MANAGEMENT OF CLOSED AND DISUSED TUNNELS
Although some tunnels are closed because they have reached the end of their serviceable
life and a decision is made not to rehabilitate them, it is often the case that still-serviceable
tunnels are closed because their use is no longer required. Closed tunnels will need to be
decommissioned and managed according to their immediate and potential future
requirements. For example, whether there is the potential for re-commissioning (possibly
for another use) in the future, or whether unchecked deterioration and collapse could
adversely affect adjacent land-use, other structures or services.
For the management of disused tunnels condition appraisal and maintenance work is less
restricted, and maintenance and repair techniques should not suffer from some of the
constraints present in operational tunnels (for example, maintenance of adequate
clearances), but there are many other problems, particularly for the health and safety of
employees and the public:
where tunnels are partially filled or closed at one or both ends, they may be
considered as confined spaces, with all the additional requirements for safety and
training that entails, unless it can be justified otherwise
working conditions may be more difficult and hazardous than in operational tunnels,
with water ingress, dirt and rubble, the presence of vermin and biological hazards and
the potential for build-up of harmful gases. Some tunnels become partially flooded or
silted up over time making access particularly difficult
access for staff and equipment to the portals and through the tunnel may be
problematic if normal means of access are removed or not maintained (eg removal of
rails from rail tunnels, overgrowth of vegetation or deterioration of roadway)
where a tunnel has deteriorated to a state where it is considered unsafe for normal
access, special precautions and equipment are required to monitor its condition and
carry out any works that may be required.
Tunnels that are closed temporarily, and might be used in the future, need to be
maintained to similar standards as fully operational tunnels so far as the condition of their
structural elements is concerned, although it may be possible to adapt the frequency and
scope of maintenance. If essential maintenance is neglected then structural deterioration
may make it very difficult or uneconomic to return the tunnel to a serviceable state in
future. Deterioration may be more rapid and severe in disused tunnels because of a
reduction in the scope and frequency of inspection, assessment and preventative
maintenance activities and greater tolerance of problems such as water ingress and lack of
ventilation.
Consideration of the potential effect of a tunnel’s structural deterioration on the volume of
ground and area of ground surface within its zone of influence (see Section 2.7.1) is
important. In this case it is necessary to identify any associated risks to people, structures,
services etc and ensure that these are suitably controlled. Typically the tunnel’s structural
integrity should be maintained at an adequate level of standard and safety.
CIRIA C671 • Tunnels 2009
91
Where the future re-use of a tunnel is highly unlikely, consideration should be given to
the most effective and economical way of mitigating any risks associated with structural
collapse. This will vary according to the location of the tunnel, its depth, proximity to
other structures and services, and the local geological conditions. Structural infilling is the
best long-term solution (see Section 5.6), but there can be technical, logistical and
budgetary reasons why this is not carried out. Where tunnels are infilled, a plentiful
source of cheap but structurally adequate fill material should be sourced and transported
to site, and placed so that the tunnel is structurally supported with no voids because
further access for inspection or remedial work will be impossible. More often, disused
tunnels are inspected and maintained in a similar way to operational tunnels, but with
different and more limited serviceability criteria so they should retain adequate structural
safety to allow the entry of staff for maintenance tasks. This may mean managing a
controlled deterioration while ensuring adequate integrity of the principal structural
elements are achieved at minimum cost because they are not revenue-earning parts of the
infrastructure.
Whatever the management objectives for the tunnel, it is often advisable or necessary to
prevent unauthorised access. This may mean sealing the portals and any other possible
entry routes. At the same time, access should be maintained in a state that is adequate for
authorised persons (eg inspectors), for carrying out maintenance and repairs and for
emergency situations. Rather than bricking up portals and incorporating small access
doors, as has been done in the past, secure metal fencing systems can provide a cheaper
and more flexible means of restricting access.
Disused tunnels need to be regularly inspected to control the risk of deterioration and
collapse. Some disused tunnels are open along public pathways and access for inspection is
relatively easy. Other tunnels such as in Monsal Dale in Derbyshire are open on a
restricted basis to authorised guided visits. This again allows inspections to take place to
ensure that the tunnels are not deteriorating past a certain level of safety.
Many disused tunnels are inhabited by bats and possibly by other protected species. This
should be taken into consideration when managing them, particularly when carrying out
remedial work, when considering infilling, or when restricting access at portals and shafts.
These animals should not be disturbed or their habitats damaged, and they should be
allowed adequate means of ingress and egress (see Section 3.6.3.2).
BS 6164 (BSI, 2001) includes methodologies for safe maintenance, renovation and repair
of tunnels, risk control and emergency planning, which may be particularly relevant to
carrying out works in closed and disused tunnels.
92
4
Condition appraisal
This chapter deals with the condition appraisal of tunnels. In this context, the term
appraisal is used in its broader sense to encompass all the activities undertaken that
determine the adequacy of the tunnel structure to perform its required functions. These
activities can include inspections to determine current tunnel condition and gather data,
site investigations to obtain more specific data, and structural assessments to evaluate the
tunnel’s structural behaviour and, for example, the influence of any changes in loading.
Due to the large number of tunnels in use and their apparent durability, unless the
loading conditions or other key features change, appraisal by inspection is commonly
regarded as sufficient to assess their serviceability and identify any special requirements
for preventative or reactive maintenance and repairs. The main objectives of these
inspections are to establish the condition of the tunnel structure, both in absolute terms
and relative to the information gathered in the previous inspections, and to collect data
necessary for any further assessment required. For example, when a sudden change in
tunnel condition or development of deterioration is observed, or where repairs or
refurbishment works are planned, it may be necessary to support the inspection with
more detailed investigations and assessment of the tunnel’s structural stability and
capacity. Structural assessment may also be required where tunnels are subject to live
loading, such as where highways pass over shallow cut-and-cover tunnels – in which case
they are often assessed as though they were bridges.
4.1
TYPES AND SOURCES OF INFORMATION
Comprehensive and reliable data is a fundamental prerequisite for effective tunnel
management, and information should be readily available in an accessible and usable
format. However, this requirement presents special challenges for tunnel infrastructure,
over and above that typically presented by other types of structure, for example:
original construction records are often unavailable, or may be inaccurate or
incomplete
only the tunnel intrados is visible, and important features are hidden from view
limiting the amount of information that can be gained by visual examination methods
there may be restrictions on gaining regular access to gather and update data.
Depending on the infrastructure type this may be limited to short duration visits and
may require special provisions for people or vehicle access, lighting, health and safety
etc
direct inspection of certain parts may require special arrangements, such as inspection
of canal tunnels from boats or drainage of canals or sewer/water services tunnels, the
use of specialist inspection techniques, such as rope-access for shafts or diving for
underwater areas, or the use of remote sensing methods (eg robot CCTV) to limit the
exposure of staff to hazardous situations.
It is particularly important to make best use of opportunities to gather data from a tunnel,
and to ensure that the scope of inspections and investigations is adequate to meet
foreseeable requirements. Return visits to gather missing data are typically costly and may
be unfeasible. Likewise, inaccurate or incomplete data can be worse than no data at all and
the validity of both historic and recent data should be considered before its use. The level
CIRIA C671 • Tunnels 2009
93
of checking should be appropriate to the degree of confidence that is required and the
likelihood and potential consequences of inaccuracy.
The principal sources of tunnel information and the types of data typically gathered are
discussed in Appendix A2.
4.2
DESK STUDIES AND EXISTING INFORMATION
Whatever the type or age of tunnel under consideration, there may be a wealth of existing
information invaluable to understanding its current situation and planning for its
management and maintenance. Sometimes, particularly for newer tunnels, the asset
owner will already have a comprehensive store of information. However, for many older
tunnels the owner’s asset knowledge is, to some degree, incomplete, and useful
information may be held by other sources. A desk study is an efficient way of starting any
investigation, with the potential to gather a lot of information very quickly and easily. It
can also provide information that no amount of costly and intrusive site investigation or
sophisticated analysis can yield, such as records relating to the tunnel’s original design and
construction, and later modifications and repairs.
Existing information may include:
original construction details (and those of any later repairs or modifications), eg
designs, drawings (particularly as constructed drawings), records of materials and
progress of work
records of condition and performance in-service (eg old inspection reports)
local geology, hydrogeology, historic land-use, mining and mineral abstraction maps
and records
location of services and records of, for example, water abstraction and leakage from
private and public utility companies
records of construction, operational disruptions, other incidents such as ground
movements, accidents, injuries and fatalities.
Although existing information provides a valuable basis for many tunnel management
activities, it is important that its limitations are recognised and care is taken not to rely on
it without proper verification, particularly where it is used to assist in making important
decisions. Records are often incomplete, or can be misleading due to errors, inaccuracies
or by omission of vital information. For example, instances have been recorded where
tunnels have not been constructed to the design, where the contractor has not provided
the specified thickness of lining or has used alternative materials. Frequently, designs were
altered to suit needs during construction where problems were encountered, for example,
by thickening the lining over short lengths of unstable ground. Also, over the long service
life of many tunnels, original features of construction may have been changed.
Where existing, and particularly historical, information is to be used, a process of
validation is necessary. Whenever using unproven information it is necessary to ask:
94
1
What is the original source of the information?
2
How was it obtained and what assumptions does it make?
3
Are these assumptions reasonable?
4
Is it logical, does it make sense, and is it what might normally be expected?
5
Can it be easily verified using other available information from a different source?
6
What are the consequences of using this information if it is incorrect?
Wherever possible any information gleaned from existing sources should be verified,
particularly where the reliability of the original source is open to question or where
circumstances might have changed since the data was collected. All information should at
the minimum be subjected to a sense check to ensure that it seems reasonable and does
not conflict with other information on the tunnel. The extent of the validation exercise
should depend on the level of risk associated with the existing data being inaccurate. For
example, when designing repair works that could affect the structural integrity of a tunnel
lining, it would normally be recommended that historic construction records be validated
by direct intrusive investigation to determine critical factors such as lining thickness and
ground contact. Frequently though, factual information can be verified simply by a walkover survey of the site or a walk-through of the tunnel. It is normally quicker and more
economical to validate existing information than to start from scratch. Despite the need to
exercise caution, existing historic records are valuable and if treated appropriately can be
used to guide and inform future investigations and assessments.
4.3
VISUAL INSPECTION
It is necessary to continually update knowledge on asset condition and performance,
typically by periodic visual inspection supported by simple assessment techniques. Also, it
may be necessary to carry out more in-depth investigations of particular features or
phenomena, and to monitor aspects of tunnel behaviour and performance over time using
more advanced techniques and instrumentation.
Effective inspection requires an understanding of the tunnel structure, its materials,
behaviour and potential causes of deterioration together with knowledge of tell-tale signs
of problems and where to look for them. Effective inspections gather detailed, accurate,
well-presented and objective information to permit others (not directly involved in the
inspection) to understand the problems, draw conclusions and take action where
necessary. Even when no action is taken after an inspection, a complete and objective
record of what was found is vital to permit the next inspection to measure or assess any
deterioration or other changes during the intervening period.
4.3.1
Advantages and limitations of visual inspection
The main advantages of visually-based inspections are that they are simple, rapid, and
relatively inexpensive, do not require any specialist equipment and minimise disruption to
the use of the tunnel. If inspections are carried out by well-trained and sufficiently
knowledgeable staff who regularly inspect the same tunnels, visual inspection can provide
a good indication of tunnel condition and any changes.
Unfortunately, inspection does have weaknesses, the main ones being reliance on visible
features and subjectivity of observations. Typically the only part of a tunnel that is visible
to inspections is its intrados surface. This is particularly a problem in lined tunnels
because the body of the lining, its contact with the ground and the ground are all hidden.
Although the early signs of structural distress and deterioration may manifest themselves
in changes that are visible at the intrados (eg cracking, bulging, loose and fallen material)
certain features, such as lining thickness and voids between its extrados and the ground,
and defects, such as separation between rings of a brick arch lining, may be difficult or
impossible to discern from visual inspection alone. Also, important but visually subtle
changes to the tunnel intrados may be overlooked or perceived as inconsequential,
particularly where more dramatic defects are present, even though these may be
CIRIA C671 • Tunnels 2009
95
longstanding and of less importance. Whether such symptoms are seen and recorded
depends on the skill, knowledge and diligence of the inspection staff and their familiarity
with the tunnel. Lack of continuity of inspectors or in inspection methods can lead to
reduction in the effectiveness of inspections and confidence in results.
The usefulness of visual inspection is very reliant on the quality of records kept, which
should provide accurate and comprehensive details of condition and defects. This is
discussed further in Section 4.3.5.
The application of new and emerging survey and monitoring techniques holds some
promise for the development of more objective intelligent inspection methods in future
(see Section 7.2). However, there are steps that may be taken at the present time to
optimise inspection procedures and the quality of results (see Section 4.3.5).
4.3.2
Types of visual inspection and inspection intervals
The regime of tunnel inspection should ensure that any deterioration in the condition is
detected in good time to allow remedial action. The intervals between inspections are
typically specified by tunnel-owning organisations to satisfy compliance with their
statutory obligations and internal policies. The requirements for inspection are set out in
internal standards. For some of the main UK infrastructure owners these are:
for Network Rail (NR) tunnels, examination types, requirements and intervals are set
out in Railway Group Standard GC/RT5100 Safe management of structures, which is
supported by several other standards
at the time of writing, London Underground (LU) tunnels, inspection types,
requirements and intervals are set out in Engineering Standard E3701 Structural assets
inspection but this is in process of being replaced by Standard 2-01304-006
the requirements of the Highways Agency (HA) for road tunnel inspections and
inspection intervals are set out in BD 53/95 (HA, 1995). At the time of writing, BD53
is under review for updating. The EU Directive on road tunnel safety became UK law
in April 2006 and requires independent inspections every six years by an inspection
entity
the requirements of British Waterways (BW) for their tunnels are given in Mandatory
procedures for the inspection of operational assets (AIP, 2005).
The terminology of, and intervals for, inspection of tunnel structures varies between the
main UK infrastructure owners, but are similar in terms of their objectives and
methodology. This is set out in Table 4.1, which is based on the requirements stated in the
documents mentioned above. Requirements for inspection of tunnel equipment and
associated elements such as shafts, cross passages and adits may vary from those given in
the table. Depending on the asset owner, there may be other requirements, such as
minimum qualifications and competence for those carrying out inspections (see Section
4.3.3).
96
Table 4.1
Current tunnel structure inspection requirements of the main UK infrastructure owners:
Network Rail (NR), Highways Agency (HA), British Waterways (BW) and London
Underground (LU)
Type
Routine
surveillance
Routine visual
inspection
Routine detailed
inspection
Non-routine
inspection
Known as
Scope and objective
Intervals1
Superficial inspection (HA)
Length inspection (BW)
Permanent way inspection
(NR)
Cursory visual check for
deficiencies that might lead to
accidents or increased
maintenance. Part of the day-today surveillance of the transport
network carried out by
infrastructure owner’s staff (not
necessarily trained inspectors) in
the course of their normal duties
When staff visit the tunnel
during their duties
General inspection (HA, LU)
Annual inspection (BW)
Visual inspection of accessible
representative parts of the
structure (including adjacent
earthworks, waterways etc) from
ground level or from other readily
available walkways, platforms
etc to identify hazards and
changes in condition and
determine requirements for
detailed inspection
Maximum interval:
(LU) 1 year
(HA) 2 years after last
General or principal inspection
(BW) 1 year after last principal
inspection
Principal inspection (HA, LU,
BW)
Tunnel examination (NR)
Close or tactile (ie touching
distance) inspection of all
accessible parts of the structure,
including adjacent earthworks,
waterways etc with provision of
special access if necessary.
Visually based but can be
supported by measurement and
simple testing (eg hammertapping) of structure to gather
additional data
Special inspection (HA)
Additional examination (NR)
(BW)
Defect advice inspection
(LU5)
Undertaken in response to a
specific need (eg where
significant deterioration or
evidence of structural distress is
seen before, during and after the
passage of abnormal loads and
after flooding and accidents such
as impacts on the structure, fires
or chemical spillage). Visual
inspection can be augmented by
specialist techniques for
investigation of structure (in situ
testing, sampling and laboratory
analysis) as required
Normal intervals:
(LU) between 1 and 12 years2
(NR) 1 year3
(HA) 6 years4
(BW) maximum interval 5
years
As required, to investigate
particular feature or gather
specific information. May be as
a result of a risk assessment
Notes
1
2
3
4
5
Stated intervals between inspections are subject to changes in asset owner policy and procedures. The reader should check
for current requirements where appropriate.
Maximum interval varies according to primary lining type: one year for flexible iron, four years for brick/stone masonry and
concrete, 12 years for cast iron. E3701 also specifies principal inspection intervals for shafts: stair (tubbing) maximum
interval of four years. For service, vent, plant, pump, cable, disused (tubbing) shafts maximum interval of eight years.
Maximum frequency for detailed (tactile) inspection of Network Rail tunnel shafts is six years. Also to a check on the
condition of chimneys and for changes in land-use during an annual walkover survey of the ground above the tunnel.
Intervals can exceptionally be up to 10 years.
London Underground also require special inspections, which are regular visual inspections carried out at short intervals for
structures awaiting repairs.
When referring to generic inspection types, this guide adopts the terminology used in the
first column of Table 4.1.
As indicated in Table 4.1 and its footnotes, infrastructure owners may have separate
requirements for the visual inspection of tunnel shafts, which vary from that of the tunnel.
The main aim of the inspection process is that the infrastructure should be maintained in
a safe and serviceable condition, and the scope, frequency and quality of inspections
should allow timely and appropriate action to achieve this aim. Aside from satisfying
CIRIA C671 • Tunnels 2009
97
statutory obligations in such respects, the period between inspections for an individual
tunnel should be determined dependent upon the findings of the previous inspection, the
tunnel’s sensitivity to deterioration, and its criticality within the infrastructure network.
Fixed-schedule inspection and assessment schemes have some negative consequences
because valuable resources are spent on tunnels that are known to be in excellent
condition whereas tunnels in poor condition may not be inspected as regularly as
necessary. A measure of flexibility is desirable, based on a proper assessment of risk, so
that resources can be directed where they will be most effective, while ensuring the prime
objectives of safety and functionality. Subject to the policy of the tunnel owner, limited
variations in inspection frequencies may be permissible depending on the use, type,
condition, deterioration and accessibility of the tunnel, and the perceived effectiveness of
the inspection. This requires justification, typically through a risk assessment process to
demonstrate the acceptability of the proposed inspection frequency. This approach is
considered advantageous, but the risks associated with increasing inspection intervals
need to be adequately assessed on a structure-by-structure basis.
Consideration may be given to increasing the period between inspections if it has been
demonstrated that:
the condition of the structure is good and there is no potential for rapid deterioration
there is a good level of confidence in the results of inspections and assessments
it is not envisaged that there will be any significant changes in use, loadings or
environment that might detrimentally affect the tunnel
the potential modes of failure of the tunnel are understood and there is adequate
confidence that the proposed inspection type and frequency can adequately identify
structural distress in advance of failure, or that the consequences of failure are low
the likelihood of incidents that might affect the structural integrity of the tunnel (eg
ground movements, water inflow, damage through vandalism) is low.
Conversely, consideration may be given to decreasing the period between inspections if it
has been demonstrated that:
the condition of the structure is poor and deterioration is continuing or there is the
potential for rapid deterioration
the level of confidence in the results of inspections and assessments is low
changes in the use, loading or environment of the tunnel are foreseen, which might
detrimentally affect its performance
the potential mode of failure of the tunnel is poorly understood and there is
inadequate confidence that the current inspection regime can identify structural
distress in advance of failure
the consequences of failure are perceived to be particularly high
the likelihood of unforeseen incidents that might adversely affect tunnel integrity is
not low.
Where risk assessments are used to justify reductions in inspection frequency, it is
particularly important that they are updated with current data, reviewed and re-assessed
at suitably regular intervals.
Access, programming and timing of inspections is discussed in Section 4.6.3.
4.3.3
98
Competence of inspection staff
Although asset managers can specify the range of information to be gathered in the course
of inspections, the quality of this information relies entirely on the capabilities and
competence of the inspection staff themselves. However, the quality of inspection and
reporting can vary considerably between staff unless they are selected by ability and
provided with formal training to equip them with the skills required to adequately fulfil
their role, commensurate with the complexity of the task, and are supported with the
necessary resources. It is also necessary that they have an adequate level of understanding
to be able to judge when emergency measures are required for safety reasons.
Some asset owners specify a minimum standard of qualification and/or competence for
their inspectors. For example, Network Rail requires tunnel inspections to be carried out
by a chartered engineer.
The basic qualities of a good inspector are (after DfT, 2005):
knowledge of safe working practices and access requirements for inspection
experience of the techniques and tools available, and an understanding of their use
and limitations
an adequate understanding of the construction, materials and behaviour of tunnel
structures.
knowledge of the causes of structural defects and deterioration of tunnel construction
materials
adequate understanding of tunnel modes of failure and the ability to recognise and
interpret features that might require urgent action
the ability to make and record objective observations accurately, clearly and
consistently.
For a novice inspector to attain these qualities and become fully effective they are likely to
require some formal training in addition to experience gained by apprenticeship to an
experienced examiner to allow the transfer of knowledge and skills. In certain situations
specialist training and skills may be required, for instance where inspections require roped
access or working in confined spaces.
4.3.4
Visual inspection procedures and techniques
Visual observation is used as the first and most basic method of obtaining key information
on a tunnel, and determining and monitoring its condition. The shortcomings of visual
inspection, discussed in Section 4.3.1, can be overcome by supplementing it with
additional simple and rapid techniques such as photography, dimensional measurement,
hammer tapping and other simple on-site actions. These can be applied in the course of
an inspection where additional information obtained would be beneficial. The range of
techniques that can be used during visual inspections are discussed in Appendix A3.
Inspection procedure is likely to vary depending on the infrastructure type, the type of
tunnel, the requirements for access and the infrastructure owner’s internal procedural
requirements. Frequently some element of familiarisation is required for inspectors, who
may not have visited the tunnel previously or recently. This will involve a review of earlier
inspection records and general information in the tunnel asset records, and may include a
reconnaissance visit. The inspectors should take care to familiarise themselves with any
particular aspects or features of the tunnel that require special attention, for example,
existing defects, areas that are sensitive to deterioration and structurally critical elements.
Inspection procedures are discussed further in Appendix A2.
CIRIA C671 • Tunnels 2009
99
4.3.5
Optimising inspection procedures and results
Successful inspections rely upon accurately making and recording relevant observations in
a systematic and objective way. This helps comparison with the observations of previous
inspections, and allows inspectors (on-site) and tunnel engineers (back in the office) to
discern current condition and identify any changes.
Attempts have been made to try to ensure greater objectivity in the inspection process by
better training of examiners, clearly prescribing a comprehensive range of observations to
be made during examinations, and wherever possible trying to make observations
quantitative or semi-quantitative, for example, by requiring measurements to be made or
observations to be assigned an index value in accordance with a prescribed rating system.
This systematised data, in standardised and often numerical form, is suitable for recording
and comparing as part of tunnel management systems and can be manipulated and
analysed far more easily than non-standardised information such as an inspector’s general
comments on condition. Each of the major UK infrastructure tunnel owners has their own
systematised procedure for condition assessment and reporting, so that requirements for
data collection are dictated by the needs of the owner. For example, Network Rail’s
structures condition marking index (SCMI) is designed to make objective and standardise
inspection information to allow it to be more easily interpreted and compared. London
Underground has a similar system where the extent and severity of the condition of the
lining is scored using prescribed inspection template forms to provide an overall condition
rating for the structure. Recommended actions and priorities are also indicated against
each identified defect.
When considering the quality of data from tunnel inspections, it is also important to
consider the influence of the human factor and its potential effect on the quality of
observation and recording.
There are several practical steps that may be taken to improve the quality and consistency
of visual inspection observations and records:
100
1
Inspection procedures and classification systems for observations should be carefully
devised and recorded in inspection handbooks with clear, illustrated descriptions and
examples, supplied to each inspector.
2
Where possible, reporting should be standardised to reduce the risk of error and/or
important data not being recorded and to help compare observations.
3
Inspection pro forma should be devised to capture the required range and detail of
information, and to prompt inspectors to view and record information in a consistent
and systematic way.
4
Inspectors should be encouraged to make liberal use of annotated diagrams,
photographs and direct measurements of the structure to illustrate and highlight
features of interest eg condition and deterioration.
5
Hand-held data-logging devices may be pre-programmed with defect types and
prompt inspectors to record observations in a comprehensive and objective way. They
allow rapid recording, potentially increasing survey productivity and helping its later
use. However, the capabilities of such devices should not limit the scope and
complexity of investigation records unnecessarily and inspectors should still be
encouraged to augment electronic records with dimensioned sketches, photographs
etc. Electronic equipment should be suitably waterproof for use in wet tunnels.
6
Adequate lighting is one of the most important requirements for tunnel inspection,
but many tunnels do not have integral lighting systems. Hand-torches are seldom
adequate because important features can easily be missed. In many situations halogen
lighting, powered by a small generator mounted on a vehicle or trolley, provides a
much better intensity and spread of light and is likely to lead to improvements in
observation. Otherwise, powerful head-torches with long-life belt-mounted battery
packs are preferred to hand-torches, because they illuminate the area being viewed,
are less cumbersome and leave both hands free for other tasks.
7
Observing and recording irregularities in tunnel intrados (eg bulges or cracks with
displacement) can be aided by illumination with incident light, ie by taking the light
source away from the observer (or camera) and directing it at a shallow angle to the
surface so that unevenness is accentuated by highlights and shadows.
8
Inspectors may use simple assessment methods for in-the-field evaluation of certain
tunnel parameters, for example, qualitative or semi-quantitative assessments of
materials condition (see Appendix A4.1). Some can be calibrated to give estimated
absolute values for masonry constituents (bricks, stone blocks or mortar).
9
It may be possible to augment the results of visual inspections by using some of the
data collected by other techniques and results from other types of investigation, for
example, laser scanning techniques that may have been carried out for other
purposes such as gauging surveys in rail tunnels.
An often overlooked factor that can influence the quality of inspection records is the
physical comfort of the inspection staff. Some tunnel environments are unpleasant and
difficult places to work, and it may be particularly difficult to concentrate on the process of
inspection while in cold, wet conditions or at the end of a long and strenuous shift with
limited welfare facilities. In particular:
warm and waterproof clothing should be provided, including gloves and good
protective boots. Also any other necessary PPE should be used so as to be comfortable
and avoid unnecessarily encumbering or restricting the inspector in undertaking their
activities
waterproof writing equipment (pens, pencils and notebooks) should be provided.
Paper-based records can be made on a clipboard protected inside a large clear plastic
map case that protects them from water but allows enough room to write in. The use
of hand-held data-loggers may be advantageous here
when working long shifts, adequate opportunities should be provided to sit, rest, take
refreshments and advantage of welfare facilities. However note that in potentially
unhygienic tunnel environments eating and drinking may be hazardous and is
prohibited by some infrastructure owners
adequate time should be allowed to complete the job of inspection thoroughly.
Oversights, mistakes and even accidents can occur if inspectors rush their task to
complete it within unreasonable time constraints.
In addition to providing basic requirements for ensuring health and safety, taking simple
measures such as these to meet the welfare needs of inspection staff is likely to result in
improved quality of observations and recorded information.
CIRIA C671 • Tunnels 2009
101
4.4
TUNNEL INVESTIGATION
4.4.1
Objectives of tunnel investigation
Tunnel investigations are typically carried out to gather information on specific aspects
related to a tunnel’s construction and performance, for instance its structure, type,
characteristics and condition of its fabric and information about the tunnel environment
(including ground conditions). This is often in response to a specific need:
to obtain detailed information on tunnel construction for asset inventories (eg to
locate hidden shafts)
to investigate the extent, severity, cause and consequence of apparent changes in
condition (eg in response to noted defects or deterioration)
to establish the effect of changes in tunnel environment (eg ground movements)
to obtain information necessary for the assessment of maintenance, repair or
refurbishment needs, and for the design of any associated works
to obtain information necessary for the assessment and design of alterations to the
tunnel in response to changes in requirements or in its use
to establish the structural condition of the tunnel before any proposed external
development that may influence it.
To meet these objectives, the investigation may need to obtain information on one specific
feature of the tunnel or often a range of features and characteristics, for example:
properties of the ground and any variations along the length of the tunnel, including
soil and rock types, physical and chemical characteristics, spacing and orientation of
fractures, faults and joints, joint fillings, presence of mineralised zones etc
for lined tunnels, the parameters of the lining – construction type, thickness, profile,
materials, condition, structural action, evidence of distress or deterioration
(movement and distortion, cracking, delamination, debonding, spalling and loss of
section etc)
invert parameters. Is there an invert and, if so, information on its construction type,
thickness and condition
what is behind the lining? Presence of voids/infill materials/water? Nature of contact
between lining and ground around whole extrados?
variations in water ingress and tunnel wetness, potential sources and paths of water
ingress
condition, capacity, performance and use of integral drainage systems
characteristics of any water entering, especially its chemical nature and any
contamination
tunnel shafts – presence of hidden shafts, potential for unknown shafts, shaft/shaft
lining condition and safety, ground stability, potential zone of influence etc.
Obtaining this information is likely to require the use of one or a range of investigation
techniques that should be selected to efficiently meet the investigation objectives.
4.4.2
Investigation strategy and reliability of results
Tunnels may pass through a variety of ground conditions, have a range of construction
methods and include different internal environments. This may not present a problem
102
when carrying out an investigation of a specific and localised feature, but may become an
important consideration when trying to characterise larger areas. A test performed at a
single locality may not be representative of the whole tunnel, and care should be taken
when extrapolating results. For example, if tests are concentrated on the worst areas, then
the results should not be considered as representative of the tunnel as a whole. It is
generally advisable to plan an investigation strategy that will encompass a representative
range of potential variation. Reliance on one type of test to determine key parameters is
discouraged in favour of a broader approach.
It may be useful to target typical, best and worst areas based on visual inspection, or to
classify the tunnel into several zones depending on its construction, condition, features
and environment (including ground conditions). The most appropriate approach will be
dictated by the nature and needs of the tunnel under consideration, the objectives of the
investigation, constraints on its scope (particularly the availability of resources and access)
and the confidence required in the results.
Several factors should be considered:
1
Are the features of interest easily identified and targeted (eg visibly damaged areas of
lining) or is the potential variation hidden – requiring a more statistically valid
approach (eg typical strength of lining materials)?
2
The size of the area or feature of interest (is it necessary to characterise the whole
tunnel, a part of it, or just a specific small area or feature?).
3
The need to draw comparisons between areas or features (eg areas in different
condition, exposed to different environments, or of different materials).
4
The potential for variation in the parameters of interest within areas, and the need to
fully characterise the range of variation in the test results.
5
The level of confidence required in the representation of the results, considering the
potential consequences of using non-representative test results.
In determining the most appropriate sampling strategy and the most appropriate rate of
testing, consideration of basic statistical aspects may be beneficial, in particular an
understanding of the concepts of populations, means and standard deviations. Further
guidance on sample numbers and the interpretation of test results is given in BS 2846-4
(BSI, 1976) and BS 6000-1 (BSI, 2005a).
Tunnels and their environments may be subject to gradual change and it is important to
appreciate that information from a single site investigation represents a single point in
time with respect to the structure. While this is adequate for some purposes, used in
isolation it cannot provide information on how parameters have changed over time, which
is frequently desirable. A single site investigation can, for example, identify a crack in a
tunnel lining, and possibly even allow its likely cause to be discerned, but taken in isolation
it is difficult to determine whether this is an inactive defect that has been stable for a long
time or whether it is recent and rapidly developing – scenarios that might prompt very
different reactions. Although there may be clues as to whether phenomena are recent or
longstanding, such as fresh surfaces on spalled materials, deep carbonation of concrete
along a crack, or the presence of thick deposits that have built up over time, such
indicators cannot always be confidently relied upon to provide adequate or accurate
information.
When devising a sampling and testing plan, there should be clear justification for carrying out each test
or sample at its particular location, and how the results will be used. Using inappropriate techniques or
obtaining unnecessary information is a waste of resources and can cause damage to the tunnel and
disruption to its normal use.
CIRIA C671 • Tunnels 2009
103
4.4.3
Techniques for tunnel investigation
Investigation of the structure of a tunnel and its environment is carried out to:
identify or verify its construction features
carry out in situ characterisation of materials and environmental factors
identify and determine the cause of damage and deterioration
to obtain samples of materials for further laboratory analysis.
For an unlined tunnel this is likely to include:
tunnel intrados geometry and dimensions
detailed geotechnical properties of the ground (eg in situ determination of bulk rock
characteristics, orientation and spacing of discontinuities, presence of faults, and
mineralised or weak and fractured areas)
obtaining samples for the laboratory determination of materials characteristics
hydrological assessment, water chemistry.
For lined tunnels, this might also include:
type, profile and thickness of lining (including any shaft linings)
nature of lining/ground contact (presence of voids, timbers, water etc)
presence of invert and its characteristics (thickness, profile)
presence of hidden construction shafts or other features
presence of any pillars or piers that might have been constructed on the lining to
support the ground, ie point loading on the lining
nature of ground behind the lining
obtaining samples of structural materials and the ground for laboratory analysis of
materials
variations in construction joint spacings in masonry-lined tunnels (see Section A4.1.4).
Techniques commonly used in the course of tunnel investigations to obtain such
information include:
104
coring and removal of core samples
use of endoscopes/borescopes
water sampling and analysis and local measurements of ingress rate
traditional and more advanced methods of dimensional measurement and surveying
(eg laser scanning and digital photogrammetry)
specialist non-destructive geophysical techniques (eg radar, thermal imaging,
ultrasonics)
semi-destructive in situ testing methods (eg carbonation depths, corrosion potentials,
strain measurements using flat-jacks or overcoring, pull-out tests to estimate strength)
removing panels of lining for analysis directly viewing the extrados and the area
beyond it
geotechnical investigation and sampling techniques.
These may be supplemented by laboratory testing of samples, using various analytical
techniques:
physical testing to determine properties such as compressive strength and modulus
chemical characterisation of materials by x-ray diffraction, thermography and other
techniques
petrographic and metallurgical examinations by microscopy
measurements of physical characteristics such as porosity and permeability
soils testing and characterisation.
More detailed information on many of these techniques is provided in Appendix A4, while
investigation and assessment techniques specifically for unlined tunnels are discussed in
Appendix A6.
4.4.4
Selection of investigation techniques
The selection of investigation techniques requires:
an understanding of their strengths and weaknesses
the specific circumstances and needs of the investigation
consideration of a range of other influences and constraints relating to the tunnel, its
environment, owner/user requirements, and health and safety and environmental
factors.
The most direct and definitive sources of information often rely on some form of
destructive testing, for example, taking core samples through the lining and subjecting
them to laboratory examination and testing. Note that careful consideration should be
given to the effect on the structure and how this damage can be repaired without causing
the structure to weaken, leak or deteriorate. Often it is possible to rationalise the number
of destructive tests by using them in combination with mildly-destructive or nondestructive techniques.
Commonly used and potentially useful techniques for investigating various parameters of
interest are identified in Table 4.2.
CIRIA C671 • Tunnels 2009
105
Table 4.2
Recommended methods for direct investigation of tunnel parameters
Parameter
106
Primary source
Secondary source
Other information
Lining thickness
Measurements taken from fulldepth core-holes (concrete,
masonry, cast iron) but with
consideration of possible water
ingress and making good
damage
Measurements taken from
retrieved cores
Non-destructive test data
(eg radar for concrete/
masonry or ultrasonic
methods for metallic
linings)
Original records of
construction (although data
should be verified). Previous
investigation results or
records from repairs etc
Lining materials
characteristics
Visual appearance
Laboratory tests on recovered
samples (eg cores or blocks of
masonry or rock, drilled or cut
samples of metals).
In situ measurement (eg
using flat-jacks for
masonry)
Estimates based on
examination or qualitative
assessment of recovered
samples
Published data for similar
types of materials
In situ stress in linings
In situ measurements (eg
using flat-jacks or by
overcoring for concrete and
masonry, ACSM stress-probe
for metallic materials)
Laboratory testing on
samples
Published data
Materials condition and
causes of materials
deterioration
Visual observations supported
by simple in situ tests (eg
hardness tapping or sounding/
acoustic energy meter for
voiding of ring separation in
masonry)
Laboratory tests on recovered
samples (petrographic
analysis, chemical and
physical testing)
Measurement of
environmental parameters
(eg groundwater chemistry)
Previous investigation
results and records of
repairs
Intrados profile
Direct measurement (traditional
or advanced surveying
techniques depending on
requirements)
Rock mass condition
Rock mass mapping
Observation of rock mass
condition and identification
of unstable rock blocks
Rock mass classification
Scan line mapping of
discontinuity properties
Undertaking RMR and Q
classifications
Ground conditions
Coring from within tunnel,
window-sampling, boreholes
from surface, penetrometer
tests, permeability tests,
piezometers, in situ ground
stress, lab tests on recovered
samples
Geological maps and survey
records, data from existing
boreholes in the area, and
hydrological survey
information
Records from original
construction and repairs.
Features of construction (eg
changes in lining type and
thickness, joint spacing and
water ingress)
Nature of ground behind
lining
Analysis of samples from
immediate area (eg by direct
coring, window sampling
behind lining or from
boreholes from surface)
Analysis of samples from
general locality
Published geological data
for locality
Records of construction and
form of construction (eg
thickness of lining and
spacing of joints)
Nature of tunnel/ground
contact
Direct observation of interface
between lining and ground by
inspection through core-holes
using an endoscope
Indirect survey using NDT
techniques such as radar or
ultrasonics, acoustic energy
meter etc
Presence of water
behind lining
Direct observation of wetness
of tunnel intrados, or through
core-holes in lining
Piezometers installed in coreholes or boreholes from
surface
NDT techniques such as
conductivity
Records of original
construction or repairs,
drawings etc
Published hydrological and
hydrogeological maps and
other data
Table 4.2
Recommended methods for direct investigation of tunnel parameters (contd)
Wetness of tunnel lining and
water ingress through it
Visual inspection can be
augmented by assessment
against a suitable scale
(see Appendix A3.2.6).
Wetness can vary with time
depending on several
factors, eg local rainfall,
leakage from services
Spot-measurements of
moisture content of
materials (see Dill, 2000) or
Meteorological records,
wetness survey using NDT
methods (eg conductivity or information from owners of
water services
thermography) or
measurement of local rate
of water ingress (by
collection)
Presence of hidden tunnel
shafts
Construction records, also
inference from observation
of intrados features, local
water ingress at crown,
distance between shafts,
presence of possible spoil
heaps above tunnel
Aerial photographs (spoil
heap detection)
NDT methods (eg radar,
resistivity, seismic and
microgravity surveys)
Intrusive geotechnical
investigation methods
Records of local history and
land-use
Aggressiveness of
groundwater
Sampling and laboratory
analysis to identify
deleterious pollutants (eg
sulfates, pH, chlorides for
reinforced concrete)
Inference from nature of
potential water sources,
adjacent ground conditions
Visual observation of
effects on the structure
It is important for both technical and budgetary reasons to consider the optimum
sequence of investigation works. Methods that provide rapid coverage such as laser
scanning of the surface, or radar surveying of the subsurface, provide a comprehensive
overview of tunnel characteristics, but often require verification by intrusive methods. It is
often best to first use the most rapid methods giving widest coverage, then use the results
to select representative locations for localised investigation methods, for example, drilling
or coring.
Most NDT/geophysical methods involve the interpretation of parameters such as electrical
conductivity or dielectric constant that do not directly relate to useful engineering
properties. The reliability of the interpretation inevitably varies from site to site because of
varying quality and quantity of data and the availability of calibration data such as coreholes or records. Asset owners should expect geophysical specialists to report confidence
levels in their findings. A discussion of the need for the specifier to understand NDT
black-box outputs is given in Turner (1997).
Considering health and safety and environmental aspects of such techniques is required,
particularly relating to working under infrastructure owners’ operating procedures, and
health and safety and environmental guidelines.
4.4.5
Optimising tunnel investigations and results
Tunnel investigations have very specific objectives. It is important that these objectives are
clearly understood and stated, and that the investigation is designed to meet them
efficiently.
An assessment of the constraints and their impact on potential strategy and methods of
investigation plays an important part in planning. In some cases what is deemed to be
technically the most suitable option may be inappropriate or impractical due to specific
site conditions. In many tunnels, physical and time constraints are important and are
often the controlling factor in the ability to carry out an investigation. The presence of
tunnel equipment and electrical cabling, the tunnel dimensions and geometry and its
structural condition all have a bearing. Investigations may cause disruption to the normal
function of the tunnel and may need to be carried out in restricted (often very short)
periods. They may require special traffic management and access provisions, and may
CIRIA C671 • Tunnels 2009
107
employ a variety of specialist techniques and sub-contractors. For a typical tunnel
inspection the majority of the cost is associated with access and traffic controls, so it is
important that the opportunity is used to its best advantage to gather all the information
required and avoid the need for repeat visits. So investigations require careful planning
and co-ordination between the various parties involved.
1
Attempts should be made to co-ordinate access arrangements so that inspections can
take advantage of tunnel closures booked for other activities where these will not
conflict. Conversely, a booked investigation closure should be opened to other parties
to take advantage, provided that their activities will not conflict with the investigation.
2
Investigations should be focused. Obtaining superfluous information results in
unnecessary cost, damage to the structure and disruption to the tunnel’s normal
function and should be avoided.
3
Investigation and testing techniques should be carefully selected with a good
understanding of their capabilities and limitations, the results they are expected to
yield, how they will be used to achieve the investigation objectives, and the level of
confidence that is required.
4
Techniques should wherever possible be used in a complementary fashion, ie their
strengths and weaknesses and the results yielded should combine to provide the
necessary range and quality of information to adequately fulfil the investigation
objectives.
5
A suitable single person or organisation should be made responsible for co-ordinating
all parties and their work.
6
Those responsible for carrying out different elements of the investigation (eg
specialist sub-contractors and testing laboratories) should have an understanding of its
overall objectives, how their activities fit into it, their responsibilities and what is
required of them.
7
Specialist sub-contractors should be carefully selected and are required to
demonstrate suitable skills and past experience. Often it is useful to involve them in
the process of specifying and planning the investigation so as to ensure that adequate
resources and support are available, and potential problems and risks are identified
and resolved at an early stage.
8
Risks to achieving the investigation objectives should be identified and measures taken
to minimise them to acceptable levels wherever practicable (eg by having backup
equipment and staff available on stand-by for critical tasks).
9
It is necessary to consider the health and safety and environmental aspects of the
works, particularly relating to infrastructure owners’ operating procedures, and
health and safety and environmental guidelines, and the appropriate training and
competence of all parties involved.
10 A clear method statement should be produced, setting out the scope of the
investigation and the parties involved and their responsibilities. It should list the
activities to be undertaken, where, when, who by and what equipment is to be used,
identify the hazards associated with the work, details of how they are to be mitigated,
and procedures in the event of unforeseen circumstances and emergencies.
Many asset owners use specialist sub-contractors to organise, carry out and interpret the
results of site investigations, including consulting engineers, materials specialists and testing
laboratories. All should be able to demonstrate the specialist knowledge required for the
task and preferably have a successful track-record of carrying out similar investigations in a
tunnel environment. A good understanding of the issues involved in working in such an
environment is important, because the associated requirements and constraints are often
different to those presented by other types of structure.
108
4.5
TUNNEL MONITORING
Monitoring is the repeated measurement of parameters at suitable time intervals to allow
comparison and assessment. This can be anything from periodic visual inspection to realtime instrumented monitoring of rapidly changing parameters. Requirements for
monitoring include:
verification of fitness for purpose of a structure
investigation of specific changes in the structure and its environment over time
monitoring the response of the structure to changes, for example, during and after
maintenance, repair and improvement works, to assess their effect on the structure
and its longer-term performance.
Monitoring results may provide the input parameters for numerical modelling both for
design and sensitivity of the structure so the type of monitoring may be dictated by the
input requirements. Structural models may also need to be calibrated by field
measurements to verify predicted responses.
This section considers the use of instrumented monitoring systems.
4.5.1
Objectives of tunnel monitoring
Monitoring is used to detect and/or measure change in one or more specified parameters.
Monitoring can be achieved by carrying out discrete repeat observations and
measurements of phenomena at suitable times, or gathering such data using a more
continuous automated approach, eg by installing suitable dedicated monitoring
instrumentation and logging devices.
Before the selection and design of instrumentation systems the first step in any monitoring
scheme is to clearly and logically define the objectives, including a precise description of
what is to be monitored, why, and what will be done with the results (this latter
consideration is particularly important but the one most often overlooked). Requirements
for monitoring systems vary, but include, for example:
to verify the continued fitness for purpose (condition and performance) of a tunnel
to investigate specific changes in the tunnel and its environment over time
to monitor the response of the structure to change, eg during works on the tunnel or
from construction works taking place nearby.
It is often desirable to supplement historical information with continuing assessments to
monitor condition and discern any changes. Many aspects of tunnel behaviour and
performance are the result of complex interactions between parameters that undergo
change over time: rates of change can vary. It is important to gain an understanding of
how the parameter of interest is affected by other variables (eg temperature effects) so that
these may be accounted for when interpreting monitoring results avoiding erroneous
conclusions. For any monitoring results to be useful the significance of observed changes
in monitored parameters and their relevance to the structure should be properly
understood. Some changes are of no consequence, whereas others may be highly
significant, and interpretation should be able to discern between these.
CIRIA C671 • Tunnels 2009
109
4.5.2
Monitoring instrumentation and techniques
Periodic inspection is the cheapest form of monitoring and is generally very effective.
However, inspection does have limitations and there are a variety of circumstances where
it is appropriate or necessary to use instrumentation to carry out specific monitoring tasks.
Frequently this involves the installation of an automated measurement system:
where access to staff is limited or presents a safety hazard
where the frequency and timing of measurement makes manual measurement
unfeasible, uneconomic or impossible
where long-term measurement is required, either due to the aims of the monitoring
exercise, or due to the nature of the parameter being monitored.
Systems are typically based on the installation of instrumentation (ie sensors and
transducers such as tiltmeters or strain gauges) onto the elements to be monitored, or
alternatively, for monitoring movement by setting up a system based on surveying
equipment and techniques.
Applied instrumentation is most useful where the parameters to be monitored are clearly
definable and suitable measurements can be made at specific locations. For example,
measuring changes in crack width over time, mid-span deflections of a beam under
changing loading conditions, or the strain developed at a specific critical point of an
element under stress. Sensors and transducers tend to have specific characteristics, and some
familiarity with their capabilities and limitations is required to use them effectively.
Traditionally, conventional survey techniques have been used satisfactorily for the periodic
measurement of long-term movements such as building settlement, but were less suitable
in potentially more dynamic situations, where short measurement cycles or instant
feedback is required, or where frequent re-measurement is required over a long period of
time. For such applications, monitoring would typically be carried out by applied
instrumentation. However, with the automation of survey instruments, the incorporation
of automatic target recognition and reflectorless measurement technology, continuous
movement monitoring using survey techniques and instruments is now a viable alternative
to applied instrumentation in certain circumstances. With fixed datum points, 3D optical
measurements using total station instruments can be used to make absolute measurements
of movement and deflection. The equipment used for such applications typically
comprises a motorised total survey station and a series of suitable prisms/reflectors for
attachment to target areas.
A discussion of instrumentation and techniques for common tunnel monitoring situations
is included in Appendix A4.4.
4.5.3
Selection and design of monitoring systems
As with one-off tunnel investigations, long-term monitoring is often carried out in
response to a specific need and so may have very specific objectives. It is important that
these objectives are clearly understood and stated, and that monitoring procedures and
systems are designed to meet them efficiently. Similar considerations apply to those when
designing an investigation (see Section 4.4.2) however there are many other issues that
should be considered:
1
110
The parameter (or parameters) to be monitored should be carefully selected and
clearly defined to meet the monitoring objectives.
2
It is necessary to consider the full range of potential factors that might influence the
parameter to be monitored (eg changes in temperature, moisture) and determine
whether these require further measurement or monitoring to allow interpretation of
the results.
3
The likely frequency of occurrence or the rate of change of measured parameters
should be considered and the monitoring system designed to accommodate this, ie is
continuous, frequent, or infrequent measurement required?
4
The system should be capable of measuring and recording the required range of
variation likely to be encountered in the parameter to be monitored.
5
The necessary frequency of data capture and analysis should be specified and
supported, ie does data need to be constantly monitored, or checked periodically, or
only at the end of the full monitoring period?
6
The likely total duration of measurement should be considered. Is monitoring
required to record a particular occurrence, or is it needed over a period of days,
weeks, months or even years? It may be necessary to characterise normal variations in
measured parameters, eg fluctuations in movements and background vibration. Does
any installation need to be temporary, semi-permanent or even permanent?
7
The power supply requirements for equipment should be considered. Is power
necessary? What type? Where from? For how long will it need to function? Can it be
self-contained (batteries) or should it be from an external source and, if so, what does
that entail?
8
The method of data capture should be considered. Will it be possible to access the
equipment to obtain data or does it need to be transmitted to another point, for
example, outside the tunnel portal (eg by cabling) or to an office location (eg by
telecommunications links)?
9
Does the system need to react to the data in any way? For example, is it necessary to
trigger alerts or alarms, or perform some other action?
10 The installed system should not cause problems with the normal function of the
tunnel, for example, by impinging on required clearances for traffic
11 The installed system should be capable of functioning adequately and reliably within
the tunnel environment (eg can it work in darkness, deal with likely temperature
variations, wetness or immersion, dirt and dust, vibration from traffic movements?
Could it survive these conditions for the whole period of the monitoring?)
12 The criticality of the data should be considered in the system design. What are the
consequences if the system fails to function as required? What is the risk of this and is
it acceptable? Is a backup system necessary or desirable? Is access available for system
maintenance and dealing with any faults?
13 Installed systems should be electromagnetically compatible with any permanent
electrical equipment in the tunnel.
14 Systems should be capable of self monitoring and advising the monitoring engineer in
the event of internal system, or data link failure.
Successful monitoring requires not only general background knowledge of
instrumentation and measurement techniques, but also a basic initial understanding of the
parameters to be measured and their likely behaviour. If the wrong type of instrument or
technique is used, or the right type is used in the wrong location, the data collected is
unlikely to fulfil the objectives of the project. In such circumstances, even where repeat
measurement is a possibility, redeployment of fixed instrumentation may be a costly and
time consuming exercise. Where monitoring movement and displacement applied
instrumentation is, in certain circumstances, less flexible than the use of survey methods,
which can often more readily be adapted to changes in circumstances or requirements on
CIRIA C671 • Tunnels 2009
111
the job, however system design and the choice of measurement techniques may be
dictated by the specific requirements and constraints associated with the current work.
The quality of instrumentation is a significant influence on the accuracy and reliability of
measurements, and the specified accuracy of a measurement system can be degraded by
several factors:
use of the wrong instrument types, or use of instruments with unsuitable range
poor selection of instrument locations
unsuitable installation methods and procedures
incorrect calibration of instruments
poorly designed connections and cabling back to loggers.
The influence of such factors may be minimised or avoided by careful system design,
including appropriate selection, calibration, installation and wiring of instruments.
4.6
PREPARING FOR INSPECTIONS AND INVESTIGATIONS
General issues relating to the planning and preparation of inspections and investigations
are discussed in Sections 4.3, 4.4 and 4.5.
4.6.1
Risk assessment
Inspection and investigation of tunnels involves exposure of those involved (and in some
cases the general public) to a variety of health and safety hazards including:
exposure to live traffic
working over or near water
falls from height
contact with services, equipment and hazardous substances
exposure to harmful gases and fumes
working in confined spaces
exposure to hazardous chemical or biological contaminants.
There may also be risks to the environment, including pollution of the air or watercourses
with harmful fumes or substances.
A risk assessment should be carried out and suitable methods of mitigation specified for
any risks that are unacceptable. A method statement that summarises all the information,
including safe methods of working specified in the risk assessment, should be prepared,
and agreed by all parties. The method statement should take into account the review of
records and reconnaissance of the structure, access requirements, health and safety and
environmental considerations (see Section 3.6). The level of detail given should be
appropriate to the complexity, circumstances and type of inspection.
The following information should be included in any method statement, (DfT, 2005):
112
details and programme of the work to be undertaken
equipment required
methods of access to be used
traffic management details
the risk assessment including safe procedures for dealing with hazards
the resources and competence of the staff to be employed
planned working times
temporary works to be employed
protection from highway, rail, waterway and other traffic
requirements for action by others
any co-ordination or notification required
any environmental impacts of the work and proposed mitigations
the health and safety assessment and measures to be taken and equipment to be
provided to protect all parties.
Health and safety and environmental considerations in carrying out work in tunnels are
discussed further in Section 3.6.
4.6.2
Access, programming and timing
Where programming of inspections or investigations is concerned, the first consideration
should be given to making advantageous use of existing access opportunities. Any
disruption to services and associated costs can be minimised by co-ordinating them with
other activities that might affect the normal use of the tunnel, for example, inspection and
investigation may be programmed to coincide with maintenance and repair works (which
might also have the benefit of providing access to normally hidden parts of the structure)
wherever such works are mutually compatible. However, where such tasks are
incompatible inspection and investigation works may be competing for access time with
other activities.
It should be a priority that the timing of inspections always satisfies regulatory requirements
and that any delay or deferral in inspection or investigation is justified by an adequate
assessment of possible increased risk to the safety of the structure and to the public.
The timing of inspections may influence the state of the structure and the nature and
quality of observations that can be made. The environmental conditions (temperature and
weather) should be recorded as a routine part of any inspection, and the current and
recently prevailing conditions may be important. For example, the moisture state of the
masonry and water ingress may be higher after rain, cracks may open more in cold
weather, and the adequacy and functioning of existing drainage provisions may be
apparent in wet periods.
Adequate access to the tunnel intrados, including any shafts, is necessary for detailed
visual observations. Access requirements are likely to be specific to the tunnel and to the
type of infrastructure, for example, the inspection of waterways tunnels brings more risks
associated with working over water and provision of a suitable boat-supported working
platform. It may be necessary to make arrangements to temporarily remove obstructions,
such as cabling equipment or protective sheeting, to allow clear access and vision. The
inspection of canal tunnels below the waterline will require their prior drainage (and can
be combined with maintenance works to clear accumulated material from the channel).
Note that the access point onto railway tracks may be some distance from the tunnel,
especially for mechanised plant and this travelling time should be built into the
programme for inspections.
CIRIA C671 • Tunnels 2009
113
4.7
LOCATION AND INSPECTION OF TUNNEL SHAFTS
4.7.1
Detection and location of unknown hidden shafts
It is important for asset owners to have a record of all shafts associated with their tunnels
so that they are safely managed. To ensure that this record is comprehensive, it is
necessary to identify any unknown and possibly hidden shafts that do not appear on
existing records and cannot be confidently detected from visual inspection of the tunnel
and the ground above it.
Network Rail have adopted a multi-phase approach to ensuring that the location of all hidden tunnel
shafts are identified (Network Rail, 2004b), as described below:
1 PHASE A: desk study that aims to find independent records of the existence of a tunnel shaft and
its location to a tolerance of ±10m. A reconnaissance walk-over survey could be included in this
phase, if deemed necessary, but otherwise is included in Phase B.
2 PHASE B: non-intrusive investigation, including walkover and walk-through surveys, and using nondestructive techniques from the ground above the tunnel (eg radar, magnetic and seismic survey
techniques) or from within it (eg radar and ground resistivity surveying), to identify features that
might be associated with a shaft.
3 PHASE C: intrusive investigation of areas where Phases A and B suggest a strong likelihood of the
presence of a shaft (using techniques such as boreholes and ground probing, trial pits, penetration
tests) to confirm its presence and location.
This process provides a method of efficiently investigating potential shaft locations with increasing
confidence until a stage is reached when confidence is sufficient to discount the existence of a shaft or
to confirm its presence and location.
More details of this methodology are given in Appendix A5.
4.7.2
Shaft inspection
Requirements for shaft inspections vary between infrastructure owners, but commonly a
tactile (touching distance) inspection is required at a specified maximum interval (see
Table 4.1). This may be different from the requirements for inspection of the rest of the
tunnel.
Shafts present problems with man-access for detailed tactile examination that may require
the provision of suitable and safe specialist access techniques (eg rope access or
steeplejacking with ladders), equipment (eg scaffolding, mobile access platforms, cranes
and inspection cages) and safety measures (eg training and precautions associated with
working in confined spaces). In situations where inspection staff are likely to be exposed
to such risks consideration should be given to the use of remote access methods to carry
out the necessary inspection tasks. The inspection of shafts and unsafe areas of tunnel
have previously been undertaken from a position of safety using CCTV surveying
equipment (for example, as described in Case study A1.8). Such techniques also have the
advantage that a permanent and objective visual record of the shaft condition is obtained,
which can be shared with others, viewed and reviewed in an office environment and
compared to previous records. Such techniques may not provide a suitable replacement
for manned-access in all situations, and are not permitted now by some asset owners. They
do have potential for further development and possible use in the future. For example,
shafts have been surveyed using high-resolution digital video cameras mounted on
telescopic masts up to 30 m in length, and LU has recently carried out trials in which a
video camera has been suspended below a helium-filled balloon that can be remotely
manoeuvred using small thrusters (Chew and Roberts, 2005).
Guidance on working at height is given in the Work at Height Regulations 2005.
114
The shaft should be examined and explored as far as is reasonably practical and safe to
obtain the data required for the design of the remedial works. This should include:
type and condition of fill with location of any voids, headings or culverts
type, condition and thickness of linings, if used, and identification of infill material
behind the linings
details of any special construction, plugs, or staging
groundwater levels, seepages or drainage measures, and the effects if these are
changed
details of all surrounding rock and superficial deposits.
Where regular detailed inspection as part of a normal maintenance regime is not practical
or feasible, such shafts present an unquantifiable risk that cannot effectively be managed
by normal means. It is recommended that an action plan be developed to either infill the
shaft with non flammable permanent fill or install safe access to enable examinations to be
undertaken.
Figure 4.1
Two views down a tunnel shaft. Water ingress and the presence of shaft furniture can
obstruct inspection and other work in shafts and should be taken into account when
planning access
CIRIA C671 • Tunnels 2009
115
Where the shaft is to be infilled, the fill material should provide adequate support to the
shaft such that there is no undue increased load on the existing tunnel lining (NR, 2004b).
4.8
INTERPRETATION OF INSPECTION AND INVESTIGATION
DATA
4.8.1
The importance of good interpretation
The safety and serviceability of individual tunnels and the tunnel stock as a whole relies on
the quality of the data obtained in the course of routine inspections and investigations,
and on the quality of the interpretation by which the condition of structural elements is
assessed and maintenance and repair needs identified.
The importance of good interpretation cannot be overstated, and there is no substitute for
a thorough understanding of tunnel structures, the factors that influence their
performance and behaviour, and the significance of observations and defects. In this way,
the knowledge and experience of inspectors, assessors and engineers has a direct
influence on the quality of tunnel management.
It is also important to remember that the interpretation can only be as good as the data
from the investigation. Great care should be taken when extrapolating data and making
judgement about parameters that were either not directly measured or where the
accuracy of the data is suspect.
4.8.2
Considerations for interpretation
It is important to appreciate that tunnels and their environments are subject to gradual
change and that information from a single inspection or site investigation represents only
the current condition. While this is adequate for some purposes, ie in verifying
construction features, used in isolation it cannot provide information on how parameters
have changed over time, which is frequently desirable. A single site investigation can, for
example, identify a crack or bulge in a lining, and possibly even allow its likely cause to be
discerned. However in isolation it is difficult to determine whether this is an inactive
defect that has been stable for a long time or whether it is recent and rapidly developing –
scenarios that might prompt very different reactions. Although there are sometimes clues
as to whether phenomena are recent or longstanding, such as fresh surfaces on spalled
brickwork or the presence of thick leachate deposits, such indicators cannot always be
confidently relied upon to provide adequate or accurate information.
Care should be exercised in the interpretation of test results from localised sampling and
testing. The fabric of the constituent elements of tunnels may be very variable so it is
important that undue weight should not be given to individual results. The data should be
seen in the context of the behaviour/performance of the structure, particularly where the
materials are inherently heterogeneous, such as with old masonry. Individual rogue
results should not be ignored as they may help in resolving the problem. Reliance on one
type of test to determine key parameters is discouraged in favour of a broader approach.
Where rates of change are important, comparison of the current state with a previous one
is necessary, and there is no option but to rely on whatever historical records may exist.
These are particularly useful where it is necessary to extrapolate observations into the
future and make predictions. Care should be exercised here because while a good
understanding of previous behaviour is extremely valuable, the past is not always the key
to the present and future. Many aspects of tunnel behaviour and performance are the
result of complex interactions between parameters that undergo changes over time, and
116
the rates of these changes can vary. It is often desirable to supplement historical
information with continuing assessments to monitor the current state and discern any
changes. Monitoring can be achieved by carrying out discrete repeat observations and
measurements of phenomena at suitable times, or gathering such data using a more
continuous automated approach, eg by installing suitable dedicated monitoring
instrumentation and logging devices (see Section 4.5).
Table 4.3 includes a variety of common observations relating to tunnel defects and
apparent condition for each of the main tunnel construction types considered here,
together with suggestions of possible causes and potential effects. It is intended to provide
assistance to staff who are experienced in the inspection and assessment of tunnel
structures, rather than providing a substitute for their experience. It is important that any
tunnel defect, particularly any evidence of change in condition or environment, is
properly evaluated by a competent person to discern its significance and assess the most
suitable course of action.
Guidance on other defects and deterioration mechanisms that may occur specifically in
cut-and-cover tunnels is included in other publication:
Table 4.3
CIRA C656 Masonry arch bridges – condition appraisal and remedial treatment (McKibbins et
al, 2006) for masonry arches
CIRIA C664 Iron and steel bridges: condition appraisal and remedial treatment (Tilly et al,
2007) for iron and steel structures.
Interpretation of common inspection and investigation observations
UNLINED TUNNEL DEFECTS
1 Loose surface material
Loose rock on tunnel roof and walls is typically the result of poor
blasting control during construction, weathering or washout of
supporting material. Falling material is normally limited to
relatively small debris but still pose a hazard to tunnel users and
equipment.
Loose material should be identified and made safe, normally by
removal or in some cases addition of a secondary lining.
2 Potentially unstable blocks/wedges
Rock blocks are bounded by discontinuities, the spacing and
orientation of which allow kinematically inadmissible blocks. There
is a risk of unstable rock falling onto or into the path of traffic
and/or damaging tunnel equipment.
Rock stability should be checked (eg by scanline survey of joint
sets) and the need for stabilising measures (eg rock bolting,
application of mesh, sprayed concrete) assessed.
CIRIA C671 • Tunnels 2009
117
Table 4.3
Interpretation of common inspection and investigation observations (contd)
MASONRY LINING DEFECTS
1 Loss of mortar from joints
Deterioration of mortar occurs through physical weathering processes
(typically in moist conditions from leaching, physical salt attack,
wetting/drying or freeze/thaw cycling) or chemical attack (eg by sulfates).
Can result in loose masonry (hazardous if overhead) and reduced area for
load transfer leading to stress concentrations.
Where mortar is extruded from joints or joints have opened up this can
indicate deformation caused by changes in the stress state of the lining and
should be investigated. (Note though that friable and extruding mortar can
also be a sign of sulfate attack, particularly if conditions are wet and a
whitish bloom of sulfate salts is visible).
2 Spalling (weathering)
Often caused by freeze/thaw cycling in areas that are wet and subject to
freezing conditions, but can also result from physical salt attack, use of
over-hard mortar with weak bricks or changes in the stress state of the
lining.
Reduces effective thickness of section, and presents a hazard from loose
material when it occurs in overhead areas. Can indicate structural distress
of lining, particularly if joints appear to have closed up. Consider possible
causes and if necessary investigate before remediation.
British Waterways use the term weathering to describe spalling with a nonstructural cause.
3 Construction joint
Not a defect, but may look similar to a vertical crack. Joints are
distinguishable by their regular toothed appearance and continuity and have
thickened/irregular mortar joints.
In wet tunnels joints are often preferential pathways for water ingress, and
as a result they may suffer from localised deterioration (mortar deterioration
and loss and spalling from freeze/thaw action).
See Section 2.3.5 for more information on construction joints.
4 Vertical (circumferential) crack
Cracks can follow the mortar joints and/or pass through masonry units
(where they are relatively weak).
This defect would not normally affect the structural capacity of the lining but
can allow water ingress and gradual deterioration. Where cracking is open,
progressive or there is an offset across it consider and, if necessary,
investigate possible structural causes.
See Section 2.6.1 for a more detailed discussion of cracking.
5 Horizontal (longitudinal) crack
Cracks can follow the mortar joints and/or pass through masonry units
(where they are relatively weak).
Horizontal cracking may have a structural cause that should be considered
and if necessary investigated, particularly if cracking is open and/or
progressive. Can allow water ingress and gradual deterioration.
See Section 2.6.1 for a more detailed discussion of cracking.
6 Delamination
Face-parallel cracking/debonding within the lining, also ring separation in
arches. Often not visible but can be detected by hammer-tapping,
investigatory drilling or some NDT techniques such as radar/ultrasonics.
Sometimes indicated by surface bulging or cracking.
Can reduce effective structural capacity and resistance of the lining to
deformation. Cause should be investigated.
118
Table 4.3
Interpretation of common inspection and investigation observations (contd)
MASONRY LINING DEFECTS
7 Bulging and distortion
Where parts of walls or arches are out of true, show bulges or other
irregularities it can be difficult to ascertain whether these are original
features or deformations in response to stress in the lining. Also
deformation can be longstanding or recent and possibly continuing.
Judgement is reliant on appearance (there may be associated deterioration)
and quality of past inspection and records. Distortions can result in local
reduction in lining capacity. Monitoring and/or investigation are advised
where structural causes are suspected.
See Figure 2.25 for further information.
8 Wet patches
Wetness affecting areas of masonry indicates water behind the lining and
general seepage through it. Water pathways are typically permeable mortar
joints or cracks between mortar and masonry units.
If previously dry areas become wet, inflow is severe, wet masonry is
deteriorating or if water is causing other problems investigation may be
necessary.
9 Localised water ingress
Ingress of water from a specific location, feature or defect. There may be a
rapid flow, in which case there is an open water pathway through the lining
(typically mortar loss from joints, cracking or sometimes tree-root
penetration).
If inflow is severe, wet masonry is deteriorating, or water is causing other
problems in the tunnel investigation may be necessary. If there is a build-up
of fine material that is being washed-out from behind the lining this may
indicate the gradual formation of voids, which can reduce lining stability.
SEGMENTAL CAST IRON LINING DEFECTS
1 Crack in radial flange at bolt hole location
Indicates radial flange overstress arising from ovalisation of lining.
2 Deformation of circle flange
Observed inward deformation of flange. Likely causes are either damage
from the time of construction, or a symptom of overstress. If the defect is
judged to be due to overstress, the problem could result in total loss of ring
capacity.
3 Corrosion
Potential reduction in capacity, depending on the depth of corrosion relative
to the thickness of the section, as well as the total area and location
affected. May be associated with leakage and rust staining.
CIRIA C671 • Tunnels 2009
119
Table 4.3
Interpretation of common inspection and investigation observations (contd)
4 Pin holes
Small pinhole leaks with rusting and staining. No significant effect on
capacity and serviceability of section but could develop into more
widespread general corrosion.
5 Missing flange (or missing part of)
May have resulted from intentional removal of material to improve tunnel
gauge. Will have the effect of reducing the section capacity depending on
the sectional area affected.
6 Corrosion at leaking joint
Typically visible as a build-up of rust and other deposits at the leaky joint. A
minor defect, but with the potential to cause general corrosion, and
progressive flaking and delamination causing gradual loss of section.
Indicates deterioration of jointing material and/or of any surrounding grout
or waterproofing systems.
7 Skin crack
In some cases the crack may cross the flanges and associated segment
displacement may be observed. Depending on the exact situation, the
segment has failed or is about to fail in shear.
SEGMENTAL CONCRETE LINING DEFECTS
1 Cracking parallel to cross-joints
Structurally this is not a defect. Cracks are likely to exist on the outer face of
the lining as well. Water will penetrate and depending on its chemical
characteristics may cause the concrete to deteriorate and lose strength.
Crack density relates to the amount of reinforcement provided.
Shallow spalling
2 Spalling
This is a local defect and can be shallow or deep (ie exposing the steel
reinforcement).
Spalling may have been originally present due to a casting defect, or may
have later occurred through impact damage, corrosion of steel or chemical
action. If spalling is due to compressive forces resulting from excessive
loading this should be investigated further.
120
Deep spalling
Table 4.3
Interpretation of common inspection and investigation observations (contd)
3 Corner spalling
This defect is due to out-of-plane construction. Under load this may result in
damage to the overstressed corner where there is contact between the
segments, which can reduce the lining strength.
4 Diagonal cracking
This is indicative of incipient compression failure. If it is localised it can be
due to a local weakness (for example, localised concrete degradation
resulting in strength loss) or a locally higher load whose nature should be
investigated.
5 Spalling at edge of cross-joint
This is probably due to original out of plane construction as in the case of
incipient corner detachment. If the spall is the only visible defect the
structure has found its equilibrium position.
Load transfer at joints may be not happening through the full depth of the
structural section and further damage is possible if the load is increased.
6 Lipping at joint
This is a construction defect that has not caused any damage. The reduced
contact area of the joint means that this is a weak spot that may begin to
exhibit damage if loading increases.
7 Circumferential crack
A circumferential crack may indicate a bursting failure due to overloading or
inadequate longitudinal reinforcement. Examples have been noted where
the segment is acting as a jamb, at cross passage or shaft openings. This
defect would not normally affect the structural capacity of the lining but can
allow water ingress and gradual deterioration.
8 Cracking parallel to cross-joints with displacement
Causes of cracking should be investigated. Concrete strength may be
reduced due to the crack formation and friction sliding at the interface.
CIRIA C671 • Tunnels 2009
121
Table 4.3
Interpretation of common inspection and investigation observations (contd)
9 Discontinuous cracking parallel to circle joints
Cause of the cracks should be investigated and this may require advice from
a materials engineer.
10 Water seepage
Can occur at joints where sealing material has failed or through cracks and
other discontinuities. In certain circumstances leaking fine cracks can self
heal (autogenous healing) and seal themselves.
If inflow is severe, concrete is deteriorating, or water is causing other problems
in the tunnel investigation may be necessary. If there is a build-up of fine
material which is being washed-out from behind the lining this may indicate the
gradual formation of voids, which can reduce lining stability. Thick build-ups of
mineral deposits can affect on tunnel clearances.
4.9
STRUCTURAL ASSESSMENT OF TUNNELS
Structural assessment is one of the activities of the asset appraisal process and is intended
to evaluate the structural capacity and performance of an asset. More specifically, the
assessment is a check that the structure meets the basic requirements stipulated by
national standards and is adequately safe and serviceable. This section offers guidance on
structural assessment for asset management but is not intended as a detailed design guide.
The assessment is a verification process, similar to that used in structural design. Asset
owners specify different requirements for the processes of design and assessment of
existing structures in their internal specifications.
Some asset owners stipulate the need for valid assessments of all structural assets,
including tunnels, and specify procedures for achieving this within their own internal
standards. Others have no such formal requirements or specifications and assessments
may be carried out on an ad hoc basis only when a special need is recognised. Before
carrying out an assessment, the assessor should ascertain whether there are any special
requirements, such as the asset owner’s internal engineering standards and procedures,
which will need to be adhered to in addition to satisfying national standards, for example,
those giving guidance on the use of structural materials.
In some cases a tunnel may fail an assessment while being free of any sign of distress. This
is a controversial situation because in the face of apparently contradictory information it
may be difficult to assess whether the tunnel is in fact structurally inadequate or whether
there is an excessive level of conservatism in the assessment parameters or the assessment
method, or in the pass/fail criteria. Until this uncertainty can be adequately resolved the
tunnel represents an unquantified risk and special management measures may be
appropriate, for example, restrictions on use, increased vigilance through inspection and
monitoring, investigation to refine the assessment parameters or changes in assessment
methodology.
Suitably experienced engineers may be able to exercise their judgement to help resolve
such issues by identifying tunnel-specific factors that might influence the assessment
results, considering the suitability of the pass/fail criteria adopted, or by modifying
122
assumptions made in the modelling, for example, the nature of the ground/structure
interaction. There is no general rule as to how to approach such a situation, so the
accuracy and reliability of the information used in the course of assessment and the
experience of the assessor are of paramount importance. Until such a time as the
uncertainty can be resolved with an adequate level of confidence, the tunnel should be
considered as structurally sensitive and managed accordingly to minimise risk.
Although infrastructure owners may have their own specific requirements, it is important
to understand that there is no codified or generally agreed methodology specifically
appropriate for the assessment of existing tunnels, and that given this situation
practitioners have developed a variety of approaches to carrying out assessments. The
approach set out here, although not the definitive method, provides a general outline of
the assessment process favoured by the authors, and includes a discussion of simple
assessment procedures for specific situations. This guide is not intended to be an
assessment manual and more details on the methods of analysis required for an
assessment can be found in Chapter 6 of the Tunnel lining design guide (BTS and ICE,
2004).
Finally in this section, recommendations are given on the consideration of structural
defects in structural models.
Except for cast iron linings, the assessment procedures described in this publication are specific
applications of the limit state design philosophy embraced by most current international design
standards, including all of those recognised in the UK.
This section includes a general discussion of the principal methods of, and approaches to, the structural
assessment of tunnels. More detailed information, intended for practicing assessors, is given in
Appendix A7. This provides a unified, rational methodology appropriate for a wide range of assessment
situations, and is compliant with existing codes where these are relevant. It also gives a more detailed
discussion of the application of limit state principles, including a method proposed as suitable for
carrying out the limit state assessment of cast iron linings, discusses the identification of the various
load combinations to be included in assessments, and offers guidance on selection of an appropriate
method of analysis and the definition of the structural resistances.
4.9.1
Assessment in principle
The assessment method depends on whether the lining is:
masonry
cast iron
cast steel
reinforced concrete.
The assessment of masonry and reinforced concrete linings should be carried out in
accordance with limit state principles. Cast iron linings should be assessed on a permissible
stress basis, due to the brittle nature of this material and the lack of appropriate partial
safety factors for limit state analysis.
From an operational point of view limit state analysis means that both loads and
resistances are factored to cover uncertainties and to provide a margin of safety (see
Appendix A7). Also, the resistances are the ultimate resistances of the structural
components.
In the permissible stress approach loads are unfactored. The check on structural
components is a simple check that the permissible stress is not exceeded at any point in
the component. The permissible stress is obtained by applying a reduction factor to the
CIRIA C671 • Tunnels 2009
123
strength of the material. Safe working stresses can be taken from BD21/01 (HA 2001). To
avoid confusion, note that formally in BD21/01 loads for cast iron permissible stress
assessment are presented as factored by 1.0 rather than unfactored.
An assessment can be qualitative, be based on the results of tests, or be analytical. It is not
practical to resort to an assessment based on tests of complete tunnel linings and so only
the qualitative and analytical assessments are used.
Both assessment methods involve the definition of a structure (in terms of geometry,
properties and conditions of the materials and relationship with the surrounding
environment), the identification of the foreseeable actions on it and the definition of the
mechanism by which these actions are resisted. All these aspects are covered, though by a
different strategy, in both the limit state and the permissible stress methods.
4.9.1.1
Qualitative assessment
An asset owner often has more than one tunnel, but the analysis of all tunnels is generally
not necessary because proving that some of them are safe equates to proving that other
similar ones are at least as safe, subject to certain qualifications (detailed below). This
approach is similar to checking only the most stressed beam in a steel structure where all
the beams are made out of the same structural section.
Qualitative assessment of a structure involves the identification of a similar structure
whose safety level is known, and inferring the performance of the tunnel under
assessment from the known tunnel. Such an approach relies on demonstrating the validity
of these assumptions.
There are two classes of qualitative assessment:
1
The two structures (as far as is significant from an engineering point of view) can be
deemed to possess the same geometry and materials and the one to be assessed will be
subject to the same or lower loads than the one used as a reference.
2
The two structures are to withstand the same loads but the one to be assessed is more
robust.
By this approach the assessment process is mainly a matter of classification of assets and
the information provided by previous assessments (if available) can be readily used.
Although qualitative assessment provides a straightforward and efficient approach to
assessing the structural safety of a large number of similar assets, it is reliant on adequate
knowledge of the individual structures and an informed understanding of the potential
significance of any differences between them. No two structures are identical and for this
approach to be satisfactory the judgement on their equivalence should be entrusted to an
engineer with the appropriate experience to consider all the parameters characterising
them.
Aspects to be considered in the assessment include:
124
tunnel shape
materials
joints
use (current and future)
ground conditions
water conditions
defects
aggressive environment
age and life
alterations
method of construction.
Comprehensive and reliable data are required for this approach, because it is necessary to
identify unexpected sources of difference that might otherwise not be apparent. A desk
study of existing information should be carried out to identify factors such as local
geological variations (for example, the presence of scour holes, faults and shear zones).
This should be supplemented by up-to-date observations from visual inspection to identify
factors such as unrecorded alterations and variations in construction, evidence of
movement or deterioration.
4.9.1.2
Analytical assessment
To set out the principles of the analytical assessment it is convenient to group the principal
different types of tunnels and then describe the analytical assessment suitable for each
group.
The first distinction to be made is between actual tunnels and structures that are serving
as tunnels but are in fact more similar to bridges. In some asset owners’ standards this
distinction is indicated by the use of terms such as deep and shallow tunnels, although this
terminology is not technically justified. It has been suggested that a more appropriate
distinction is between bored and cut-and-cover tunnels, but again this does not clearly
distinguish between those structures that behave as tunnels and those that behave as
bridges. The decision as to whether an asset belongs to one category or the other is made
by an appropriately experienced and competent assessor based on the specific
circumstances of the structure.
Tunnels that can be considered to act as bridges are shallow tunnels and in most cases cutand-cover construction. In such cases the side walls can be regarded as abutments to decks
supporting highways or even buildings. As these decks have been built over a long period
of time their construction form spans a wide range of possibilities, for example, (in
chronological order):
masonry vaults
cast or wrought iron girders supporting brick jack arches
reinforced concrete
steel (generally concrete encased for fire and corrosion protection).
When these decks support highways or railways they should be assessed as bridges in
compliance with the codes. In some instances they support structures/buildings. This will
almost certainly require a bespoke analysis. Often, for small buildings, the decks have been
used as a base for pad foundations and there is no relation between the structure of the
building and the structure of the deck. In such cases the loads at building foundation level
should be applied as point loads to the decks factored as appropriate for the specific deck
element being assessed.
CIRIA C671 • Tunnels 2009
125
In the remainder of this section the general principles established by the main asset
owners as described in their internal specifications, are given for tunnels that act as
tunnels. The main construction types addressed are: masonry linings, cast iron linings and
reinforced and plain concrete linings. Steel linings are unusual in UK and will be briefly
addressed in the section on cast iron linings.
Note that the analytical assessment may be very sensitive to parameters like ground
pressure coefficients (Ka, Kp, Ko) and stiffness of the lining. The assessor should give
careful consideration to the choice of the appropriate parameters. Checks on the
sensitivity of the model to a range of possible parameter values may be appropriate, and
will add confidence in the assessment results.
Masonry linings
Masonry is the oldest structural material. Elastic methods of analysis are not suitable for
such a complex material as masonry so limit methods of analysis are used by engineers.
The upper and lower bound theorems of limit analysis form the basis for the assessment.
In particular, according to the lower bound theorem, it is not important to find the actual
configuration of the internal forces in a structure under a given set of external actions. If
at least one distribution of internal forces compatible with the material strengths and
equilibrating the applied loads can be found then the structure can be considered as safe.
This approach was originally introduced by Jacques Heyman for masonry arch bridges
and masonry domes (Heyman, 1966). Heyman assumed that masonry was infinitely
strong in compression. Refinements to the original Heyman approach to allow for noninfinite resistance of masonry in compression have been introduced more recently
(Crisfield and Packham, 1987).
The method consists of finding a line of thrust equilibrating the applied loads that can be
actually developed within the masonry structure.
In the case of a tunnel lining, the applied loads are those from the ground and water on
the lining (actions from inside the tunnel can also be taken into account if necessary).
When the actions on the lining have been defined, a line of thrust balancing them can be
sought by graphical or analytical methods. The tunnel is considered safe if the line of
thrust found is compatible with resistance of the lining at all cross-sections. This amounts
to a check that at each cross-section the line of thrust is contained within the cross-section
by a sufficient amount to ensure that the axial force at the section can be balanced by
stresses lower than or equal to the strength of the masonry. This is shown in Figure 4.2 for
a circular tunnel.
126
Figure 4.2
Diagram illustrating the application of the limit analysis method to a masonry
tunnel lining in principle
The actions of ground and water on the lining can be conservatively assumed as the action
on a rigid wished in place impermeable tunnel (often referred to as full overburden).
Alternatively, the actions can be modified to allow for relaxation of the soil due to the
inelastic deformation of the tunnel at collapse. Guidance on the effects of relaxation and
the geotechnical properties of the ground should be obtained from a suitably competent
geotechnical engineer. The compressive strength of in situ masonry should ideally be
determined by appropriate testing or in the absence of test results by adopting suitably
conservative values based on guidance given in available standards, for example, BS 56281 (BSI, 2005b) and EN 1996-1-2 (BSI, 2005c).
Some masonry linings do not have an invert. In such cases it should be checked that soil
and water will not penetrate into the tunnel by using classical geotechnical methods with a
sufficient degree of safety. Foundations to the lining should also be checked in this case.
4.9.1.3
Cast iron and steel linings
Cast iron linings can be assessed using elastic methods of analysis that are vital when
dealing with grey cast iron. This is the only form of cast iron used in tunnel linings until
recent decades. More modern ductile (spheroidal graphite) cast iron may be treated as for
steel. Due to the brittle nature of grey cast iron (especially in tension) and lack of
standards for limit state analysis, a permissible stress approach is preferred. Guidance on
this can be found in BD21/01 (HA, 2001) and in LU Engineering Standard E3322. The
assessment involves the determination of the axial force and the bending moment at any
cross-section of a lining ring (see Chapter 2 for construction details of this type of lining).
Closed form solutions have been given for tunnels in soils that can be assumed as linearelastic (Curtis, 1974 and Einstein and Schwartz, 1979). These solutions are valid for an
isolated tunnel in a homogeneous medium with a uniform surcharge at surface level.
Empirical rules to make allowance for joints are also available. For other cases an
estimation to the elastic solution can be found by numerical methods (finite elements,
finite differences or boundary elements).
CIRIA C671 • Tunnels 2009
127
It is usually necessary to calculate the axial force and the bending moment at the crown
and the horizontal axis of the tunnel lining only. Once these have been determined the
corresponding point in the axial-bending plane can be plotted on a chart together with
the interaction diagram of the cross-section of the cast iron elements forming a lining ring.
An interaction diagram is a curve (a sequence of straight lines) enclosing all points
representing safe stress conditions for the lining. The interaction diagrams can be easily
constructed based on the cross-sectional properties of the lining and the permissible
stresses for the material. Failure of bolts can also be superimposed on the same diagrams.
The procedure for establishing whether the lining passes the assessment once the axial
force and the bending moment have been determined is shown in Figure 4.3. In the
figure the points below the dotted lines are representative of bolt failure. The part of the
elastic interaction diagram corresponding to tensile axial force is unrealistic for tunnel
linings. From the diagram it is evident that bolt failure at joints is unlikely to be a principal
failure condition. Bolts are in fact used mainly for construction reasons. The capacity of
the ribs to transfer the actions of the bolts to the skin should also be checked.
Figure 4.3
Assessment of cast iron linings
Material properties of the cast iron should be determined by reference to BD21 (HA,
2001) for older grey cast iron, by testing, or by referring to the relevant literature (taking
into account the information on the manufacturer if marked on the components).
Steel linings can be assessed by using the same methodology but replacing the elastic
interaction diagrams with plastic ones and using axial forces and bending moments
derived applying loads factored by the appropriate partial safety factors. The partial safety
factors for loads can be obtained from BS 5950-1 (BSI, 2000). Information on the analysis
of steel castings can be found in SCI (1996).
Simplified rules for carrying out a rapid preliminary estimation of axial forces and
bending moments acting on a metal lining are given in Section 4.10.2 on multilevel
assessment procedures.
Concrete linings
Concrete linings can be pre-cast or in situ, reinforced or plain concrete. Even linings that
are considered to be plain concrete will normally contain a nominal quantity of
reinforcement. In all cases the assessment procedure is not dissimilar from that of cast
iron linings discussed previously, except that in this case limit state analysis is carried out
in accordance with BS 8110 (BSI, 1985), which gives guidance on the use of structural
128
concrete in construction. The main difference is that factored loads are used for the
determination of bending moments and axial forces, which are compared with the
ultimate resistances of the cross-sections determined by elasto-plastic analysis rather than
by a permissible stress approach.
The interaction diagrams for reinforced concrete columns found in BS 8110-3 (BSI,
1985), can be used for the assessment of tunnel linings. Alternatively the engineer can
derive these using well-established techniques (De Vivo, 1998). There is a wealth of
commercially available proprietary software for their determination.
The closed form solutions and the numerical techniques as suggested for the
determination of axial forces and bending moments in cast iron linings are applicable to
concrete linings as well.
When assessing concrete linings, particular care should be taken in the assessment of
joints, although joint details in concrete linings are not adequately covered by the available
standards. Joints are often designed to limit the contact area between consecutive
elements to make them act like pins. In this way the bending moments attracted by the
lining are kept to a minimum and the soil is supported mainly by hoop action. The
adverse effect is that large bursting forces may arise at the contact locations.
These joints are often designed based on test results and it is difficult to define a reliable
analytical technique for their assessment. In this situation some assessors use the general
rules for the design of pre-stressed reinforced concrete end blocks.
In general, if the loading regime on the lining has not been altered from the original
design and no sign of degradation is visible, joints should be safe. There is no consensus
on how to approach an assessment of joints under new conditions and there is no general
guidance available now.
4.9.2
Multi-level assessment procedure
It is common practice, and appropriate, to minimise the analytical effort required if it can
be demonstrated that relatively complex and refined methods of analysis are not necessary
in particular circumstances. This is especially convenient when a large number of assets
are to be assessed. Note that in contrast to design, in an assessment it is not an issue if the
structure is more robust than required. Guidance is given on simple and robust
approaches to analysis that will provide conservative and adequately reliable results.
In the simplest case a qualitative assessment, based on comparison with similar structures
already assessed, may be adequate.
Assessments are intended to provide a pass or fail result and, beyond this, how close the
assessment effects are to the assessment resistances is not of great concern. So if it can be
proven that a simplified model is more conservative than a more rigorous one, this simple
model can be used for the assessment. More refined methods of analysis may only become
necessary when failure is predicted by the results of the simple model. This presents an
efficient approach to the assessment of tunnels with a high degree of safety.
This approach leads to a multi-level assessment procedure starting with a very simple and
conservative analysis and is refined only if a more rigorous analysis offers the likelihood of
an assessment pass. A multi-level approach is also convenient for reducing the risk of
errors as the simple models assist the interpretation of results from complex models,
which can help to identify errors or modelling inadequacies.
CIRIA C671 • Tunnels 2009
129
Because simple models are based on very pessimistic assumptions, they are useful where
the parameters required for a more sophisticated assessment cannot be derived with an
adequate degree of certainty.
Simplifications can be introduced into an analysis either by simplifying the determination
of the assessment effects or by simplifying the determination of the assessment resistances,
although the latter does not generally result in a great simplification.
The first simplification that can be introduced is the reduction of the problem to 2D by
enforcing symmetry. This excludes variability of loads along the longitudinal axis of the
tunnel and precludes explicit representation of the construction process, making the
wished in place assumption preferable. In this case, the stresses in the ground during
construction can still be simulated by resorting to an axi-symmetric model, but this
involves some degree of expertise and is not recommended except where validating the
results of 3D models. Details like openings and junctions cannot be directly assessed in 2D
because they lack the necessary symmetry, but in most cases these elements can be assessed
by combining the results from several 2D models, as is discussed further in this section.
In 2D the simplest possible analysis that can be carried out on a tunnel involves the
application of the stresses in the ground to the lining extrados, calculated as if the tunnel
were not there. This is a conservative method because the beneficial effects of ground
structure interaction are neglected. Note that loads internal to the lining (for example,
from traffic) cannot be applied in this way. This approach can be used to rapidly assess the
order of magnitude of hoop stresses. Note also that there are some practical difficulties
connected with the boundary conditions when applying this method, in particular that
most commercial software cannot find a solution for models that have no restraints but are
under loads that are in balance. If the geometry of the structure and loads are
symmetrical this can be circumvented by using a reduced model in which the symmetry
conditions are exploited by introducing appropriate fictitious restraints (see Figure 4.4).
Figure 4.4
Exploiting the symmetry conditions to avoid boundary condition problems
If such symmetry does not exist, fictitious restraints should be introduced. As these should
not induce any forces in the structure under the given load conditions (the associated
reactions should be nil) they should be carefully chosen to avoid any redundancy while
still yielding an inherently stable structure. The analyst should be fully aware of the
limitations of the software being used when dealing with these issues.
A more refined level of analysis involves considering the ground as a Winkler’s springs
bed (a boundary condition imposing, at any point of the lining, a reaction proportional to
the displacement) and apply to the lining the ground loads acting on its extrados surface
130
(calculated again as if the tunnel was not there). Guidance on selection of an appropriate
stiffness for the Winkler’s springs can be found in the literature and in this approach all
loads can be applied to the tunnel (O’Rourke, 1984 and Duddeck and Erdman, 1985). It
is important to note that this method is not necessarily conservative, as overestimation of
the Winkler’s springs bed would result in underestimation of the assessment effect on the
lining. This applies to all models allowing for ground/structure interaction.
In the cases above preliminary analysis of the stresses in the ground without the tunnel is
required. Any surcharge is taken into account in this preliminary phase.
The next step is to model the ground as a continuum (this can be linear or, with additional
complexity, nonlinear) in which case ground and lining are modelled together and the
surcharge is applied directly to the ground. When the lining is modelled as a continuum,
rather than by using a beam type idealisation, concentrations of stresses at sharp corners –
typically the side wall and invert junctions – can suggest that the lining is locally
overstressed. These local stress concentrations are usually accommodated in the real
structure by plasticity or local damage that allows scope for stress redistribution. Plasticity
and damage can be introduced into the material model to find a more realistic stress
distribution automatically. Alternatively the analytical picture can be clarified by
examining the stress distribution in the lining cross-section just away from, and either side
of the junction across the thickness of the lining. This should give a more credible stress
distribution, which can then be faired around the corner, smoothing notional local peak
stresses down by using an equivalent rectangular stress block as given in BS 5628-1 (BSI,
2005b) and BS 8110-3 (BSI, 1985).
For shallow tunnels with masonry lining (where they act more like bridges than tunnels),
limit analysis methods in the Heyman fashion can be used (Page, 1994) or the semi
empirical MEXE method can be used if the shape of the masonry vault is within its
applicability limits (Hughes and Blackler, 1997).
The different levels of analysis defined can be done by using closed form solutions or by
numerical analysis.
As previously noted, closed form solutions exist for:
circular linings for application of full load pre-existing in the ground before tunnel
excavation
models using Winkler’s springs
modelling the ground as a continuum provided this is homogeneous, isotropic and
linear elastic.
Some available closed form solutions are listed in Table 4.4.
Table 4.4
Closed form solution for analysis of tunnel lining
Method
Tunnel shape
Type
Muir Wood, 1975
Circular
Elastic continuum
Curtis, 1976
Circular
Elastic continuum
Einstein and Schwartz, 1979
Circular
Visco-elastic continuum
ITA, 1998
Any
Ground modelled as Winkler’s springs bed
CIRIA C671 • Tunnels 2009
131
Alternatively analytical assessment can be approached numerically by the use of any of the
finite elements, boundary elements, finite differences and discrete element analysis
software packages available commercially. Although such methods require a more
sophisticated approach they may be justified in the following situations to:
not impose constraints on the tunnel geometry
permit the use of complex constitutive models for the materials
allow 3D models that can be solved only numerically.
Some features, such as junctions, do not possess the necessary symmetry to be analysed in
2D, so either numerical analysis should be used or simplified methods may be
appropriate. At the junction, the hoop force mechanism of resistance is not possible in the
lining because of the opening. The hoop forces that would exist if the opening was not
there should then be redistributed either into the portions of lining adjacent to the
junction/opening or in a framing structure if present. In the first case the width of lining
to be considered in the redistribution on each side of the opening can be estimated by
suitable 2D analyses. Bending of the lining spanning between the complete rings on the
sides of the opening should also be checked. This procedure is not very rigorous and
involves some engineering judgement but the results should be relatively safe if ductile
materials are involved. Special care should be taken if brittle materials are involved and in
such circumstances more refined analysis or the application of a factor to allow for the
inaccuracy of the analysis is recommended.
4.9.3
Structural defects
Defects in tunnel linings can be due to deterioration of materials, imperfect construction
or manufacture, misuse of the structure, impact and fire. Defects can affect both the effects
and the resistances. When assessing the influence of defects in the course of a structural
analysis the following points should be considered:
132
deterioration can be taken into account by using reduced properties for the
deteriorated materials. The properties should ideally be determined directly by
testing either in situ or on samples of the material taken from the structure.
corrosion can be allowed for by using reduced structural sections taking into account
the loss of material due to corrosion
imperfect construction can result in deviation from the design shape or misaligned
joints. This defect is difficult to deal with because surveying the tunnel would give the
deformed shape under the actual load. The derivation of the undeformed shape, for
comparison with the design shape of the tunnel, is not easy. Sometimes the size of the
strains in the lining is such that the current shape is mainly the result of defective
construction. If this can be proved then the current shape obtained from a survey can
be used in the analysis as the undeformed initial shape
misalignment of joints results in localised contact stresses or overstressing of bolts.
This generally causes only local damage, for example, spalling of concrete at joint
edges. Specific checks (if required) include finite element modelling of the contact at
the joint or use of simplified methods based on the lower and upper bound theorems
for ductile materials
misuse of the structure and impact should be dealt with by methods selected on a
specific basis
structures damaged by fires should be assessed using the material properties of the in
situ material after the fire event (see Section 2.6.3). Loss of section due to removal of
damaged materials should be noted.
Further information on defects and their potential structural influence can be found in
Section 4.8 of this guide. It should be noted that some defects are of no structural
significance and do not require any attention. The notes in Table 4.3 give further
guidance on this.
4.10
REPORTING ON AND INTERPRETING ASSET CONDITION
Following completion of site inspections, and other investigations or assessments that may
have been carried out, it is necessary to evaluate the results and determine the need for
any further investigation, monitoring or repair work (other than planned maintenance).
Further investigation and/or analysis of deteriorated or damaged elements may be
required to assess current reserves of strength and factors of safety, and also to estimate
the time that the repairs should take to be completed. In some cases it may be sufficient to
initiate a monitoring programme, or it may be necessary to monitor for a further period
of time for a reliable diagnosis (Swannell, 2003).
4.10.1
Reporting inspection and investigation results
Section 4.3.5 discusses some methods of improving the objectivity, reliability and efficiency
of making and recording inspection results.
The records from a visual inspection should normally include, at a minimum:
General information:
general details of the tunnel (its name, asset reference, location, construction type and
any changes, length etc)
the aim of the inspection and its scope, including identification of any parts of the
structure that were not inspected and why
the date and details of the previous inspection, including reference to the documents
being used and acknowledgement of any special requirements identified for this
inspection
the methods used in the inspection (whether touching-distance or not, description of
access methods and lighting provisions, details of any in situ testing and assessment
techniques)
details of environmental conditions (including weather on the day of the inspection
and preferably a general comment for the weather prevailing over the preceding few
days, in particular temperature and rainfall)
details of any problems encountered, particularly where these might affect the
inspection results (eg with access, obstructions, lack of time, other activities taking
place in the tunnel, difficult working conditions).
Objective inspection results:
concisely annotated tunnel plan with the results of the inspection, including locations,
and notes on features of interest such as defects
supplementary information, typically comprising clear sketches, and annotated with
detailed observations, measurements and photographs as appropriate
the results of any in situ testing carried out (eg hammer-tapping, hardness testing)
comments on any issues highlighted by the previous inspection that needs future
attention
comments on any apparent changes since the previous inspection.
CIRIA C671 • Tunnels 2009
133
Subjective/interpretative elements:
a statement on the overall condition of the tunnel
identification of those areas considered to be worst with explanatory comments
a statement of any issues that require immediate or urgent attention
identification of elements or circumstances that require special attention in future
inspections
a statement of the need for follow-up inspection, investigation or monitoring and its
urgency (ie should be carried out urgently, or before the next planned inspection, or
during the next planned inspection)
identifying any other issues relating to the asset and its performance in general,
particularly, under the inspector’s duty of care as a responsible person, where they
might affect the safety of tunnel users or staff (eg evidence of unauthorised entry,
presence of hazardous objects or materials, defective or damaged elements of tunnel
systems).
Ideally reporting should be carried out to a predefined format, preferably using proforma
sheets, which might include a simple base-plan of the tunnel for inspectors to identify the
location of features of interest and make concise notes. These locations should be crossreferenced to proforma, which includes further and more detailed information (sketches,
descriptions, photographs etc).
Asset owners will have their own requirements for the procedures of inspection, condition
assessment and reporting, and of the roles, responsibilities and required competencies of
the individuals involved.
For inspection results, the suitability of including items that require some interpretation,
such as recommendations for future inspections and their urgency, or identification of the
worst areas of the tunnel, will largely depend on the nature of the problems themselves
and the competence of the inspection staff. It is necessary for all information, and
particularly subjective elements such as recommendations, to be reviewed by an
appropriately knowledgeable and experienced engineer capable of interpreting the data,
identifying any issues that require further action and deciding what action is most
appropriate. However, it is important that inspectors are given an opportunity to present
their own interpretation of their observations based on their first-hand experience, and
experience of similar situations. Inspectors may be in a better position than an engineer
back in the office to see the bigger picture and to make deductions based on many
disparate observations. This might not be readily apparent to a reviewing engineer
presented only with a few objective descriptions. It is useful for both parties to contribute
their own strengths and experiences to the matter of interpretation.
Inspection results are likely to be more useful and reliable if inspectors are able to take an
active part in the interpretation of their own observations, so they should be suitably
competent. Quality of results is also aided by consistency of staff, methodology and
records between inspections. Also, it may be improved and assured through suitable
checking and auditing procedures.
4.10.2
Initial evaluation and identification of sensitive structures
From the results of inspections, clear structural failures (eg deformed linings, fractured
beams) and many forms of structural deterioration can be readily identified and evaluated
accordingly. Traditionally this has proved to be a good basis for carrying out reactive
repairs, but the ideal is to move from a reactive to a proactive, preventative regime of
maintenance. This is a greater challenge, requiring a more detailed evaluation of changes
134
in tunnel condition and a good understanding of its performance, the factors that can
potentially affect it, and how these changes might manifest themselves. The difficulty
comes in identifying those tunnels or parts of tunnels that are undergoing slow stress
changes (eg through consolidation of ground or gradual weakening of structural fabric).
Visible evidence of increasing structural stress may not be manifest at an early stage, but
could potentially result in failure at a later date.
Knowing where to look and what to look for is very important, and this relies on carrying
out an initial engineering evaluation of each tunnel based on its current condition and
other available information. This is one area in which analytical assessment can be a useful
tool, because it has the potential to highlight those tunnels, areas and elements that are
the most structurally sensitive. The results can be used to focus and improve the
effectiveness of surveillance of condition, allowing problems to be identified at an early
stage, potentially before they would be picked up by a routine inspection. The aim is to
identify and monitor the critical indicators of condition.
Based on engineering assessment results, tunnels which are identified as not specially
sensitive, operationally or structurally, may be subject to a continuing regime of routine
procedures for condition assessment, for example, periodic visual inspections at
prescribed minimum intervals (see Section 4.3.2).
Tunnels or parts of tunnels that are identified as being particularly structurally sensitive or
operationally critical should be further assessed to determine the optimum strategy for
evaluating their condition and identifying and responding proactively to any significant
changes. This may involve carrying out additional surveillance over-and-above the
routine, general inspection, which is adequate for less sensitive tunnels, for example:
detailed engineering evaluations, which may require further investigation of the
tunnel’s structure or the condition of specific sensitive elements and may involve
structural assessment and analysis
special inspections that might be at a greater frequency than routine inspections and
focus on specific indicators of condition and performance
periodic or continuous measurement of specific parameters, for example, deflection
and distortion, using suitable techniques and instrumentation.
Whether or not a tunnel is classified as being especially critical, improvements in the
collection and evaluation of data on condition and performance are likely to result in
more effective asset stewardship, underpinning attempts to move toward a more proactive
and efficient system of tunnel management.
4.10.3
Interpretation of results
The safety and serviceability of tunnels depends on the quality of the data obtained in the
course of routine inspections and investigations, and on the quality of the interpretation
by which tunnel condition is assessed and maintenance and repair needs identified.
Interpretation can only be as good as the data from the investigation.
The importance of good interpretation cannot be overstated, and there is no substitute for
a thorough understanding of tunnel construction, the factors that influence tunnel
performance and behaviour, and the significance of observations and defects. In this way,
the knowledge and experience of tunnel inspectors, assessors and engineers has a direct
influence on the quality of tunnel management.
Great care should be taken when extrapolating data and making judgement about
CIRIA C671 • Tunnels 2009
135
parameters that were either not directly measured or where the accuracy of the data is
suspect. For example, older tunnels will have been constructed using imperial units, so in
this context metric dimensions and whole rounded numbers in metric measurements
should be checked.
Tunnels and their environments are subject to gradual change but information from a
single inspection or investigation represents only the condition at that time. This is
adequate for some purposes, but used in isolation it provides no indication of how
parameters may have changed over time. Where an inspection identifies a potentially
significant change in condition, interpretation relies on the results of previous inspections,
and the importance of a repeatable and consistently high standard of observation and
recording (along with the consistency of inspection methodology and the inspection staff
themselves) becomes clear.
Care should be exercised in the interpretation of test results from localised sampling and
testing, because the tunnel’s structural fabric may be variable and so it is important that
undue weight should not be given to individual results. Individual rogue results should
not be ignored, because they may provide important insights.
Further discussion is included in Section 4.4.2 on investigation strategies.
4.10.4
Condition ratings
Each of the major UK infrastructure tunnel owners has its own systems and procedures
for condition assessment and reporting, so that requirements for data collection are
dictated by the needs of the owner. As discussed in Section 4.3.5, such systems are
designed to make objective and standardise inspection information to allow it to be more
easily interpreted and compared. Typically, the tunnel condition is assessed by considering
the extent and severity of any defects and an overall condition rating is awarded to the
structure. Recommended actions and priorities can be indicated against each identified
defect. A typical condition classification system would have simple condition grades, for
example, from A to E, where A represented an asset in an ideal condition and E
represented a serious safety concern. This type of classification system provides both an
absolute and a relative measure of condition that can be used in several ways:
to identify tunnels where current condition is unsatisfactory or even unsafe
to rank tunnels and defects within tunnels in terms of their priority for further
assessment and/or remedial actions
to provide a benchmark against which asset condition (of individual tunnels and the
tunnel stock as a whole) can be monitored over time, and the success of management
and maintenance policy can be evaluated.
Other information on the tunnel, its past performance and maintenance history is also
considered. Using this information, current tunnel performance and condition is assessed
against serviceability criteria that are assigned by the asset owner. These criteria will
include standards for safety as well as structural and operational performance and will
vary according to the infrastructure type and owner policies and objectives.
The condition classification indicates the relative level of concern with the asset. There are
two generic types of asset concerns:
136
1
Safety.
2
Performance.
Safety concerns are associated with defects or hazards that could potentially cause failure
of the asset or may endanger life or cause significant disruption to the service.
Performance concerns relate to defects or hazards that, although not causing a risk of
structural failure, may be detrimental to the normal operation of the tunnel and related
infrastructure. However, if performance concerns are not addressed they could eventually
deteriorate and become safety concerns in the future. The main difference is that action
on safety concerns should be undertaken urgently or immediately, whereas serviceability
concerns could potentially be addressed as part of routine maintenance or left to a later
date (although this may not be the most efficient way of dealing with them). Where safety
concerns are identified in the course of condition assessment, there should be suitable
procedures for dealing with these urgently, by assessing the risk and taking appropriate
mitigation measures such as an increased frequency of inspections or some other method
of monitoring, undertaking temporary or permanent remedial works, or restricting
tunnel use.
CIRIA C671 • Tunnels 2009
137
5
Selecting and carrying out works on
tunnels and shafts
5.1
SELECTION, PLANNING AND PREPARATION FOR WORKS
5.1.1
Planning and programming
The decision to carry out works within a tunnel may be based on a routine maintenance
regime (see Section 3.4), in response to a deterioration of condition, or to address a need
to improve or alter its structure. Continuing deterioration should ideally be picked up
during routine inspections and dealt with as part of routine maintenance wherever
possible, so that the tunnel’s condition and performance is not allowed to get worse.
Apart from routine maintenance works, which should be detailed by the tunnel owner or
asset steward in written procedures and maintenance documents, it will be necessary to
prepare structural designs and specifications for more extensive repair works or for
replacement of tunnel linings.
Generally, where works are of a structural nature the tunnel owner or asset steward
should take advice from a specialist professional engineer or suitable contractor who is
deemed competent and has a proven track record in carrying out similar works, unless
the required design capabilities are available in-house.
Structural designs should consider the stability of the tunnel and the safety of operatives at
all stages of any proposed demolition and reconstruction, and for proceeding
incrementally in multiple stages where access for work is restricted, as is normally the case
for transport tunnels. This may require careful planning and co-ordination if acceptable
rates of working are to be achieved. The design should also consider the effect the works
may have on adjacent tunnels or underground excavations.
If the available access periods are very short compared to the total time required for the
works, so that the works need to be carried out over such periods, a relatively high
proportion of available time may be taken up in activities such as:
safety procedures
gaining access to and from the work site
transporting plant and materials
setting up and dismantling temporary access equipment
making safe any incomplete works at the end of each session.
In such situations productivity is likely to suffer and the time required to complete the
work, and its cost, may be significantly increased. Such piecemeal working is likely to be
uneconomic compared with closing the tunnel for a single period and accepting the
resultant disruption to its normal service. So there may be other factors that influence the
decision such as technical and health and safety.
Another aspect to be considered at an early stage is how the works will be carried out and
what procurement measures need to be put in place.
138
Working in tunnels typically requires specialist skills and techniques, and a good
understanding of safety procedures and safe methods of working. Contractors who are
unfamiliar with working in a tunnel environment may lack the experience necessary to
understand the particular constraints they are likely to be working within and may fail to
foresee potential problems at the planning stage. Potential contractors should be carefully
assessed before being allowed to tender for this type of work. Ideally they should be able
to demonstrate a good track record in similar work, with a professional, flexible and cooperative approach, remaining focused on achieving project aims even when unforeseen
circumstances require changes to the work scope (Swannell, 2003).
The tunnel owner or asset steward may have a maintenance team available to carry out
routine maintenance works, or provisions in place such as call-off or term contracts with a
suitable contractor. An important consideration when carrying routine maintenance works
using external contractors is that the tunnel owner or asset steward should have direct
control or management during the works. This may include direct supervision or
direction of the works under a suitable contractual arrangement that allows the works to
be carried out unhindered by cost or quality constraints, such as target cost or total
reimbursement contracts. Where more extensive remedial or strengthening works are
required, project specific contracts may need to be procured under a conventional
engineer design and contractor build or contractor design-and-build route. Again direct
supervision of such projects should ideally be carried out by the tunnel owner or asset
steward, or appointed engineer.
In any case, contracting arrangements that clearly assign responsibilities to each party and
encourage co-operative working are beneficial.
5.1.2
Managing risk
Unforeseen circumstances, such as encountering unanticipated conditions or restrictions,
or the discovery of hidden structural details, are not uncommon when carrying out works
on tunnels. This can cause serious disruption to programme and increase costs, resulting
in significant variation to contracts. Although thorough research and a carefully designed
and executed site investigation can help to minimise such risks, the full extent of repairs
cannot always be determined in advance, and unforeseen circumstances can still arise to
challenge planned methods of working. Where uncertainty remains it is important to
reduce the potential consequences of unforeseen problems by ensuring that all parties
involved adopt a flexible approach and that good channels of communication are
established at an early stage in any project. It is worthwhile giving consideration to possible
alternative construction details, work scopes or methods based on the most likely scenarios
and ensuring that contracts accommodate such variations wherever it is practical.
Generic risks that should be assessed as a part of contracts for tunnel works include:
design
construction
health and safety and environmental
programme
economic
incidental and indirect.
These risks apply to most if not all construction and repair works, so they require careful
consideration when carrying out tunnel works because their likelihood and their potential
consequences may be greater than with other types of structure.
CIRIA C671 • Tunnels 2009
139
To optimise the safety and efficiency of tunnel works, and to ensure that they satisfy
objectives, it is vital that all the appropriate preparatory work is undertaken:
5.1.3
identify and establish causes of deterioration and defects (through inspections or
examinations on-site)
undertake a desktop study to collect information on the history of the structure
including previous inspection reports, works completed etc and to identify
parameters important to design and for further remedial works
identify the need for further site or structural investigations or examinations to
confirm the structural dimensions and the properties of the fabric of the tunnel and
the geological and geotechnical setting
hazards and risks should be identified and assessed, and measures used to mitigate
risks and ensure that any residual risk is acceptable and ALARP
consideration of the requirement for and effects of temporary works and working
conditions
ensure that material specifications are compatible with existing fabric of the tunnel
assess the adequacy of the existing fabric of the tunnel at each stage of the works
consider the immediate structural and engineering consequences of the works and
their effects on the long-term performance of the tunnel
contractual requirements to complete the works should be considered if not already
in place, and enforced, eg use of term or call-off contracts, design and build contract,
or use of owners own maintenance staff
site access should be planned and secured and all necessary permits and
authorisations obtained including approval of conceptual and final designs with
consideration of shared access between other users or asset owners
temporary works should be approved and clearances checked
staff should be appropriately trained and have the necessary skills and experience
(competence) to undertake the work.
Selection of techniques
Selection of the most appropriate remedial works should include consideration of several
factors (Broomhead and Clark, 1995):
type of fault to be repaired
ease of access
environmental, health and safety, heritage considerations and constraints
available clearances
length of possession times/lane closure requirements
cost of repair options
expertise required to execute repairs and contractor availability
performance, long-term durability and maintenance requirements of repairs
purpose of repair and ability to satisfy requirements
obstruction of future inspections.
Also the following should be taken into account when considering the scope and method
of the works adopted:
140
existing condition of area or areas to be repaired and areas adjacent to the part to be
repaired
history of the tunnel, including type and method of construction and previous
remedial works carried out
construction type and details
location, condition and status of any shafts or adits
the proximity of nearby tunnels or underground excavations whose behaviour could
be influenced by the works
foundations type and extent, including invert if present
geology and cover of the surrounding material, including mineral workings
groundwater regime, including effects of water, drainage and known watercourses
compatibility of materials
other works being carried out simultaneously
aerodynamic effects, including cross-section requirements (kinematic envelope)
climatic effects
presence of services within the tunnel or buried utilities
the condition of tunnel equipment, track and any other infrastructure.
When designing works reference should be made to the following documents and
standards applicable to tunnel repair measures, materials to be used and methods
adopted:
5.1.4
Specification for tunnelling (BTS and ICE, 2000)
Tunnel lining design guide (BTS and ICE, 2004)
appropriate British and European Standards.
Method statements and risk assessments
Tunnel repair often has to be carried out under strictly controlled access arrangements,
with restricted or confined space working areas, working at heights and in difficult
environmental conditions. Often it involves using potentially hazardous equipment and/or
materials. So it is extremely important that the preparations are made and the works
planned for properly in compliance with current health and safety requirements and, if
applicable, the tunnel owner’s or asset steward’s specific procedures.
Fundamental to this is the requirement for detailed method statements and risk
assessments to be completed covering all stages of the work, with appropriate contingency
and emergency measures included no matter how minor the works are. This should also
consider the tunnel environment during the works and the requirement for any special
precautions such as temporary ventilation to remove potentially noxious gases and
maintain respirable dust to harmless levels, and workplace noise level assessments.
More extensive repair and refurbishment works may often require partial demolition of
existing tunnel linings, so operatives may have to work below unsupported roofs. Where
possible this should be avoided by planning the work using remotely operated equipment,
such as hydraulic scaling and water jetting, and remotely controlled sprayed concrete and
bolting machines. If this is not possible temporary protective canopies may be appropriate
in some cases. Safe working conditions should be ensured at all stages by careful and
detailed planning and rigorous control on site (Swannell, 2003).
CIRIA C671 • Tunnels 2009
141
Where risk assessment has identified the potential for ground or water pollution, suitable
mitigation measures should be incorporated into the working methods. This is especially
important where potential pollutants and hazardous materials are being used, and where
groundwater pollution may affect aquifers and sensitive ecological sites.
Risk assessment should similarly consider all stages of the process, including tunnel
stability during demolition. Consideration should be given to the monitoring of surface
and/or other underground infrastructure, and other public-safety measures that may be
appropriate during the work.
5.1.5
Completion of works and beyond
On completion full details of the works should be retained in the tunnel’s asset files for
future use. This does not just extend to that information required in compliance with
current health and safety CDM regulations, but all works no matter how minor. Such
information would include as-built drawings and sketches, material performance records
and COSHH statements and details of monitoring installations or procedures.
Where appropriate, and particularly where the performance of a repair or alteration to
the structural elements of a tunnel is uncertain, for instance where a novel technique or
material has been used, or a well-tried one is used in new circumstances, it is important to
have a monitoring system in place to continuously or periodically measure and record
this. This may involve a periodic visual check carried out in the course of routine
inspections or the long-term installation of monitoring instrumentation. Procedures
should be set up to capture and review this information and make use of it for immediate
or future needs.
5.2
TUNNEL REPAIR MEASURES
There are three main categories that work may be undertaken on a tunnel during its life:
1
Routine (preventative) maintenance.
2
Remedial repair (to maintain structural integrity).
3
Strengthening, alteration, enhancement or improvement (includes replacement of
complete lining elements).
These are described in Table 5.1, along with comments on the purpose and applicability of
a range of associated activities, with references to the appropriate parts of this guide for
further information.
142
Table 5.1
Repair techniques for tunnels
Technique
Deals with
Applicability*
Routine (preventative) maintenance
Reference*
Section 5.3
Tunnel cleaning
Assists inspection of asset,
identification of defects, prevention of
corrosion or degradation
Ms, Mt, C, U
Section 5.3.1
Drainage cleaning/
improvement
Inadequate groundwater drainage
Ms, Mt, C, U
Section 5.3.2
Vegetation removal
Potential damage to brickwork and
concrete
Ms, Mt, C, U
(portal structures)
Section 5.3.3
repointing
Restoring of lining components
Ms
Section 5.3.4
Application of protective
coatings
Corrosion protection to iron and steel,
protective coatings to concrete and
masonry structures
Mt, C, Ms
Section 5.3.5
Stemming and water control
Controlling water ingress
Ms, Mt, C, U
Section 6.2
Joint caulking/re-caulking
Controlling water ingress
Mt, C
Section 6.3.1
Remedial repair (maintaining structural integrity)
Section 5.4
Patch repairs
Strengthening of lining components/
controlling water ingress
Ms, C
Section 5.4.1.1 (Ms)
Section 5.4.3.1 (C)
Crack repairs (including
flange strapping)
Strengthening of lining components/
controlling water ingress
Ms, Mt, C
Section 5.4.1.2 (Ms)
Section 5.4.2 (Mt)
Section 5.4.3 (C)
Ring separation repair,
including pinning and
grouting
Strengthening of lining components
Ms
Section 5.4.1.3
Welding structural steel
work
Strengthening of lining or structural
components
Mt
Section 5.4.2.2
Grouting
Controlling water ingress/structural
strengthening
Ms, Mt, C, U
Section 6.4
Replacement and strengthening
Section 5.5
Replacement or
Strengthening existing
Tunnel linings
Replacement of existing lining
completely or relining to strengthen
the existing lining (eg due to severe
deterioration)/increasing structural
capacity/stabilising unlined tunnel/
tunnel enlargement (increase
structural gauge)
Ms, Mt, C, U
Section 5.5.1
Underpinning
Instability of foundations
Ms
Section 5.5.2
Invert repair:
strengthening/replacement
Stabilisation of structural or unlined
invert
Ms, C (non-segmental)
Section 5.5.3
Rock stabilisation (rock
mass reinforcement)
Instability of surrounding rock
U
Section 5.5.4
Treatment of shafts
Section 5.6
Maintenance and repair
General deterioration
Ms, Mt, C, U
Section 5.6.2
Shaft filling
Safely decommissioning
Ms, Mt, C, U
Section 5.6.3
Shaft grouting
Safely decommissioning/
consolidating shaft filling
Ms, Mt, C, U
Section 5.6.4
Shaft capping
Safely decommissioning/sealing and
providing support
Ms, Mt, C, U
Section 5.6.5
Note:
*Tunnel types: Ms=masonry linings (brick and/or stone), Mt=metal lining (cast iron/ steel), C=concrete lining, U=unlined
CIRIA C671 • Tunnels 2009
143
In the majority of cases, works other than routine (preventative) works may alter the
fabric and behaviour of the lining. In these circumstances full or partial engineering
assessments should be made of the existing tunnel lining using the principles described in
Section 4.9 to determine its current behaviour and the potential immediate and long-term
effects of carrying out the works.
5.3
ROUTINE (PREVENTATIVE) MAINTENANCE
There are several basic maintenance activities that should be carried out regularly on any
tunnel to maintain its performance, prolong its serviceable life and reduce its requirement
for more significant remedial works over time. Failure to carry out regular basic
maintenance is a short-sighted approach and a false economy. Basic cyclic maintenance
should be seen as a routine and beneficial for tunnel management, rather than an
unnecessary and avoidable drain on valuable resources. Tunnel owners should ensure that
sufficient budget and resources are available to carry out routine maintenance.
Access to many tunnels is restricted and owners and contractors alike should be
encouraged to find ways of improving efficiencies in their methods of working, and make
use of the limited time that may be available to do the work. With long tunnels under
restricted working conditions (for example, engineering hours on London Underground
tunnels, possession requirements on Network Rail tunnels) the means of access to the site
may play an important part in planning works: considerable time can be spent in
travelling to and from the worksite. Especially in railway tunnels, consideration should be
given to the use of track trolleys or engineering trains as mobile platforms, and the use of
hand-held, lightweight portable tools and access scaffolds.
Consideration should also be given by the asset manager to co-ordination of work
activities. It may be more cost-effective to carry out several routine maintenance work
packages and take advantage of a planned possession or tunnel closure required for more
significant works, rather than trying to complete routine works in restricted hours.
Routine maintenance typically comprises minor and minimally disruptive activities aimed
at preserving the tunnel’s structural fabric in good condition and keeping it in a state in
which it is performing as intended.
Although the specific regular maintenance activities required for individual tunnels will
vary depending on their nature, condition and environment, activities that should be
considered on a cyclic basis include:
making minor local repairs to the fabric of any tunnel lining
ensuring any drainage is working effectively
ensuring tunnel services and equipment (eg ventilation) are working effectively
cleaning of the tunnel and drainage systems
monitoring the tunnel environment (air quality, lighting etc)
cleaning of secondary lining, ie vitreous enamel panelling
other activities aimed at preventing continued deterioration of the tunnel lining
(depending on tunnel lining type).
Regular maintenance activities are extra to, or may result from, routine visual inspections
or examinations aimed to ensure that the structure is performing adequately, that there
has been no significant change in condition (including monitoring known defects) and
that there are no external factors that may detrimentally affect the tunnel or its function.
144
The tunnel owner or asset steward should also be responsible for controlling unnecessary
or potentially harmful works by others who may use the tunnel to carry services. This
should also include internal users such as signalling, power and communication services
often found in transport tunnels. Fixing to, or cutting or alterating the tunnel lining
should be strictly controlled and only carried out after a careful assessment has been made
as to whether the works will have any effect on the tunnel structure or ancillary
equipment or services contained within.
Control measures should also be carried out on works by others above or adjacent to the
tunnel. For instance London Underground impose an exclusion zone around their
tunnels with no bored piling within a 3 m horizontal distance from a tunnel and 6 m
above.
It is important to keep a detailed record of all maintenance and repair work carried out
on a tunnel, preferably including good before and after photographs and measurements
where appropriate. This information is a valuable part of the tunnel’s history and is useful
when investigating the cause and significance of new defects, and budgeting and
programming future maintenance requirements. Such information should be maintained
in the tunnel inventory files and databases.
The following sections discuss some recognised routine maintenance works. It will be the
asset manager’s responsibility to determine which specific works are required, how often
they are to be carried out and for allocating resources and budgets to complete the works.
5.3.1
Tunnel cleaning
Cleaning tunnel linings is considered vital by some tunnel owners to assist with the
inspection process as actual or potential defects may be masked by a layer of dirt or
hidden by accumulated rubbish. Frequent cleaning of a tunnel and the surrounding area
may also be beneficial for the health and well being of tunnel operator’s staff who may
spend a considerable amount of time in the structure. In circumstances when cleaning is
not carried out, the soiling may contribute to the deterioration of the tunnel lining so
cleaning may be desirable for its preservation, for example, where sulphur-bearing soot
deposits have been left by steam or diesel locomotives in masonry tunnels (see Figure 5.1),
or where bio-fouling of sewer tunnels leads to the production of potentially corrosive
products. However, like any other work carried out in a tunnel cleaning can be difficult,
expensive and disruptive, and the tunnel’s fabric can be damaged if done incorrectly, so
some asset owners do not clean their tunnels unless there is a specific need.
CIRIA C671 • Tunnels 2009
145
Figure 5.1
Thick accumulation of soot on a rail
tunnel crown. Note that the area from
the haunches and below is clear of soot
– this is because of longstanding water
seepage in this area that has prevented
it from accumulating
As well as the accumulation of dust and grime on the surfaces of tunnel elements, rubbish
and litter may build-up and, because these may be flammable, present a fire hazard.
Transport tunnels and underground structures used for stations, and especially tunnels
with open inverts below a deck, are prone to the collection of rubbish, particularly where
the invert is used for air flow as part of the ventilation system. In tunnels where this has a
tendency to occur all such materials should be periodically removed.
The frequency of carrying out the cleaning may depend on the use of the tunnel, access
restrictions, ambience concerns (eg the passage of public through the tunnel) and
environmental issues. For example, it is generally not practicable to clean sewage tunnels
on a regular basis as access to the tunnel is very restricted due to use. In this case, access
may only be possible when any routine inspection programme is done and the sewer
isolated for this purpose. Apart from the more unpleasant aspects of inspecting a sewage
tunnel, the sewage may be causing deterioration of the tunnel lining, and cleaning should
be considered vital to complete a thorough inspection. Tunnel owners will need to decide
how often the tunnel is cleaned to satisfy these or other recognised factors.
A variety of techniques and proprietary cleaning products are available and are often
actively marketed by their producers/applicators. However, every structure and situation is
unique and there may not be a single technique or product that can be relied upon to
achieve the desired result while avoiding undesirable effects. The selection of the most
suitable technique will depend on a variety of factors including the type and existing
condition of the lining, the nature of the material or deposits needing to be removed and
the acceptability of the change in appearance (and possible irreversible damage) that
might result.
There are three main groups of cleaning methods (Mack and Grimmer, 2000):
1
146
Water methods soften the dirt or soiling material and rinse the deposits from surfaces.
2
Chemical cleaners react with dirt, soiling material or paint, allowing it to be rinsed off
surfaces with water.
3
Abrasive methods mechanically remove the dirt, soiling material or paint and may
also be followed with a water rinse.
General guidance for tunnel cleaning includes:
clean where there is a good reason and a definite benefit
use the gentlest method possible, commensurate with achieving the desired result
cleaning should only be carried out by experienced contractors
areas with particular historic or aesthetic value (eg listed portals) should be approached with
particular care and trials carried out.
Often it is possible to adequately clean brick and masonry lined tunnels by soaking using
low-pressure water spray followed by light brushing. Extreme care should be taken if
using high pressure water jetting as the technique has the potential to erode weaken
mortar and dislodge loose brick and masonry units. High-pressure jetting should only be
carried out after an assessment that the technique will not have any lasting effect on the
lining, and be carried out by skilled operatives who are fully aware of the potential harm it
could cause.
With masonry tunnel linings the application of wrong cleaning materials, chemicals, or
techniques can have disastrous results and leave the masonry surface in a weakened and
disfigured state. For instance limestone and sandstone masonry units can be damaged by
acidic treatments.
Transport tunnels are particularly prone to grime and dust from engine emissions and
brake dust. During the mid-1970s through to the late 1980s and early 1990s London
Underground carried out extensive tunnel cleaning of the brick lined tunnels forming
part of the Circle Line. Some of these tunnels date back to the 1860s to 1880s where steam
trains were used. Grime and soot deposits accumulated on the tunnel linings leaving a
heavy encrustation of up to 50 mm thick. The most successful method to clean the tunnel
was using high pressure water. Compressed air and grit blasting were attempted but
found to be less efficient or uneconomical.
Typically at risk during tunnel cleaning would be any operational services (signals, power,
communication systems etc) that may be carried within the tunnel together with fragile
secondary lining or finishes. Usually there will be insufficient time, or other practical
reasons why, removal or re-positioning of services before cleaning can be carried out, or
time to install elaborate protective measures. Care is also required when using such
techniques along the joints of segment lined tunnels so as not to disturb any caulking
compounds or material that may be present to prevent water ingress through the joint.
Special precautions may also be required to avoid disturbing hazardous materials that may
be present such as asbestos based sheeting used in secondary linings, segmental lined
caulking compounds and fire protection coatings. The material being cleaned down may
also be considered hazardous and require special precautions for its collection and
disposal.
Caution is advised in the use of sand blasting/cleaning techniques, which can only be used
on more resilient lining surfaces such as cast iron and only with care on concrete or
masonry linings. Such techniques are considered more intrusive, requiring greater
protection measures to be done on ancillary structures and tunnel equipment, and time
allowed for cleanup operations at the end of each shift should the tunnel need to be
brought back into service.
CIRIA C671 • Tunnels 2009
147
Further guidance on cleaning different materials in tunnel linings can be found in BS
6270-3 (BSI, 1991a) for metals, and in BS 8221-1 (BSI, 2000a) for brick, natural stone and
concrete.
In rail tunnels the track may be supported on ballast, which should be maintained in good
order to avoid contamination from the cleaning operations.
During any tunnel cleaning operation that involves water, detergents or chemicals of any
kind, careful consideration should be given to the potential for these materials (in the
normal course of the works, or accidentally) to enter and contaminate drainage systems,
groundwater and watercourses, and the possible environmental impact they might have.
This may merit the consideration of alternative materials or working methods to mitigate
risk, or to minimise the consequences of accidental discharges. Equipment in the form of
sumps and settling tanks may need to be used or built into a tunnel drainage system to
intersect any potentially contaminated fluids resulting from the cleaning operation for
separate treatment. Even plain water used in a cleaning operation may become
contaminated with either grit or grease, which would need to be filtered out before
discharge or disposal. Particulate contaminates would also cause a problem for any
pumped system used to remove water or fluids in a cleaning process as it is likely to cause
increased wear and tear on the pumps, or blockages.
The health and safety of the tunnel cleaning operatives would also need to be taken into
consideration through adequate risk assessment of the materials being used for, and
resulting from, the cleaning process and the use of suitable and adequate personal
protective equipment (PPE) and work methods.
The inspection of unlined rock tunnels benefits from having a clean floor policy where all
accumulated rock debris is recorded and cleared after every inspection as a means of
monitoring spalling rock. This may provide an early warning of a deep seated instability.
5.3.2
Drainage maintenance
Effective management of water is fundamental to the long-term serviceability of tunnels.
Where provision has been made for drainage or management of water, whether as part of
the original structure or added later, it is important that it is maintained, for example, by
ensuring all drains and drainage paths are kept clear and functional. Management of
water ingress through existing water management systems is usually more economical
than trying to stop it completely.
Often during the construction of tunnels, local drainage measures are installed to adapt to
ground conditions found during excavation. Drainage measures may include weep holes
at the base of side walls and collection pipes. It is important that such installations are
maintained and, if necessary, improved upon to manage the ingress of water most
effectively. The removal of such systems by grouting up weep holes etc should not be
considered as it is very likely that the equilibrium of the water flow around the tunnel will
be upset resulting in water ingress through previously dry sections of the tunnel. More
importantly the loading on the tunnel lining may increase due to increased hydrostatic
water pressure and this could lead to overstress and possible damage or even collapse.
Drainage channels in the tunnel invert should be kept clear of debris and fines to maintain
the flow of ingress water to pumped sumps etc.
148
Figure 5.2
Guttering and downpipe system that has been installed to channel
water ingress from a tunnel wall into the invert drain, but has not
been maintained so that it is no longer effective
Most old brick lined or masonry tunnels do not incorporate any kind of waterproofing
system. However, given the nature of the materials used the structure is usually permeable
and water can drain through the lining. The use of lime-based mortars make the lining
breathable – allowing it to dry out where there is adequate ventilation, rather than
remaining in a permanently saturated state. So care is needed in the selection of
appropriate mortars for repointing of brick and masonry to maintain the drainage
characteristics of the tunnel lining and avoid accelerated deterioration.
5.3.3
Management and removal of vegetation
The effects of vegetation are more likely to be seen in the areas around tunnel entrances,
shafts and locations where tunnels have very shallow cover. Masonry structures are more
likely to be prone to the adverse effects of vegetation than other forms of construction but
segmental tunnel linings (eg concrete and cast iron tunnel linings) may still be affected.
Tree roots penetrating masonry tunnel linings several tens of metres below ground level,
causing damage and allowing the free ingress of water has been recorded.
Plants, tree roots and accumulated moss have the potential to disrupt and displace the
fabric of a tunnel, block drainage channels and retard the drying out of wet masonry.
Ideally the vegetation should be completely removed from the structure, and monitored
in the adjacent area. The vegetation should be cleared away from all parts of the structure
and the roots raked out. It may also be beneficial to treat any remaining roots with a
suitable herbicide, although the potential environmental impact of the use of such
materials should be considered. Vigorously growing plants and shrubs immediately
adjacent to the structure should also be cleared away because their roots may penetrate
the tunnel lining, which affects structural components of a tunnel where it surfaces, eg
portal structures at tunnel entrances. Vegetation may also obscure tunnel structures and
hinder inspection.
The best time to clear or control vegetation is during the spring. Care should be exercised
where flora on structures may include rare and protected species or provide homes for
protected fauna.
5.3.4
Repointing of masonry-lined tunnels
Loss of pointing and jointing mortar loosens masonry units, brick or stone, which may
present a hazard to traffic and members of the public using the area below. Loss of mortar
CIRIA C671 • Tunnels 2009
149
from joints also reduces the ability of the masonry to transmit and evenly distribute forces,
focusing stress in localised areas and potentially leading to cracking and distortion.
Repointing may be required to prevent progressive deterioration of the masonry fabric
and should be dealt with as part of a routine maintenance programme.
Repointing is most frequently required close to tunnel portals and open shafts in exposed
locations and/or subject to severe weather conditions, particularly wetness and freezing.
Tunnel ventilation is an important factor, and tunnels on a coincident alignment to the
prevailing wind may also be more affected by the elements than a tunnel running at an
oblique angle. Frost damage can affect the first 300 m to 400 m length of tunnel from a
portal (Haack, 1991).
Waiting until the majority of the pointing in a lining has completely deteriorated or fallen
out before carrying out repointing is not advisable, because other damage may have
already occurred to the structure. Ideally repointing should be carried out in dry
conditions, particularly where lime-based mortars are used, and at a time where there is a
low risk of exposure to freezing temperatures.
Care should be exercised where gaps in mortar may contain protected species such as
bats.
For some tunnel owners a loss of mortar to a depth of over 20 mm may trigger repointing
(Railtrack, 1996). In cases where mortar shows signs of deterioration and the decision is
made to repoint, the joint should be cleaned out to a depth at least twice the width of the
joint, or to a maximum depth of 18 mm to 25 mm from the finished face.
When cleaning out for repointing care should be taken to avoid damaging the
surrounding masonry units. Hand tools (quirks and long necked jointing chisels with
parallel faces) are normally adequate where the old mortar is weak but for treating larger
areas in a limited time, high pressure water jetting may be used if carried out with care.
Where joints are thin and dense mortar has been used it can sometimes be difficult to
remove, but the use of mechanical tools should only be considered when necessitated by
the scale of the work. Where such methods are necessary appropriate equipment should
be used by skilled operatives to prevent damage to the masonry units. Cutting out using
angle grinders is not advised as the risk of damage to the masonry units is too great.
Where deterioration of jointing mortar is extensive resulting in voids and friable mortar
deep in the joint and considered beyond repair by repointing techniques, pressurised
compressed air or mechanical repointing may be necessary. The loose and very soft
mortar should be removed back to more solid material to a depth of up to 100 mm for
brickwork and potentially more for stonework depending on the size of the masonry
units. This may cause loosening of the facing course of brickwork and masonry units, and
care should be taken to avoid their damage or displacement. A suitable mortar can then be
injected to fill the joints under pressure using compressed air or mechanical pointing
equipment to pressurise the mortar and force it through a hose to a gun nozzle. The
operator should then build up mortar in layers from the back of the joint to the front in
one continuous operation to avoid cold joints.
Mechanical pointing equipment and techniques (in which grout is injected into the joint)
have been successfully used for deep-pointing of masonry rail and canal tunnels for almost
50 years (Sowden, 1990). The resulting finish can be less satisfactory, but this is typically
not a problem in most tunnels and if necessary the surface can be re-finished once the
mortar starts to set. This method does not seem to be in use at the present time, largely
because the equipment is unavailable, but its potential benefits may merit its revival for
use in tunnel maintenance.
150
5.3.5
Application of protective coatings
5.3.5.1
Metal tunnel linings
The traditional method of corrosion protection of iron and steel to prolong the lifespan of
the material is the application of a protective coating. The effects of corrosion can be very
damaging not only to the metal component through partial to complete loss of section, but
also to surrounding or connecting members.
In practice structural iron and steel should have been protected as stipulated by the
various standards, codes of practice and regulations at the time of construction, though
this protection may not be effective for the full lifespan of the element. So cleaning and retreatment of the metal is often required, although clearly this can only be applied to
exposed and accessible surfaces.
Cast iron by virtue of its method of manufacture has good resistance to corrosion as, when
cast, silica from the moulding sand fuses and coats the surface of the casting to form a
barrier against oxygen. So it may require little or no treatment in the long-term. However,
should environmental conditions change or the surface be scratched or cracked corrosion
can set in. Cast iron can be weakened by graphitisation, which requires little oxygen as it
can be brought on by the presence of brackish or acidic groundwater. Wrought iron has
reasonable resistance to corrosion and fares better than carbon steel.
The requirement for replacement of protective coatings to iron and steel components in
situ should be identified through regular inspection. However, difficulties exist where
structural iron and steel work is covered or embedded. Intrusive inspection techniques
may be necessary to examine the condition of the metalwork and assess the need for
protection. In some cases this may not be practicable and the effects of corrosion may not
present themselves until the damage is done and remedial works required.
When applying a protective coating to metals, either to a new component or to existing
sections in situ, surface preparation is critical. Cleaning can be carried out using grit or
shot blasting or high pressure water techniques, with consideration given to protecting the
fabric of the tunnel and any services, and the effects on health and safety of operatives,
particularly when working in situ. Depending on the condition of the metal finish, hand
cleaning using light tools such as wire brushes and scrapers may be all that is required to
ensure an adequate surface for the application of a protective coating. The cleaning
should remove any rust, salts, loose or flaking paint and other contaminants (dirt, grease
and oil) that may be present. Removal of oil-based contaminants may need detergents.
Protective coatings in the form of primers, barriers coats and bituminous coatings, should
comply with the requirements of BS EN ISO 12944 (BSI, 1998a). When selecting an
appropriate coating compatibility with other finishes or materials used, or to be bonded to
the metal, should be considered.
Coatings can also be used to increase the fire rating of metallic elements, and there are
several technologies available. Passive fire protection materials insulate metallic elements
from the effects of the high temperatures that may be generated in fire and can be divided
into two types:
1
Non-reactive: the most common types are boards and sprays.
2
Reactive: intumescent coatings are the best example.
CIRIA C671 • Tunnels 2009
151
A discussion of these materials, their characteristics and use is beyond the scope of this
document. More information on the effect of fire on structural materials is included in
Section 2.6.3.
5.3.5.2
Concrete, brick and masonry linings
Brickwork and masonry can be protected from weathering and other environmental
effects by the application of a cementitious-based render. Natural cement, a hydraulic lime
cement formed from burnt limestone and clay, has many benefits when used as a
protective coating including fast setting, high early strength, good resistance to the
passage of water, the detrimental effects of salts, chemicals, acid and alkalis and high
corrosion protection properties. Another method of protecting concrete, brick and
masonry is through the application of a proprietary water repellent coating. These are
typically silicone based products and have a high resistance to alkalis, good adhesion
properties for paints and are water-based and environmentally friendly.
The use of coatings on these lining materials has many potential pitfalls. If used in the
wrong situation or incorrectly selected, specified and applied, coatings can fail rapidly or
may cause damage to the tunnel’s structural fabric – particularly where old masonry (brick
and stone) is concerned. An important consideration when waterproof coatings are used is
that the substrate should be dry and not entrap water, which could exacerbate
deterioration. Also coatings often require a sound and regular surface, and on surfaces
that are dirty, damaged or weakened (as is the case in many tunnels that have been in
service for long periods of time) this may cause problems. It is recommended that advice is
sought from an independent and suitably experienced specialist before using such
treatments, rather than relying on manufacturer’s claims.
Care should be exercised when applying waterproof or protective coatings not to block
any permanent drainage paths, drains or water outlets that are installed to prevent the
build-up of water behind the lining (see Section 5.3.2).
Damage to concrete, brick and masonry can also be caused by corrosion of fully or
partially embedded iron or steel beams or structural work used in the construction of the
tunnel lining. Typically the corrosion product occupies between five and 10 times the
original volume of un-corroded material, which can severely damaged adjacent brick,
masonry or concrete, cracking and outward displacement of the surrounding material.
5.4
REMEDIAL REPAIRS
The following sections discuss a range of techniques that are either directed to repairing
existing, or to locally replacing with new, tunnel linings. Through external effects they
have either failed or are in such an advanced state of deterioration that their failure would
have catastrophic (ie collapse) or operational effect on the tunnel. Deterioration in
components of old tunnels may also be because they have outlived what would be
considered their original design lives (if there was such a consideration during the original
planning and design of the tunnel) with deterioration occurring through wear and tear.
Repair usually implies reinstatement of lost or damaged, structural or weakened material
with the same material. However strengthening also implies using extra sound material,
which may be the same as the original material or another material, to share in the
support of the load directly. Strengthening of tunnel linings is dealt with in more detail in
Section 5.4.5.
152
Tunnel repair works are usually beyond the preventative maintenance works discussed in
Section 5.3, which are aimed at preserving the tunnel’s structural fabric in good condition
and are considered as minor and minimally disruptive activities. Tunnel lining repairs and
more extensive strengthening or tunnel lining replacement works are aimed at providing
a long-term solution. This could involve more extensive works, possibly requiring greater
possession or closure of the tunnel to be done, and may need external specialist designers
and contractors to be employed.
The need for a good understanding of the cause, severity and extent of deterioration
before devising and carrying out schemes for tunnel repairs cannot be overemphasised.
This allows the extent of the work to be determined in advance and minimises the risk of
encountering unexpected situations or requirements. When carrying out repairs on any
type of structure being unprepared is undesirable and potentially costly, but in tunnel
repairs this is especially so. Unanticipated problems (for example, logistical problems with
resources and materials, difficulties with techniques used, slow rates of progress,
unforeseen constraints and underestimated scope or work) can have considerable negative
impact on the success of tunnel repair contracts , and could lead to operational disruption,
spiralling costs and contractual disputes.
5.4.1
Masonry linings
The repair of masonry may be categorised by the following:
patch repair
crack repair
ring separation repair.
Table 5.2 describes and gives advice on using the various types of repair to address
common defects in masonry linings.
CIRIA C671 • Tunnels 2009
153
Table 5.2
Summary of typical defects of brick and masonry tunnel linings and possible remedial
solutions (adapted from Railtrack 2002)
Defect
Description
Possible repair type
Solution
Soft/spalling of
brickwork or
masonry
groundwater or soluble salts from
certain bricks may result in sulfate
attack causing progressive cracking
and crumbling of the mortar joints
condition arises when bricks and
Patch repair
masonry units are subject to constant
water saturation
freeze-thaw action may cause frost
damage with units flaking and
becoming friable.
Loose brickwork or
masonry units
this will occur with continuous
washout or perished joints
a breakdown of the mortar between
bricks or masonry units can occur
and local areas may become
unstable to the extent that they
become loose or even fall out.
Deep repointing or
patch repair
Drummy/hollow
brickwork
brickwork sounds hollow when
tapped with a hammer (sound
brickwork should ring bright)
Patch repair and ring
separation
dull or flat tone often indicates
potential hollow brickwork due to ring
separation or spalling.
Ring separation
Irregular
profile/bulging
154
may be the result of poor
construction – lack of cross bonding
or joint mortar between rings
a symptom of washout or perished
joints
may be the result of flexure of the
lining due to live loading/vibration/
fluctuating ground and groundwater
stresses
multiple separations may occur
between successive courses
Ring separation
may be found in structural linings
where distortion of the surface profile
results from eccentric loading or
voiding
could be an indication of lining failure
or ring separation (a great concern)
bulging or tunnel distortion may have
an effect on the structural gauge of
Patch repair and ring
the tunnel
separation
bulging may be encountered at
meeting point of construction joint
may indicate presence of water – see
wet patches
bulging or irregular profile may be the
result of intrusive vegetation or
external forces acting such as heave
or lateral loading.
initially nothing other than monitoring
spalling
ensure adequate drainage and prevent
use of de-icing salts (on adjacent roads or
footpaths), prevent contact from airborne
sulfates through use of protective coatings
as condition deteriorates, replacement of
defective brickwork, masonry or mortar
may be required, using temporary
formwork or centring and selection of
appropriate mortar
in severe cases full structural repair or relining.
deep repointing may be possible after
temporarily pinning the masonry units in
place, though replacement of masonry
may be the most appropriate course of
action
in vertical walls the area to be replaced
should be controlled. It is recommended
that an area no more than 1 m² is
removed and replaced at any one time.
Care and attention should be given to
supporting the surrounding brickwork or
masonry with the aid of temporary form
work or centring.
trial hole/NDT/distress testing to
determine either ring separation or
spalling
clean off spalling brickwork until fresh
brickwork found and test for drumminess
– if hollow assume ring separation
if ring separation assumed investigate by
intrusive or non intrusive methods – may
require an area of brickwork to be
removed to determine very narrow
separation
treat for ring separation.
trial hole/NDT/distress testing to
investigate and confirm the defect
depends upon severity of the ring
separation
pinning and grouting may suffice for
sidewall, lower haunch repair and
localised crown repair
in severe cases re-lining may be
warranted.
trial holes/NDT to investigate and confirm
extent of the defect
monitoring irregularities in tunnel profile
to ascertain extent and rate of movement
and help prioritise remedial works or
further actions
water pressure may be relieved by
installing weep points in the structure or
carry out maintenance of the drainage
system (cleaning out blocked weep holes
etc)
voiding may be prevented by pinning and
grouting the affected area
in extreme cases, removing and replacing
the affected brickwork, with the aid of
temporary support, may be necessary
in very severe cases full structural repair
or re-lining may be required.
Table 5.2
Summary of typical defects of brick and masonry tunnel linings and possible remedial
solutions (adapted from Railtrack 2002) (contd)
Brickwork/masonry
cracks (general)
cracking of any tunnel lining may result due
to any of the above defects
dormant or residual cracking may be the
result of past movement or defects that
have since been treated.
Crack repair
Transverse cracks
occur laterally along the length of the tunnel
created by rotational movement of the lining
associated with ring separation in masonry
cracking may be attributed to differential or
eccentric loading of the lining
increased external loading (or unloading) ie
adjacent tunnelling works, foundation piling
settlement of the tunnel invert or ground
bearing walls.
Crack repair
Longitudinal cracks
Impact damage
Wet patches
caused by shear action or dynamic loading
may occur at junction between two
separately stiffened sections (main tunnel
and headwalls or portal structures etc)
through differential thermal action
(expansion/contraction) etc
increased external loading (or unloading). ie
adjacent tunnelling works, foundation piling
settlement of the tunnel invert or ground
bearing walls.
Crack repair
tunnel lining struck by road vehicle/train etc
scouring/gouging of brickwork or masonry.
Patch repair
in flat exposed structural elements, wet
patches may be the result of poorly
protected brickwork or masonry
this may be the result of a broken or
blocked down pipes or gutter systems used
to deflect/manage water inflow
in extreme situations, including active water
ingress, may be an indication to the
presence of a nearby burst water main
wet patches may also occur as the result of
natural fissures in the host rock allowing
groundwater to permeate the tunnel lining
(refer to irregular profiling or bulging).
investigation required to determine cause of
failure
monitoring cracks to ascertain extent and
rate of movement and help prioritise
remedial works or further actions
repointing inert cracks using mortar or, for
live cracks, flexible/elastic filler (mastic etc)
cross stitching cracks with metal dowels
enforcing ground improvement remedial
measures
in extreme cases strengthening of lining may
be required through re-lining, construction of
an extra inner ring or lining, or underpinning.
investigation required to determining the
cause of failure
monitoring cracks to ascertain extent and
rate of movement and help prioritise
remedial works or further actions
repointing inert cracks using mortar or, for
live cracks, flexible/elastic filler (mastic etc)
combination of cross stitching accompanied
with localised brick replacement or pinning
and grouting
in extreme cases strengthening of lining may
be required through re-lining, construction of
an extra inner ring or lining, or underpinning.
depends upon severity – localised repair to
replacement of individual components
through to full structural repair or re-lining.
investigation required to determining the
cause and extent.
treat cause were possible
repair/unblock/maintain drains and gutter
systems
for a broken water main, instigate repairs to
the main (report to service provider).
However, if suspect broken water main the
tunnel should be monitored closely for signs
of distress
install water management system(s)
in some cases relieve water pressure by
installation of weep holes
grout fissures/soil to prevent water ingress
pinning and grouting masonry
install waterproof lining.
Water control
measures
CIRIA C671 • Tunnels 2009
depends upon the severity
monitoring cracks to ascertain extent and
rate of movement and help prioritise
remedial works or further actions
repointing to replace lost or damaged mortar
for inert cracks, but for live cracks flexible/
elastic filler (mastic etc) might be better.
Removal and re-casing the affected brickwork
with the aid of temporary support in extreme
cases
cross stitching cracks with metal dowels.
155
5.4.1.1
Patch repairs
Patch repairs are mainly required where there is excessive localised deterioration of the
mortar and/or the masonry or brick units and it is necessary to reinstate the structural
integrity of the tunnel lining.
A patch repair in brick or masonry tunnel lining is the replacement of several adjacent
bricks or masonry units in one area. For example, this may involve the partial removal of
one or more courses of a brick or masonry lining in a localised area that may be showing
signs of deterioration and replacing with similar materials or alternative materials to
maintain structural continuity of the lining.
When carrying out a repair on old masonry or brickwork it is preferable to replace like for
like, ie replacement with what was taken out using the old bricks and new mortar,
provided the brick units themselves do not show any signs of deterioration, otherwise
replacement using new or recycled bricks. However, alternative methods of repair may be
deemed more applicable in terms of:
Figure 5.3
availability (of materials, local knowledge, skills base etc)
cost (repairs using masonry are considered to be expensive compared with alternative
methods such as shotcreting)
time (masonry repairs may not be the fastest method of repair)
suitability (of design).
Several visibly distinct phases of patch repair to an old rail tunnel lining
Selecting sympathetic materials
Using materials that result in an overly-strong or overly-stiff repair (relative to the
surrounding original fabric) could result in a hard spot in the tunnel lining. This may
affect local stress redistribution, potentially causing premature failure or distress of the
surrounding area.
156
see Tables 2.8 and 2.9 for typical properties of bricks used in existing railway
structures. Durability of the brick and masonry components is an important
consideration, including resistance to frost and soluble salt content. High strength
bricks generally exhibit low levels of water absorption and so often (though not
always) have better frost resistance. Water absorption characteristics also determine
how a brick will react to environmental changes, such as temperature and humidity.
New dry bricks will have the potential to absorb a great deal of moisture, and may
expand sufficiently to cause further compressive stress to that generated by external
loading. When carrying out a repair the bricks should be stored in the local
environment for a reasonable period to stabilise
bricks produced today are of a standard size, typically 215 mm × 102.5 mm × 65 mm.
However, in the past a much wider variation in size of bricks was manufactured,
particularly in their height. When renewing brickwork consideration should be given
to matching the existing brick size as the bed joints may have to be deeper or thinner
than those in the original brickwork giving the impression of poor workmanship
manufactured and natural stone masonry units, should comply with BS 5628 (BSI,
2005b), and now BS 5628 and EN 771-6 (BSI, 2001d). Where possible they should be
selected on the basis of proven durability and resistance to weathering in a similar
climate and exposure condition to the masonry to be repaired
it is generally good practice to ensure that the new jointing mortar is weaker than the
masonry unit being used in the repair. New mortar should also have adequate
permeability to allow the brickwork to breathe and for moisture to evaporate through
the joints rather than through the masonry units. For old masonry tunnel linings
consideration may need to be given to the use of lime-based mortars, particularly
hydraulic limes.
For further information on the selection of materials for masonry repair see McKibbins et
al (2006) and the Concrete Society (2005).
Carrying out repairs
Where individual bricks are replaced, temporary formwork is not necessary as timber
wedging is generally sufficient to maintain the individual units in position. Similarly, it is
generally recommended that relining vertical faces of a tunnel lining can be carried out
without the need for temporary works provided the extent of repair is limited to 1 m².
Where brickwork is to be renewed over the crown of the tunnel, some form of temporary
formwork or centring may be required to support the brickwork and is only removed
once the mortar has attained sufficient strength. Formwork should be suitably designed
taking into account expected loading conditions and clearance requirements if the tunnel
is to be kept operational. Certain authorities or asset owners may require a temporary
works design certificate to be issued for the formwork.
a
Figure 5.4
b
Typical patch repair to two courses of brickwork (a) with pinning detail (b). Note the use of a
centering rib to support the repair
CIRIA C671 • Tunnels 2009
157
Figure 5.5
Carrying out patch repairs using temporary supports
Once the defective brickwork is broken out, the exposed surface of the underlying ring or
leaf of brickwork should be inspected to determine if further repairs are required. Before
carrying out further, deeper, remedial works it is important that full consideration is given
to the stability of the tunnel lining and that appropriate experienced and qualified staff
are involved in the decision making.
Where possible, when carrying out a repair, the bond of the original brickwork should be
maintained and the new brickwork keyed into the surrounding brickwork. No extra
stiffening of the lining should be introduced through further bonding between successive
rings forming the arch of the lining. If this is done the structural behaviour of the lining
may be altered. If the brickwork is to be tied between successive rings on a repair of more
than one brick ring thickness, the use of stainless steel brick fishtail ties is one suitable
method of tying back new brickwork, installed on a diamond pattern with a nominal pitch
of 400 mm × 400 mm.
Repairs to stonework require similar considerations to those discussed for brickwork, but
employ slightly different techniques and labour skills, so are likely to require the use of a
specialist contractor with suitably skilled and experienced masons. Local repairs, which
involve the replacement of a small number of masonry units or damaged parts of units
only, can be achieved either by replacement with new stone or by piecing in to repair the
damaged areas only. Stonework repairs are pinned back to the original fabric of the lining
in a similar way to brickwork repairs. The use of plastic repairs where mortar is used to
replace original stonework, or the use of bricks to replace stone should be avoided
wherever possible, because the results can be unsightly and repairs may fail prematurely
or damage adjacent masonry fabric.
For extensive works such as re-casing the tunnel crown, detailed programming of the
works will be necessary. This will allow the number of possessions or closures to be
determined to install any temporary works and complete the works. During the planning
stage consideration should be given to the overall area of brickwork to be replaced and
any limitation placed on the area of brickwork or masonry removed and replaced at any
one time.
It is important that when temporary works are to be installed within any operational
tunnel that sufficient clearance for traffic is always maintained. Permission should be
obtained from the appropriate authority before installation works start, and consideration
given to the design of the temporary works for any restrictions that may be in place.
158
Alternative repair techniques
Alternative repair methods for brick and stone masonry may include such materials as
sprayed concrete (discussed in Section 5.4.8) or cast in situ concrete.
5.4.1.2
Crack repairs
Cracking of masonry tunnel linings should not be repaired until the cause has been
adequately established and, where necessary, dealt with. Crack repairs are only worthwhile
if the cause of the cracking is unlikely to recur, or if provision is made for future
movements. Superficial repairs to cracking involve sealing the surface of the crack to
prevent the ingress of moisture and deterioration of the adjacent materials, but do not
restore structural connection between the masonry either side of the crack.
Longstanding inactive cracks can be repaired using mortar materials that should not be
too hard or brittle, or else small movements are likely to result in a recurrence of the
cracking and failure of the repair. Cracks that are expected to experience further
movement, for example, through cyclic moisture or thermal variations, can be treated as
joints and sealed with a flexible material that will accommodate the anticipated range of
movement. If a crack is acting as a drainage path then a permanent pipe or other means
of drainage may be incorporated into the repair. The pipe may be connected to a water
collection system if dripping or flowing water cannot be tolerated (see Section 6.2).
Where cracks have confined themselves to the mortar joint lines they can be repaired
using normal pointing methods as in Section 5.3.4. However, cracks that pass through the
masonry units themselves are more difficult to treat, and patch repairs may be necessary.
Alternately the crack may be stitched using a stitching bars installed diagonally through
the crack at suitable intervals or, in the case of brick linings, along the brick courses over
the length of the crack. Several proprietary systems are available on the market that could
be used. The advantage of stitching is that it offers an element of reinforcement to the
cracked masonry that repointing does not achieve.
Figure 5.6
Installation of stitching bars along a crack
CIRIA C671 • Tunnels 2009
159
Box 5.1
Assessing the nature of a crack
Where cracking and distortion of a masonry lined tunnel has occurred, it is very important that this is not
simply covered up by repointing, because this can hide serious structural problems. An assessment should
be carried out of the likely cause of the problem and the remedial measures done to rectify it.
Where cracking is known to be longstanding and non-progressive, repointing may be considered but
detailed records of such defects should be made before and after remediation, including drawings with
measurements and photographs.
Cracks that experience movement, for example, through cyclical moisture or thermal variations, can be
treated as joints and sealed with a flexible material to accommodate the movement. However, such
cracks should be carefully assessed to ensure that they do not represent deep seated failure
mechanisms of the tunnel lining, localised or otherwise, which may require more extensive forms of
treatment such as localised re-lining or strengthening works.
5.4.1.3
Ring separation repair
The primary causes of ring separation in multi-ring brick arches are:
erosion and washout of mortar between courses
overstressing the lining from increased imposed loads
flexure of the tunnel lining, possibly caused by eccentric loading or uneven stresses
distribution within the lining brought about by voiding around the tunnel and
fluctuating ground and groundwater stresses
poor original construction
vibration and live loading
settlement of the lining walls (foundation failure)
damage by impact
delamination of previous repairs.
Failure of the inner course or courses of brickwork or masonry as a result of ring
separation can range from loosening or fall out of localised areas or individual bricks or
masonry units, to complete collapse of one or more courses over larger areas. However, as
with all defects in tunnel linings, the root cause of failure should be established, including
carrying out a thorough investigation, before deciding the most suitable repair method.
Ring separation is often associated with water ingress through the lining, and as
mentioned, washout of the mortar between courses is a primary cause of ring separation
or voiding occurring within the brickwork linings.
Brickwork and masonry repairs of ring separation comprise interstitial grouting of the
void or separated courses with a suitable grout or, in extreme cases, complete re-casing of
the lining in the failed areas. When re-casing, or re-building successive rings of brickwork
or masonry, consideration should be given to the factors discussed for patch repairs
regarding formwork design.
If grouting is to be carried out it is advisable that this is done in conjunction with pinning
of the brickwork or masonry. The advantages of pinning include:
160
tying two or more courses together to jointly resist grouting pressures
forming a composite action between pinned rings, with the pins acting as shear
connectors
pinning individual loose brick and masonry units.
Drilling holes and installation of pins should be carried out through the centre of the unit
on a regular staggered pattern, as close as 400 mm centres, but subject to the extent of
ring separation and grouting that needs to be carried out.
Figure 5.7
Brick lining pinning for grouting ring separation
Several proprietary products are available. Some are installed in holes of a larger size than
the pin and the annulus between the pin and the hole filled with either a cement, epoxy
or polyester grout. In others, the pins may be driven into smaller diameter holes than
themselves. The advantages of driven pins are speed of installation, no requirement for
annulus cement or grout to hold the pin in, and the added interlocking strength to the
brickwork to resist grouting pressures. Propriety systems also combine driven pins and
interstitial grouting techniques in one operation. The driven pin provides mechanical
reinforcement to the ring where separation is occurring, while the grouting, using
chemical grouting methods (see Section 5.4.1.2) is used to control water ingress through
the brickwork. The spiral design of the driven pin allows access for the resin grout into the
interstices of the structure.
Grouting of voided or delaminated brick or masonry linings is usually carried out using
cementitious grouts, often with additives or fillers such as pulverised fuel ash (PFA), or
resin grouts. However, the selection of the grout will depend on whether the grout is to
provide structural strength and bond to the brickwork or for prevention of water ingress,
or a combination of both.
Provision should be made in grouting works for cleaning the grouting equipment and
pipes during or at the end of working shifts – bins or skips may be necessary to collect
residual grout fines and contaminated water.
5.4.2
Metal tunnel linings
Deterioration or defects in metal tunnel linings or metal components forming a lining will
be limited to corrosion or distortion (that may result in cracking or fracture) or a
combination of both.
CIRIA C671 • Tunnels 2009
161
The following sections discuss a range of options for treating iron or steel linings or
components forming a lining. The techniques are aimed at either repairing or replacing
deteriorating or defective sections in situ. Work to strengthen iron or steel tunnel lining
components may use the same techniques as those for repair, but will be designed to
improve the structural capacity of the section by installing extra components.
Strengthening works to iron and steel tunnel lining components are not discussed further
but factors affecting repair should be considered when carrying out such works.
5.4.2.1
Cast iron lining repairs
Flange strapping and pan-plates
Figure 5.8
Typical example of a damaged circle joint flange of a bolted cast iron lining
For defects such as cracked flanges, flange straps can be installed to strengthen areas using
the existing bolt holes either side of the defect (see Figures 5.9 and 5.10).
Figure 5.9
Typical example of flange strapping
in cast iron lined tunnels
162
For a corroded or cracked segment pan a similar repair can be used by bolting a plate
onto the original, sound part of the pan (see Figure 5.10). If such a repair is to be
considered, a watertight connection should be sought if there is a likelihood that
groundwater seepage will occur. The components used for the plate repair should also be
treated to prevent corrosion.
Ordinary bolts (Grade 4.6, 8.8 etc) are recommended to secure straps or plates to cast
iron. The use of high strength friction grip bolts is not recommended as these can create
tension in jointed members. They are, however, acceptable for jointing plates to steel
sections. New bolt holes in cast iron should not be drilled using percussive drilling
equipment due to the risk of shattering the brittle metal.
Figure 5.10
Typical example of plate repair to cast iron tunnel segment pan
If properly designed, it is likely that such repairs will result in a structural component of
comparable stiffness to that of the original, undamaged lining. At the very least they will
go some way to strengthen the lining if it is required, or in the case of the pan plate,
prevent water ingress. However, these repairs require an element of pre-design and preconstruction to form the individual components. The requirement to prefabricate
elements may make a repair more costly and possibly require multiple visits to first survey
and then install the repair.
If the damage to the lining is due to over stressing from external loading regimes, ie
change in ground stress or increased loading due to nearby piled foundations, it is
unlikely that such repairs would be sufficient to strengthen the linings as a permanent
solution. When this is the case, it may be necessary to install structural members to
strengthen the existing lining, provided the works do not interfere with clearance lines for
the operation of the tunnel. In the extreme case, it may be necessary to locally re-build the
tunnel lining with a lining of greater structural capacity.
Caution is needed in the use of different metals in repair of metallic linings due to
bimetallic (galvanic) corrosion. Bimetallic corrosion occurs where different metals of
significantly different nobility come in contact and form a galvanic couple, particularly in
wet or damp environments. The result is usually a localised area of corrosion, such as
around fixings using nuts and bolts, rivets or welds. To prevent bimetallic corrosion of
adjoining metals the follow precautions may be used:
CIRIA C671 • Tunnels 2009
163
isolating the metals electrically using insulators
isolating the metals from the environment using coating etc
choosing metals that are close together in terms of nobility
application of cathodic protection.
See Baeckmann (1997), BSI (1991b) and BSI (2001a) for further guidance on cathodic
protection.
Metal stitching
The repair of cast iron tunnel and shaft linings by welding is difficult, if not impossible to
carry out in situ, especially in an underground environment where vital cables and track
side services may create an operational restriction. The safety requirement for the use of
hot work may also be too onerous, with the potential for fire or the generation of harmful
fumes in confined spaces. To effectively weld cast iron the metal needs to be preheated to
very high temperatures, typically about 480°C for brazing and 700°C for welding. Welding
cast iron, if not carried out under strict control, can cause further in situ stresses within the
lining resulting in distortion, hardening and extra cracking.
Methods employed in the past to repair cast iron have included bolting on plates and
straps over the damaged or cracked section of the lining, as discussed in the previous
section. An alternative in situ repair technique for cast iron is the proprietary lock and
stitch metal stitching system that originated in the USA and has a been used for repairing
high-pressure castings such as water pumps, compressors, engine blocks and gear box
casing. The technique provides a high strength watertight repair without the need for
welding. The metal stitching system uses a series of interlocking studs and involves the
drilling of the cast iron using hand-held drills to form a threaded hole to take the stitching
stud. The holes are cut with a specially tapped spiralhook thread, which engages with the
studs to effectively draw the metal into the threads. When used with the patent lock, a
strap device that bridges the crack to prevent separation, a complete repair can be
achieved that is likely to restore the lining material close to its original strength.
The material used for the metal stitching components is compatible with the cast iron in
terms of thermal expansion properties and corrosion compatibility.
The repair system is potentially useful for the repair of cracked linings. In the case of
badly corroded sections where there is a complete loss of section it may be possible to cut
out the corroded section and replace with a new section using the metal stitching
technique to bond the new to the old.
164
Figure 5.11
Metal stitching process (courtesy Lock N’ Stitch UK Ltd)
A typical repair uses the lock plate and stitching stud (see Figure 5.11). The locks are first
installed normal to the crack and the stitch holes are then drilled along the length of the
crack between the locks (see Figure 5.12).
Figure 5.12
Example of metal stitching of cast iron tunnel lining
Stud sizes can range between 5 mm and 30 mm diameter, depending on the width of the
crack and are drilled the full depth of the material to be repaired.
The result is a completely watertight repair with the ability to take induced stress and
prevent further cracking.
CIRIA C671 • Tunnels 2009
165
5.4.2.2
Wrought iron and steel repairs
Welding wrought iron and steel
Welding is generally the most effective and practicable option for the repair or
strengthening of steel used in tunnel linings, and with some limitations (Bussell, 1997) can
be used on wrought iron, although some asset owners do not condone this. Welding in situ
may be carried out using electrical arc welders or oxy-acetylene gas welding techniques.
The choice of welding technique in an underground situation will depend on issues such
as the materials to be welded, access, environmental conditions and tunnel owner’s or
asset steward’s stipulations (health and safety/operational constraints). If welding is to be
carried out, consideration should also be given to the practicalities of welding in situ,
including the risk to tunnel services that may be present. For example, cables may require
removal or displacement which, if old themselves, could cause damage to cable sheathing.
The use and/or storage of welding equipment and gas bottles underground, welding in
confined spaces and possible risk of fire are other issues that will need to be taken into
consideration.
If water is present it may not be possible to weld using conventional equipment without
the water being controlled and kept away from the area being welded. Other precautions
to be taken include the removal of toxic materials that may be present and affected by the
welding. This may include the presence of lead paint and plastic coverings that would
emit potentially noxious gases and be a health and safety risk.
When deciding on the correct welding technique it is important to consider the age of the
structure. Also, in certain circumstances wrought iron may not be distinguishable from
steel without proper analysis. So specialist advice should be sought. For welding to be
successful consideration needs to be given to the chemical composition, mechanical
properties and metallography of the metal being welded, combined with the appropriate
choice of welding electrodes and welding practice.
In principle (Bussell, 1997):
wrought iron can be butt-welded successfully, but fillet welds are likely to fail by
lamellar tearing
with the right choice of welding materials and procedures old steel can be welded
successfully.
Old or early steel is considered to be steel manufactured before 1906, the date when BS
15 (BSI, 1908 revised 1962) was first introduced for the specification of structural steel for
bridges and general building construction. It was superseded by BS 4360 (BSI, 1968).
During the repair the quality of the weld should be continually checked. The final weld
should be smooth with no notches. If necessary, grinding should be undertaken to
maintain smooth flowing contours. On completion of the repair the weld should be
inspected for smoothness and quality. Examinations for the presence of defects should also
be carried out using NDT methods, such as ultrasound and/or magnetic particle
inspection. Typical acceptance criteria that may be applied to NDT can be found in BS EN
1011-1 (BSI, 1998b) and BS EN 1011-2 (BSI, 2001b).
Detailed guidance on site welding iron and steel structures is given in SCI (2002).
166
Figure 5.13
Strengthening repair of buckled steel section lintel used in an opening of a
cast iron lined tunnel due to structural defect (courtesy Tubelines Ltd)
Bolted repair to wrought iron and steel
Repairs and strengthening of wrought iron and steel are often carried out using bolted
connections, for example, bolting flange plates to the underside of beams. Connections
between components are made using high strength friction grip bolts in accordance with
BS 4604 (BSI, 1970). The design of connections should be in accordance with BS 449
(BSI, 1969).
For the connection to work, the high strength friction grip bolts are tightened to a
specified minimum shank tension so that the load is transferred across the joint by friction
rather than by having the bolt working in shear. When using friction grip bolts to make
connections to wrought iron, care is required to check for bearing due to the lower yield
strength of the iron. The bolts are usually tensioned to the required load using a torque
wrench or the use of load indicating washers, which provide a fool-proof and economical
solution.
5.4.2.3
Alternative repair solutions
One alternative repair solution is the use of composite carbon fibre plates that may be
attached to the structural members to strengthen them. Recent strengthening work using
this method has been successfully carried out on cast iron beams in a jack arch tunnel
lining found on London Underground subsurface lines, where composite carbon fibre
plates were glued to the underside of the beams. The advantages of this system of repair
included reduced man-handling requirements due to the plates being very light in weight
compared with steel plates and speed of installation. Similar repairs on other projects
using steel plates glued to structural steel and reinforced concrete have also been carried
out.
For further information on plate-bonding and other types of repair to metallic elements,
see Tilly et al (2007) and Cadei et al (2004).
Replacement of cast iron linings with stainless steel has been carried out near Old Street
Station on London Underground, where the original cast iron linings had deteriorated
CIRIA C671 • Tunnels 2009
167
sufficiently from aggressive groundwater to warrant replacement. This form of lining is an
exceptional case and would not usually be considered due to cost. Details of this work is
given in Case study A1.9 on Old Street Tunnel.
5.4.3
Concrete tunnel linings
There are a range of methods available for localised or more general repair and
improvement of concrete in tunnel linings. The most appropriate method will be chosen
based on factors similar to those of concrete repair (for example, the cause and nature of
deterioration), and in tunnels the practical problems associated with carrying out the
repairs is often particularly important.
Concrete repair usually involves the removal and replacement of damaged or defective
concrete with one of a variety of repair materials, or using extra materials on an existing
concrete element. However it also covers using a wide variety of other techniques to
improve durability and slow deterioration where it is already occurring such as the
application of protective coatings, the use of corrosion inhibitors and the installation of
systems for cathodic and sacrificial anodic protection.
The cost of repairs often depends on the provision of access and labour, with the cost of
materials less significant. The selection of the most appropriate techniques and materials
should not be compromised for minor cost savings.
A good understanding of the condition of the deteriorated element, the cause and severity
of deterioration is a prerequisite for successful repair. The most common cause of
deterioration in reinforced concrete is reinforcement corrosion, but deterioration can be
caused by a range of other factors including physical impact, chemical attack or fire.
Remedial work will have one of four aims:
1
Where deterioration has not started but the risk of deterioration or insufficient
durability has been identified, preventative work may be carried out to improve
durability and delay damage.
2
Where reinforcement corrosion has been initiated (detected by testing) but concrete
deterioration has not become manifest, measures can be taken to reduce corrosion
rates and delay the onset of damage.
3
Where deterioration is continuing and damage has occurred, it can be repaired and
extra measures taken to arrest or reduce the rate of future deterioration.
4
To reinstate or improve lining strength, either to remedy the effects of deterioration
or to provide further structural capacity to an under-strength element.
Table 5.3 illustrates the range of methods available for use in each of these four scenarios.
168
Table 5.3
Principles and available methods for prevention and repair of deterioration to structural
concrete
Purpose
1
Preventative
measures
Principle
Available method
Improve inadequate concrete durability
2
Corrosion
minimisation
To reduce the rate of continuing
reinforcement corrosion and delay the
onset of concrete damage
surface treatment/impregnation
surface coatings
migrating corrosion inhibitors.
migrating corrosion inhibitors
cathodic protection (impressed current or
sacrificial anode)
realkalisation
chloride extraction.
Repairs to
damage
To reinstate physical damage (whether
caused by corrosion, physical impact,
chemical attack or fire) and minimise
continuing deterioration
3
concrete repair/reinstatement (patch repair,
sprayed concrete, flowable concrete, recasting
with formwork) combined with preventative
methods as in 1 and 2.
Structural
strengthening
To reinstate or improve element capacity,
either as a response to loss of strength
through deterioration or to changes in
requirements
4
replacement of structural steel and recasting
concrete
plate-bonding
increasing section by adding concrete or mortar.
While a detailed review of concrete repair methods is beyond the scope of this publication,
an overview of concrete repairs and other types of repair (including preventative,
strengthening and enhancement methods) is given in the following sections, as they apply
to tunnel repairs.
5.4.3.1
Concrete repairs
Concrete repairs may involve the replacement of existing damaged or defective concrete,
or the use of further concrete to increase the section of an element. Where local repairs
are required, the former is often the most appropriate method, involving the application
of patch repairs by hand. Where more general, widespread repair or structural
improvement is necessary, the use of sprayed concrete is preferred, although the
associated reduction in the internal cross-section of the tunnel may not be acceptable in
some situations.
In carrying out any repair the objective is to provide adequate protection to the existing
lining components to enable the lining to function as a load bearing structure. Unless load
is removed from structural elements before repair, the repair will only contribute to
resistance of extra loads. The ability of the repair to resist loading will depend on a range
of characteristics including its compressive strength, elastic modulus and bond strength
with existing concrete. Differential shrinkage and creep should also be considered.
Suitability and compatibility of materials and their influence on the structural behaviour of
the lining elements is an important consideration.
Reference should be made to BS EN 1504, the comprehensive new British Standard for
products and systems for the protection and repair of concrete structures. This standard
includes 10 parts:
Part 1 includes definitions
Parts 2 to 7 include product specifications
Part 8 addresses product conformity
Part 9 provides a methodology for assessment and repair selection
Part 10 addresses site execution and quality control.
CIRIA C671 • Tunnels 2009
169
Parts 9 (BSI, 1997) and 10 (BSI, 2003b) are of particular relevance to those determining a
strategy for and executing repairs on concrete structures.
Where repairs involve the removal of material from structural members, an assessment
should be made of the extent to which the tunnel’s structure will be affected. Any
necessary temporary support and/or limitations on extent of repair at any one time should
be specified.
Local repairs (patch repairs)
Patch repair involves the local removal and replacement of damaged or defective areas of
concrete, usually those that are cracked and spalling. This can be achieved using
mechanical breakers but increasingly (and preferably) by hydro demolition for all but very
small repairs. The boundary of the repair should ideally encompass all defective material
and extend a short distance into sound concrete. In chloride-contaminated concrete it is
common to specify removal of all concrete containing more than 0.3 per cent total
chloride, although in practice it is sometimes necessary to relax this figure to avoid
excessive replacement. Generally it is best to remove concrete from behind reinforcement
to allow the latter to be completely encapsulated by the repair material, particularly where
chlorides are present. Saw-cuts are then made at the boundaries of the repair to provide
clean and well-defined edges. Any reinforcement within the repair should be surfacecleaned and treated according to the requirements of the repair material. Where
corrosion has resulted in significant loss of section in structural steel, it may be necessary
to add reinforcement to reinstate capacity.
There are a wide variety of generic and proprietary concrete repair materials available
and selection of appropriate materials is an important factor in achieving a successful
repair. This should be based on a consideration of the function of the repair (for example,
whether structural or non-structural, cosmetic, or to protect reinforcement from chlorides
or carbonation), and any constraints (for example, on the size and shape of the repair, its
location, or on the period available for application or hardening/curing). Compatibility
between the physical and chemical properties of the repair material and its substrate is an
important factor, in particular shrinkage/expansion, strength and stiffness and coefficient
of thermal expansion. Where repairs are extensive or performance particularly critical,
trials and testing should be considered.
As well as plain cementitious materials (concretes and mortars) a wide range of polymermodified cementitious materials and resin-based materials are available, which can be
used to meet specific repair requirements. A consideration of the characteristics of these
materials and their suitability for differing applications is beyond the scope of this
publication, and specialist advice should be sought when selecting and specifying repair
materials. Further guidance is given in BS EN 1504-3 (BSI, 2005d) and some information
on the principal methods and materials is included in Table 5.4.
170
Concrete repair methods and materials
Flowable grout
or concrete
Repair mortars are most commonly used for patch repairs with limited size, less than 1 m area and up to
about 30 mm in thickness. For thicker repairs and of use in larger areas, proprietary repair concrete or
design mix repair concrete may be used.
Generic Portland-cement based materials can be used but need to be compensated for shrinkage. Bond
strength may be limited (though this may be improved by making mechanical connections with sound
concrete or by cutting back beyond reinforcement). Depending on the repair location and dimensions it may
be difficult to place and compact.
More expensive, proprietary polymer-modified cementitious materials are frequently used where there are
particular requirements for high bond strength, good chemical resistance, low permeability, rapid setting and
curing and limits on the thickness of repair. They can also be applied to vertical or overhead surfaces without
formwork and built up in thin layers to form a thicker repair, although this can introduce discontinuities that
potentially reduce durability. Resin-based repair materials will degrade at much lower temperatures than
cement-based materials, so caution is needed where they are required to reinstate structural elements that
require fire resistance.
For large repair areas flowable grout or self-compacting concrete can be used and does not require
compaction/vibration in situ. These materials are carefully poured into formwork to produce the required
repair shape and are useful for rebuilding badly damaged elements.
Sprayed
concrete
Hand-placed mortars or concrete
Table 5.4
Sprayed concrete is frequently used for tunnel lining repairs where it is necessary to replace large areas of
defective concrete or to thicken and strengthen existing structural elements. It has the advantage that it can
be applied rapidly over large areas in a single operation. The concrete can either be dry mix or, for reduced
rebound, wet mix and can be modified (for instance, by the incorporation of polymers) to meet performance
requirements. It is generally considered good practice to include a light steel fabric within the repair. Skilled
and experienced operatives are required and it is not feasible to produce very thin (<100 mm) layers or to
treat small areas.
For further guidance refer to TRL AG43 (TRL, 2002) and the Concrete Society TR38
(Concrete Society, 1991).
Local patching of deteriorating concrete does not protect the area beyond the repair, and
may even lead to accelerated deterioration of adjacent areas (through the incipient anode
effect). Depending on the original cause of the problem, this approach often requires an
acceptance of further patching to treat future damage on an ad hoc basis as it appears. To
reduce future requirements, it may be useful to consider combining patch-repairs with
other types of preventative or corrosion-reducing treatment (discussed in Section 5.4.3.2).
Figure 5.14
Typical patch repair of pre-cast concrete tunnel lining
CIRIA C671 • Tunnels 2009
171
Crack sealing and repair
Before carrying out crack repairs it is important to ascertain the cause of cracking because
this can indicate structural problems and the need to carry out structural repairs rather
than simply sealing the cracks. Where tunnel linings have cracked due to continued
overstressing, (ie active cracking due to external effects such as increased ground loading)
then re-construction of the tunnel lining or strengthening, as discussed in Section 5.4.4,
may be necessary. Cracks in pre-cast concrete segmental linings that are inert, for
example, those resulting from damage due to transport or handling during construction,
or from shrinkage, and no longer showing signs of propagation, may be sealed to preserve
durability and/or prevent water seepage. Where cracking is the result of reinforcement
corrosion, the underlying cause should be determined and it may be appropriate to breakout defective concrete and repair using the patch-repair techniques discussed earlier.
Crack sealing and structural repairs are generally carried out by injection of a flowable
repair material through a series of injection ports inserted along the crack’s length.
Typically the injection is carried out under pressure. A wide variety of materials, including
SBR, acrylic and co-polymer emulsions, as well as various types of low-viscosity resins are
potentially suitable for crack sealing and repairs. Resin-injection is a specialist activity and
requires careful selection of materials and experienced specialist contractors for good
results.
Selection of a suitable material will depend on a variety of factors, including crack width,
the requirement for accommodation of future movement, the need for restoring loadtransfer, and whether any moisture is present in the substrate. Where considerable extra
movements are anticipated, crack-sealing materials are likely to fail and consideration
should be given to chasing out a groove at the surface of the crack and using a flexible
sealant capable of accommodating the movement.
If the crack is a source of water ingress, the crack should also be sealed by injection of a
suitable resin-based grout using those techniques described in Section 6.4.
Further guidance on the principles and methods for crack injection of concrete structures
are included in DD ENV 1504-9 (BSI, 1997), under Principle 1 (Protection against ingress
and waterproofing) and Principle 4 (Structural strengthening). Specifications for injection
products and systems are included in BS EN 1504-5 (BSI, 2004a).
Figure 5.15
172
Example of cracking in pre-cast expanded concrete tunnel lining
(crack highlighted by paint marking during inspection)
5.4.3.2
Other types of treatment and repair
Depending on the requirements, other types of treatment and repair can be carried out
independently of concrete repairs or in combination with them. They are used to slow the
progress of and/or minimise the effects of reinforcement corrosion and can be used as
preventative measures or, where deterioration has already occurred, to improve future
durability or structural performance.
Surface treatments
These fall into two main categories – coatings and impregnations which aim to limit the
ingress of deleterious substances such as chloride ions and carbon dioxide into the
concrete, and migrating inhibitors, which aim directly to reduce the corrosion rate of
otherwise depassivated reinforcement.
Because they can only readily be applied on the lining intrados, the usefulness of coatings
and impregnating treatments in tunnels depends on the source of the deleterious
substances. For example, they might be useful in protecting concrete from chlorides in deicing salts used within a highway tunnel, but will not be effective if the chlorides are
present in groundwater at the extrados. The majority of surface impregnators are based
on silanes, siloxanes and silicone resins and their hydrophobic action allows the concrete
to breathe while reducing the ingress of moisture and dissolved contaminants. These types
of material are useful in protecting against intermittent wetting and drying, but not
intended for use where water is ponded or under pressure. They have a considerable
history of use in highways structures and guidance on their use and application is given in
the BA 33/90 (HA, 2007).
For further information on surface treatments for concrete refer to the Concrete Society
(1997). Also guidance on the principles and methods for the surface treatment of concrete
structures are included in DD ENV 1504-9 (BSI, 1997), and specifications for products
and systems are included in BS EN 1504-2 (BSI, 2004b).
Another type of surface-applied treatment is the use of migrating corrosion inhibitors
(MCIs), which modify the chemistry of the concrete in contact with reinforcing steel to
inhibit steel dissolution and prevent or slow the progress of corrosion. They are soluble
salts that are applied in solution to concrete surfaces, but to be effective they should
diffuse through the cement paste in the vapour phase and accumulate at a suitable
concentration at the reinforcement surface. This presents problems where the concrete is
saturated above a certain level, or in relatively good quality concrete which is not
sufficiently permeable. This type of treatment is not yet able to demonstrate a proven
record of success but it may be of benefit in certain situations.
In tunnels, one advantage of surface treatments is that they are relatively rapid to apply,
although pre-cleaning may be needed to remove surface dust and grime from dirty
concrete surfaces. The health and safety implications of using chemical treatments in
confined spaces needs to be carefully considered and risks controlled through, for
example, suitable application techniques and assisted ventilation.
Cathodic protection (CP)
Cathodic protection is an electrochemical treatment that can prevent corrosion of
reinforcement in structures where the concrete is contaminated with chlorides (or, in
certain circumstances, where concrete is carbonated). It involves generating a small
electric current between the reinforcement (which acts as a cathode) and an anode.
Typically the anode is a conductive metallic coating or mesh at the concrete surface
CIRIA C671 • Tunnels 2009
173
through which a low voltage DC current is passed (known as impressed current cathodic
protection or ICCP). Less commonly, the anode is provided by an electrochemically active
metal (in sacrificial or galvanic anode cathodic protection, SACP) although this type of
system offers less control over the process. Embedded discrete anodes can be inserted into
holes drilled into the concrete surface, or included in repaired areas of concrete to
provide local protection to reinforcement.
The principal advantage of CP treatment is that only damaged areas of concrete need be
repaired, and there is no need to try to remove all chloride-contaminated (or carbonated)
material, which can sometimes be unfeasible. However, CP systems need to be custommade for individual structures, requiring careful design and installation by specialists
followed by a requirement for continuing maintenance and monitoring of performance.
Applied in tunnels, costs are likely to be significantly higher than for more accessible
structures but despite relatively high initial costs, in the long-term this type of protection
may be an economical alternative to repeated phases of local concrete repair.
Cathodic protection is dealt with in BS EN 12696 (BSI, 2000b) and further information
and guidance is given in the Concrete Society TR36 (Concrete Society, 1989a).
Chloride extraction and realkalisation
These are specialist treatments that can be used for treating chloride-contaminated and
carbonated concrete respectively. They are similar to cathodic protection because they rely
on applying a small electric current to the concrete. Both involve the installation of
external anodes at the concrete surface and associated electrical supplies, and require
careful monitoring during the treatment process.
Chloride extraction is a one-off treatment which attempts to remove chloride ions by
causing them to migrate out of the concrete into an external anode, under the influence
of an electrical field. Potentially this can effect a significant reduction in chloride
concentration in the concrete and restore passivity of reinforcement.
A similar impressed current method can be used to restore concrete’s protective alkalinity
to reinforcement, where this has been lost through carbonation. Realkalisation reverses
the fall in pH value which occurs on concrete carbonation, re-establishing the passivity of
reinforcing steel.
These types of treatment are not suitable for all situations, and there are limitations and
risks (the potential for causing hydrogen embrittlement of reinforcement and promoting
alkali-silica reaction of the concrete) associated with their use, which require careful
assessment in advance. They are specialist treatments and need to be installed and
operated by experienced specialist contractors. It is important to appreciate that they are
one-off treatments and once completed the ingress of chlorides/progress of carbonation
will begin again. Further measures may be appropriate to prevent similar problems
recurring within the required serviceable life.
Treatment times are usually between four to six weeks for chloride extraction and one to
two weeks for realkalisation (Broomfield, 2004). Access for installation, reduction in
clearances, provision of power and access for monitoring during and after application
(although this can be achieved remotely) are particular considerations for use in tunnels.
174
For further information, see Broomfield (2004), and the European Standard BS DD CEN/
TS 14038-1 (BSI, 2004d).
Plate-bonding
Plate-bonding techniques, using either steel or fibre-reinforced composite materials, can
be used to strengthen structural elements against bending or shear forces. They may be
used in combination with conventional concrete repairs to restore capacity to damaged
elements. These techniques rely on bond strength with the original concrete or repairs, so
careful inspection of the substrate concrete to identify weakened or delaminating areas is
vital. Remedial work should follow to address any inadequacies and careful preparation of
the substrate surface before applying adhesives and bonding. Further guidance is given in
the Concrete Society TR55 (Concrete Society, 2004). The principles and methods for
plate-bonding of concrete structures are included in DD ENV 1504-9 (BSI, 1997), and
specifications for products and systems are included in BS EN 1504-4 (BSI, 2004c).
In tunnels plate-bonding techniques have the potential advantage that they can increase
strength and resistance to certain forces without significantly adding to lining thickness
and reducing clearances within the tunnel. They may also be useful where strengthening
using sprayed concrete is unfeasible for this reason.
5.5
STRENGTHENING AND STRUCTURAL IMPROVEMENT
During the life of a tunnel it may be necessary to carry out more extensive repairs to
improve the condition of the tunnel lining and may involve the replacement or
strengthening of structural elements. Such works may be necessary because the tunnel
lining has deteriorated beyond the point where routine maintenance works can keep it in
an operationally safe condition, where the loadings acting on the tunnel lining have
changed, or where a change of use is intended.
For many tunnel owners, routine maintenance works may come under annual budgets set
aside for the overall management of the tunnel. However, improvement works beyond
routine maintenance can mean greater expenditure, and the works themselves are likely
to have a greater impact on the operation of the tunnel. Complete tunnel closure is the
most efficient environment for carrying out such works, but may not be operationally or
economically viable. Careful planning and the selection of appropriate techniques and
methodologies is critical for major works of this type, where the cost of the physical works
may be less than the incidental costs associated with suspension or disruption of services.
Carrying out the work during partial or multiple short duration closures may be feasible.
5.5.1
Replacement and strengthening existing tunnel linings
Many options exist within a tunnel where the existing lining is severely deteriorated,
damaged or otherwise inadequate for resisting imposed loadings. These could include the
replacement of the existing lining with a new lining, strengthening the existing lining with
the installation of a new permanent secondary structural lining, or the replacement of
individual components of the existing lining (as with segmental lined tunnels). The most
commonly used replacement and lining strengthening techniques are summarised in
Table 5.5 with an indication of the likely application for different tunnel lining types and
operational use.
CIRIA C671 • Tunnels 2009
175
Replacement of the lining in its entirety may be the only viable option if there are
clearance restrictions that would not allow the installation of a secondary structural lining
or other support system that may interfere with the operation of the tunnel.
Where there is sufficient clearance, and the existing lining has deteriorated to such an
extent as to become inadequate, a more cost-effective option may be the installation of a
secondary, permanent structural lining installed within the existing tunnel. Secondary
lining may also be used to improve the structural capacity of the existing lining if
conditions affecting the existing lining have changed significantly and the existing lining is
operating at or close to its structural capacity. In this case the secondary lining may not
need to be designed to take the full, expected loading but could use the inherent
structural capacity of the existing lining.
The option to reline a previously unlined tunnel showing signs of instability may also be
available and the principles used in design should be the same as for a new tunnel.
When considering replacement or strengthening works as an option, the works
themselves are likely to have a significant impact on the operation of the tunnel. They may
involve complete closure, taking it out of service for the duration of the works.
Alternatively through the design of suitable temporary works the tunnel may remain in
operation during the works while providing protection to the tunnel users and workers
alike, thereby minimising closures. Other considerations when planning the works include
the need to remove or reposition important services and live cables that may be present,
requiring a long lead-in before starting the tunnel remediation works.
The method or methods employed to replace or strengthen a tunnel will depend on the
type of tunnel lining and its particular function. The tunnel owner or asset steward may
also prefer to employ specific techniques that may be subject to whether the works are to
be carried out by the tunnel owner’s or asset steward’s own work force, or are to be
contracted out under a specific design or design and build contract. Whatever the
technique used the following should be considered during the design and implementation
stages:
176
foundations of the existing lining may require checking for the new loading
arrangement. New strip or raft foundations may be required to support the new
lining system
proposed new foundation arrangement may be affected by the presence of existing
buried services beneath the structure
the presence of extensive tunnel services within the tunnel. Significant and disruptive
works may be required to move or alter services to allow the replacement or relining
of the tunnel
temporary stability of the lining during works should be considered at the design and
construction stages, allowing for excavation for installation of new foundations,
installation of a replacement lining or strengthening an existing lining
areas of extensive bulging (in brick or stone masonry linings) may need to be removed
before the installation of any relining works. Temporary support of the area may be
necessary where the lining has been removed before construction of the new lining
provision should be made to channel groundwater that may be seeping though the
existing lining, either temporarily or permanently, in the form of weep pipes, ducting
and/or waterproof membranes
the appearance of the structure from within may be significantly affected, but unless
the tunnel is of historic importance is unlikely to be a concern
when relining to strengthen, inspection of the existing structure will not be possible
following installation of the lining system. But this is not important if the existing
structure becomes redundant
the design life of the replacement or strengthening works may be specified by existing
standards or should be determined by the requirements of the asset owner/steward
environmental conditions affecting the replacement or strengthening works, either
external or internal, corrosion protection from aggressive ground water or use of
road salts during winter
specialist fabricators manufacturing either cast iron, steel or concrete liners (or
alternative lining material types) will need more lead in time scheduled into the
scheme programme
environmental impacts.
The following sections discuss further considerations for more commonly used techniques
in either replacing or strengthening existing tunnel linings.
5.5.1.1
Replacement of tunnel lining
Replacing an existing tunnel lining with a new lining should only be considered if the
existing lining is beyond repair and there are severe clearance restrictions that would
prevent installation of the secondary structural lining or other support system within the
existing tunnel. Replacement may be the only option available if the existing tunnel lining
is functioning below the required capacity or the clearance needs of the tunnel are to be
increased, requiring over excavation of the existing tunnel profile.
When considering replacing an existing lining, it may be an option to remove the entire
lining and replace with a new lining, or structural components such as foundations or
structural inverts (as in the case of non-circular tunnels) may be re-used with the new
lining built upon them. The latter can only be considered if the existing foundations and
invert are sound and have sufficient structural capacity to take expected loadings from the
new lining.
An important consideration when replacing an existing lining is that the ground may
require continuous temporary support during removal before installing the replacement
lining. In these circumstances there are advantages in using segmental lining techniques
which limit the amount of open, unsupported ground and can provide almost immediate
support and protection to workers carrying out the works.
An example of this solution is the refurbishment of the Blisworth Canal tunnel in 1984,
replacing 900 m of severely distorted masonry lining with a pre-cast concrete segmental
lining. The method adopted was to replace the 5 m wide by 5.5 m high, 0.5 m thick brick
lined tunnel with a 6 m internal diameter pre-cast concrete segmental lining. This was
achieved by the hand excavation of an 8 m diameter by 8 m long chamber at one end of
the 900 m long section in which a tunnel shield was erected. The shield was then driven
along the tunnel alignment with the brick lining ripped out by an excavator mounted
within the shield. The new segmental lining was erected immediately behind the tail of the
shield.
CIRIA C671 • Tunnels 2009
177
5.5.1.2
Tunnel strengthening
Many techniques exist to strengthen an existing tunnel lining either over the entire length
of the tunnel or over discrete sections. The options available may depend upon the size of
the tunnel and availability of man or machinery. The main reason for selecting
strengthening works is that the existing lining may have inadequate structural capacity
and is exhibiting significant or advanced signs of distortion or deterioration.
Techniques to strengthen an existing tunnel lining may include the installation of a
secondary, permanent lining constructed in situ within the tunnel using for example,
reinforced or unreinforced cast in situ concrete or sprayed concrete, or the installation of a
prefabricated lining system. The main disadvantage with relining is that the structural
clearances within the tunnel will be reduced, impacting on the operational requirements
for the tunnel. But if the reduction of clearance is not important, then relining may be
preferable. Generally it provides the least disruption and is the quickest option available
when compared with replacing the original lining. In many circumstances in the past,
relining has been successfully carried out while the tunnel has remained in operation.
Another consideration is that strengthening works may permanently conceal the original
lining, which may be important if dealing with a tunnel of historic importance or a listed
structure (for example, the renovation of Brunel’s Thames Tunnel by London
Underground, see Case study A1.3). With the existing lining being covered there may be
no future opportunity available for inspection of that lining. However, if the relining is
designed to completely replace it, the need to inspect the original lining would be
reduced.
Depending on whether the condition of the existing lining is sound and provides an
element of structural capacity, a secondary lining need only be designed to improve the
existing lining giving extra support and reducing the structural thickness of the relining
to a minimum.
Included in this section is the use of ribs with possibly some form of lagging to support a
tunnel lining showing signs of unacceptable distortion. Ribs and lagging may be used in
an emergency or as a temporary measure to provide support while other permanent
works are being done. They may also be used as a permanent solution but this may need
to be restricted due to the limited design life and nature of the materials being used, eg
mild steel ribs and timber lagging. If used as a temporary or permanent solution, as with
any other material or technique with a limited design life, an inspection programme
should be done to monitor the lining for signs of distortion or deterioration on a suitably
frequent basis.
Ribs, steel sets and lattice girders are also commonly used in conjunction with sprayed
concrete to provide a temporary or permanent in situ tunnel lining (see Figure 5.16).
178
Figure 5.16
Use of ribs and sprayed concrete to provide a secondary lining (Richards and Haider, 1987)
The use of ribs, steel sets and lattice girders with or without lagging is not restricted to use
in previously lined tunnels but can be used in unlined tunnels as a form of primary lining
to provide structural support.
Prefabricated tunnel liners are also included in this section as they provide a secondary
support mechanism that can be installed within an existing tunnel showing signs of
distress or deterioration. The ideal situation for use of prefabricated liners is where a
reduction in clearance within the tunnel is acceptable and where access is limited, time
consuming, expensive or impractical. Although the existing structure is often ignored in
the design of the liner, liners may also be installed as a pure strengthening measure taking
into account the residual strength of the existing lining.
Prefabricated systems include steel or pre-cast concrete liners, which are installed in
immediate contact with the intrados of the existing tunnel lining. Any gaps between the
liner and the existing structure are grouted to provide continuous support to the original
lining. Other types of liners available include glass reinforced plastic (GRP) and glass
reinforced cement (GRC) panels.
The new linings may be designed to be free standing, taking the imposed loading within
the arch of the lining. Alternatively they may be designed to have an extra means of
support, such as anchors tied back within the parent rock or soil surrounding the tunnel
or structural ribs on the intrados of the new lining.
Slip-form liners have been developed and are used in the water and sewage industry to
reline pipes that require rehabilitation. Several techniques are available and some are
applicable for use in man-entry pipes up to about 2.5 m diameter. Those applicable for
this size of pipe or small diameter tunnel include:
continuous pipe liners – the installation of a continuous length of liner pulled through
the existing pipe. A loose fitting liner that is sometimes grouted around the annulus
to provide support
CIRIA C671 • Tunnels 2009
179
discrete pipe liners – short sections of pipe installed within the pipe to form a
continuous lining
spiral wound liners – ribbed plastic strips either spirally or helically wound to form a
continuous liner held in position by either the expansion of the liner or grouted in
place.
Materials used to form the pipe liners are generally polyethylene or polypropylene
plastics, or in the case of discrete pipe sections glass reinforced cement (GRC), glass
reinforced plastic (GRP), concrete or plastic reinforced concrete (PRC).
Further guidance and advice on renovation techniques used in sewers can be found in
(Read, 1996) and Atkinson (2000).
5.5.1.3
Replacement of structural elements
If a structural element such as an individual segment or series of segments forming a ring
of a segmental lined tunnel, suffers damage or deterioration, it may be replaced with
identical or similar units.
In some cases this may not be easily achieved as the rings may have been compressed
during construction giving a tight interlocking structure. The joint surfaces may also be
corroded as in the case of steel or cast iron linings, or designed with a step or other
interlocking joint system. However, where it is considered possible and practicable,
replacement with a segment piece or structural element of the same design as the original
will return the lining to its intended load capacity.
When carrying out works to replace structural elements several factors should be
considered:
180
replacement should be carried out in small areas only in a systematic manner to
ensure stability of the intact lining. Temporary formwork or propping may be
required to support the lining around the section being removed
with a segmental lined tunnel, care should be exercised when removing or breaking
out the defective segment pieces so as not to damage the adjacent sound segments
the replacement of segments in a segmental lined tunnel, this should be done on an
individual basis, ideally with only one segment piece removed at a time
removal of an entire ring of a segmental lined tunnel may require temporary ground
support or ground treatment to prevent collapse of the ground, depending on the
nature of the ground being supported by the tunnel.
Table 5.5
System/
technique
Summary of tunnel lining replacement and strengthening techniques
Description
Application
Engineering/
implementation
aspects
Tunnel use
Segmental
linings
(includes cast
iron, steel –
including
stainless steel,
pre-cast
concrete,
although other
composite
materials may
be used such
as GRP)
Replacement
and
strengthening
replacement of
existing lining, the
existing lining is
removed and replaced
with a new lining
designed to take full
expected ground and
groundwater stresses
strengthening of an
existing lining by the
installation of a
secondary structural
lining within the
tunnel, designed to
take full or partial
ground and
groundwater stresses
lining of a previously
unlined tunnel is
carried out as with a
new tunnel and
designed to take full
expected ground and
groundwater stresses.
Ms, Mt, C, U
Cast in situ
concrete lining
Replacement
and
strengthening
replacement of
existing lining, the
existing lining is
removed and replaced
with a new lining
designed to take full
expected ground and
groundwater stresses
strengthening of an
existing lining by the
installation of a
secondary structural
lining within the
tunnel, designed to
take full or partially
expected ground and
groundwater stresses
lining of a previously
unlined tunnel is
carried out as with a
new tunnel and the
designed to take full
expected ground and
groundwater stresses.
Ms, Mt, C, U
CIRIA C671 • Tunnels 2009
replacement technique ideally
suited to circular segmental linings
that provide continuous support to
the ground during installation
can be installed within an existing
segmental lining provided
reduction in clearance is
acceptable
designed to take full expected
ground and groundwater stress if
replacing or strengthening of an
existing lining, unless used to
improve the structural capacity of
an existing lining
All types of
existing unlined tunnels may
tunnels
require further excavation to
accommodate the new lining
can provide a watertight lining
solution immediately once installed
if used to partially replace the
existing lining, existing foundations
or inverts will need to be checked
for structural capacity for new
lining loadings
new lining is either expanded to
make positive contact with the
ground or existing lining, or grouted
to fill any void.
may be used in all manner of
applications where man or plant
access is available
used to improve structural capacity
of existing lining as well as
preventing continued deterioration
of the existing lining through
corrosion, erosion, chemical attack,
groundwater seepage etc
formwork required but can limit
application on awkward tunnel
profiles
may be reinforced (steel mesh or
bars) or unreinforced
use of existing foundations or
inverts will need to be checked for
structural capacity for new lining
loading (including loading from
temporary formwork)
All types of
may be used in conjunction with
tunnels
rock bolts or anchors etc to
improve structural capacity
may require contact grouting
between the existing and
secondary lining to overcome
shrinkage effects
generally lengthy process including
assembly of formwork and
concrete curing. However, pressure
placement systems are available
for high compaction and total void
filling, which when used with quick
and easily assembled formworks
can speed up construction time
areas of water seepage will need to
be controlled before placing in situ
concrete, although pressure placed
concrete systems can sometime
overcome these problems and
provide a water tight lining.
181
Table 5.5
Summary of tunnel lining replacement and strengthening techniques (contd)
Sprayed
concrete lining
Replacement
and
strengthening
replacement of
existing lining, the
existing lining is
removed and replaced
with a new lining
designed to take full
expected ground and
groundwater stresses
strengthening of an
existing lining by the
installation of a
secondary structural
lining within the
tunnel, designed to
take full or partially
expected ground and
groundwater stresses
lining of a previously
unlined tunnel is
carried out as with a
new tunnel and
designed to take full
expected ground and
groundwater stresses
Ms, Mt, C, U
Slip-lining
(includes
continuous
pipe, discrete
pipe and spiral
wound liners)
Strengthenin
g
Strengthening to an
existing lining that involves
pushing or pulling a new
lining into an existing lining
Ms, Mt, C
182
may be used in all manner of
applications where man or plant
access is available. Can be used in
non-man entry pipes using robotic
equipment
used to improve structural capacity
of existing lining as well as
preventing continued deterioration
of the existing lining through
corrosion, erosion, chemical attack,
groundwater seepage etc
may be reinforced (steel mesh or
fibre – steel, glass fibre or
synthetic fibre) or un-reinforced
may be used in conjunction with
lattice girders, steel ribs or sets,
rock bolts or anchors etc to
improve structural capacity
may be used in conjunction with
waterproof membranes to provide
watertight lining
use of existing foundations or
inverts will need to be checked for
structural capacity for new lining
All types of
loading.
tunnels
rapid set and early high strength
shotcrete available speeding up
installation and construction time
suited to awkward shaped tunnels
without the need for formwork
dry and wet mix applications are
available, the selection of which
may depend upon site conditions
(eg distance to work site from
batching location etc)
areas of water seepage will need to
be isolated and controlled.
rebound shotcrete will need to be
contained to aid cleanup process mitigated by the use of low
rebound shotcrete
small portable plant can be used
for mixing and application
robotic equipment can be used
which aids quality control,
minimises health and safety issues
and used in non-man entry tunnels
cost-effective and convenient
mainly used in circular tunnels or
large pipes
primarily used to line the tunnel or
pipe to prevent continued
deterioration through corrosion,
erosion, chemical attack, or to
prevent water ingress or leakage
not used to improve structural
capacity
loose fitting liners require annular
grouting but may have the
disadvantage of reducing the bore
or clearance of the tunnel or pipe
friction build-up with spiral wound
liners could limit installation length
access to end of pipe may be
required – may involve a temporary
lead-in trench to locate equipment
used to insert the lining
quick insertion
may be jointed into one long
continuous string or jointed before
insertion or within the tunnel or
pipe if space is restricted
can accommodate large radius
bends
any flow in tunnel or pipe will need
to be stopped before insertion
Mainly used in
the water and
sewer
industries but
could be
applied to
other,
generally
small
diameter,
circular
tunnels
Table 5.5
Summary of tunnel lining replacement and strengthening techniques (contd)
Ribs and lagging
Strengthening
installation of steel
ribs or sets of light
weight lattice girders
with or without
lagging (timber or
steel plate) to provide
support and restrain
movement of the
tunnel lining
Ms, U
not strictly a relining technique but
may be built within a tunnel to
provide extra support and help
restrain movement of the existing
tunnel lining
may be used as a temporary
support measure
lattice girders are easy to handle
and install compared to steel sets
made up from rolled steel sections
or square hollow sections
ribs need to be shaped to the
profile of the tunnel (easily done
with lattice girders but more
difficult with steel sections)
in most cases several ribs are
installed along the length of the
tunnel (spacing dependant upon
specific conditions) and can be
connected together with tie rods to
improve stability
may be used with some form of
lagging (typically timber but steel
plates can be used). Wedges are
used to provide continuity with the
existing lining
ribs and lattice girders can be
used in conjunction with shotcrete
as a secondary lining for increased
structural capacity.
Typically
used in
masonry and
unlined
tunnels
Notes:
Application for tunnel types: Ms=masonry linings (brick and/or stone), Mt=metal lining (cast iron/ steel), C=concrete lining,
U=unlined
5.5.2
Underpinning of masonry-lined tunnels
Underpinning involves the construction of new substructure supports under existing
brickwork or masonry lining foundations where the ability to transfer the imposed loads
from the structure to the formation has deteriorated or failed. This failure may be
attributed to structural defects, changes in loading regime, subsidence and time related
consolidation settlement of the ground below the foundations or a combination of these.
Underpinning could also be employed if there is a future risk of settlement from
proposals to increase imposed loadings, for example, the increase in ground loading due
to land redevelopment above the tunnel.
Before starting, the designer should be satisfied of the mode of failure of the formation
material that is causing settlement and/or instability. Key considerations and actions at the
design phase could include:
initial dimensional and structural surveys to establish location, extent and size of
structural deterioration/failure. Particular attention to be given to visual signs of
distress in the tunnel lining such as bulging, vertical stepped cracks and fractures
desktop study of structural records to determine the existing form of substructure
construction and available geological/geotechnical information, both to aid the
planning and specification of the site investigatory works, if required, and design of
the remedial works
identification of services and utilities likely to impact on the planning of any ground
investigation that may be required and/or the development of a permanent
underpinning solution
CIRIA C671 • Tunnels 2009
183
trial pitting to confirm the dimensions, construction and condition of the existing
substructures
if insufficient geotechnical information is available, do a ground investigation to
establish formation material type, depth to interfaces between varying materials and
to provide samples for laboratory testing
confirmation of precise location of buried services
confirmation of type, size and rate of structural failure by precise levelling and/or
from tell-tales
range of laboratory testing with required outputs for design including strength,
angles of internal friction and settlement/consolidation (eg tri-axial or shear box and
oedometer testing as appropriate for material type)
groundwater monitoring by piezometers should be considered to establish the
position of the natural water table and to allow hydrostatic pressure to be included in
design calculations, also the planning of construction and associated temporary works
chemical analysis of soil and ground water samples to be used for sulfate/acid resisting
mix designs, where required.
Once acquired the data should be presented in a geotechnical interpretive report. Along
with the desktop information, this will provide the basis for assessing the cause of existing
settlement problems and/or the potential for settlement in the future.
At an early stage following review of the geotechnical interpretive report the following
criteria should be considered when selecting the underpinning method:
depth at which the required bearing capacity can be established
nature of material to be excavated
total length of structure to be underpinned
available programme for completion of the works
physical site constraints (limited plan area, headroom etc).
There are no definitive rules that state the depth at which traditional strip footings and
piles should or should not be employed, as each site and the associated ground conditions
may be judged individually. Also, the following criteria will assist in reaching the most
appropriate solution:
where consolidation settlement is occurring in a clay horizon, strip footings alone
would be ineffective as the loads will merely be transferred to a lower level permitting
the cycle of settlement to begin again
underpinning measures should ideally be taken down to relatively incompressible strata
true actual cause of settlement should be established (eg clay shrinkage, de-vegetation,
subsidence, overloading).
Note that where the failure is deep seated due to mining subsidence or other
circumstances, then traditional methods of underpinning on their own are unlikely to be
effective without further extensive remedial works to stabilise the ground.
The following fundamental criteria should be considered at the design stage:
184
imposed loading from structure
live and imposed ground loading
exposure conditions and cover to reinforcement used in the underpinning works
chemical composition of formation stratum and ground water to allow specification of
durable chemically compatible mix
expected design life of the remedial measures
connectivity between eccentric pile caps or strip foundations and foundation structure
where continuous strip foundations are to be used the designer should specify the
maximum length of foundation that can be underpinned at any one time, and should
also indicate a hit and miss sequence for the construction
the designer should develop a reinforcement schedule that will achieve structural
continuity in conjunction with the intended construction sequence
consideration should be given to the specification of mixes that are capable of
achieving high early strength, allowing installation periods to be reduced, especially
when working under closures of limited possessions
for piled solutions the designer should specify the length and diameter of the pile that
will be governed by the structure loadings, allowable bearing pressure of the
formation, end surface area and shaft surface area
the designer should provide details for the repair of any cracked brickwork or
masonry tunnel lining that occurred as a result of historic settlement.
Other forms of underpinning techniques include ground improvement by chemical or
cement injection. Again these may be carried out either from outside the tunnel or, more
commonly, from inside. For guidance on ground treatment methods see CIRIA C514
(Rawlings et al, 2000) and CIRIA C573 (Mitchell and Jardine, 2002).
Following completion of the works a monitoring regime should be done to measure any
continuing settlement or movement of the tunnel lining. The result from the monitoring
regime should be reviewed at intervals over a period established by the design engineer to
ensure that the underpinning measures have been effective.
The most common form of underpinning techniques employed to remedy failing or failed
substructures include mass concrete continuous strip foundations and piles. These are
briefly described in the following sub-sections.
5.5.2.1
Continuous strip foundations
This method involves the excavation of working pits longitudinal to the foundations being
underpinned at predetermined intervals that will not compromise stability of the
structure. The pits are filled with mass concrete to the underside of the existing
foundation and the process repeated at designed intervals until the desired extent of
underpinning is achieved. The process can then be repeated for the remaining length of
the structure to be stabilised in strict accordance with the specified sequence.
This method is considered suitable where:
the depth to competent formation is relatively shallow
material below the existing foundation can be feasibly excavated and adequately
supported with temporary works to prevent collapse
the water table is below the level of the proposed excavation
the structure is able to tolerate temporary undermining to form the underpinned
strip foundation.
CIRIA C671 • Tunnels 2009
185
The concrete should be placed as soon as possible once individual pits have been
excavated. The concrete is placed by pump and/or hand tools and compacted in the
normal manner using appropriate vibrators and agitators to the underside of the existing
foundations. At this stage it is important that the cast in situ concrete is left to set and
shrink (note that this may take time, during which the old foundation is not supported).
Once setting and shrinkage have taken place the void between the cast in situ concrete and
the existing foundation can be filled with a suitable non shrink grout. Pressure grouting
using a cementitious grout could also be employed to ensure that all voids at the interface
between old and new structures are removed.
Following completion the shutter can be removed and the working pit backfilled with a
specified fill using appropriate means of compaction.
Other key aspects to note in the continuous strip foundation underpinning technique
include:
5.5.2.2
development of approved design for the temporary support of the working pits. For
guidance see CIRIA R97 (Irvine and Smith, 1983)
planning of hit and miss sequence for excavation and installation of underpinning
concrete foundations
careful excavation of working pits and portion of failed formation underneath
existing foundations to proposed new formation level
health and safety aspects of excavation using hand-held tools
shuttering is normally installed between working pit and proposed underpinning to
contain pour
concrete mix to be checked for compliance with specification via laboratory and field
testing.
Piling methods
Piling offers a suitable solution where the competent stratum is at a depth that would
preclude the use of strip foundations. Depending on ground conditions various piling
methods can be employed with cased or uncased bored piles being the most common.
Displacement piles are not normally favoured due to the potential for vibration and
displacement of the ground to adversely affect the structure being stabilised. Bored piles
are also favoured from the standpoint of accessibility as the installation rigs are capable of
operating in locations of low headroom.
186
Figure 5.17
Underpinning a tunnel portal structure by piling to prevent structural movement
The two main installation techniques that can be used for piled underpinning solutions to
successfully address problems of access are:
1
Where the existing foundation of the tunnel is of a significant width or there are
significant restrictions on the surface, it may be more practical to install mini-or micro
piles (commonly known as root piles) from within the tunnel. These piles can be
drilled using low percussive equipment from vertical to a steep raking angle through
the tunnel lining and foundations into the formation. Once formed, galvanised or
stainless steel bars are inserted into the holes and grouted. These piles are particularly
useful where limited space is available for installation and where ground conditions
prohibit hand excavation. They are also particularly appropriate if the area adjacent
to the base of the lining is congested with services and utilities.
2
The installation of piles from the surface with the pile cap formed adjacent to the
foundation in plan on the extrados of the tunnel. The piles are capped with a
reinforced concrete pile cap that extends under the existing foundation. This method
involves excavation to the underside of the foundation of the tunnel to form the pile
cap in a manner similar to the construction described for continuous strip footings to
ensure structural stability is maintained throughout construction. Care is required to
be able to locate the line of the tunnel linings by precise surveying methods and strict
control on verticality during installation.
When carrying out piled underpinning solutions the major steps and points for
consideration during design and construction include:
the appointment of competent specialist piling contractors at the earliest opportunity
during the design phase. By involving this expertise early the designer will be able to
specify the pile type with buildability input
through previous experience and expertise the piling contractor will be able to review
the site constraints and ground conditions to develop the most appropriate and costeffective means of installing the piles
physical site constraints may limit the available space for installation. The pile size,
numbers and arrangement will need to be designed around an installation rig that
feasibly can operate within the site constraints where early strength is required, or
where piling through loose material, it may be prudent to consider using permanent
CIRIA C671 • Tunnels 2009
187
casings that will provide improved structural performance and will ensure that pile
does not suffer localised necking during installation. If necking occurs, little or no
cover is afforded to reinforcement leading to potential corrosion and early failure of
the pile
in water bearing sandy soils boring should proceed carefully to prevent draw down of
sand into the hole. Pumping of ground water from the hole is not recommended as
this will lead to settlement caused by boiling at the base of the hole.
When this cantilevered pile cap method is employed the piles are installed to the designed
arrangement in advance of excavation for the pile cap. Once installed the piles are
exposed and reduced to underside of pile cap level. It is recommended that the pile is
dowelled with stainless steel kicker bars to provide durability in the structural connection
between the pile cap and piles.
5.5.3
Invert repair (strengthening/replacement)
Structural inverts to masonry and concrete lined tunnels are an integral part of the lining
and may be installed at the time of construction from brick, cast in situ concrete or laid as
pre-cast units. In shallow tunnels no structural invert may be present with the tunnel walls
founded on continuous strip foundations. Where the tunnel is located at depth the
structural invert may be required to accommodate uplift forces, although there are
exceptions, for instance in old brick lined tunnels through soft ground located above the
groundwater table.
Circular segmental lined tunnels are not considered in this section as they have no
separate invert, with the whole lining acting as one unit within the ring to resist stresses
imposed by ground and groundwater.
Inverts fail because of increased uplift forces, possibly from rising groundwater or longterm redistribution of ground stresses. Many early tunnels have suffered from flat inverts,
which have later failed. For example, flat inverts were used in canal tunnels to reduce the
amount of puddle clay that was used to hold water in the canal.
Where a tunnel invert has failed or is showing signs of excessive deterioration or distress,
either strengthening or replacement repair works will need to be carried out.
Invert repair work, whether routine maintenance, or replacement, or structural
strengthening works, is likely to be very disruptive and may severely affect the operation
of the tunnel through full or partial closure of the tunnel. It may be possible in certain
circumstances to affect repairs to the invert in stages on one half of the tunnel at any one
time, thereby keeping the tunnel partially open with minimal disruption to its operational
use. Canal and water or sewer tunnels will most likely require complete closure to carry
out invert repairs.
Recent repairs to failed inverts have been the complete removal of the original invert and
replacement with a new reinforced or mass concrete invert with a greater curvature than
the original invert to take expected radial and horizontal stresses. Earlier examples of
invert repairs in masonry tunnels included the removal of the original invert and
replacement with another masonry invert, also on a greater curvature.
Relining the existing invert with a reinforced concrete overslab may also be an option
provided there is sufficient headroom clearance to accommodate the new invert, and that
the foundations to the sidewalls of the tunnel can take the imposed uplift stresses.
188
a
Figure 5.18
Details of a replacement invert (a)
details of an overslab invert (b)
b
An important point to be considered at an early stage of any repair or strengthening
works to the invert of the tunnel is the control required to maintaining the stability of the
existing tunnel during the works. This may also be necessary when carrying out routine
maintenance works such as localised patch repairs to the invert. Temporary works to
support the tunnel may be required during excavation of the invert depending on the
extent of the repair being carried out. Excavation of the failed invert may need to be
carried out in transverse bays of restricted width (eg 2 m wide bays), with the full
reconstruction completed and specified concrete strengths achieved for each bay before
proceeding to the next bay. A programme of monitoring is recommended during the
repair reconstruction period to check the stability of the tunnel and adequacy of any
temporary support at each stage of the works.
5.5.4
Rock stabilisation: unlined tunnels
Remedial measures for unlined tunnels range from local removal of loose rock through to
construction of new structural linings. Measures include:
scaling of loose material and block removal (with consideration of the global
discontinuity patterns and potential for loosening further areas of rock on removal)
rockfall protection shelters (at portals and within a tunnel as a passive measure)
spot bolting of individual blocks and defects
localised pattern bolting
systematic rock bolts to reinforce zones of poor rock mass
rock bolts with rockfall protective mesh to protect against spalling and sloughing
between bolts
sprayed concrete for areas of progressive deterioration or as a structural lining
ribs or steel sets
lattice girders
cast in situ concrete lining.
CIRIA C671 • Tunnels 2009
189
The application of each of these methods is outlined in the following sections. The
detailed description and design of the support measures is beyond the scope of this guide,
refer to BS 8081:1989 (partially superseded by BS EN 1537: 2000), the DMRB (HA, 1999)
and Hoek (2007).
Further guidance on the description, design and use of rock reinforcement in
underground excavations and structures can be found in Douglas and Arthur (1983),
Chapter 14 of Hoek (2007), Hoek, Kaiser and Bawden (1995) and DMRB (1999).
Scaling
Scaling, or baring down removes loose surface rock in a controlled manner to prevent
material falling on tunnel users. Most loose surface material should have been removed at
the time of tunnel construction, but there may be occasions where this was not done
thoroughly and loose rock may remain. This is particularly true where careless blasting
practices were used. Loosening of the exposed rock surface may also occur through
weathering, water ingress and stress release.
Scaling should be considered only in places where the remaining rock will be sound and
the rocks to be removed are not providing any support. There is evidence that ad hoc
scaling that has been routinely carried out in the past may have caused further rock
instability. It is suggested that geotechnical observations and discontinuity measurements
be undertaken before any scaling works.
The tunnel infrastructure and services within the tunnel should be protected before
scaling takes place. The required frequency for carrying out scaling works often cannot be
assessed, but it can be undertaken as part of the inspection process for the tunnel. The
need for inspection and scaling should be considered with regard to any known history of
falling material and potential for damage or harm to users of the tunnel. During any
scaling operation, care should be taken to ensure that over scaling does not create
problems through loosening previously tight areas. Areas to be scaled should first be
sounded and scaling activities restricted to areas where the rock sounds hollow.
Scaling can be undertaken using picks or pry-bars from mobile scaffolds. Fallen material
can be collected from the scaffold platform for removal. The use of mechanical peckers
and remotely controlled plant should be considered to protect the health and wellbeing of
the operatives. Larger loose blocks can be carefully removed by prising down onto
platforms constructed close to the rock face to avoid them falling and possibly damaging
the floor or services in the immediate area.
Ideally tunnel floors should be cleaned of scaled material to ensure any future rockfall can
be noted from newly fallen debris (see Section 5.3.1 on tunnel cleaning) as an aid to
identify areas for scaling later.
Rock fall protection structures
Rock fall protection structures, in the form of canopies, are a passive measure aimed at
providing protection from falling material but not for supporting the rock mass in any
way. They should only be considered where there is an acceptance of continued fall of
rock and debris. There is also long-term maintenance responsibility involving the removal
of rock and debris that may accumulate on the structure. When designing a rockfall
protection structure, both impact loads from falling material and static load from the
accumulation of rock and debris should be considered.
190
For the majority of infrastructure tunnels the loss of clearance would be unacceptable and
the introduction of a rockfall protection structure may not be cost-effective when
considering the structure’s continuing maintenance.
Rock bolts, dowels and cables
Rock bolts and dowels have been used for many years to support underground
excavations and are probably the most frequently used rock support measure today. A
wide variety of bolt, dowel and cable types have been developed to meet specific support
requirement.
Rock bolts
Rock bolts typically consist of steel bars anchored and/or grouted into the rock mass with a
face plate and conical washer at the rock surface to distribute the load and support mesh.
Other materials used in the manufacture of rock bolts include glass reinforced plastic
(GRP). GRP rock bolts have the advantage of being very flexible and can be bent on a
tight radius for installation in small diameter tunnels. They may also have a longer design
life as they have a much better corrosion resistance than steel bolts. However, there may
be an issue with long-term creep. GRP bolts are also not capable of taking shear stress and
this should be taken into consideration when designing a rock support system. Steel bolts
can withstand some shear component, but where significant shear is anticipated, dowels
should be used.
Rock bolts are tensioned to provide active support to the rock mass. They can be used to
support individual unstable rock blocks or in wider areas installed in a pre-determined
pattern to reinforce the rock mass.
Figure 5.19
Support of unlined tunnels using rock bolts (after CIRIA, 1983)
CIRIA C671 • Tunnels 2009
191
Three types of rock bolt are commonly used:
1
Mechanically anchored bolts that include an expansion shell that tightens against the
drill hole wall to fix the bolt as illustrated in Figure 5.20 (a).
2
Cement grouted bolts that are anchored over a specified length at the end of the bolt
within the drilled hole and the remainder of the shaft of the bolt isolated to allow
tensioning, and re-tensioning if necessary, as illustrated in Figure 5.20 (b).
3
Resin bonded bolts where the end of the bolt is anchored by a fast setting resin to
provide the initial anchor to tension the bolt. The remainder of the bolt is bonded to
the rock by slower setting resin, as shown in Figure 5.20 (c).
The advantages and disadvantages of each type are summarised in Table 5.6.
Table 5.6
Summary of different rock bolt types
Bolt type
Advantages
Disadvantages
Mechanically anchored
Speed of installation
Economical
Limited to hard rock
Can work loose in weak rock
Cement grouted
Economical
Good corrosion protection
Difficult to install
Dependant on skilled operatives
Resin bonded
One step installation
Fast setting
Very strong bond under suitable conditions
Can be used in most rocks
Good for difficult access situations
Easy to install in up holes
Expensive
Can slip in clayey ground
Limited shelf life of materials
Hole diameter critical for good resin mixing
Corrosion protection not as good as cement
grout
Temperature sensitive
Rock bolts can be used to hold rockfall protective mesh to provide support to areas of
small loose blocks or spalling rock surfaces between bolt locations, and to provide greater
overall support. Steel straps can also be installed between bolts to provide extra support.
Dowels
Rock dowels comprise a steel bar grouted into a drill hole with a face plate and nut at the
rock surface. The entire length of the bar is grouted but no tension is applied. Dowels rely
on movement in the rock mass to mobilise load and so are usually only installed shortly
after construction of a tunnel. They are of limited use in remediation of ageing
infrastructure tunnels as the majority of movement will already have taken place and
further movement is likely to be undesirable. However, they can be used to help provide
support to localised defects within the rock mass and hold up rockfall protective mesh.
Two other types of dowel are available, both relying on friction produced directly between
the dowel and the rock without a grout or resin bond. Split sets are C-shaped tubes that
are compressed into slightly undersized drill holes. The spring action of the compressed
tube provides radial pressure to hold the dowel in place. Swellex type bolts are installed
into slightly oversized holes and then expanded against the drill hole wall using water
pressure. Both types are quick to install but corrosion protection is difficult, although
modern versions are plastic coated to provide protection.
Cable bolts
Cable bolts are used extensively in the mining industry, particularly where support is
temporary or sacrificial. They comprise long lengths of steel wire in various cross-sectional
shapes grouted into drill holes. Civil engineering applications are usually restricted to
192
large underground openings, for example, in hydroelectric power stations where the
larger spans usually require rock support to be taken further back into the rock mass. One
advantage of cable bolts for use in rock strengthening works within tunnels is that they are
flexible and can be used in applications with restricted access.
Sprayed concrete
Sprayed concrete has been used to support underground excavations since the 1930s but
has only come into prominence since the 1970s. Also known as shotcrete, sprayed concrete
comprises cement, sand and fine aggregate applied to the rock surface pneumatically and
compacted under high velocity. There are two principal forms:
Dry mix, where the components are mixed in a hopper and fed using compressed air to a
hose where water is added at the nozzle.
Wet mix, where all components are mixed in a large container and delivered to the nozzle
hydraulically.
Both forms of sprayed concrete have similar properties, although the dry mix equipment
is more compact and portable and may be more suited to small scale tunnel remediation
applications. Where large volumes of sprayed concrete are required, for example, if a long
length of tunnel requires treatment, wet mix sprayed concrete may prove more
economical.
Advantages in using sprayed concrete include:
formwork is not required, which can be cost prohibitive and impractical where
difficult access exists. Sprayed concrete operations can often be accomplished in areas
of limited access
adaptable in application to non-regular surfaces and tunnel profiles
where thin layers or variable thickness are required, or normal casting techniques
cannot be used
can be used in temporary or permanent support applications
excellent bonding of sprayed concrete to other materials is often possible
sprayed concrete is applied using pneumatic plant thereby producing good
compaction and penetration into surface irregularities
within limits, sprayed concrete is self-supporting and can be used in overhead
applications
cost savings are possible because sprayed concreting requires only small plant for
manufacture and placement compared to other support types requiring concrete
delivery
limited closure of the tunnel is possible because the plant used is small and mobile.
Often areas of work can be continued in later shifts, for example, if the tunnel is kept
open during the day or traffic hours, and the works are carried out at night in
engineering hours.
Sprayed concrete can be used in conjunction with rock bolts and rockfall protective mesh
or weld mesh and with ribs or structural steel sets to replace lagging to provide extra
support (see Section 5.5.2).
Recent developments include using silica fume, which acts as a pozzolan to add significant
strength to the sprayed concrete. Steel or synthetic fibres can be added to strengthen the
CIRIA C671 • Tunnels 2009
193
concrete and provide ductility. This can remove the need for conventional weld mesh
although this is still used where rock quality is very poor and sprayed concrete adhesion is
difficult.
The design of sprayed concrete support is imprecise, largely based on rock mass
classifications. Further guidance is provided in Hoek (2007).
mechanically anchored rock bolt
cement grouted rock bolt
mechanically anchored rock bolt
resin bonded rock bolt
Figure 5.20
Examples of different types of rock bolts
5.6
TREATMENT OF TUNNEL SHAFTS
Shafts, whether open, closed, in use or disused, need to be managed through inspection,
maintenance and remedial treatment in the same way as the main tunnel. Where they are
no longer in use and present a maintenance liability, they can be treated to make them
safe while minimising or eliminating any ongoing liability, for instance by capping and
filling.
Treatment of a shaft requires a study of the geology and hydrology of the area where it is
located to identify features that may be significant to its stability and the selection of the
most appropriate treatment method. Consideration should be given to the engineering
characteristics of soils and rocks around the shaft and the qualities, quantity, movement
and pressure of water within them (NCB, 1982). The nature and condition of the shaft
lining and any existing capping and/or infill should also be determined. Wherever filling,
capping or other treatments that can impose extra loading to any connecting tunnel lining
194
are to be considered, the load distribution and adequacy of the support provided by the
tunnel lining should be checked, including the details of its structural connection with the
shaft. The most appropriate method of treatment will require consideration of all these
factors alongside asset owner policies, aims and future strategy.
More detailed advice on the treatment of shafts and the development of land above them
is included in Healy and Head (1984) and in NCB (1982). The latter document also
includes details of the theoretical behaviour of shaft fill material and is now under revision
by the HSE to provide more up-to-date guidance. In the meantime, HSE publication The
design and construction of water impounding plugs in working mines (HSE, 2003)
includes information on design and construction considerations that are relevant to shaft
treatments.
The following sections provide a discussion of access methods for working on shafts, and
techniques for their maintenance, repair and sealing/filling.
5.6.1
Access for working
Access systems for working within shafts vary according to the nature of the shaft and its
environment, and the specific requirements of the work. The principal factors influencing
the selection and design of such measures include:
shaft geometry and internal diameter, layout, depth and verticality
the presence of obstructions (steelwork, garland drains, downpipes etc)
the scope of the remedial works
safety requirements.
Suspended platforms are frequently used for work within shafts. There are many specialist
access equipment suppliers who have the expertise to assist with the design and
fabrication of such systems, and they should be compliant with the Lifting Operation
Regulations and Lifting Equipment Regulations 1998 (LOLER) (HSE, 1998a) for manriding platforms and the Provision and Use of Work Equipment Regulations 1998
(PUWER) (HSE, 1998b). The design should be suitable for transporting staff, equipment
and materials to the necessary locations within the shaft.
Desirable features for such platforms include:
a minimum of two suspension ropes and winches
winch controls located on the platform for use by the crew
a service hoist for delivery of materials and removal of spoil
adjustable guide wheels for stability against the lining
guards/toeboards/skirts
a fall arrest system
provision of emergency access and egress measures.
Other regulations, such as the Confined Spaces Regulations (HMSO, 1997), may specify
further requirements on detailed working methods and safe access, egress and emergency
provisions.
CIRIA C671 • Tunnels 2009
195
5.6.2
Shaft lining maintenance, repair and decommissioning
Maintenance and repair techniques for shafts are similar to those used for the lining of the
main tunnel, as discussed in Sections 5.3 and 5.4. However there are some issues and
methodologies that require special consideration.
Methods of treatment to ensure the safety of shafts include:
regular inspection, maintenance and repair of the existing lining
in situ capping or employing a prefabricated cover to prevent subsidence
consolidation of the shaft by filling and/or grouting
consolidation and capping
filling and plugging using cement-based grout.
The types of repairs envisaged will influence the specification and design of the access
system required (see Section 5.6.1).
The design and construction of long-term remedial measures such as caps, plugs and shaft
fillings should follow best civil engineering practice, and with a step by step approach,
involving people with a range of competencies, to build a safe structure that will remain
secure for the life of the tunnel. Health and safety of those involved in investigation and
construction works must be carefully considered and adequate measures taken to identify
and control the associated risks.
5.6.2.1
Deteriorating cross-members
If present, deteriorating beams, decking, cross-members etc may need reinforcement,
removal or replacement, particularly where structural timbers are found because these rot
and disintegrate over time, potentially resulting in instability. Internal elements may also
hinder access to other parts of the shaft lining for inspection or repair.
5.6.2.2
Water ingress
Vertical shafts intercept permeable strata above a tunnel, so they frequently act as points
of water ingress. Water can pass through the lining and into the shaft, or travel down
behind the shaft lining where the ground has been disturbed by its construction.
Often water ingress through shafts has been dealt with by installing garland drains to
capture and channel the flow before it entered the main tunnel, but the difficulties
associated with maintaining equipment in shafts mean that such measures have fallen into
disrepair and become ineffective. Where drainage systems are used it is important to make
provision for regular maintenance.
Simple collector drains manage water ingress but do not reduce it, and the prolonged
passage of water through the lining may result in its gradual deterioration and weakening.
An alternative solution is to try to reduce the permeability of the water-bearing strata
behind the lining by grout injection. Drainage holes can be drilled into the permeable
strata from the base of the treated area and pipes installed to channel the water into
garland drains or some other collector system that would have to be maintained. Grouting
to control water ingress is discussed further in Chapter 6.
196
5.6.2.3
Shaft lining stability
Prolonged water ingress through a shaft lining can weaken it and result in washout of
fines and void formation in the adjacent ground. This may result in a gradual loss of
interaction between the ground and the lining and reduced frictional shear support,
leading to shaft instability and increased loading on the tunnel lining at the shaft eye.
Where this is a concern, consideration should be given to a combination of grouting to fill
voids and reduce the permeability of the ground behind the lining and the installation of
alternative water pathways in the form of drainage holes through the lining (and into a
suitable collector system). Also, rock bolting or soil nailing techniques could be used to key
the lining into stable ground and improve ground/structure interaction.
5.6.2.4
Relining
An alternative to piecemeal repair and like-for-like replacement is the provision of a new
shaft lining constructed within the existing shaft annulus. A method used in relining sewer
tunnels involves the installation of preformed glass reinforced plastic (GRP) or glass
reinforced cement (GRC) panels that are fixed within the shaft leaving a gap of around 50
mm between the preformed panel and the existing annulus. The annulus void is then
grouted up through injection points in the preformed panels to form a rigid composite
internal lining. Weep-holes can be included to relieve water pressure behind the new
lining. An example of this type of repair carried out on sewer tunnels described in Case
study A1.7.
Another method that could be considered is relining with sprayed concrete. Although
many shafts have too small a diameter to allow man-operated construction, it would be
possible to use a robotically controlled spray nozzle for this purpose. More information on
sprayed concrete linings is given in Section 5.4.8.
Relining will inevitably increase loads on the tunnel at the shaft eye unless more support
(as discussed in the previous section) is provided.
5.6.3
Filling
The decision to fill a redundant shaft should not be taken lightly, filling does reduce
dangers from falling debris and is likely to reduce maintenance requirements. However,
there are many potential disadvantages (NR, 2004b):
shaft filling is an expensive operation regardless of the material used
it could be assumed that maintenance will stop and the shaft, with its fill, will revert to
solid ground. Unfortunately this is not the case and, after filling, it will be impossible
to inspect or work within the shaft
filling with any material, even polymeric foam, will increase the vertical loading of the
shaft lining on its supports because shear support between the lining and the fill is
often negligible
filling an open shaft is likely to reduce ventilation in the tunnel, which may increase
the general level of dampness and increase the likelihood of build-up of gases
(although it may also reduce frost damage to the tunnel lining)
open shafts are an important alternative method of access for emergency services and
equipment. They also provide emergency escape and evacuation routes. Filling open
shafts removes this valuable facility.
CIRIA C671 • Tunnels 2009
197
Wherever possible a thorough structural inspection of the shaft should be carried out
before a decision to fill is made. Careful records should be kept and some structural
repairs should be considered, especially if the lining and the fill is to be supported above
an operational or disused but unfilled tunnel. It may be necessary to replace existing
beams, curbs or buntings with more durable materials, and to introduce extra,
intermediate support beams at stages down the shaft. The load bearing capacity at the
base of the shaft can be improved by incorporating steel or concrete beams before
plugging the bottom of the shaft.
There are two ways to support the infill: either on the shaft lining and base support or by
providing a column beneath the shaft, which stands on the tunnel floor. The latter method
can only be considered when passage through the tunnel is no longer required.
A variety of materials can be used to infill a shaft, including general granular fill, hardcore
and demolition waste, waste from industrial processes and a variety of cementitious and
chemical grouts. The selection of material depends on the circumstances applicable to the
individual shaft. These circumstances include the local bearing capacity of the shaft lining
or the base plug, whether the tunnel is still operational, the type of ground and the extent
of water ingress. Fill materials should be stable, inert, and normally locally available and
inexpensive, and should fulfil other technical requirements depending on the specific
application, eg grading, porosity and bulk density.
Build-up of water in the shaft should be avoided, and fill materials should be nonabsorbent. Water should be allowed to run through the infill or down the inside of the
shaft, to the bottom where it will pass into the tunnel via holes provided in the plug or
past the column. Where water flow is anticipated through the fill, measures should be
taken to prevent the washout of fines and loss of stability eg by collecting or diverting it, or
using open-graded large size granular material, which can drain freely.
Care should be taken to ensure complete filling without voids, depositing material down
the centre of the shaft and not at its edges. Where shafts are partly infilled below
obstructions such as old staging, the obstructions may be removed if accessible, or peagravel may be used to flow into voids and further stabilised by grouting if necessary. The
volume of fill material used should be recorded and checked against the volume of the
shaft to check the adequacy of filling, and access should be provided below any capping to
allow regular measurements of the level of fill to reveal whether cavities may have formed.
The position of the shaft should be permanently marked inside the tunnel and at the
ground surface.
5.6.4
Grouting
Grouting is recommended to consolidate existing shaft filling where more stability is
required, or where voids are present. This may be achieved by drilling and casing a
central hole down through the fill to the shaft base, then grouting the hole in stages as the
case is withdrawn. Alternatively, a perforated injection pipe may be used inside the cased
hole, injecting using packers as the case is withdrawn. This allows the casing to be
withdrawn immediately on completion of the hole, increasing the productivity of the
drilling rig. Mixtures of cement with fly-ash, PFA or bentonite may be suitable as a grout
material.
198
Figure 5.21
Grouted plug remedial measure
for deteriorating shaft lining
(Healey and Head, 1984)
Depending on the nature of the fill and the grout material used, shafts of up to 2.5 m
diameter can be treated from a single central hole using this technique. Larger diameter
shafts may require multiple holes and staged withdrawal of the casing to ensure complete
filling (Healy and Head, 1984).
Another option for injection into a shaft is the use of a chemical grout, such as
polyurethane. This tends to be an expensive system to use but if properly carried out can
give a very satisfactory result (NR, 2004b). The use of expansive grouts needs to be
carefully considered because these can exert considerable pressures when used in
confined areas. The operation requires careful specification and is best conducted by
experienced contractors.
5.6.5
Capping
Shaft capping at the surface has the advantage that the shaft (and any fill material) is
protected from the environment. The most common method of covering a shaft is the
construction of a prefabricated reinforced concrete cap to span the shaft void. The size of
the cap should sufficiently exceed the internal diameter of the shaft and be adequately
supported, preferably at or below rock-head if present. Where this is not possible, the cap
may be supported on competent ground with extra support measures, for example,
plugging, grouting (inside and outside the shaft lining) piling and diaphragm walling etc
depending on the situation.
Figure 5.22
Potential failure mechanism of a shaft cap located at rock head level (Healey and Head, 1984)
CIRIA C671 • Tunnels 2009
199
Where a cap is to be constructed at the base of a shaft, it should be designed to support
any shaft fill material and provisions for water drainage should be included.
Shaft caps should include access covers to help inspection of shaft condition or, where the
shaft is backfilled, to allow the surface level of the fill to be checked for signs of settlement.
Shafts may have been infilled with materials that decay to produce potentially hazardous
gases, particularly domestic or industrial refuse. Covers to such shafts can be sealed using
gypsum, cement or resin-based products to restrict the emission of gases to the
atmosphere and prevent their migration to and accumulation in confined areas. It may be
necessary to install vent pipes to prevent excessive gas build-up, and these may need to be
fitted with flame arrestors and where appropriate protected by lighting conductors.
The location of caps should be recorded and their central position permanently marked
on the ground surface and within the tunnel so that they can be easily found and
identified at a later date.
200
6
Water ingress and control
6.1
GENERAL CONSIDERATIONS
Bauer (1985) stated that:
“Probably the most important factor which governs the success or failure of a (new)
tunnel project is the proper control of groundwater.”
This statement is probably valid in the long-term operational maintenance and use of
existing tunnels as well.
The majority of tunnels are constructed deep below ground surface, often below the
groundwater table and/or below rivers or other bodies of water. Water permeation is
widespread and potentially serious problems can arise if left unchecked. Water ingress is a
key factor in most mechanisms that result in gradual deterioration of a tunnel’s structural
fabric but can equally affect the ancillary structures and services contained within the
tunnel. Also for pedestrian tunnels, dripping water, slippery surfaces and high humidity
levels can be a problem. So minimising and controlling water inflow is usually a significant
concern to tunnel owners. The maintenance of tunnel drainage systems and other
preventative measures are just part of the overall strategy to be considered when dealing
with the control of groundwater. This section gives advice on the wider problems of
dealing with water ingress and considers some of the measures that can be done to
minimise and control it.
Water ingress can occur in all types of tunnel construction even those designed to be
watertight. Tunnels built below the groundwater table and lined with segmental linings of
concrete or cast iron were usually designed with gaskets and seals to prevent water ingress
through joints. Some tunnels may have been constructed with a waterproof membrane
either external or internal to the primary lining at key points on the structure. However,
post-construction tunnel movements from effects such as settlement, redistribution of
ground stresses, tidal effects and vibration can lead to opening of joints, and cracking or
rupture of waterproof membranes. Other effects including poor workmanship during
construction and shrinkage of annular grout can also lead to water ingress. In masonrylined tunnels water seepage often occurs slowly over large areas through permeable
mortar or fine cracks at masonry/mortar interfaces, but can eventually lead to local
washout of mortar through the full lining thickness, the resulting holes allowing much
more rapid flow.
In lined tunnels groundwater can enter by one or more of the following paths:
through the construction joints and interfaces in the lining structure
through cracks that have formed in the lining structure
through permeable areas of the lining materials
service entry or exit points, cable ducts etc.
Wet patches on a tunnel lining are not always the result of water ingress. Considerable
surface condensation can occur at cold spots on the lining that, for example, can be caused
by some environmental anomaly, particularly in cast iron tunnels where humidity is high.
CIRIA C671 • Tunnels 2009
201
Numerous measures to minimise and/or control water ingress are available but the
method chosen in any one instance will depend on the extent and rate of seepage present
or water inflow. Localised treatment may be acceptable where a single leak occurs.
However, where water ingress is extensive and over large areas, options may lead to
extensive grouting or complete relining or provision of a secondary drained liner at the
intrados.
There are effectively four distinct sources of water ingress:
1
Groundwater.
2
Surface (rain) water.
3
Water from underground water mains.
4
Sewage or waste water.
Although frequently inconclusive, attempts should be made to identify the source of the
water ingress before carrying out remedial measures, as this may influence the most
appropriate response. Where ingress arising from leaking or burst water mains, sewage or
waste water can be positively identified it may be possible for the utility owner to stop it at
source and avoid carrying out remedial measures in or around the tunnel. Sampling and
analysis of the water entering a tunnel may also identify contaminants that might be
hazardous to the health and safety of maintenance operatives and tunnel users, or
potentially aggressive to the tunnel lining or any proposed remedial measure.
Rising groundwater tables in urban areas resulting from a significant reduction in
groundwater abstraction is a concern for asset owners with tunnels potentially at risk,
especially where consideration was not given to keeping the tunnel watertight when
originally designed (Simpson et al,1989).
Water ingress is one of the main causes of deterioration of tunnel linings and of ancillary
structures and equipment. It may also be detrimental to services contained within the
tunnel (signals, power and communication systems) and in the case of railway tunnels,
may affect the track support structures, ie cause deterioration of sleepers and the ballast if
present, particularly if there is poor or inadequate track drainage. In some tunnels water
ingress can be tolerated to a certain degree, for example, those on the canal system.
Leakage from tunnels carrying sewage or foul water may have a detrimental effect on the
environment and cause pollution of groundwater. For many ageing tunnels the control of
water ingress (or leakage) is a continuing maintenance liability and its treatment is often
dictated by the policy and operational requirements of the asset owner.
The effects of water ingress on the tunnel lining can range from minor surface corrosion,
in the case of metal linings, to weakening and ultimately complete deterioration with the
associated risk of collapse should it be left unchecked. In broad terms water ingress may
result in the following processes either singly or a combination of several:
202
corrosion of cast iron and steel linings resulting in loss of section and reduction in
structural capacity
corrosion of steel reinforcement in concrete lined tunnels or structures (cracked
linings or linings with poor quality concrete)
corrosion of bolts in segmental lined tunnels
in unlined tunnels, erosion or washout of interstitial material from discontinuities
leading to rockfall
disruption of services contained within the tunnel and associated fastenings/ brackets.
For example, signal, power, ventilation and communication systems, which in the case
of rail tunnels are often in direct contact with, or are near to the intrados of the
tunnel lining
deterioration of caulking and packing material may lead to increased water ingress
and distortion/movement of adjacent segments
fines being washed in from behind tunnel linings – has the potential for ground
settlement and/or formation of voids resulting in eccentric loading on linings
increased corrosion due to stray current (in railway tunnels)
softening and erosion of mortar in masonry tunnels leading to loosening of the
masonry units and reduction in structural capacity
deterioration of mortar, masonry units and concrete through physical salt attack or
expansive sulfate reactions and, in areas exposed to freezing temperatures, the effects
of repeated freezing and thawing
the formation of ice sheets and icicles can reduce clearances within tunnels.
Remedial actions for controlling or cutting water ingress can be classified under two basic
types:
1
Passive control – controlled diversion of a leak that does not reduce the water flow but
either partially contains and/or channels it along a specific route into a drainage
system.
2
Active control – sealing a leak by an appropriate method to stop or reduce the flow to
an acceptable rate.
The general range and common use of these options relative to the type of structure
requiring remedial action are summarised in Table 6.1.
Table 6.1
Summary of passive and active water ingress control measures
Type of control
Passive
Active
Active measure – treated
separately in Section 6.4
Lining type
Drip tray (including guttering and down pipes)*
All types
Secondary lining panels* – to deflect or channel
ingress water
All types
Weep pipes
Masonry and concrete
Channelling
All types
Caulking, grummets and bolt holes
Cast iron and pre-cast segmental
concrete
Secondary structural lining* – including membranes
Masonry or concrete (other types
may also be appropriate)
Surface sealing
Masonry and concrete
Dewatering – internal or external to the tunnel
All types
Grouting – cementitious or resin grouting
All types
Note:
*
Method may only be applicable if there is sufficient space and does not interfere with the safe operation of the tunnel.
Before carrying out any groundwater ingress (or egress) control measure a good
understanding of the cause and likely effects is required. This may involve collecting data
pertaining to the construction of the tunnel and detailed investigations/inspections on-site.
CIRIA C671 • Tunnels 2009
203
Care is required when selecting active measures as remedial works to treat seepage may
result in water being chased along the tunnel, with seepage starting or worsening in other
areas. With careful planning this risk can be minimised. For example, leak-sealing by
grout injection requires consideration of the best injection points and sequence of work.
In some cases it is a good idea to pre-inject areas each side of the water ingress to
compartmentalise the area and reduce the possibility of pushing the water elsewhere. If
any treatment might result in the increase of hydrostatic pressure behind the tunnel
lining, an assessment of the potential effect on the tunnel lining may be necessary.
6.2
PASSIVE MEASURES
The principle behind passive measures is management of ingress water, not preventing it
from occurring. These measures include the installation of weep holes, drip trays,
guttering and secondary lining panels to deflect the water away from the area it is
affecting.
The main criterion for choosing to install a passive measure is that the ingress or presence
of water in the environment would not be detrimental to any structures, services or
equipment in the tunnel and is not likely to cause deterioration of the tunnel lining. If the
presence of ingress water can be tolerated, but it needs to be channelled away for safety or
on environmental grounds, or is marginally affecting ancillary structures or services, then
a permanent long-term water management system may be sufficient. However, there may
be occasions where the installation of a drip tray may be required as a short-term solution
if there is a safety concern before active or other measures are implemented.
The one disadvantage with passive measures is that there is usually a residual
maintenance responsibility to ensure that the drainage paths and installed drip trays,
guttering and panelling etc are kept clean and free of debris to ensure the continued flow
of ingress water. Although the initial cost of installing passive measures is often (but not
always) less than that of taking active measures to reduce ingress, the cost of maintenance
is sometimes underestimated. The maintenance, which is often neglected, can be more
costly and problematic in the long-term. In some situations a combination of active and
passive measures can help reduce maintenance requirements and costs.
In railway tunnels that have ballasted tracks, the ballast is often used as a drain (with or
without a drain pipe installed) and the water allowed to drain from the tunnel or to sumps
for collection and removal later by pumps. If the ballast is allowed to become
contaminated this will reduce its effectiveness as a drain.
Many road tunnels have open inverts below the road deck. The inverts may be used as a
collecting sump for ingress water, which in the case of a tunnel under a river or other
body of water will be led to a pumping station at, for example, the mid-river point.
One feature of open inverts, particularly in road or railway tunnels, is the amount of
rubbish or debris that often accumulates in them. The accumulation of rubbish and debris
if not cleared away on a regular basis may have the potential to block drains and pumping
equipment.
The choice of design and means of installing passive measures cannot be defined in detail
as they will vary according to the different types of structure that may be affected, and the
location and magnitude of the water ingress. However, some basic points include:
204
location – would the installed system reduce clearances to an unacceptable level?
ease of installation
rate of flow of the water ingress
use of low-maintenance materials, eg stainless steel (depending on the nature of the
local environment) or uPVC guttering (or similar) for short-term solutions
ease with which the installed system can be cleaned and maintained
external effects or forces, eg design against buffeting by passing traffic
the ability to be taken down to aid inspection of the tunnel lining
appearance – would an installed system be tolerated by the users/owner/members of
the public?
The following sections discuss some of the main methods that can be employed to manage
the ingress of water.
6.2.1
Drip trays (including guttering and down pipes)
These measures collect water from a localised area of seepage that may be causing a
problem on the structure below or to any ancillary structure or service in the area. For
example, water dripping onto the current line of a railway or onto signalling equipment.
The water is then allowed to drain away in a controlled manner, typically to the tunnel
invert or any drainage system that may be present.
Drip-trays are usually fixed tight under the leaking area so it is often difficult to see the
condition inside the tray. Ideally the drip tray and associated down pipes should be
detachable to allow inspection and cleaning, and be contoured to prevent structural
clearance problems. They should also have high edges and an oversized down pipe to
avoid blockage and overflow (see Figure 6.1).
6.2.2
Secondary lining systems (or drainage membrane)
Transport tunnels (road, railway station etc) and tunnels used by members of the public
often have secondary lining systems installed to improve the appearance of the space
being used and to conceal the primary lining. To overcome persistent water ingress issues
the secondary lining system may also be specifically designed to deflect and divert ingress
water away from the space below. The panelling may also be used to hide otherwise unslightly passive drainage systems such as drip trays, guttering and down pipes.
Vitreous enamelled panels may be installed either directly onto the primary tunnel lining
or to a frame work that may be free standing or attached to the tunnel lining. Other
systems include the installation of a waterproof membrane lining system (eg uPVC, HDPE
and geotextiles) attached directly to the primary tunnel lining that acts as drainage layer
to collect ingress water and divert it off to a sump or drainage system. Proprietary
drainage membranes may be partially bonded to the structure or sheet membranes fixed
to the surface.
There is a risk with membranes that they may become punctured either during
installation or during use of the tunnel, particularly by others fixing internal systems
(communication, cabling and signalling equipment) or signage.
CIRIA C671 • Tunnels 2009
205
a
Figure 6.1
Drip trays to control water ingress
before (a) and after installation (b)
b
When planning to install a secondary lining or membrane system several issues need to be
considered including:
availability of space
ease of installation
use of maintenance free materials
external effects and forces
presence of equipment and services fixed to the inside lining of the tunnel.
The ease of keeping the secondary lining clean and free of debris also needs to be
considered, in particular the drainage trough or collector drain that may be installed at
the base of the sidewalls.
The main disadvantage with the installation of a secondary lining system is that it will
cover and obscure the main tunnel lining, which could inhibit periodic inspection and
maintenance works. Although provisions can (and should) be included to easily inspect the
main tunnel lining behind the secondary lining, through the installation of inspection
hatches or removable panels, particularly at critical locations and those that require
frequent inspection and/or maintenance works, the use of such a system requires a
positive commitment to carry out thorough periodic inspections and accept further
difficulties involved in gaining access.
206
a
b
Notes:
These figures are taken at the lay-by area to improve ambience and provide a means to deflect water ingress through the installation
of vitreous enamelled panels.
White and blue coloured vitreous enamelled panels used to match original paint scheme.
Figure 6.2
Reconstruction of the lining of the Mersey Tunnel (a) exposed painted cast iron and stainless
steel support members for panels during installation works (b) partially completed section of
secondary lining with panels (courtesy of Mott MacDonald)
6.2.3
Weep holes (and pipes)
The installation of weep holes can relieve water pressure from an area of seepage,
allowing water to be diverted off to a particular location. They can be used in conjunction
with other passive measures including drip trays or channels (see Section 6.2.4).
Weep holes should only be installed where there is no risk of erosion behind the tunnel
lining by the ingress of fines. They may also be installed during or after grouting works
(see Section 6.4) to relieve the pressure of the groundwater acting on a section of the
lining or repair.
Weep holes and pipes may also be connected either singly or in series with many other
weep pipes to a plumbing system to divert the water to the invert or drainage system.
Garland drains are used in shafts in conjunction with weep holes to collect and drain away
seepage.
6.2.4
Channelling
Common to many of those passive measures described in Section 6.2.3 may be the need to
install or create channels to allow the water to drain off in a particular direction towards a
specific drainage system or sump. This may include a chased channel, for instance at the
base of a wall where water seepage is occurring, to remove the water in a controlled
manner rather than let it pond on the floor. However, even with a basic water channel or
pathway, periodic maintenance may still be required to ensure that the water is free
flowing. Channels should have adequate capacity for coping with surges and gradual
reduction in effectiveness through silting-up between maintenance visits.
6.3
ACTIVE MEASURES
Active measures to control the ingress of water should be used when permeation of water
cannot be tolerated for structural, operational, environmental or safety reasons. Active
measures are primarily used to seal the tunnel lining preventing water permeation from
occurring. However, active measures may also be implemented to reduce the flow of water
to an acceptable level, followed by passive measures to manage the residual flow.
CIRIA C671 • Tunnels 2009
207
Active measures can be expensive and disruptive, so careful planning is required, based
on sufficiently detailed information on the source, nature and extent of water permeation.
Other information on the structure, the adjacent ground and constraints on working will
often need to be considered. The works should be implemented using the highest
specification available with respect to the materials used and quality of expected
workmanship. Though some measures are relatively simple and straightforward and may
be carried out by less-skilled operatives, others will inevitably require skilled engineering
input into the design and the use of experienced specialist contractors to complete the
work. In practice it is quite usual to find the cost of materials to be a very small percentage
of the total cost of the works, so no compromise should be made in the selection of
materials.
The following sections discuss some of the more common active measures used to prevent
or limit the permeation of water through a variety of tunnel types. In some cases the
methods employed may be adopted for different lining systems, ie caulking joints on cast
iron or pre-cast concrete segmental lined tunnels.
6.3.1
Caulking, bolt holes, grummets and grout holes
Groundwater is able to permeate a bolted segmental lined metallic or pre-cast concrete
lining in several ways:
through flange joints where the caulking has failed
through bolt holes (linked to failure of caulking)
through pre-formed grout holes
through cracks in the tunnel lining.
In more modern segmentally lined tunnels the extrados of the tunnel is often grouted
during construction. The grout is there to serve two purposes:
1
To allow uniform ground stress to act on the tunnel lining by filling any voids that
may be present.
2
To help make the tunnel watertight.
An exception is in tunnels with expanded segmental linings, for example, on the London
Underground Victoria Line where no provision was made for grouting as the linings were
designed to be expanded to make contact with the excavated ground and only used in
areas that were considered dry.
Typically a water and cement grout mix has been used, although aggregates such as pea
gravel and pulverised fuel ash (PFA) have been included to bulk-out the mix. One of the
main disadvantages with a cement based grout is shrinkage unless additives are included
to inhibit shrinkage. If the grout shrinks it may allow a water path to form between the
grout and the extrados of the tunnel lining. Should the lining joints leak in any way then
seepage will occur.
6.3.1.1
Caulking joints in segmental linings
There is no single design of the flange joint with either metallic or pre-cast concrete
linings. For example, flange joints in cast iron linings may be machine finished or unmachined (although these were much earlier, smaller diameter tunnels built through dry
ground), with or without a gasket or caulking groove. If caulking grooves are present they
208
may be shallow or deep, with or without pockets at bolt locations and may vary in width.
During planning of a tunnel, the allowance for a gasket or seal around the flange joint
may have been included where the tunnel was known to be passing though wet ground
conditions. Gaskets or seals may include bitumen felt, fibre reinforced bitumen, neoprene
rubber or in recent tunnels, hydrophilic gaskets. Hydrophilic gaskets are made from a
variety of compounds that swell when in contact with water. As the technology behind
these materials continues to develop, alongside the traditional bitumen, bentonite types,
newer materials make use of nanotechnology, with the hydrophilic particles fitting within
the base material matrix allowing it to expand more without loss of integrity.
Caulking materials used will also vary enormously depending on the design and age of the
tunnel. Lead caulking has been used in conjunction with cast iron tunnel linings since the
19th century for tunnels in most types of ground conditions, but primarily wet ground.
Earlier tunnels could be caulked with a rust compound consisting of iron or steel filings
mixed with sal ammoniac, and on occasions, sulphur, to produce a rapid hardening
mixture that accelerated rusting. The mixture is introduced into the caulking groove to
form a solid mass that expands through the oxygenation process and tightens-up the
joint, making it watertight.
Cement based caulking compounds are mainly used for caulking concrete segments,
although they have been used in the past on cast iron tunnel linings in dry ground
conditions.
Caution should be exercised when working on caulked joints where there may be
materials hazardous to health. As well as lead-containing products, asbestos cord
impregnated with cement was often used as a caulking material in tunnels built in the
1960s and 1970s. When working on or replacing old cementitious caulking compounds
precautions must be taken to determine the presence of asbestos. If present the risk from
asbestos containing materials must be managed and minimised, and this may entail their
encapsulation or safe removal by a specialist contractor in compliance with current
regulations.
On occasions a combination of different caulking materials would have been used, such as
lead initially caulked into the caulking groove followed by a cementitious caulking or rust
compound. In the Greenwich subway (1899–1901) and Rotherhithe tunnel (1904–1908) a
combination of lead strip or lead wire caulking in the bottom of the caulking groove
finished off with rust compound provided excellent results in watertightness. However,
one of the main drawbacks with the caulking compounds described above is that they are
relatively inflexible and will not accommodate significant tunnel lining movements.
CIRIA C671 • Tunnels 2009
209
Figure 6.3
Seepage from circumferential joints in a pre-cast
concrete lined tunnel
Re-caulking can be a time consuming and expensive operation. If re-caulking is
undertaken it is vital that the original material present in the caulking groove is cleaned
out fully before replacement with a suitable material (either lead and/or low-shrinkage
cementitious compounds). It is important not to overlook critical areas such as a deep
caulking groove or pockets located behind the tunnel bolt positions. If the defective
caulking is not removed in these locations the process of re-caulking may not be
successful. To carry out these works efficiently it may be necessary to remove the tunnel
bolts to ensure the old caulking can be removed from behind and replaced. Consideration
should be given to a combined programme of bolt (or grummet) replacement combined
with leak sealing to provide a practical and economical solution. The volume of water
ingress can influence the order of works. The leak sealing exercise should be done before
replacing the caulking as the existing caulking serves as a dam to prevent a
disproportionate loss of grout, and to provide a dry work surface for the replacement
caulking. When the water flow is high it may be necessary to make temporary damming of
the joint and inject to reduce the volume of ingress. After that, caulking can be carried
where necessary.
There are many proprietary caulking compounds available from suppliers that may be
used to re-caulk segmental lined tunnels. These include asbestos-free cementitious
compounds, which can be applied in damp or dry conditions, but are generally inflexible
once set. The selection of the compound and technique to seal the joint will depend upon
the rate of water inflow through the joint, and most good specialist contractors should
have their own techniques for leak sealing. If water inflow is severe, then grouting behind
the joint may be the only option (see Section 7.4) to prevent or reduce the water inflow
before re-caulking. The use of rapid-hardening mortar may be used on moderate inflows
followed by caulking.
The selection of caulking compound should also consider whether there will be a risk to
further movement of the rings. If movement is likely or suspect, then a flexible caulking
compound can be used such as butyl rubber or mastic compounds.
210
6.3.1.2
Sealing bolt holes
Where seepage occurs at bolt holes an active measure would be the replacement of the old
bolt and grummets with a new bolt set ensuring the bolt is tightened to a suitable torque.
However, the leak at the bolt hole should be injected before replacing the bolt.
Gel grummets are manufactured from jute yarn impregnated with a waterproof gel and
used for waterproofing bolt holes in all types of bolted segmental linings. Oyster
grummets are manufactured from low density polyethylene and are used to waterproof
bolts holes in cast iron and steel bolted segmental linings.
Likewise, if no bolt is present in a weeping bolt hole, then a new bolt and grummets
should be installed to prevent further water ingress.
6.3.1.3
Sealing grout holes
Seepage from grout holes is a very common fault, especially when the grout plugs have
been removed. Seepage can occur because of shrinkage of the grout post construction
and/or due to settlement or movement of the tunnel.
Active measures to prevent seepage of ingress water through grout holes would be the
replacement of purpose made grout plugs (cast iron or other complementary material, eg
neoprene plugs for concrete segments). Leaching grout holes may also be grouted with a
suitable polyurethane and/or acrylic grout (see Section 7.4 for details of grout types that
may be used) and plugged off with a suitable quick setting epoxy mortar.
Figure 6.4
Typical seepage from cast iron
segmental lining grout hole
6.3.2
Surface sealing
In some circumstances slow seepage through concrete and masonry structures can be
treated by the application of a surface sealant, though this is seldom the best or most
effective method. There are three main types of sealant used:
CIRIA C671 • Tunnels 2009
211
1
OPC and very fine silica sand and active chemicals.
2
Ordinary cementitious waterproof slurries.
3
Flexible surface coatings.
With the first example the OPC and silica sand are just a carrier, the sealant properties
come from the active chemicals, silicates or phosphates, which penetrate the concrete or
masonry against water pressure, absorbing and sealing in the water. On concrete
structures, penetration of at least 5 mm to 7 mm is possible, with cracks up to 0.5 mm
width being sealed.
The use of ordinary cementitious slurries is limited as the OPC forms part of the seal. The
coating is only useful where no further cracking or movement of the structure is likely as
the coating is inflexible and brittle. The slurry does not penetrate into the substrate, other
than filling porous areas and cracks that are open at the surface.
The third category is generally a proprietary flexible sealant, plastic and latex materials,
which is applied directly to the surface of the structure to form a flexible coat. As with the
OPC slurries, they do not penetrate the surface.
The risk inherent in using impermeable surface sealants is that a build-up of hydrostatic
pressure may de-bond the coating if the surface of the structure has already deteriorated,
and may result in damage to weakened brick, stone or concrete substrates. Also there is a
possibility of the water passing through fine cracks in the coating as a vapour, resulting in
the formation of salt crystals behind the coating, causing it to fail. This may be less of a
concern for shallow structures but the risk may be reduced for deeper structures, where
practical, by the installation of pressure relief weep holes through the lining. These holes
may need to be channelled off to the invert in a manner discussed above in Section 6.2.3,
if the presence of localised water seepage from the relief holes cannot be accepted. In
considering such a method the possibility of the holes closing through a build-up of silt
and salts should not be overlooked. However, it is always preferable to grout through the
structure into the positive water side.
All surface sealants require a clean and sound substrate for optimum performance.
6.4
GROUTING
6.4.1
Grouting technique selection
Grout may be a cementitious (cement-based) particulate grout or a chemical (polymeric)
resin. The grout is either injected into the ground close to the tunnel or within the body
of the tunnel lining structure to control groundwater and fill permeable areas and larger
voids if present.
When grouting is carried out in the ground it can be achieved in one of two ways:
1
Grouting without ground fabric displacement or replacement (interstitial or fissure
grouting).
2
Grouting with ground fabric displacement or replacement (mainly used in soils).
Figure 6.6 illustrates the main grouting techniques associated with these two principles.
212
Figure 6.5
Summary of grouting techniques
Of these two grouting techniques, the first is principally used to prevent water finding its
way to the outside of the tunnel or underground structure where it may then find a
passage into it. Grouting that displaces or replaces the ground fabric is generally used to
strengthen the ground and improve the ground/structure interaction. However, this
constitutes ground structure improvement and is beyond the scope of this document.
Interstitial or fissure grouting may offer many advantages over other passive or active
methods of controlling water ingress as discussed in Sections 6.2 and 6.3, in that:
the structure remains uncovered allowing inspections to be carried out without
removing sheeting or covers
correctly installed, using the appropriate materials, the interstitial grouting procedure
will provide an effective means of preventing water permeation, without the
requirement for maintenance over a considerable period of time
the method uses comparatively few staff and small plant making mobilisation easy and
small works economically viable
there is no requirement to remove track or road surfaces
the work can be halted at any stage and re-started at a later date without detrimental
effects, cutting operational delays on critical structures
structural maintenance is not adversely affected by the presence of appropriately
installed grouts
by preventing water percolation, deterioration to mortar courses is arrested
grouting may be the only practicable option
by controlling water permeation other water sensitive maintenance work can be
carried out.
Choosing the appropriate grout is fundamental to the success of waterproofing. Grouts
can also be used in combination. For example, large voids can first be filled with
cementitious grout and tightened up with a chemical resin such as acrylic. Twocomponent low foam polyurethane grouts can also be used for smaller void filling in
combination with acrylic resin to stop the water flow.
CIRIA C671 • Tunnels 2009
213
The extent and purpose of the grouting works dictate the grouting technique to be used.
The treatment works should be designed specifically for each project taking into account
several factors that will influence the success of the work including, but not limited to:
geological and hydrogeological setting of the tunnel to be treated
results of any ground investigation and in situ (permeability, ground stress etc) testing
and laboratory testing (of the soil/rock/groundwater)
groundwater chemistry
whether the tunnel is lined or unlined, and if lined the lining type (segmental, cast in
situ, lining material, ie masonry, cast iron, steel, concrete, and lining geometry etc)
methods of grouting and types of grout available
site constraints such as access, length of possession or closure of the tunnel for any
works
possible effects of grouting on adjacent structures
health and safety and environmental constraints in using particular grout types.
The different grouting techniques and grout types applicable to different ground types
are summarised in Table 6.2.
Table 6.2
Grouting techniques with relevant ground types (adapted from CIRIA C514, Rawlings et al, 2000)
Type of ground
Grouting technique
Permeation
Fissure or rock grouting
(includes void grouting)
Soil
Gravel, coarse sand and sandy gravel
k > 5 × 10-3 m/s
Pure cement suspensions,
cement based suspensions
Sand, medium sand
5 × 10-3 m/s < k < 1 × 10-3 m/s
Micro-fine cement
suspensions, solutions
Fine sand, silt (silty clays)
5 × 10-4 m/s < k < 1 × 10-6 m/s
Specific chemical
(chemical/resin grouts)
Fissured rock
Faults, cracks, karst
e > 100 mm
Cement-based mortars or cement-based
suspensions (clay filler)
Cracks, fissures
0.1 mm < e < 100 mm
Cement-based suspensions or micro-fine
cement-based suspensions
Microfissures
0.05 mm < e < 0.1 mm
Micro-fine and ultra-fine cement-based
suspensions
e < 0.05 mm
Silicate gels
Specific chemicals (chemical/resin grouts)
Notes:
K = Coefficient of permeability, e = fissure width
Grouting is not restricted to pumping grout under pressure but can be achieved using
vacuum pump systems to draw the grout in. Vacuum grouting has been used successfully
in conjunction with low viscosity chemical grouts to seal very fine cracks or apertures in a
variety of applications including tunnel linings.
Further guidance on components of grouting systems and the basis for grouting design is
provided in CIRIA C514 (Rawlings et al, 2000).
214
Provision should be made in grouting works for cleaning the grouting equipment and
pipes during or at the end of working shifts – bins or skips may be necessary to collect the
residual grout fines and contaminated water that result.
6.4.1.1
Cementitious grouts
Cementitious grouts come in an enormous range of types. Suitable grouts for injecting
would include the following properties:
stability, ie the ability to remain in suspension under grouting pressures and not set
prematurely
anti-washout
low strength, typically 15 – 20 N/mm²
highly thixotropic
early initial set
similar modulus to lime mortar (when used for masonry tunnels)
shrinkage compensated
easy to use and handle on-site
minimal toxicity.
Cementitious grouts are generally composed of Ordinary Portland Cement (OPC) mixed
into a slurry with a water/cement ratio of the order of 0.1 to 0.4. Proprietary cementitious
grouts are often formulated with additives such as accelerators and plasticizers to control
setting times, and improve flow characteristics. Rapid hardening and sulfate-resisting
cement can be used to replace OPC where conditions dictate. Sand can be added to
cementitious grout suspensions as a filler when the system of voids or fissures to be
grouted becomes wide. The maximum sand size and distribution is chosen to match the
size of the voids or fissures to be grouted and available grout pump equipment, lines and
fittings.
Other additives, such as bentonite and pulverised fuel ash (PFA), can also be used to bulk
out the grout to reduce the cost of grouting large volumes as well as to improve
performance.
The advantages of bentonite/cement grouts include: lower slurry weight, increased slurry
volume, reduced grout viscosity and lower heat generation during cement curing.
However, there are several disadvantages that may be detrimental when used as a grout
for the control of water ingress, including lower strength, higher porosity and increased
fracturing. There is also a misconception that the use of further bentonite to cement helps
reduce the amount of cement shrinkage during curing.
The advantages of using a PFA/cement grout include: reduced grout bleed, improve
pumpability and flow characteristics, reduced permeability and lower density. PFA may be
used on its own as a low strength filler grout for the treatment of cavities.
Cementitious grouts are, however, limited in their use due to four fundamental factors:
1
Penetration capabilities are limited by their constituent particle sizes and high internal
friction, which prevents them from sealing interstitial spaces in the range required for
general water penetration treatment (see Table 6.3).
CIRIA C671 • Tunnels 2009
215
2
The comparative rigidity of cementitious grouts reduces their long-term in situ
durability where structural movement takes place.
3
Open mortar joints and poor or non-existent pointing allow grout to escape from the
structure if the joints are steeply angled. This becomes worse in wider joints.
4
Setting times for cementitious grouts are measured in hours, with chemical grouts in
seconds. So cementitious grouts are more prone to washout by moving groundwater
before they have set.
Despite these limitations cementitious grouts can be used to great advantage for void
filling where the interstitial space is comparatively large, say over 4 mm, and are
economical.
Cementitious grout being a particulate grout material is generally too coarse to be used to
seal tight leaks, but this has been overcome to a certain extent by the development of
microfine cements that can be used to penetrate much finer cracks (see Table 6.3).
Table 6.3
Joint aperture range for various cement grouts
Cement type
6.4.1.2
Particle size of grout
Practical joint aperture range
Ordinary Portland Cement
80–100 microns
> 400 microns
High early strength cement
40–60 microns
> 200 microns
Microfine cement
10–12 microns
> 50 microns
Chemical (resin) grouts
In tunnels chemical (resin-based) grouts are well suited for injection within structural
elements, into joints, cracks or fissures in concrete, rock or brickwork, or as a curtain
formed against the outside of the structure, either to form a low-permeability layer on the
lining extrados or to compartmentalise water ingress by forming a water-stop.
Materials and their characteristics
The most common materials used in chemical resin grouting are:
1
Polyurethane resin.
2
Acrylic resin.
Polyurethane resins are available in a range of grades from single-component materials
that react rapidly with water and are highly expansive, to two-component grouts with
lower foaming ratios giving a higher compressive strength and generally greater flexibility
in the final product. The equipment needed for injection under pressure is electrically,
pneumatically or manually operated. Polyurethane is primarily used as a void-filler and
can displace water that causes a load on the structure. Most polyurethane is, after
polymerisation, rigid and no longer reacts to water. If the ground (or structure) moves it
will either tear or compress inelastically, potentially allowing water ingress later on. In
either case it will not return to its original form.
The single-component polyurethanes are mostly water reactive and can expand up to
sixty times their original volume, but it is worth noting that where 1 litre of liquid resin
has expanded into 60 litres of foam, the resulting material contains 59 litres of air and has
a very low wall-thickness and strength. The two-component polyurethanes are more
216
controllable but require specialist two-component pumping equipment. Most of these
react through use of a chemical additive rather than through contact with water. These
materials can be controlled to produce a more modest expansion at a specified rate, with a
considerably higher strength than the high-foaming types.
With the single-component resins the amount of foam rise depends to some extent on the
amount of available moisture. A lot of water can cause a greater expansion, a lack of water
can result in little or no foaming, in which case should the unreacted material escape from
the structure somewhere the moisture in the air can cause it to react, potentially resulting
in large mounds of foam stuck to the tunnel wall or growing on the tunnel floor.
Low-foaming high-strength two-component resins are, for example, suitable for grouting
behind cast iron or concrete segmental structures where water may be present within any
voids or fissures between the lining and the excavated ground. Polyurethane resin
injected into these voids can form the first part of a combination grouting action. The
polyurethane should not rely upon the presence of water for it to react, otherwise the
initial foaming reaction can cut off the water supply to the resin behind it. This can result
in voids being filled with un-reacted resin, which could leak out of the structure again
causing large amounts of polyurethane foam reacting in the moist tunnel air (possibly at a
later date). To overcome this tendency water can be added to the resin shortly before
injection but in practice this is difficult to control properly. Care should be taken to ensure
that the separate chemical components are not allowed to spill onto cables or equipment
and the polymerised foam will stain concrete, which may be a consideration where
aesthetics are important.
As it is seldom possible to accurately evaluate the volume of the voids, consideration
should also be given to the development of pressure generated during the reaction of
expansive resins in enclosed spaces, and its potential effect on the structure if injected into
voids behind the lining (particularly for masonry linings).
The other main type of resin used in injection is acrylic-based. These are two-component
materials of low strength and extremely low viscosity, again requiring the use of specialist
two-component pumping equipment. They can be used in sealing earth structures, dams
or embankments but are more suited to injection into cracks or fissures in rock, joints and
cracks in concrete or masonry. The resin is highly controllable, and can be made to react
within a few seconds or up to half an hour or more. Acrylic resin has an extremely low
viscosity, almost that of water, and can be injected safely at very low pressures while still
being capable of penetrating the very fine cracks and paths that water can take. There are
different techniques for injecting into tunnel segments depending on their type, the resin
being able to travel within the segment joint searching out the real water ingress point
that can be some metres away from where it shows on the surface.
Unlike polyurethanes the individual components of acrylic resins can be neutralised in the
event of spillage through dilution with water.
Structural retention of very low viscosity acrylics can be a problem especially on open
structures such as masonry tunnel linings as the material is easily lost close to the point of
injection and can be difficult to control, though measures can be taken to minimise this
risk. Acrylic resins after reaction are prone to shrinkage if allowed to dry out, but contact
with water will lead to swelling and re-expansion. It is this ability that enables the resin to
prevent water ingress. Depending on the quality of the product and the environment, in
the long-term there may be some irreversible shrinkage through desiccation, which can
potentially reduce its effectiveness. Also freeze-thaw action can break down acrylic resins if
the thermal cycles penetrate deep enough to overcome the anti-freeze properties of the
product.
CIRIA C671 • Tunnels 2009
217
Application
Polyurethane resins usually (but not always) have a relatively high viscosity requiring high
injection pressures. They are all too commonly misused for sealing small cracks and joints,
often with the loss of much resin and an unsatisfactory (and unsightly) result. Great care
needs to be taken when injecting through, or into, small joints or cracks in concrete or
masonry as any overpressure at the head increases the chance of blowing the surface off.
As a generalisation, it can be said that whereas large voids are the domain of polyurethane
materials, cracks, joints and honeycombing are often better treated using an acrylic resin.
Together polyurethanes and acrylics are compatible for combination grouting when, after
the voids behind the structure have been filled with polyurethane, acrylic resin is injected
in the structure to seal any leaking joints or cracks.
Weak brickwork can be strengthened with injection at about four fifths depth into the
structure using a microfine cement grout, although the engineering implications of
injecting cementitious grouts and their potential effect on the lining behaviour should be
carefully considered in advance. When enough cement grout has been injected to secure
the structural stability, back-grouting with acrylic resin can be used to reduce water
permeation degrading the structure. Although this may use a little more than usual
volume of resin the risk of deterioration of the structure through further water seepage
can be greatly reduced.
If grouting is considered necessary the tunnel owner or asset steward should seek
appropriate engineering advice. Leak-sealing is generally carried out by specialist
contractors, and many of these have their own preferred methods and materials that they
will tend to use in any situation they come across. While this offers the benefit of
familiarity and experience, it can also lead to the use of techniques that are inappropriate
for the specific circumstances. It is always advisable for clients or their engineers to gain a
good understanding of and ideally become involved in the process of deciding exactly
what and where specialist contractors intend to inject, the technique they intend to use,
why this has been chosen, what were the alternatives and why these have been rejected.
Leak-sealing works should be thought through from first principles to find the most
appropriate solution using the full range of techniques and material available, rather than
presuming the use of one particular set of methods and grouts. Consideration should be
given to the type and nature of structure to be treated, the potential for structural
movements and the influence of (and on) ground and groundwater conditions. Ideally the
works should include a trial on a section of the tunnel to be grouted to establish the
optimum grouting procedures and grout mixes.
It is not the intention of this document to give detailed opinions or advice as to what are
the best methods, as this can vary considerably. Most leak-sealing situations are unique
and require their own well-considered and reasoned solution where possible from past
experience in similar circumstances.
Aside from technical considerations, selecting the right specialist contractor for the job and
selecting the most appropriate form of contract and method of measurement/ payment
process, is likely to have a significant influence on the success (and eventual cost) of the
works.
218
Figure 6.6
Grouting operation in progress in a pre-cast concrete segmental lined tunnel
Health and safety and environmental considerations
It is important to consider the health and safety and environmental hazards and risks
associated with the use of resin grouting materials, and to ensure that they are properly
managed through appropriate and approved method statements and risk assessments.
This requires careful consideration of the material safety data, manufacturer’s product
specifications and COSHH assessments. There are many proprietary chemical grouts
available from a range of suppliers, and it is important to understand that the different
products may also have different chemical bases. The potential toxicity not only of the
fully reacted material but also the base components (which might, by accident, be spilled
or otherwise enter groundwater in an unmixed state) should be assessed before their use.
It is worth gaining a proper understanding of the manufacturer’s information on material
safety data sheets. Blanket statements such as no known toxic effects or not known to be
environmentally hazardous may not fully represent the potential risks of using such
materials and a more detailed assessment may be appropriate. This is particularly the case
where there is a high risk or significant potential consequences of pollution (see the
section on the Hallandsås Tunnel in Case study A1.18).
6.4.2
Grouting masonry-lined tunnels
Water permeates brick and masonry lined tunnels in several ways including:
General seepage
This may be a large area of seepage where it may not be possible to directly see a source,
although open joints between bricks or masonry and deeply eroded and weakened mortar
are characteristic of an area suffering deterioration. Isolation of the source of water
ingress may be difficult over time because water paths can change and may only occur
during periods of rainfall. Also, the original construction of the lining may be highly
permeable and capable of admitting water. The origin and direction that water takes
within the structure is rather complex and may not be easily determined.
CIRIA C671 • Tunnels 2009
219
Joints
Other than the mortar joints between the individual masonry units, construction joints
between lengths of brickwork are particularly permeable to water ingress because the
jointing and mortar-packing between the dog-toothed brickwork is generally worse than
that of the surrounding brickwork. All joints between or within structures are prone to
suffer from water ingress as they are a point of weakness.
Cracks
Cracks usually provide a simple and direct water path through the structure. In brick and
stone masonry structures, cracks are normally aligned along the mortar joints except in
cases of high shear or in very weak materials, which may crack though the individual
bricks or stone units. The direction that a crack takes within the lining can be difficult to
determine by examining only the lining surface, and this may have an impact on how and
where to inject grout in to reduce water seepage.
Structural defects
These are sometimes difficult to recognise from a visual survey as there may be no direct
evidence of these visible at the surface but they may act as water paths within the
structure. For example, defects such as joint failures that could include the separation
between rings within the arch of a tunnel lining.
Point leaks
Such leaks occur as single points in relatively dry areas of the structure and may be
centred on a joint or other defect. In such cases the location of the original source of the
water may not be predicted with any certainty. Point-leaks may also be associated with and
occur through fixing bolts, weep holes, repairs, cable ducts, brackets, supports or any
other fixing to the structure.
Figure 6.7
Series of longstanding point leaks
from the lower part of the arch in
a brick lined tunnel, made clear by
the thick deposits of carbonate
that have built up on the brickwork
220
Each of the types of leak described requires a different form of treatment if an economical
and successful solution is to be achieved. Within a structure a range of different leaks will
require treatment using a variety of techniques. Where joints, cracks and point leaks are
present these are injected first where they cross areas of general interstitial water ingress.
This is because grouting of these features will form localised grout masses that will act as
barriers for the general grouting and prevent any potential loss of grout from that feature.
The following sections discuss in more detail the procedures used to grout the different
types of water ingress that may be present within a brick or stone masonry lined tunnel
structure.
6.4.2.1
Grouting procedures
General seepage
General areas of seepage require systematic interstitial grouting. The distribution of voids,
porosity and water pathways within the structure is not easy to predict and so a
generalised injection pattern needs to be adopted.
As with cracks, the depth of the grout injection hole, which is controlled by the structural
thickness of the lining, will determine the penetration and spread of the grout. If short
holes are drilled the probability of intersecting discrete water paths will be reduced. A
balance between injection hole depth and effective grout penetration is achieved by
applying the 60 per cent of the structural thickness rule of thumb, ie injection holes are
drilled to 60 per cent of the structural thickness. With grout penetration being limited by
the depth between the bottom of the hole and the extrados of the structure, ie 40 per cent
of the structural thickness, grout injection points are then spaced at 80 per cent of the
structural thickness (twice the remaining structural thickness from the bottom of the hole).
This is easy to remember and provides a reliable guide.
There are two exceptions to this rule:
1
Where the back of the structure is in contact with a layer or barrier impermeable to
grout.
2
Where the lining is very thick the grout injection holes need not be drilled to a depth
exceeding 500 mm as it is assumed that sufficient grout would be injected to form an
effective barrier within the available structure mass. Also, the spacing between grout
injection holes should be limited to a maximum of 700 mm.
Some structures may require grouting to greater depth for reasons other than the
prevention of seepage. For example, if aggressive groundwater is present grouting of the
full thickness of the structure may be required to provide a degree of protection to the
overall structure.
Injection holes should be drilled at right angles to the surface but within reasonable limits,
the angle is not critical to the success of the work. However, they should be placed in
staggered rows to provide full coverage. During injection some grout escape through to
the (internal) surface of the structure is inevitable. Care should be taken to avoid
uncontrolled grout flow and excessive loss, which is not only costly, messy and potentially
polluting, but could result in a void being left unfilled.
The grout injection hole pattern described above is applicable to all types of structural
element and is not adjusted for curvature or orientation.
CIRIA C671 • Tunnels 2009
221
Figure 6.8
Section through a multi-ring brick arch illustrating the positioning of
the access holes relative to the structure
Figure 6.9
Elevation showing a typical access hole pattern. Note that access
holes are drilled through brick centres in staggered rows
Cracks and joints
Cracks and joints are treated by cross-drilling grout injection holes to intercept the plane
of the crack or joint. To be sure of intersecting a crack the practice of drilling a sequence
of angled holes from alternate sides of the crack will ensure that at least one set of holes
will intersect the crack. If the hole-drilling exercise reveals the orientation of the crack
then drilling can be continued on the appropriate side. The injection hole should be
drilled to pass the crack at 60 per cent of the thickness of the structure, though this rule
can be waived when the thickness exceeds around 800 mm. The centres between the
injection holes are often based upon experience of the operative but rarely go less than
300 mm and can go up to 600 mm. It is assumed that grout will fan outward from the
injection point equally in all directions (except where the crack varies in width, in which
case the flow of grout is likely to be unequal. Holes are never drilled directly at right
angles into the crack or joint.
Injection holes, as far as practicable, should be drilled through the centre of a brick or
stone block to ensure that an adequate seal can be made between the outside of the
injection hole and the injection lance or packer. Watch out for hollow bricks and blocks in
ancillary structures. Where high water flow is encountered the injection holes can be
drilled to divert water away from the mouth of the crack or joint allowing temporary seal
of hydraulic cement or similar being made over the crack or joint preventing the grout
escaping during injection.
222
Figure 6.10
Section through cracked masonry arch showing typical grout access hole
layout (note structural thickness and type of grout used to determine the
access hole centres)
Where seepage is occurring from a construction joint in brickwork or masonry tunnel
linings, it should receive the same treatment as a crack. Joints are likely to follow a more
predictable plane and so do not need to be drilled from both sides.
Figure 6.11
Elevation and structural drawing of an access hole pattern
for sealing a joint in brickwork. Note the mouth of the joint
should be pointed with hydraulic cement to prevent grout
escaping through the inside face
Point leaks
Isolated point leaks, including man-made holes through the structure such as those for
fixing bolts, cannot usually be grouted effectively by drilling directly into the leak point
because it may miss the water path. By drilling a pattern of grout injection holes around
the leak on three to four sides sufficiently deep to reach the mortar joints around the
water path is usually acceptable. The grout injection holes need not be more than two
courses deep and are usually drilled through the centre of the masonry unit.
CIRIA C671 • Tunnels 2009
223
Figure 6.12
Closely spaced access holes to deal with point leaks
6.4.3
Metal or pre-cast concrete segmental lined tunnels
The treatment of the flange joints in cast iron segmental lined tunnels and segment joints
in concrete lined tunnels where the caulking has failed and is a source of water ingress,
and the treatment of seepage from bolts holes and grout hole have already been discussed
in Section 7.3.1. However, they may be supplemented by grouting should the water
ingress be persistent or the treatment or repair not be successful. The treatment of cracks
in segmental linings may also be accomplished in a similar manner.
Low-viscosity polyurethane or acrylic grouting techniques using the grouts discussed in
Section 6.4.1.2 are considered to be the most appropriate solution, as they are able to
penetrate very fine cracks. They also can be used with fast setting times to create a
watertight bulb behind the lining close to the joint or crack.
It will be necessary to carry out localised, but systematic grouting where the water seepage
is occurring. If the seepage is occurring at a bolt hole then, depending on the bolt type
and lining design, various methods to seal the joint of the bolt hole are available. For
seepage at a grout plug, a single hole close to the hole is drilled through the segment and
grout injected behind the lining. A similar procedure is carried out for joints or cracks
that are a source of water ingress, but with a series of holes drilled though the segment
along the line of the caulking or joint at hole centres equal to the depth of the flange or
segment thickness. Any drilling into a concrete lining requires careful consideration
before the work to avoid causing other damage.
Re-grouting behind the tunnel lining to tighten up an original annular grout may be
considered an option but caution is advised because without proper investigation of the
problem and careful planning this may result in an expensive and time consuming job
with no guarantee of success. To be successful the main water paths and cause of the
seepage should be first identified and the grouting operation directed toward sealing
these. However numerous difficulties exist in identifying the primary water paths as water
seepage can occur at locations remote from these paths through very fine cracks in the
lining and cracks or fissures in the annular grout. Without a well-designed sequence of
grout injection this process could result in chasing water seepages through the tunnel in
areas that were otherwise dry. The risk can often by minimised by careful planning.
If this form of remedial grouting is considered, cementitious, particulate grouts would be
used with or without sand aggregates, in areas where known voids exist behind the tunnel
lining. By necessity, the grouting operation would use the existing grout holes as it is not
recommended to drill extra holes in the lining as these could become a seepage path in
the future. Where there are no discernable or very narrow voids, chemical grouts or the
224
use of micro-fine cement grouts may be required to ensure successful penetration of the
cracks, fissures or voids.
When using existing grout holes they will need to be drilled out before grouting to allow
the new grout to penetrate the possible voids or fissures within the existing grout. Strict
control of this type of work is necessary, and the use of experienced contractors. Control
measures include maintaining consistency in the mix details of the grout, quality of
workmanship, limiting grout pressures and grout takes (high grout pressures can cause
damage to the tunnel lining and surround ground or structures in the area) and a
systematic approach to the work to ensure adequate grout coverage. Ideally trials should
be carried out to determine the most appropriate approach to be adopted.
6.4.4
Concrete-lined tunnels
Water seepage through non-segmental lined cast in situ concrete lined tunnels, either mass
concrete or reinforced concrete tunnel linings, generally occurs through isolated cracks or
construction joints. The treatment of cracks and construction joints to prevent water
ingress can be dealt with in a similar manner to cracks in brick and masonry lined tunnels
discussed in Section 6.4.2.1 using an appropriate cementitious or chemical grout (see
Sections 6.4.1.1 and 6.4.1.2). The treatment of cracks to prevent water ingress may need
to be carried out in conjunction with structural repairs (see Section 5.4).
Seepage through the concrete matrix is generally not as common as water ingress through
cracks and joints, and will generally only occur in poor quality, permeable concrete.
Treatment may be carried out in a similar manner to treatment of general seepage in
brick or masonry tunnels as discussed in Section 6.4.2.1 and if necessary, in conjunction
with grouting of voids or fissures behind the tunnel lining as discussed in Section 6.4.5.
Where poor quality and permeable concrete is present the remedial grouting works may
need to be carried out in conjunction with patch repair works.
Isolated point leaks including man-made holes through the structure, such as fixing bolts,
cable or pipe ducts cannot usually be grouted by drilling into the leak point. There are
several different techniques available such as drilling a pattern of holes around the leak on
three to four sides sufficiently deep to almost reach the built in element although this
requires great care and carries an element of risk. Often the concrete around the element
needs to be broken out and an injection aid cemented in place. After the cement fill has
cured it can be injected. In new tunnels the problem will not occur as there are many
water stop elements available for building in during construction. The same applies for
the ingress of water through cable ducts as there are more advanced systems available for
sealing, using new hydrophilic materials, other than the common cement, bitumen, putty
or balloon methods now used.
6.4.5
Void grouting behind linings
The presence of voids behind a tunnel lining may have a significant impact on the tunnel
structure not only in terms of its structural stability but also as a likely cause for water
ingress through the lining.
Voiding behind tunnel linings may occur for many reasons, including:
voids left during excavation and construction of the tunnel lining. These may be open
or partially filled with rubble and/or caved in material and timber from the original
headings
lack of, partial or ineffective annulus or contact grouting behind linings, including
segmental, cast and built in place linings
CIRIA C671 • Tunnels 2009
225
washout of fines into the tunnel causing voiding in soft ground tunnel conditions
erosion of fissured rock discontinuities, or dissolution (as in limestone rocks)
existing voids within the ground, such as karstic limestone.
Whatever the cause of the voiding, the tunnel lining may show signs of distress due to
flexure and eccentric loading. Voids behind a tunnel lining may also be a source of water
ingress, with the void acting as a reservoir, even where the tunnel is above the natural
groundwater or perched water table, as the void may be recharged by rainfall.
Voids should be avoided for the structural integrity of the tunnel lining. Grouting would
typically be carried out to fill the voids behind the lining to ensure that ground loading on
the lining is evenly distributed and to prevent water ingress. If voids are present and it is
not practically possible to fill them, the tunnel lining may require localised extra support
to stabilise it from distortion and distress.
The selection of grouting technique will depend on the extent of voiding and anticipated
void or gap size, the properties of the soil or rock to be grouted and the purpose of
grouting, ie for structural stability of the lining or to prevent water ingress.
Depending on expected void sizes behind the tunnel lining, in areas where voids are
known to exist, various types of remedial grouting can be considered. Where grouting is
to be carried out to prevent water ingress, reference should be made to Tables 6.2 and 6.3,
which give an indication of the most appropriate grout and grouting technique to be used.
Cementitious, particulate grouts could be used with or without sand aggregates, but they
have a high dead weight when wet and their transient and longer-term influence on
structural stability should be taken into account. However there are lightweight materials
available and each problem should be considered as unique to find the optimal solution.
By necessity, the grouting operation would use the existing grout holes as it is not
recommended to drill extra holes in the lining that may become a seepage path in the
future. Where there are no discernable or very narrow voids, chemical grouts or the use
of micro-fine cement grouts may be better to ensure successful penetration of the cracks,
fissures or voids. A combination of grouts and grouting technique may also be required.
For example, in large voids the primary grouting may be with cementitious grouts, with or
without fillers, followed by secondary injection of a chemical grout to tighten up the
ground and prevent minor seepages.
Ideally trials should be first carried out to determine the most appropriate approach to be
adopted. Care should be taken to avoid a sudden inflow of high pressure mud and water
as the structural lining is penetrated and voids filled with water are tapped.
6.5
ALTERNATIVE MEASURES
6.5.1
Groundwater lowering (dewatering using well-points)
While this technique is not known to have been used as a method of controlling water
ingress in existing tunnels, groundwater lowering by the installation of abstraction wells,
or well-points has been used during the construction of new tunnels and other
underground structure. The technique may have an application in the short-term in
extreme cases of groundwater ingress while other ground or tunnel lining treatment
measures are done, or in the long-term as a permanent solution.
The technique involves the artificial lowering of the groundwater close to the tunnel by
drilling a series of wells or well-points on one or either side of the tunnel to a level deeper
226
than the tunnel invert and pumping out the groundwater using either surface or
submersible pumps. The aim is to reduce or cut the pressure head of the water ingress in
the tunnel.
The primary considerations for the design of a well-points system for dewatering are:
the depth of the tunnel below the existing groundwater table
well spacing, depth, number of wells and estimated pumping capacity required
geology and hydrogeology, including identification of aquifers and varying ground
permeability
soil subsidence and damage risk to adjacent structures
use of surface or submersible pumps
fines handling (within the pumping system)
treatment of any perched water tables
an acceptance of long-term maintenance and operational costs.
The amount of the drawdown of the groundwater can be effectively controlled by
selecting the appropriate pump size and number of wells in the area to be dewatered.
Careful consideration is required of the effects any dewatering technique will have on
buildings and structures in the immediate area, which if used on an existing tunnel, will
include the tunnel.
The other main disadvantage of using well-pointing in the long-term as a solution to
control water ingress in an existing tunnel is the running and maintenance costs that are
likely to be involved. Any system of wells installed to abstract the groundwater will need to
be constantly monitored and maintained, probably making it uneconomical as a remedial
measure.
Further information on the design and construction of groundwater control systems and
methods using well-pointing and other pumping techniques (sumps etc) can be found in
CIRIA C515 (Rawlings et al, 2000).
6.5.2
Electro-osmosis (dewatering)
Electro-osmosis is based on the principle of electrolysis using two electrodes (positive and
negative), which are inserted into the ground. The positive anode is located immediately
within or adjacent to the structure where the water seepage is occurring, while the
negative cathode is placed some distance away from the structure. A low voltage direct
current is then passed between cathode and anode. By placing a low voltage charge
between negative and positive electrodes, the water becomes ionised causing the water
molecules within the capillaries of the structure and surrounding ground to travel towards
the negative electrodes, which has the effect of preventing the water from intruding back
into the structure.
The movement of the ionised water in the capillaries can be both stronger than the
capillary action and the driving head for the water ingress. For example, proprietary
systems have been proven successful in preventing the penetration of water with
groundwater pressures up to 600 m head (60 bars).
Electro-osmosis systems have been used successfully in waterproofing basement,
diaphragm walls and other underground structures in Norway, Hong Kong and the USA
including the Oslo Central Railway Station for the Norwegian State Railways and on
CIRIA C671 • Tunnels 2009
227
several New York underground stations built in the 1930s. Some consider electro-osmosis
to have potential for development for use in controlling water ingress in tunnels, but the
technology is not proven in such circumstances now. Installation costs are likely to be high
but once installed an electro-osmosis system could be left with a minimal operating cost,
although regular monitoring would be important to ensure that the system remains
effective.
228
7
Recommendations and future needs
7.1
RECOMMENDATIONS FOR GOOD PRACTICE
At a strategic level, recommendations for the management of tunnel infrastructure are as
follows:
1
In the past a reactive approach to infrastructure management has frequently
prevailed, but this is now viewed as being a disruptive, inefficient and uneconomic
approach and not consistent with achieving sustainable transport and distribution
networks. There is considerable benefit in adopting a planned and more proactive
approach, setting out policies that aim to meet the long-term objective of preserving
the serviceability of ageing tunnel infrastructure well into the future.
2
If long-term objectives are to be achieved, it is necessary to develop and carry out
effective management procedures that support them in the short and medium term.
These should be geared toward identifying the maintenance needs of tunnels and
developing and justifying maintenance plans that make efficient use of resources. This
requires application of current good practice in the core activities of tunnel
inspection, assessment, maintenance, repair and improvement, and continual
assessment and feedback to ensure that procedures are refined.
3
Sufficient resources should be allocated to enable long-term tunnel asset management
objectives to be realised in an efficient way, ie to fund the clearance of any existing
maintenance and repair backlog, to carry out preventative works where appropriate,
and to achieve an overall steady state of fully serviceable condition for tunnel
infrastructure.
The proactive approach to management and maintenance is based on a good
understanding of a tunnel’s past history, its current condition and its likely future
requirements. This information can be used to formulate plans for preventative
maintenance, in which tunnel condition is maintained at the optimum level, rather than
reactive maintenance, which is carried out in response to unacceptable tunnel
performance and can be more costly and disruptive.
To provide the necessary support for achieving these strategic aims at the operational level
of tunnel management, recommendations are:
1
Those involved with the management and maintenance of tunnels should recognise
and develop an understanding of their special characteristics and needs, as distinct
from other types of structure (see Section 3.2). They should be more effective in
ensuring their continued serviceability and supporting good practice in asset
stewardship (see Section 3.4).
2
Good tunnel management decisions require the development and implementation of
reliable systems to manage current information, inspection, assessment, maintenance
and repair. Existing tunnels can then be kept in good condition and their capacity
fully used, minimising unnecessary and expensive unplanned works, reducing
environmental impact of closures and diversions, and avoiding increased repair and
early replacement costs (see Section 3.5).
CIRIA C671 • Tunnels 2009
229
230
3
A huge amount of valuable tunnel-related information has been lost for various
reasons, and to the detriment of these assets and the infrastructure as a whole. This
information represents an invaluable resource for future asset management decisions
and strategy. There is a need to research and collate existing information and, in
future, ensure records are kept up-to-date. Use of electronic information
management systems is a good way of doing this, but they should be updated
regularly and protected from obsolescence, and maintained alongside hard copies of
information rather than replacing them (see Section 3.5.1).
4
Efforts should be made to improve the quality and objectivity of visual inspections
because the resulting information provides the basis for all other activities. This can be
achieved through:
careful selection, thorough training and certification of inspectors (see Sections
3.6.2 and 4.3.3)
use of inspection methodologies that promote the accurate and objective
recording of data and allow reliable assessment and comparison of tunnel
condition, eg hand-held data-loggers and techniques for image capture (see
Sections 4.3.4 and 4.3.5)
optimisation of inspection programmes to direct resources where they are most
needed without compromising the safety and serviceability of any part of the
tunnel stock (see Sections 3.4.3.3 and 4.3.2)
5
Where more information is required on tunnel structure and performance, tunnel
investigation and monitoring should be carefully planned and executed to efficiently
obtain accurate and reliable data within necessary constraints. Selection of the most
appropriate investigation and assessment techniques is important, and requires some
knowledge and experience of their potential strengths and weaknesses, their practical
application and the nature and reliability of their results (see Section 4.4).
6
Assessing engineers should be aware of the capabilities and limitations of available
assessment techniques and understand how the parameters required for analysis are
influenced by the specific construction, materials and defects of tunnels. The
significance of hidden construction features and materials deterioration should be
appreciated and investigation of these factors undertaken where necessary to improve
confidence in results (see Sections 4.4.2 to 4.4.5).
7
Although preventative maintenance is often overlooked or given a low priority, it is
likely to have considerable benefit in the long-term. Asset managers should establish a
proactive regime of preventative maintenance for all tunnels to reduce the rate of
deterioration and deal with small problems before they become significant. Wherever
feasible, maintenance and repair should deal with the causes and effects of
deterioration (see Section 3.4.2).
8
Effective maintenance planning is necessary to ensure that tunnels remain in a fully
serviceable condition while optimising efficiency and minimising disruption to normal
services. Closures need to be planned in detail to ensure best use of the time available.
Where tunnel access is at a premium, it is advisable to co-ordinate all foreseeable
repair, maintenance, renewal and other works to the tunnel structure and associated
equipment. Where necessary, diversions of traffic or services should be planned in
advance and carefully managed (see Sections 3.4.3.6 and 5.1.1).
9
When selecting maintenance, repair and improvement techniques the potential effects
on the tunnel’s long-term performance should be carefully considered. Where changes
or additions are intended to work in a composite fashion with the existing structure, the
techniques and materials used should be compatible with it and not change its structural
action (see Section 5.1.3). There may be benefits in producing standard designs and
details, based on agreed good practice, for common types of tunnel repairs.
10 The environmental and ecological (and, where appropriate, heritage) impact of
tunnel management and maintenance works should also be considered and measures
taken to minimise undesirable effects. For example, by preventing and controlling
pollution, damage or disturbance of protected species, re-use and recycling of
materials and consideration of the relative energy efficiency of alternative repair
solutions (see Sections 3.6.3 and 3.6.4).
11 For any activity carried out in a tunnel a safe system of work should be generated to
mitigate all reasonably foreseeable risks. If this is not possible due to lack of necessary
information then sufficient investigation should be carried out to supply that
information, which could mean intrusive investigation in some cases (see Sections
3.6.1 and 5.1.4).
12 Maintenance, repair and improvement works in tunnels often require a different
approach from similar works carried out on above-ground structures. When
procuring such works it is important to engage experienced specialist contractors who
are familiar with the particular requirements of working in tunnel environments and
can demonstrate an awareness of the particular health and safety issues, working
practices and procedures that are likely to be required (see Section 5.1.1).
13 Unforeseen circumstances and variations in the original scope of works often cause
problems in the execution of tunnel works, and these can be minimised by adopting
practices that encourage co-operative working and being prepared for possible
changes. It is particularly important for all parties to maintain a flexible and cooperative approach during the works and to be proactive in seeking out potential
problems at an early stage and devising suitable solutions. Setting up the right project
team, suitable contractual and working arrangements, and thorough planning are
particularly crucial to the success of tunnel projects (see Section 5.1).
7.2
AREAS REQUIRING FURTHER RESEARCH AND FUTURE
NEEDS
Specific requirements for future research and development to assist with the management
and maintenance of tunnel infrastructure include the following:
1
The development of smart integrated asset management systems that may consider a
wider variety of factors than is possible now. These should help proactive maintenance
planning. The next generation of management systems should be mutually
compatible with existing systems and allow tunnel owners to buy into an over-arching
system that can deliver best value for the owners, society and the environment.
2
Improvements in the methods of capturing, recording and retrieving tunnel
condition information. For example, the use of hand-held computers/data-loggers as
part of the inspection process with direct access to database(s) of previous tunnel
condition information for direct comparison to prepare condition inspection reports
on-site.
3
The development of more efficient investigation techniques, in particular nondestructive investigation and monitoring techniques, to the point where they can be
applied routinely and efficiently to obtain adequately reliable data on tunnel
construction, condition and performance, and particularly the presence, location and
condition of hidden shafts and of ground conditions, voids and the presence of water
behind tunnel linings.
4
Further development of rapid tunnel scanning/surveying equipment could provide a
relatively quick and economical method of recording a detailed visual record of the
tunnel intrados, including accurate dimensional measurements. This information can
be incorporated in a database, providing a very comprehensive virtual reality survey
CIRIA C671 • Tunnels 2009
231
that can be easily viewed and used to assess and communicate information without the
need to enter the tunnel. Such a system has many potential advantages for improving
tunnel management, as well as providing a valuable source of information to assist
with improving safety and emergency planning.
232
5
Proper guidance (now lacking) is needed for the structural assessment of existing
tunnels, either in the form of a formal standard or an agreed methodology. Ideally
this would be based on an improved understanding of interaction and stress transfer
between the ground and tunnel linings and the effects of different construction
methods and features. Potentially the development of improved methods of
assessment could enable routine health checking of tunnels to identify those that are
particularly sensitive or susceptible to structural distress. This also allows
improvements in the design and work carried out in remedial and improvement
works.
6
To avoid the subjectivity intrinsic to condition appraisal based on traditional
inspection methods, there is scope for the development of intelligent
inspection/assessment techniques, which provide a more objective view of the changes
in tunnel condition over time. This would involve the periodic collection of specific
objectively measurable parameters. The significant parameter is the tunnel profile,
because this is directly linked to structural performance and can be determined
rapidly, with considerable accuracy and at limited cost using modern surveying
techniques. Other parameters, such as the stress state of the lining and/or ground,
may also be suitable. Analysis of this data could provide a more objective view of the
changes in the tunnel condition and performance over time. The relationship
between these parameters could be explored by using advanced algorithms, for
example, neural networks, and refined through gradual accumulation of further data.
A key requirement for this methodology to be of any use would be a very clear
definition of the measurement procedure and consistency in its execution. The aspect
of intelligent inspection and monitoring has considerable potential for further
development.
7
Improved understanding of the deterioration mechanisms that affect tunnel lining
materials, the potential rates of deterioration, effects on their physical characteristics
and their influence on tunnel performance. This would allow more reliable prediction
of maintenance and repair requirements for the structural fabric, and identification of
preventative maintenance opportunities.
8
With the benefit of hindsight, it has become apparent that inappropriate repairs have
been carried out in the past, with the consequence that further work has been
required to bring affected tunnels back into full serviceability. There is a need for
more data on the performance of both established and novel methods of repair and
improvement on tunnels, and independent assessment to establish their efficacy, in
particular their long-term effects. Opportunities should be taken to monitor repaired
tunnels by appropriate methods to provide the necessary data.
9
Low energy and sustainable maintenance and repair strategies and solutions should
be sought to reduce the environmental impact of tunnel ownership and maintenance,
and help to ensure their prolonged serviceability without resulting in an excessive
drain on limited resources.
References
ABDUNUR, C (1995)
“Direct assessment and monitoring of stresses and mechanical properties in masonry arch
bridges”
Arch Bridges, C Melbourne (ed), Thomas Telford, pp 327–335 (ISBN: 0-72772-048-1)
ADDIS, B and TALBOT, R (2001)
Sustainable construction procurement. A guide to delivering environmentally responsible projects
C571, CIRIA, London (ISBN: 978-0-86017-571-1)
ALLEN, G, ALLEN, J, ELTON, N, FAREY, M, HOLMES, S, LIVESEY, P and
RADONJIC, M (2003)
Hydraulic lime mortar for stone brick and block masonry
Donhead Publishing, UK (ISBN: 978-1-87339-464-9)
ANDERSEN, T and PAASKE, B J (2002)
“Fire safety in tunnels and selection of tunnel concept”
In: Proc 23rd ESReDA seminar on decision analysis: methodology and applications for safety of
transportation and process industries, European Safety, Reliability & Data Association, Delft
University, Netherlands
ASHURST, J and ASHURST, N (1988)
Mortars, plasters and renders in conservation
RIBA Publications Ltd, London (ISBN: 978-0-90786-602-2)
ASHURST, J (1997)
The technology and use of hydraulic lime
The Building Conservation Directory <www.buildingconservation.com>
ASTM (1991a)
Standard test method for in situ compressive stress within solid unit masonry estimated using flatjack
measurements
Standard C 1196-91, ASTM International, PA, USA <engineers.ihs.com>
ASTM (1991b)
Standard test method for in situ measurement of masonry deformability properties using the flatjack
method
Standard C 1197-91, ASTM International, PA, USA <engineers.ihs.com>
ATKINSON, K (2000)
Sewer rehabilitation and renovation techniques
Subterra Developments, 2000
BADDOO, N R (1996)
Castings in construction
P172, The Steel Construction Institute (SCI), Berkshire (ISBN: 978-1-85942-041-6)
BAECKMANN, W V, SCHWENK, W and PRINZ, W (eds) (1997)
Handbook of cathodic corrosion protection
3rd edn, Gulf Professional Publishing, Elsevier, Oxford (ISBN: 978-0-88415-056-5)
BARTON, N, LIEN, R and LUNDE, J (1974)
“Engineering classification of rock masses for the design of tunnel support”
Rock Mechanics and Rock Engineering, vol 6, 4, December, Springer Wien, pp 189–236
CIRIA C671 • Tunnels 2009
233
BASSETT, R H, KIMMANCE, J P and RASMUSSEN, C (1999)
“An automated electrolevel deformation monitoring system for tunnels”
Geotechnical engineering, 137, 3, Institution of Civil Engineers, Thomas Telford, London,
pp 117–125
BAUER, G (1985)
“How to control groundwater in tunnelling projects”
Tunnels and Tunnelling, 17 June, pp 55–57 <http://www.tunnelsonline.info/>
BATES, W (1984)
Historical structural steelwork handbook
The British Constructional Steelwork Association Ltd, London (ISBN: 0-85073-015-5)
BCT (2002)
Bats, development and planning in England
Specialist support series, Bat Conservation Trust <www.bats.org.uk>
BEARD, A and CARVEL, R (eds) 2005
The handbook of tunnel fire safety
Thomas Telford, London (ISBN: 978-0-72773-168-5
BELL, F G (1975)
Site investigations in areas of mining subsidence
Butterworth-Heinemann, London (978-0-40800-164-9)
BIDDISCOMBE, P (2005)
3D Laser scan tunnel inspections keep expressway infrastructure project on-schedule
Trimble white paper, Trimble Engineering and Construction Group, OH, USA
BIENIAWSKI, Z T (1979.)
Engineering rock mass classification
Wiley-Interscience Pub1ication, New York: John Wiley & Sons
BROOMHEAD, S F and CLARK, G W (1995)
“Strengthening masonry arches”
Bridge modification, B Pritchard (ed), Thomas Telford, London, pp 174–184
(ISBN: 978-0-72772-028-3)
BRE (1992)
Review of excavation methods and their implications for the near-field barrier of a deep underground
repository
PECD 7/9/588, Building Research Establishment and Mott MacDonald (unpublished)
BRE (1995)
Damage to structure from ground-borne vibration
Digest 403, Building Research Establishment, Garston (ISBN: 1-86081-002-0)
BRE (1995)
Masonry and concrete structures: measuring in situ stress and elasticity using flat jacks
Digest 409, Building Research Establishment, Garston (ISBN: 978-1-86081-033-6)
BRE (1997)
Measuring the compressive strength of masonry materials: the screw pull out test
Digest 421, Building Research Establishment, Garston (ISBN: 978-1-86081-139-5)
BRE (2000)
Corrosion of steel in concrete, part 2: Investigation and assessment
Digest 444, Building Research Establishment, Garston (ISBN: 1-86081-361-5)
BRE (2005)
Concrete in aggressive ground
3rd edn, Special Digest 1, Building Research Establishment, Garston
234
BROOMFIELD, J P (2004)
Electrochemical realkalisation of steel – state of the art report reinforced concrete
Technical Note 9, Corrosion Prevention Association (CPA), Surrey (ISBN: 0-95449-940-9)
THE BRITISH TUNNELLING SOCIETY and AND THE INSTITUTION OF CIVIL
ENGINEERS (2000)
Specification for tunnelling
Thomas Telford Ltd, London (ISBN: 978-0-72772-865-4)
BRITISH TUNNELING SOCIETY (2004)
Tunnel lining design guide
Thomas Telford Ltd, London (ISBN: 978-0-72772-986-6)
BULMAN, T N C, SEAGRAVE, J H, MANNING, H D, BURTON, R A H, JAEGER, C,
COLLINGRIDGE, V H and CUTHBERT (1955)
“Investigations into the design of pressure tunnels in London Clay”
ICE Proceedings, vol 4, 4, Thomas Telford, London, pp 456–471
BURGESS, N, FAGENTS, J and PATERSON, J (2002)
“Northern Line Tunnel reconstruction at Old Street”
ICE Proceedings – Transport, vol 153, 1, Thomas Telford, London, pp 1–11
BUSSELL, M (1997)
Appraisal of existing iron and steel structures
Publication 138, The Steel Construction Institute (SCI), Berkshire (ISBN: 1-85942-009-5)
CADEI, J M C, STRATFORD, T J, HOLLOWAY, L C and DUCKETT, W H (2004)
Strengthening metallic structures using externally bonded fibre-reinforced composites
C595, CIRIA, London (ISBN: 978-0-86017-595-7)
CBDG (2002)
Guide to testing and monitoring the durability of concrete structures
Technical Guide 2, Concrete Bridge Development Group, Surrey
CHAKRABARTI, B, YATES, T and LEWRY, A (1996)
“Effect of fire-damage in natural stonework of buildings”
Construction and Building Materials, vol 10, 7, October 1996, Elsevier Science Ltd, pp
539–544
CHEW, C and ROBERTS, K (2005)
“Inspection techniques for the underground environment”
In: Proc conf Best practice in tunnels management, October (unpublished)
CHOQUET, C, SCHWENZFEIER, A and LAMONT, D R (2005)
“Channel Tunnel – monitoring strategy for the tunnel lining”
In: Proc 31st ITA-AITES World Congress Tunnel: underground space uses, analysis of the past and
lessons for the future, May 2005, Istanbul, vol 2, pp 1211–1214
CRADDOCK, P (2002)
Guide to site welding
P161, The Steel Construction Institute (SCI), Berkshire (IBN: 978-1-85942-052-2)
CONCRETE SOCIETY (1989a)
Cathodic protection of reinforced concrete
TR36, The Concrete Society, Surrey
CONCRETE SOCIETY (1989b)
Analysis of hardened concrete: a guide to test procedures and interpretation of results
TR32, The Concrete Society, Surrey
CONCRETE SOCIETY (1990)
Assessment and repair of fire damaged concrete structures
TR33, The Concrete Society, Surrey
CIRIA C671 • Tunnels 2009
235
CONCRETE SOCIETY (1991)
Patch repair of reinforced concrete subject to reinforcement corrosion
TR38, The Concrete Society, Surrey
CONCRETE SOCIETY (1997)
Guide to surface treatments for protection and enhancement of concrete
TR50, The Concrete Society, Surrey
CONCRETE SOCIETY (2000)
Diagnosis of deterioration in concrete structures - identification of defects, evaluation and development
of remedial action
TR54, The Concrete Society, Surrey
CONCRETE SOCIETY (2004)
Design guidance for strengthening concrete structures using fibre composite material
2nd edn, TR55, The Concrete Society, Surrey
CONCRETE SOCIETY (2005)
Mortars for masonry: Guidance on specification, types, production and use
Publication 159, The Concrete Society, Surrey
COVENTRY, S and WOOLVERIDGE, C (1999)
Environmental good practice on site
C650, CIRIA, London (ISBN: 978-0-86017-650-5)
COVENTRY, S, WOOLVERIDGE, C and HILLIER, S (1999)
The reclaimed and recycled construction materials handbook
C513, CIRIA, London (ISBN: 978-0-86017-513-1)
CRISFIELD, M A and PACKHAM, A J (1985)
A mechanism program for computing the strength of masonry arch bridges
Research Report 124, Transport Research Laboratory, Crowthorne
CURTIS, D J (1974)
“Visco-elastic tunnel analysis”
Tunnels and Tunnelling, November <http://www.tunnelsonline.info/>
CURTIS, D J (1976)
“The circular tunnel in elastic ground, discussion”
Geotechnique, 26, Thomas Telford, London, pp 231–237
DEERE, D U (1989)
“Rock quality designation (RQD) after 20 years”
MS: waterways experimental station, Contract Report GL-89-1, US Army Corps Engrs
Vicksburg
DE VIVO, L and ROSATI, L (1998)
“Ultimate strength analysis of reinforced concrete sections subject to axial force and
biaxial bending”
Computer Methods in Applied Mechanics and Engineering, vol 166, 3–4, 27 November 1998,
Elsevier Science B V, pp 261–287
DEPARTMENT FOR TRANSPORT (2005)
Management of highway structures: a code of practice (paperback)
HMSO, (ISBN: 978-0-11552-642-8)
DILL, M J (2000)
A review of testing for moisture in building elements
C583, CIRIA, London (ISBN: 978-0-86017-583-4)
236
DoT (1986)
Report on the derailment and fire that occurred on 20th December 1984 at Summit Tunnel in the
London Midland Region of British Railways
Report by the Railway Inspectorate on behalf of the Department of Transport, HMSO
DOUGLAS, T H and ARTHUR, L J (1983)
A guide to the use of rock reinforcement in underground excavations
R101, CIRIA, London (ISBN: 978-0-86017-212-3)
DUDDECK, H and ERDMANN, J (1985)
“Structural design models for tunnels in soft soil”
Underground Space, vol 9, 13 May, pp 246–259, USA
EA (2000)
Works in, near or liable to affect watercourses
PPG5, Environment Agency, Bristol <www.netregs.gov.uk>
EA (2001a)
Guide to the prevention of pollution
PPG1, Environment Agency, Bristol <www.netregs.gov.uk>
EA (2001b)
Working at construction and demolition sites
PPG6, Environment Agency, Bristol <www.netregs.gov.uk>
EA (2002)
Maintenance of structures over water
PPG23, Environment Agency, Bristol <www.netregs.gov.uk>
EDGELL, G (ed) (2005)
“Testing of ceramics in construction”
Testing in Construction Series, vol 2, Whittles Publishing (ISBN: 1-870325-43-5)
EINSTEIN, H H and SCHWARTZ, C W (1979)
“Simplified analysis for tunnel supports”
Journal of Geotechnical Engineering Division, vol 105, 4, April, ASCE, Reston, USA, pp
499–518
ELLIS, P (2001)
Guaging lime mortars
Building Conservation Directory <www.buildingconservation.com>
FERRO, M L (1980)
“The russack system for brick and mortar description: a field method for assessing
masonry hardness”
Technology & Conservation, vol 5, 2, summer, pp 32–35
GNANAKRISHNAN, N and LAWTHER, R (1990)
“Performance of masonry walls exposed to fire”
In: Proc conf 5th North American masonry, vol 3, University of Illinois, Champaign/Urbana,
3–6 June 1990, pp 901–914
GREATHEAD, J H (1895)
“The City and South London railway; with some remarks upon subaqueous tunnelling by
shield and compressed air”
ICE Proceedings (minutes), vol 123, 1896, Thomas Telford, London, pp 39–73
GRIPPER, C F (1879)
Railway tunnelling in heavy ground
E & F N Spon, London
CIRIA C671 • Tunnels 2009
237
GUTHRIE, P M, WOOLVERIDGE, A C and PATEL, V S (1997)
Waste minimisation in construction – site guide
SP133, CIRIA, London (ISBN: 978-0-86017-482-0)
HAACK, A (1991)
“Water leakages in subsurface facilities: required watertightness, contractual matters and
methods of redevelopment”
Tunnelling and Underground Space Technology, vol 6, 3, pp 273–282
HAIDER, G, RICHARDS, L R, THORBURN, S, TURNER, A J, HOWARD, A J,
PULLER, M J and HUGHES, F (1987)
“Netherton Canal Tunnel: Recent repairs and failed sections”
ICE Proceedings, vol 82, 4, Thomas Telford, London, pp 1231–1239
HAMMOND, M (1981)
Bricks and brick-making
2nd (rev) edn, Shire Publications, Buckinghamshire (ISBN: 0-74780-067-7)
HAY, D and FITZMAURICE, M (1898)
“The Blackwall Tunnel”
ICE Proceedings, vol 130, 1897, Thomas Telford, London, pp 50–79
HEALEY, P R and HEAD, J M (1984)
Construction over abandoned mine workings
SP32, CIRIA, London (ISBN: 978-0-86017-218-5)
HENDRY, A W (1990)
Masonry properties for assessing arch bridges
CR244, Transport Research Laboratory, Wokingham, UK
HENDRY, A W (1998)
Structural masonry
2nd edn, Palgrave-MacMillan, Hampshire (ISBN: 978-0-33373-309-7)
HEYMAN, J (1966)
“The stone skeleton: Structural engineering of masonry architecture”
International journal of solids and structures, vol 2, Elsevier, Oxford, pp 249–279
HEYMAN, J (1982)
The masonry arch
1st edn, Ellis Horwood Ltd, Cirencester
HIEATT, M, ELLIS, J and LOCKE, K (2005)
“Experiences in the inspection and maintenance of a raw water tunnel over thirty years
of operation”
In: Proc conf Underground construction: tunnel maintenance and remediation, 26–27 October
2005, Olympia, London
HIGHWAYS AGENCY (1990)
Impregnation of concrete highway structures
Advice Note BA 33/90, Highways Agency, London
HMSO (1997)
The confined spaces regulations
SI 1997/1713, The Stationary Office (ISBN: 0-11064-643-6)
HOEK, E (2007)
Practical rock engineering
<www.rocscience.com/hoek/PracticalRockEngineering.asp>
HOEK, E and BRAY, J (1981)
Rock slope engineering
3rd edn, Institution of Mining and Metallurgy, Spon Press (ISBN: 978-0-41916-010-6)
238
HOEK, E and BROWN, E T (1990)
Underground excavations in rock
1st edn, Institution of Mining and Metallurgy, Spon Press (ISBN: 978-0-41916-030-4)
HOEK, E, KAISER, P K and BAWDEN, W F (1995)
Support of underground excavations in hard rock
1st edn, Taylor & Francis, Balkema, Rotterdam (ISBN: 978-9-05410-187-1)
HSE (2003a)
Safety and health in mines research advisory board
Annual Review 2002, Safety and health in mines research advisory board (SHMRAB),
<www.hse.gov.uk>
HSE (2003b)
The design and construction of water impounding plugs in working mines
Health and Safety Executive, Norwich <www.hse.gov.uk/mining/circulars/waterplu.htm>
HUGHES, T G and BLACKLER, M J (1997)
“A review of the UK masonry arch assessment methods”
Structures and Buildings, vol 122, 3, Institution of Civil Engineers, pp 305–315
HUGHES, T G and PRITCHARD, R (1994a)
“In situ measurements of dead load and live load stresses in a masonry arch”
Engineering Structures, vol 20, 1, Elsevier, Oxford, pp 5–13
HUGHES, T G and PRITCHARD, R (1994b)
“An investigation of flat jack flexibility in the determination of in situ stresses”
In: Proc 10th int conf Brick/block masonry, Calgary, Canada, pp 569–578
INGASON, H and LONNERMARK, A (2005)
“Heat release rates from heavy goods vehicle trailer fires in tunnels”
Fire Safety Journal, vol 40, 7, October, Elsevier, Oxford, pp 646–668
IRVINE, D J and SMITH, R J H (1983)
Trenching practice
2nd edn, R97, CIRIA, London (ISBN: 978-0-86017-192-8)
IStructE (1996)
Appraisal of existing structures
2nd edn, Institution of Structural Engineers, London
ISRM (1978)
Suggested methods for the quantitative description of discontinuities in rock masses
EUR 8, International Society for Rock Mechanics, Lisbon, Portugal
<http://www.isrm.net/gca/?id=51>
ISRM (2007)
The complete ISRN suggested methods for rock characterisation, testing and monitoring: 1974–2006
Ulusay, R and Hudson, J A (eds), International Society for Rock Mechanics, Lisbon
ISRM (1987)
Suggested methods for rock stress determinations
EUR 4, International Society for Rock Mechanics, Lisbon, Portugal
<http://www.isrm.net/gca/?id=51>
ITA (1987)
“Civil engineering records of underground structures”
Tunnelling and Underground Space Technology, vol 2, 3, Elsevier, Oxford, pp 307–309
ITA (1998)
“Working Group on general approaches to design of tunnels: Guidelines for the design of
tunnels”
Tunnelling and Underground Space Technology, vol 3, Elsevier, Oxford, pp 237–249
CIRIA C671 • Tunnels 2009
239
JARDINE, F M (ed) (2001)
Response of buildings to excavation-induced ground movements. Proceedings of the international
conference held at Imperial College, London, UK on 17–18 July 2002
SP199, CIRIA, London (ISBN: 978-0-86017-810-1)
KAY, P (1990)
A guide to railway research and sources for local railway history
SSG Publications Ltd, Teignmouth
KIMMERANCE, J P, OSBOURNE, N H and MAY, M R (1999)
“The design, application and performance of in-tunnel monitoring systems”
In: Proc 5th int symposium on field measurements in geomechanics, Singapore, C F Leung, S A
Tan and K K Phoon (eds), Balkema, Rotterdam, pp 401–406
KIRKLAND, C (2002)
“The fire in the Channel Tunnel”
Tunnelling and Underground Space Technology, vol 17, Elsevier, Oxford, pp 129–132
KODUR, V K R (1999)
Fire performance of high-strength concrete structural members
Construction Technology update no 31, December, NRCC, Ottawa, Canada
KODUR, V K R and SULTAN, M A (2004)
“Structural behaviour of high strength concrete columns exposed to fire”
In: Proc int symposium on high performance and reactive powder concretes, Sherbrooke, Quebec,
September 1998, Report NRCC-41736, pp 217–232
LAWSON, R M and NEWMAN, G M (1990)
Fire resistant design of steel structures – A handbook to BS 5950: Part 8
Publication 080, The Steel Construction Institute (SCI), Berkshire (ISBN: 1-87000-448-5)
LIE, T T and WOOLLERTON, J L (1988)
Fire resistance of reinforced concrete columns: test results
Internal report no 569, NRCC, Ottawa, Canada
MACK, R C and GRIMMER, A (2000)
Assessing cleaning and water-repellent treatments for historic masonry buildings
Preservation brief 1, National Parks Service, US Dept of the Interior, Washington DC
MASTERS-WILLIAMS, H, HEAP, A, KITTS, H, GREENSHAW, L, DAVIS, S, FISHER, P,
HENDRIE, M and OWENS, D (2001)
Control of water pollution from construction sites. Guidance for consultants and contractors
C532, CIRIA, London (ISBN: 978-0-86017-532-2)
MCDOWELL, P W, BARKER, R D, BUTCHER, A P, CULSHAW, M G, JACKSON, P D,
MCCANN D M, SKIPP, B O, MATTHEWS, S L and ARTHUR, J C R, (2002)
Geophysics in engineering investigations
C562, CIRIA, London (ISBN: 978-0-86017-562-9)
MCKIBBINS, L D, MELBOURNE, C, SAWAR, N and SICILIA, C (2006)
Masonry arch bridges – condition appraisal and remedial treatment
C656, CIRIA, London (ISBN: 978-0-86017-656-5)
MCQUEEN, L B (2005)
Mechanism of spalling stress failure in Sydney Tunnels
Golder Associates (unpublished)
MITCHELL-JONES, A J (2004)
Bat mitigation guidelines
IN13.6, Natural England (ISBN: 1-85716-781-3) <www.english-nature.org.uk>
240
MITCHELL, J and JARDINE, F M (2002)
A guide to ground treatment
C573, CIRIA, London (ISBN: 978-0-86017-573-5)
MOLESWORTH, G L (1862)
Pocket-book of useful formulae and memoranda for civil and mechanical engineers
E & F N Spon, London
NCB (1975)
Subsidence engineer’s handbook
National Coal Board
MUIR WOOD, A M (1975)
“The circular tunnel in elastic ground”
Geotechnique, 25, 1, Thomas Telford, London, pp 115–127
MUIR WOOD, A M (2000)
Tunnelling: management by design
1st edn, Taylor & Francis, London (ISBN: 978-0-41923-200-1)
MUNCKE, M and ZUBER, P (2004)
“Safety in railway tunnels: Current UIC recommendations and European Regulations”
Fire Safety in Tunnels, no 324, October, ITA/PIARC
NCB (1982)
The treatment of disused mine shafts and adits
Mining Department, National Coal Board
NETWORK RAIL (2004a)
Data gathering for the management of tunnels
Network Rail good practice guide (internal guidance document, not in public domain)
NETWORK RAIL (2004b)
Location and treatment of tunnel shafts
Network Rail good practice guide (internal guidance document, not in public domain)
NEVILLE, A M (1995)
Properties of concrete
4th edn, Prentice-Hall, Pearson Education Limited, UK (ISBN: 978-0-58223-070-5)
NEWTON, J, WILLIAMS, C, NICHOLSON, B, VENABLES, R, WILLETTS, R and
MOSER, B (2004)
Working with wildlife
C587, CIRIA, London (ISBN: 978-0-86017-587-2)
NHBC (2003)
Building near trees
Section 4.2, National House Building Council Standards
NCSCCMI (1994)
Assessing the condition and repair alternatives of fire-exposed concrete and masonry members
Fire protection and planning report, National Codes and Standards Council of the
Concrete and Masonry Industries, Skokie, Illinois
O’ROURKE, T D (1984)
Guidelines for tunnel lining design
ASCE, Reston, USA (ISBN: 0-87262-402-1)
PATTERSON, D and PERRY, J (1998)
“Geotechnical data and asset management systems for highways”
In: Proc conf Maintenance engineers, Nottingham, 9–11 September, Surveyor and Municipal
Journal, London
CIRIA C671 • Tunnels 2009
241
PEARSON, S and PATEL, R G (2002)
Repair of concrete in highway bridges – a practical guide
HA Y100533, Transport Research Laboratory, Berkshire
PROCTER, D and FILLINGHAM, K C (1983)
Structural tests on a renovated brickwork sewer at St Helens, Merseyside
WRc External Report ER107E, June 1983
PERRY, J and WEST, G (1996)
Sources of information for site investigations in Britain
TRL 192 (revision of TRL Report LR403), Transport Research Laboratory, Berkshire
(ISBN: 1-84608-191-2)
PRICE, D G, MALKIN, A B and KNILL, L J (1969)
“Foundations of multi-storey blocks on coal measures with special reference to old
workings”
Journal of Engineering Geology, vol 1, 4, pp 271–322, Geological Society of London
PRITCHETT, I (2003)
“The modern face of lime”
Journal of Building Limes Forum, vol 10 <www.buildinglimesforum.org.uk/>
RAILTRACK (1996)
Guide to tunnel management
(Internal guidance document, not in public domain)
RAWLINGS, C G, HELLAWELL, E E and KILKENNY, W M (2000)
Grouting for ground engineering
C514, CIRIA, London (ISBN: 978-0-86017-514-8)
READ, G F, VICKRIDGE, I (1996)
Sewers: repair and renovation
X171, Butterworth Heinemann, Elsevier, Oxford (ISBN: 978-0-34054-472-3)
RILEM TC (1994)
“Strength/elasticity tests on masonry based on the flat jack”
RILEM Recommendations for the testing and use of constructions materials, paper LUM D3
(1991), E & F N Spon, London, pp 506–508 (ISBN: 2-35158-011-7)
ROACH, M J and BRUNEL, M (SIR) (1998)
“The strengthening of Brunel’s Thames Tunnel”
ICE Proceedings, vol 129, 2, Thomas Telford, London, pp 106–115
ROBERTS, T M, HUGHES, T G, DANDAMUNDI, V R and BELL, B (2006)
“Quasi-static and high-cycle fatigue strength of brick masonry”
Const. and Building Mats, vol 20, Elsevier B V, Oxford, pp 603–614
ROCK, T and IRELAND, T (2005)
“Refurbished road tunnels meet modern demands by application of new technology and
operating procedures”
In: Proc conf Underground construction: tunnel maintenance and remediation, 26–27 October,
Olympia, London
ROSE, D (1982)
“Revising Terzaghi’s tunnel rock load coefficients”
In: Proc 23rd US symposium on rock mechanics, AIME, New York, 1982, pp953–960
RPA (2000)
Risk reduction strategy and analysis of advantages and drawbacks for acrylamide
Department of Environment, Transport and Regions, March 2000
242
RUSSEL, H (2004)
“ITA Guidelines for structural fire resistance of road tunnels”
Fire safety in tunnels, ITA/PIARC Routes-Roads magazine, no 324, October
SCHROEDER, F (2002)
“The influence of bored piles on existing tunnels – a case study”
Ground Engineering, July 2002, vol 35, no 7
SHIRLAW, N, WEN, D and ONG, J (2006)
“The importance of the choice of gasket type and grout mix in segmental tunnel linings”
In: Proc int conf and exh on tunnelling and trenchless technology in the 21st century, 7–9 March,
Malaysia
SIMMS, F W (1896)
Practical tunnelling
4th edn, Crosby Lockwood, London
SIMPSON, B, BLOWER, T, CRAIG, R N and WILKINSON W B (1989)
The engineering implications of rising groundwater levels in the deep aquifer beneath London
SP69, CIRIA, London (ISBN: 978-0-86017-303-8)
SMITH, M R (ed) (1999)
Stone: Building stone, rock fill and armourstone in construction
SP16, Geological Society Engineering Group, London (ISBN: 1-86239-029-0)
SNH (2004)
Bats in buildings. A guide for building professionals 2004
Scottish Natural Heritage <www.snh.org.uk/pdfs/species/BatsBuildings.pdf>
SOWDEN, A M (1990)
The maintenance of brick and stone structures
1st edn, Taylor and Francis, London (ISBN: 978-0-41914-930-9)
STARKEY, A, IVANOVIC, A, RODGER, A A and NEILSON, R D (2003)
“Condition monitoring of ground anchorages by dynamic impulses: GRANIT System”
Meccanica, vol 38, 2, March, Springer, The Netherlands, pp 265–282
STEELE, K (2004)
Environmental sustainability in bridge management
IP14/04, Building Research Establishment, Watford (ISBN: 978-1-86081-732-8)
STEELE, K, COLE, G, PARK, B, CLARKE, G and HARDING, J (2003)
“Environmental impact of brick arch bridge management”
Structures and Buildings, 156, SB3, August, Institution of Civil Engineers, pp 273–281
SWANNEL, N G (2003)
“Tunnel inspection, repair and maintenance – an international perspective”
In: Proc 3rd int seminar for safety of infrastructure, Korea Infrastructure Safety and Technology
Corporation, Seoul
SWANNELL, N G (2003)
“Managing tunnel sustainability”
Tunnels and Tunnelling International, vol 35, 5, May, Polygon Media, Sevenoaks, UK, pp
51–53
TEMPLE, B P and KENNEDY, A (1989)
Masonry structures project: The engineering properties of old bricks
Report TM-CES-128, British Rail Research
CIRIA C671 • Tunnels 2009
243
TERZAGHI, K (1946)
“Rock defects and loads on tunnel supports”
In: Rock tunneling with steel supports, R V Proctor and T L White (eds), 1, Youngstown, OH:
Commercial Shearing and Stamping Company, pp17–99
TEUTONICO, J-M (ed) (1997)
The English Heritage directory of building limes
English Heritage, Donhead Publishing, Shaftesbury (ISBN: 978-1-87339-421-2)
THE STATIONARY OFFICE (2007)
Design manual for roads and bridges
Volume 1
BA 28/92 Evaluation of maintenance costs in comparing alternative designs for
highway structures
Section 1 Part 2
BD 36/92 Evaluation of maintenance costs in comparing alternative designs for
highway structures
Section 1 Part 1
Volume 3
BD 53/95 Inspection and records for road tunnels
Section 1 Part 6
Volume 2
BA 33/90 Paints and other protective coatings. Impregnation of concrete highway
structures (superseeded by BD 43/03)
Section 4
BD 43/03 Paints and other protective coatings. The impregnation of reinforced and
prestressed concrete highway structures using hydrophobic pore-lining impregnants
Section 4 Part 2
TILLY, G P, MATTHEWS, S J, DEACON, D, DE VOY, J and JACKSON, P A (2007)
Iron and steel bridges: Condition appraisal and remedial treatment
C664, CIRIA, London (ISBN: 978-0-86017-664-0)
TUNNELKOMMISSIONEN (1998)
Miljö i Grund och Botten - Erfarenheter från Hallandsåsen, Slutrapport från Tunnelkommissionen
Final report, SOU 1998: 137, Tunnel Commission on the incident at Hallandsås.
Available from: <http://miljo.regeringen.se/propositionermm/sou/index.htm>
TURNER, M J (1997)
Integrity testing in piling practice
R144, CIRIA, London (ISBN: 978-0-86017-473-8)
TYLER, M and LAMONT, D R (2007)
“Construction health and safety”
1st edn, Construction law handbook, Thomas Telford, London, pp 467–532
UIC (2001)
Safety in railway tunnels; recommendations for safety measures
International Union of Railways, Ernst Basler & Partners Ltd, Zurich
ULM, F J, ACKER, P and LEVY, M (1999)
“Chunnel fire. II: analysis of concrete damage”
Journal of Engineering Mechanics, vol 125, 3, March, ASCE, Reston, USA, pp 283–289
UNDERWOOD, B D and REES, C W (1985)
“Brick sewer renovation”
The Public Health Engineer Journal, vol 13, 1, January, Stirling Publishing, London
VILLAESCUSA, E, SETO, M and BAIRD G (2002)
“Stress measurements from oriented core in Australia”
In: Proc 5th int workshop on the application of geophysics in rock engineering, 7 July, Toronto,
Canada
244
WALKER, J R et al (1860)
“Description of the works on the Netherton Tunnel branch of the Birmingham Canal”
ICE Proceedings (minutes), vol 19, 1860, Thomas Telford, London, pp 277–282
WELTHAM, A J and HEAD, J (1983)
Site investigation manual
SP25, CIRIA, London (ISBN: 978-0-86017-196-6)
WRC, 2001
Sewerage rehabilitation manual
Water Research Centre (this reference is currently undergoing revision)
CIRIA C671 • Tunnels 2009
245
Further reading
HAACK, A, SCHREYER, L and JACKEL, G (1995)
State of the art of non-destructive testing methods for determining the state of a tunnel lining
Report to International Tunnelling Association Working Group (No 6) on maintenance
and repair of underground structures, Tunnelling and Underground Space Technology, vol 10,
4, Elsevier, Oxford, pp 413–435
ITA (1991)
“Report on the damaging effects of water on tunnels during their working life”
Report of International Tunnelling Association Working Group (No 6) on maintenance
and repair of underground structures, Tunnelling and Underground Space Technology, vol 6,
1, Elsevier, Oxford, pp 11–76
ITA (2001)
“Study of methods for repair of tunnel linings”
Report of International Tunnelling Association Working Group (No 6) on maintenance
and repair of underground structures, final report and documents. Available from:
<www.ita-aites.org>
RICHARDS, J A (1998)
“Inspection, maintenance and repair of tunnels: international lessons and practice”
Tunnelling and Underground Space Technology, vol 13, 4, Elsevier, Oxford, pp 369–375
246
Regulations and standards
REGULATIONS
Work at Height Regulations 2005 (SI No 735)
<http://www.opsi.gov.uk/si/si2005/20050735.htm>
The Lifting Operations and Lifting Equipment Regulations 1998 (‘LOLER’) (SI No 2307)
(ISBN: 0-11079-598-9)
Provision and Use Of Work Equipment Regulations 1998 (‘PUWER’) (SI No 2306)
(ISBN: 0-11079-599-7)
The Construction (Design and Management) Regulations 2007 (CDM 2007) Approved
Code of Practice (ACOP) (ISBN: 978-0-71766-223-4)
STANDARDS
BS 15:1906 Standard specification for structural steel for bridges and general building construction.
Revised BS 15:1961 Mild steel for general structural purposes
BS 4360:1990 Specification for weldable structural steels (withdrawn).
Replaced by BS 7613:1994, BS 7668:1994, BS EN 10029:1991, BS EN 10113:1993 (parts
1–3), BS EN 10155:1993, BS EN 10210-1:1994
BS 449-2:1969 Specification for the use of structural steel in building. Metric units
BS 4604-1:1970 Specification for the use of high strength friction grip bolts in structural steelwork.
Metric series. General grade
BS 2846-4:1976 Guide to statistical interpretation of data. Techniques of estimation and tests
relating to means and variances
BS 5400-9.1:1983 Steel, concrete and composite bridges. Bridge bearings. Code of practice for
design of bridge bearings
BS 8110-3:1985 Structural use of concrete. Design charts for singly reinforced beams, doubly
reinforced beams and rectangular columns
BS 8081:1989 Code of practice for ground anchorages
BS 1452:1990 Specification for flake graphite cast iron (no longer current)
BS 6270-3:1991 Code of practice for cleaning and surface repair of buildings. Metals (cleaning
only)
BS 7361-1:1991 Cathodic protection. Code of practice for land and marine applications (no longer
current)
BS 5930:1999 Code of practice for site investigations
BS 8221-1:2000 Code of practice for cleaning and surface repair of buildings. Cleaning of natural
stones, brick, terracotta and concrete
BS 5950-1:2001 Structural use of steelwork in building. Code of practice for design – rolled and
welded sections
BS 6164:2001 Code of practice for safety in tunnelling in the construction industry
CIRIA C671 • Tunnels 2009
247
BS 5950-8:2003 Structural use of steelwork in building. Code of practice for fire resistant design
BS 6000-1:2005 Guide to the selection and usage of acceptance sampling systems for inspection of
discrete items in lots. General guide to acceptance sampling
BS 5628-1:2005 Code of practice for the use of masonry. Structural use of unreinforced masonry
BS EN 1011-1:1998 Welding. Recommendations for welding of metallic materials. General
guidance for arc welding
BS EN 12696:2000 Cathodic protection of steel in concrete
BS EN 1537:2000 Execution of special geotechnical work. Ground anchors
BS EN 12954:2001 Cathodic protection of buried or immersed metallic structures. General
principles and application for pipelines
BS EN 1011-2:2001 Welding. Recommendations for welding of metallic materials. Arc welding of
ferritic steels
BS EN 1990:2002 Eurocode. Basis of structural design (see also National Annexe issued 2004)
BS EN 1504-10:2003 Products and systems for the protection and repair of concrete structures.
Definitions. Requirements. Quality control and evaluation of conformity. Site application of products
and systems and quality control of the works
BS EN 13636:2004 Cathodic protection of buried metallic tanks and related piping
BS EN 1504-5:2004 Products and systems for the protection and repair of concrete structures.
Definitions, requirements, quality control and evaluation of conformity. Concrete injection
BS EN 1504-2:2004 Products and systems for the protection and repair of concrete structures.
Definitions, requirements, quality control and evaluation of conformity. Surface protection systems for
concrete
BS EN 1504-4:2004 Products and systems for the protection and repair of concrete structures.
Definitions, requirements, quality control and evaluation of conformity. Structural bonding
BS EN 1996-1-2:2005 Eurocode 6. Design of masonry structures. General rules. Structural fire
design
BS EN 1504-3:2005 Products and systems for the protection and repair of concrete structures.
Definitions, requirements, quality control and evaluation of conformity. Structural and non-structural
repair
BS EN 771-6:2005 Specification for masonry units. Natural stone masonry units
BS EN 15112:2006 External cathodic protection of well casing
BS EN ISO 12944:1998 (Parts 1 to 8) Paints and varnishes. Corrosion protection of steel
structures by protective paint systems. General introduction
BS DD CEN/TS 14038-1:2004 Electrochemical realkalisation and chloride extraction treatments
for reinforced concrete. Realkalisation
BS DD ENV 1504-9:1997 Products and systems for the protection and repair of concrete structures.
Definitions, requirements, quality control and evaluation of conformity. General principles for the use
of products and systems
248
A1
Case studies
CIRIA C671 • Tunnels 2009
249
Case study 1: Remedial treatments to
Folkestone rail tunnels
By Chris Levy of Mott MacDonald
Client
Network Rail
Consultant
Mott MacDonald
Contractor
Skanska Construction UK
1.1
RE-LINING AND WATER MANAGEMENT AT ABBOTSCLIFFE
TUNNEL
Abbotscliffe Tunnel is located on the coastal railway between Folkestone and Dover, and
was constructed c1840 by Sir William Cubitt for the South Eastern Railway Company. It is
one of several tunnels along the line that include Martello Tunnel, to the west, and
Shakespeare Tunnel, to the east.
Figure A1.1
Abbotscliffe tunnel portal
Abbotscliffe Tunnel is a single-bore of horseshoe section, with approximate dimensions of
7.3 m internal width and 6.0 m from the rail level to crown. It is 1775 m long and
accommodates two tracks, electrified by 3rd rail (see Figure A1.1). The tunnel lining is
constructed in brick, typically yellow and red stocks laid in English bond in the walls and
stretcher bond in haunches and crown.
Abbotscliffe Tunnel was excavated through Lower (Grey) Chalk (see Figure A1.2). The
vertical cover above the crown varies up to a maximum of 110 m. The tunnel is relatively
close to the cliff face, with a distance that varies from 25 m to 85 m. Also conventional
vertical shafts, horizontal galleries were driven between the cliff face and the tunnel to
assist with disposal of the chalk. All the galleries are unlined and many are open, offering
the opportunity to inspect the existing rock mass.
250
Figure A1.2
Geological section and corresponding view of west portal
This area of coastline has a complex geology. Ground movements in the area have had a
significant effect on the performance of the railway line, most notably in Folkestone
Warren, which lies between Martello and Abbotscliffe Tunnel. The tunnel lining has
experienced distress over many years, exacerbated by the redistribution of stress in the
surrounding ground caused by the regression of the cliffs.
The repair works described here were initiated by growing concerns over the structural
integrity of discrete lengths of Abbotscliffe Tunnel and Shakespeare Tunnel. Distress had
manifested itself within the tunnel linings in the form of longitudinal fractures and
general degradation, ie joint loss, ring separation and deep spalling (see Figure A1.3).
More pronounced degradation was taking place within Abbotscliffe tunnel in an area with
high levels of seasonal water ingress, known as the Lydden Spout.
Figure A1.3
Fracturing in the masonry lining wall in Abbotscliffe tunnel
CIRIA C671 • Tunnels 2009
251
1.1.1
Scheme development
The project was progressed in many stages:
1.1.2
initial desk-based studies – collation of historic and archive information. This exercise
provided limited background information. Typically there was little record of the
condition of the tunnels, except in the past 25 years, and there was no information on
remedial works that had been carried out after construction
site investigations – undertaken to gather more information on the condition of the
lining and to characterise the rock surrounding the tunnels
option development – investigation and evaluation of deterioration mechanisms and
options for remedial treatments using numerical analysis techniques. An option
selection process was carried out taking into account constructability, possession
strategy and cost to decide on the preferred solutions for remediation
preliminary design – the preferred solution was developed and form A, approval in
principle, was produced together with a more accurate estimate of cost
detailed design and preparation of tender documentation
the remedial treatments were undertaken in an 18 week block between May and
September 2005, in conjunction with other repairs in Shakespeare and Martello
Tunnels.
Site investigations
With only the opportunity to gain limited possessions of the track, two site investigations
were undertaken. The first at the beginning of the study to evaluate the general condition
of the tunnel, and the second during option development, to corroborate information
used to justify option selection and establish data required for detailed design. The site
investigations comprised:
252
100 mm dia. cores drilled through the lining to determine:
lining thickness
brickwork consistency
interface condition between brick and chalk
voiding behind lining
quality of chalk at rear of lining
laboratory investigation, including petrographic analysis, of selected brickwork
samples to assist in identifying the quality of materials and causes of degradation
100 m dia. cores, up to 6 m long, to collect intact rock samples to characterise the
chalk by fault-logging and prescribed geotechnical tests
televiewing (using downhole viewing techniques) to record condition of and
discontinuities in the chalk seen from within the core holes. This provided extra data
to that obtained from the samples retrieved from the cores
window panels of about 500 mm × 500 mm, (see Figure A1.4) removed the full depth
of the lining. The locations were targeted at particular fractures in the tunnel to
identify whether cracking was occurring at depth within the section. These also
allowed direct observation of brickwork quality through the lining, the nature of its
interface with the chalk, and the characteristics of the chalk itself
trial pit excavations to determine footing geometry of the walls (see Figure A1.5).
Further small diameter drilling was undertaken from the excavated pit to determine
invert level and lining thickness.
Also, to intrusive techniques, the opportunity was taken to record visual features and
defects in the tunnels:
defect mapping was produced from detailed inspections of the tunnels
specific defects were recorded in detail
geotechnical logging of discontinuities in the chalk exposed within the unlined adits
was carried out
samples of groundwater and soot on the tunnel lining were collected and later
analysed to identify potential effects on tunnel durability.
Figure A1.4
Window panel through lining exposes chalk at extrados
Figure A1.5
Trial pit through ballast to expose footings
CIRIA C671 • Tunnels 2009
253
1.1.3
Feasibility studies
Historic information indicated that originally the brick lining was installed for protection
and not as a load bearing structure. However, the distress seen during the site
investigations suggested a structural influence resulting from a changed loading regime. A
systematic technical appraisal was undertaken to identify and test failure mechanisms that
could potentially generate the defects seen in the lining.
Standard empirical methods were used to give predictions of rock load, based on certain
rock classification parameters. Also more complex finite element and finite difference
programs that allowed non-linear modelling of the soil mass were used to produce
predictions of soil/structure interaction. Preliminary analyses included:
Terzaghi (as modified by Deere, 1989 and Rose, 1982)
UNWEDGE™
FLAC™
LUSAS™.
A variety of mitigation techniques were developed and evaluated using a risk-based
appraisal technique to determine the most appropriate remedial solutions, taking account
constructability, cost and use/possession strategies.
1.1.4
Remedial treatment
Interpretation of the site investigations led to two specific treatments being designed to
cater for the defects in Abbotscliffe Tunnel. Key features of the scheme were to repair/
refurbish the following areas:
Lydden Spout
Fractured lengths.
Lydden Spout is a length of 80 m in the tunnel where intermittent, but extreme, water
ingress events result in disruption to rail operations (see Figure A1.6). Also, the results of
the site investigations suggested that the tunnel lining was being loaded to an extent that
produced cracking within the section. It was postulated that flushing of the joints in the
rock mass during these water ingress events may have caused loosening and dislocation of
chalk blocks and settlement onto the lining. The works at Lydden Spout were designed to
strengthen the existing lining (with rock dowels and sprayed concrete underlining)
incorporating a water resistant membrane to prevent groundwater entering the tunnel
above track level.
Nine lengths of wall were found to exhibit longitudinal fracturing. The site investigations
(and associated analysis) that were undertaken did not confirm a unique mechanism to
account for the fracturing. The preferred remedial solution was chosen to mitigate the
effects of several potential loading mechanisms. The designed remedial works comprised
installing rock dowels, removing delaminated brickwork, and introducing a sprayed
concrete underlining.
Lydden Spout
The design philosophy for Lydden Spout was for a new structural and water-resistant
lining that is integrated with the existing brick lining and rock mass (using rock dowels) to
gain full benefit from their combined inherent strength. The final option consisted of a
system of structural reinforcement through rock doweling and a secondary lining of
254
sprayed concrete, in conjunction with a water-management scheme that includes a
sprayed water-resistant membrane and enhanced invert drainage.
Figure A1.6
Lydden Spout, February 2006
The construction sequence was considered to be the key to the performance of the
proposed design. The rock dowels serve a dual purpose: to provide face support to the
existing brick lining and prevent the unravelling of the lining during the removal of
sections of brick. It would also act as permanent support to the new composite lining
during tunnel operation. The rock dowels, were designed to provide adequate support to
the existing lining during construction as well as to resist elevated groundwater pressure
during Lydden Spout flow events.
Figure A1.7
Remedial treatment at Lydden Spout
CIRIA C671 • Tunnels 2009
255
The removal of sections of brick from the existing tunnel lining was necessary to maintain
clearance to the train. One course of bricks (equivalent to about 125 mm of lining
thickness) needed to be removed from the sides and shoulders of the tunnel to
accommodate the placement of layers of water-resistant membrane and sprayed concrete.
The sprayed waterproof membrane was applied over the intrados of the tunnel lining.
Fibre-reinforced sprayed concrete was then applied to match the profile of the original
tunnel where brick lining had been removed and as an underlining on the crown to form
a lining that is 125 mm in thickness throughout.
Hydraulic modelling was undertaken to assess the potential water pressure that could be
encountered during a Spout event. This included the potential for an increase in pressure
due to sealing the tunnel against water ingress. Under-track drainage was enhanced to
cater for the potential increase in flows through the unlined invert of the tunnel.
The remediation measures were designed using computer finite element analysis and
more traditional algorithms such as rock mass quality, Q, and support requirements. The
final scheme is illustrated in Figure A1.7.
Fractured lengths
Treatment of fractured lengths adopted a similar solution to that used at Lydden Spout
with the exception that the water-resistant membrane was omitted and no extra invert
drainage was installed. Rock dowels were of a reduced length, 5 m, as it was considered
that these sections of tunnel would not be subjected to the same intensity of hydrostatic
groundwater pressure that could occur at Lydden Spout.
1.1.5
Implementation
The works were carried out in a blockade of the line between May and September 2005.
The contractor adopted a 12 hour shift pattern to work 24 hours a day, seven days a week,
for the duration of the blockade.
430 linear metres of the tunnel were identified for treatment, located in 10 sections
distributed throughout the length of the tunnel. This included an 80 m stretch
encompassing the area affected by water ingress associated with Lydden Spout.
Principal activities were:
installing rock dowels
removing the inner skin of brick lining in the walls
applying fibre-reinforced sprayed concrete
at Lydden Spout, installing sprayed water-resistant membrane and under-track
drainage.
Various pre-site trials were specified and carried out in advance of the blockade. A sprayed
concrete trial was undertaken in a purpose made mock-up of the tunnel to test spraying
technique and was done in conjunction with material tests on the concrete mix. Also, the
rock dowel installation was trialled in a section of cliff close to one of the portals of the
tunnel. Load tests were specified for the dowels, the installation gave the contractor the
opportunity to test different equipment and techniques, and the supervision team to view
procedures and workmanship.
256
Preliminary works, on starting the blockade, were to install temporary services, lighting, a
compressed air main, power and ventilation. Initial cleaning of the intrados for the whole
length of the tunnel was specified with a view to confirming the extent of repair areas.
Physical protection to permanent infrastructure elements was introduced as works
progressed. However, the works did benefit from track renewal, which was programmed
at the end of the blockade, and included replacement of the ballast.
Initially the rock dowels were installed. Cementitious grout was used and the steel dowels
were wrapped with grout socks to contain the grout. Permanent face plates on the dowels
provided support to the brick and were tightened onto the brick face to prevent
unravelling during removal of the inner skin.
The brick walls were excavated using a rotating cutter head mounted on a road/rail
vehicle. This provided an efficient method of removal for the relatively large areas that
needed to be treated. However, the machinery required skilled operatives to ensure a
satisfactory finish and to prevent over-cutting.
The new underlining was spray-applied with concrete supplied from a batching plant on
site and transported into the tunnel in mixer drums mounted on flat bed trailers. The
spraying operation was generally carried out for one whole shift. In Lydden Spout, an
initial smoothing coat was applied to all the brickwork to provide a suitable substrate for
later application of the sprayed water-resistant membrane.
All plant was track-mounted and was introduced from each portal. Figures A1.8 to A1.11
show the primary activities.
Figure A1.8
Drilling for rock dowels
CIRIA C671 • Tunnels 2009
257
258
Figure A1.9
Rotating cutter head removing brick
Figure A1.10
Excavated wall panel
Figure A1.11
Applying sprayed concrete to crown
The scheme required a total of 2500 rock dowels to be installed, 400 m³ of brick
excavation and 650 m³ of sprayed concrete. Unfortunately the full scope of works could
not be completed in the planned blockade and a further blockade will be required in the
future.
1.1.7
Monitoring
A system of permanent monitoring was designed and installed in the tunnel.
Instrumentation was limited to several representative sections of tunnel, with arrays of
electro-levels, vibrating wire strain gauges and multi-point borehole extensometers. A
typical installation is shown in Figure A1.12.
Also, at Lydden Spout, sensors were installed to identify when water flows, and a flume
was constructed in an adjacent drainage outfall to measure the discharge flows.
The system incorporates data capture at two data-loggers, located at the portals, which are
programmed to transmit the data records to a remote site for monitoring. The dataloggers are powered by a combination of solar panels and small wind powered generators.
Figure A1.12
Typical array of monitoring instrumentation
1.2
PATCH REPAIRS AND LINING REPLACEMENT AT MARTELLO
TUNNEL
1.2.1
Introduction
Martello Tunnel’s lining is a single-bore horse-shoe profile with a brick invert (see Figure
A1.13). It is 500 m long and accommodates two tracks, electrified by 3rd rail.
Remedial works were required to repair general degradation of the brick lining from the
environmental effects of water ingress combined with freeze/thaw action. Martello Tunnel
was reputed to be one of the worst tunnels for water ingress within the Network Rail
Southern Zone, and was subject to substantial deterioration of the inner skin of brick.
This case study presents the development of a plan for patch repairs and provides details
of the implementation.
CIRIA C671 • Tunnels 2009
259
Detailed design of the works was undertaken in the latter part of 2004 and was the subject
of a Category III check. The repairs were undertaken in an 18 week blockade from May
to September 2005, in conjunction with the structural repairs in Abbotscliffe and
Shakespeare Tunnels, as previously described.
Figure A1.13
View of Martello tunnel portal (a) and details of its lining profile (b)
1.2.2
Reasons for patch repairs
Patch repairs are implemented in response to degradation and erosion of face brickwork
that can lead to:
brick and debris falling onto the track
exposure of potentially poor brickwork behind the inner skin
ring separation as face loss releases support to brickwork above (particularly prevalent
in the haunch and crown where stretcher bond in typically used)
reduction in structural capacity.
Patch repairs are required to stabilise the lining and to reduce the rate of further
deterioration.
1.2.3
Existing techniques
Methods and procedures have been developed for repairing brick arches and they can be
found in standard drawing and specifications. These are often adopted for repairs in
tunnels. They rely on excavating the affected brick and replacing with brickwork,
theoretically with properties and characteristics to match the original.
Perceptions of these techniques are:
260
the repairs often promote accelerated deterioration in the surrounding original
brickwork
the repairs induce cracking and crushing in the existing brickwork at the perimeter of
the patch
the workforce with the skills required for good quality repair in brick is declining
methods required to provide a good key with the original brickwork are often not
undertaken effectively
methods are slow – limiting productivity during possessions.
1.2.4
Proposed patch repairs
For the Folkestone tunnels the opportunity was taken to consider, develop and carry out
methods for a repair technique that would mitigate against some of the perceived
shortcomings mentioned here.
A search for alternative materials yielded none with characteristics that would match the
original brickwork. There was even difficulty in establishing the properties of the existing
to provide a suitable specification for a match. In the case of the Folkestone tunnels, the
type of brick that would need to be specified under codes of practice now available would
be very different to the quality of the stock bricks used in the original construction.
The choice of free-form structural materials is limited and of the concrete alternatives,
sprayed concrete offers proven installation techniques, though it is recognised that the
application relies heavily on skilled operators. It has the advantages that it:
is quick to install
produces good productivity
has good repeatability – through the use of pre-bagged or pre-blended materials.
The application also gives the benefit that dowel bars can be installed, and inspected, in
the repair area before sprayed concrete is applied. This should give greater assurance that
adequate bond to the substrate is achieved (to prevent later delamination).
Having chosen a material with disparate properties to the brick, the interaction that may
develop around the perimeter of the repair was considered. The different material
characteristics of the sprayed concrete and the brick will induce stresses at the interface
whether the influence is extra load, thermal changes or changes in moisture content of the
materials. In the Martello Tunnel there was little evidence that extra load would be
induced and the effects of moisture changes were considered to be small. The review was
limited to the effects of thermal changes.
The difference in the thermal characteristics of the materials can theoretically result in
stresses developing along the interface, which could lead to distress in the surrounding
brick. To avoid this situation a compressible strip was considered for the perimeter of the
patch repairs. A designed material was sought that could be introduced to maintain some
load transfer compatibility with the adjacent brick, but the interaction between the
different materials is complex and the inherent uncertainties in material properties
overshadowed the reliability of analytical results. After careful consideration, a 10 mm
thick rubber strip, complying with BS 5400-9 (BSI, 1983) was specified around the
perimeter of the patch repairs.
Dowels were specified through the rear interface (between brick and concrete) to provide
a path for shear transfer and for security of the patch in the event of debonding.
Brick removal is traditionally carried out by hand using mechanical tools. Extensive use is
precluded because of hand arm vibration syndrome (HAVS), and health and safety
guidance puts limits on productive working time. Road-header type equipment is now
available for breaking out brick and can be mounted on road/rail vehicles for easy access.
This provides an effective alternative to hand demolition, but does require experienced
operators to achieve the necessary control on the cutting head. Evidence from site was
that, handled correctly, this type of machine is not too aggressive and was able to remove a
specified depth of brickwork without compromising the remaining courses.
CIRIA C671 • Tunnels 2009
261
1.2.5
Prioritisation of repairs
To determine the extent of repairs required, a touching distance survey of the lining was
carried out during site investigations. Areas of deterioration were categorised by depth of
loss of brick, and the following bands were chosen:
Also, an exercise to establish rates of degradation was carried out. This consisted of a
review of previous historic records, particularly annual inspections, which dated back to
1964.
0–25 mm
Shallow
25–50 mm
Moderate
50+ mm
Deep
Condition matrices (see Figure A1.14) were prepared for each year that had an annual
inspection record and the results compared, with the aid of colour-coded keys, to
determine trends in deterioration. The results were used as a guide to predicting future
deterioration rates. It is recognised that the condition surveys are influenced by many
variables, such as subjective recording and seasonal weather conditions, which make
accurate assessment impossible. This approach does allow a qualified forecast of
serviceable life, which is useful in the absence of more reliable information.
Figure A1.14
Example condition matrices. Inspection June 1964 (a) and Inspection February 2002 (b)
The study indicated that in wet areas the maximum rate of degradation was 50 mm of face
loss over a 40 year period. In drier areas a maximum of 25 mm of face loss could be
inferred over the same period. This information allowed best use of resources by
prioritising repairs. Based on the assumption that the deterioration rates will be similar
over the next 40 year period it was decided to treat:
262
all moderate and deep spalling defects in wet areas
only deep spalling defects in areas that were intermittently wet.
1.2.6
Re-lining
On applying these criteria to the defect mapping (from the condition survey) it became
apparent that individual patch repairs were not going to be economic for substantial
lengths of the tunnel, leading to consider re-lining portions of the tunnel. As the most
severe deterioration occurred in places where there was water ingress, measures to control
the water also needed to be arranged.
A range of techniques to control water were considered. Principal options are summarised
in Table A1.1.
Table A1.1
Options for water control
Technique
Comment
Treat surrounding ground
Unreliable in terms of quantity and effectiveness
Treat brickwork
Difficult to achieve, and water is likely to displace to other zones
Application of water-resisting layer
Difficult to apply effectively in areas of running water. Water is likely to
displace to other areas
Allow water ingress and collect
Maintains steady state
Promote draining of surrounding ground
Unknown quantities of water may be released (available drainage capacity
limited by existing pipes)
The method chosen for the Martello Tunnel was to allow water to continue to pass
through the lining and collect and channel it into the existing drainage system.
Areas suffering the most severe degradation were generally on the walls and gauge
clearance was limited at these points. So there was a need to remove the inner
deteriorated skin of brick in the walls to install the water collection system and repair.
A cavity sheet drain was chosen for water collection. Applied over the whole area of
excavated brick it does not require site evaluation of specific locations of water ingress and
it enabled application of the sprayed concrete without risk of wash-out. The cavity sheet
feeds into a newly installed bespoke channel gutter, laid to falls, and discharges into
downpipes through fabricated connections. The downpipes feed into the refurbished,
existing tunnel drains.
Areas of the crown were also found to be suffering from water ingress, though generally
there was little degradation of the brick face. However, water dripping onto the track bed
and windscreens of trains is considered a hazard and treatment of these areas was
included in the scope of works.
A similar system of cavity drain sheeting was adopted covered with 125 mm of sprayed
concrete to give protection and support. In the crown area an underlining solution
(without excavation of brick) was adopted (see Figure A1.15), as clearance to the train
gauge is not critical.
Throughout the repaired and over-sprayed areas stainless steel dowels were installed to
provide permanent physical connection between the original brick lining and the new
sprayed concrete. They also provided support to a steel reinforcing mesh, installed within
the sprayed concrete to aid application and reduce initial shrinkage cracking.
CIRIA C671 • Tunnels 2009
263
Figure A1.15
Re-lining in Martello Tunnel
1.2.7
Implementation
For the patch repairs a temporary frame was installed around the perimeter of the repair
(see Figure A1.16). This provided a straight edge to assist with saw-cutting and the bolts
that secured the frame act as both temporary and permanent pinning of the bricks around
the repair. The brick was excavated, followed by installation of stainless steel “L” dowels,
fixing the mesh, introduction of the perimeter rubber strip and lastly the application of
sprayed concrete.
Figure A1.16
264
Framing to patch repair
Figure A1.17
Brick excavation in Martello
Figure A1.18
Martello wall panel
In the re-lined sections, excavation of brick was limited initially to 2.5 m advances. As
confidence was gained in the techniques and the temporary monitoring showed no
perceptible movement of the lining, this was increased to 5 m to improve productivity.
The top of the excavation was framed with a steel flat to temporarily restrain the
remaining brickwork. Excavation was carried out using a Schaeff rotating head mounted
on a small excavator (see Figure A1.17). With careful operator control the exposed
substrate required only a small amount of secondary trimming, which was carried out with
hand-held mechanical tools. The cavity drain was installed and pinned in place with shotfired nails. The stainless steel dowels were drilled and anchored with two-part resin, and
sealed around the penetration of the membrane. Drainage gutters were fixed at the base
of the cavity sheet (Figure A1.18). Finally, the reinforcement mesh was installed and the
sprayed concrete was applied.
CIRIA C671 • Tunnels 2009
265
1.2.8
Conclusions and observations
The patch repairs and re-lining were carried out effectively on-site, though certain aspects
required significant attention such as setting-out (to ensure clearances were maintained),
fastening of the membrane, and the detailing at water collection points.
The mechanised excavation was successful, though a limited amount of local trimming by
hand was required to finish the surfaces. It proved faster than breaking out the complete
area by hand.
The brick removal and the sprayed concrete application generate dust and debris.
Effective protection of the ballast is essential to prevent contamination. Over-spray and
rebound needs to be cleared up immediately to avoid more activities with the potential for
HAVS.
While the works for this project were carried out in a blockade, it is considered that, with
the correct investment in mechanical plant, similar solutions could be implemented within
shorter possessions of the track.
266
Case study 2: Investigation and
treatment of ground instability and water
ingress at Blackheath Tunnel
By Leo McKibbins, Mott MacDonald
Client:
Network Rail
Consultant:
Mott MacDonald
Main contractor:
Sir Edmund Nuttall Ltd
Specialist subcontractors:
System Geotechnique Ltd (groundworks),
Datum Monitoring Services Ltd (monitoring)
2.1
BACKGROUND
Blackheath tunnel in east London is a 1681 yards long masonry-lined rail tunnel,
constructed in 1849. Anecdotal evidence suggests that throughout its life it has suffered
from water ingress along much of its length. This was particularly severe at a location
about halfway along the tunnel, where water spouted several feet into the tunnel under
pressure through gaps in the mortar. At some time in the past plastic sheeting had been
installed to deflect the water into the tunnel drainage system (Figure A1.19).
In 2000 a member of the pubic reported the sudden appearance of a depression in the
ground surface above the tunnel, in an undeveloped grassy area of ground adjacent to a
road and paved footpaths. The depression was circular in plan, about 0.5 m deep and 2 m
in diameter. Train movements were immediately suspended while an inspection was
carried out of the tunnel below. There was no evidence of structural distress in the tunnel
lining, but it was noted that the subsided area was directly above the area of severe water
ingress inside the tunnel. The tunnel was returned to service with temporary restrictions
on train speed and the subsided area above ground was fenced off. Over the following few
weeks the depth of subsidence increased gradually before stabilising at about 1.5 m.
Figure A1.19
Water streams from the base of one of the plastic sheets used to deflect its flow down
the tunnel wall rather than spouting into the running area of the tunnel
CIRIA C671 • Tunnels 2009
267
The evidence pointed to the possibility of an unknown and hidden construction shaft with
unstable fill, and, following an initial investigation by another consultant, Network Rail
involved Mott MacDonald to investigate the situation and provide an options study to
examine and assess the feasibility of potential remedial solutions.
2.2
TUNNEL AND GROUND INVESTIGATIONS
Mott MacDonald’s investigations involved various activities, principally to determine the
ground conditions in the area of the subsidence and adjacent to the tunnel at depth. The
main aims were:
to investigate the cause of the subsidence at the ground surface
to assess the undisturbed ground conditions
to assess the cause of the localised water inflow into the tunnel
to obtain sufficient information for design of remedial measures for both the water
ingress and surface subsidence problems.
The investigations consisted of a desk study and several phases of site investigation,
undertaken both from above ground and from within the tunnel, to gather the necessary
data.
2.3
DESK STUDY
A comprehensive desk study was carried out and revealed a wealth of useful information
including local topography, geology and hydrogeology, the tunnel’s construction and
history, and more:
268
original drawings showed the details of the masonry lining and the depth of the
tunnel below ground surface at the subsided point (about 20 m). Six original
construction shafts were recorded, and the subsidence had occurred about midway
between two of these
nearby borehole data showed that the tunnel passed through the Thanet Sands
(generally clayey, pebbly sands, silts and marls). The ground at the tunnel crown was
thought be from the Lambeth Group (silty sands) and, at the ground surface, the
Harwich Formation (sands with pebble beds and variable cementation)
historical records, including items in local newspapers of the time, suggest that water
ingress had been a problem in this tunnel since its early years in service, and in 1861
it suffered a partial collapse, although the location was not recorded
hydrological records indicated that the general level of the water table in this area was
about 5 m to 10 m below the tunnel footings
a water main was identified within 10 m of the suspected shaft location, although
when contacted the local water authority stated that there was no evidence of any
significant leakage from their mains water services in this area
Blackheath has historically been quarried and is known to contain unidentified
caverns and underground workings that have caused subsidence elsewhere in the
area, but no records were found of any such natural features or land-use
corresponding to this location.
2.4
SITE INVESTIGATIONS
The potential sensitivity of the ground in the subsided area meant that the site
investigation was carried out in a series of carefully sequenced phases to minimise
disturbance. Initially, work from the ground surface was carried out only in periods where
train movements were suspended. One of the first activities was to identify any large voids
which might allow sudden collapse of the ground using geophysical techniques followed
by physical probing. The least disruptive techniques were used first, followed by
increasingly intrusive techniques once the risk of triggering further settlement or damage
to the tunnel lining could be shown to be acceptable. Load-spreading working platforms
were used to support heavy plant. The range of tests in the site investigation phases
comprised:
a surveying exercise, using linked above ground and below ground traverses between
portals, accurately established the relative spatial locations of features above and
below ground
geophysical techniques, ground-penetrating radar (GPR) and microgravity surveys,
were used to examine the ground around the subsided area and known construction
shafts for comparison
trial-pits and trenches were excavated to examine the ground in the area of
subsidence and adjacent to it, and to identify any leaks from nearby water services
boreholes were sunk off the axis of the tunnel to establish the geological sequence and
investigate the ground. Piezometers were installed to allow groundwater monitoring
using stuffing boxes to control water ingress, coring, endoscope inspection and
window-sampling techniques were used to investigate the lining, the ground beyond
it and the presence of water. The cores were sent for analysis by physical testing and
petrographic examination
the rate of water ingress into the tunnel in this area was estimated using a simple
collector system. Water samples were taken for laboratory analysis to try to establish
the possible source. An assessment of the capacity, condition and flow rates of the
tunnel drainage system was made
an array of 23 dynamic penetrometer tests were carried out from the ground surface
to a depth of about 16 m to identify voids and changes in ground conditions in and
around the subsided area. The results were used to produce a 3D map of subsurface
conditions (see Figures A5.5 and A5.6).
The above ground investigations were carried out in a quiet residential area, so the
generation of noise and disturbance and restrictions on periods of working were discussed
in advance with the local authority. Measures to control these within agreed limits were
put into place, and Network Rail held a public information meeting to inform concerned
local residents of the situation.
2.5
INVESTIGATION RESULTS AND CONCLUSIONS
From the results of the site investigation the following conclusions were drawn:
the tunnel passes through the Thanet Sands. Although its invert is well above the
regional groundwater table, shallower groundwater (possibly a perched water table)
was locally present at the base of the Blackheath Beds (comprising highly permeable
coarse gravels) located several metres above the tunnel crown
generally this water is prevented from percolating down to the tunnel lining by the
CIRIA C671 • Tunnels 2009
269
relatively impermeable Lambeth Group (predominantly clays), which is typically
present between the top of the Thanet Sands and the base of the Blackheath Beds.
But at the location of the subsided area there was an irregular funnel of very weak
ground, consisting of poorly consolidated gravels, present between the ground
surface and the tunnel lining (see Figure 2.3). Within this area the Thanet Sands and
the impermeable Lambeth Group appeared to be absent. This localised feature acted
as a vertical drain for the groundwater perched at the base of the Blackheath Beds,
channelling water down to the tunnel lining and through it into the tunnel
the flow-rate of water into the tunnel over the worst affected 4 m length of sidewall
and haunch was considerable, estimated at about 5 lt/s, and did not appear to exhibit
any noticeable seasonal variation, even after a prolonged period of dry summer
weather. Although the original source of the groundwater at the base of the
Blackheath Beds was not known, it was considered possible that it had leaked from
local services
during the ground investigations some extra local subsidence occurred at the ground
surface, overlapping but slightly offset from the location of the original subsidence.
This showed how sensitive the ground in this area was to disturbance (Figure A1.20)
the masonry tunnel lining in the area of the water inflow was locally in a poor
condition. Although the bricks themselves appeared to remain in a reasonable
condition, much of the mortar was missing or severely deteriorated from the
longstanding passage of water.
Figure A1.20
Site investigation resulted in
some additional subsidence at
the ground surface, affecting
an area of about 1 m², which
subsided by around 300 mm
270
Although the initial appearance and form of the surface subsidence strongly suggested the
presence of a previously unrecorded hidden construction shaft, the investigations did not
provide any reliable evidence for this. Rather they indicated that there was a funnelshaped zone of very weak ground, centred on the subsided area, where there appeared to
have been a disturbance of the normal geological sequence between the ground surface
and the tunnel lining. This suggested that a local ground feature (either natural or
induced) was causing both the subsidence and the continued water ingress in the tunnel.
Certain characteristics of the lining, including local reduction in joint spacing and an
apparent local increase in the masonry thickness, along with the disturbed appearance of
the adjacent ground, indicated that this location might have suffered from a partial
collapse at the time of construction. This was supported by the historic records obtained as
part of the desk study indicating that such an incident had occurred, although at an
unknown location.
Figure A1.21 shows an idealised sketch cross-section of the tunnel and adjacent ground,
illustrating the main features identified by the investigation.
Figure A1.21
Idealised cross-section through tunnel at location of water ingress showing inferred
ground conditions and water pathway between perched water table and tunnel
The subsidence event appeared to be a sudden expression at the ground surface of the
gradual migration of voids within the poorly consolidated ground in this area. This was
possibly exacerbated by the washout of fines from the longstanding movement of water
through this permeable funnel and into the tunnel.
The ultimate source of the water in the perched water table could not be confidently
determined. Although there remained a strong suspicion that a significant proportion of
the water entering the tunnel was from one or more mains supply sources this could not
CIRIA C671 • Tunnels 2009
271
be conclusively demonstrated based on the results of laboratory testing of water samples.
Discussions with the local water authority and their later investigations of water loss from
their local network failed to resolve this matter, precluding the possibility of dealing with
the water ingress at source.
Although at this point in the investigation many uncertainties remained regarding the
ground conditions, tunnel lining parameters and the cause and nature of the subsidence
event, it was decided that enough information had been gathered to proceed with a
consideration of potential remedial options and the selection of a preferred option.
2.6
EVALUATION AND SELECTION OF PREFERRED OPTIONS
Several outline remedial options were identified, which could potentially remedy either
one or both of the problems of subsidence at the ground surface and continued water
ingress into the tunnel. These included:
construction of a reinforced concrete or geotextile cap just below the ground surface
to span the weak and disturbed ground
ground improvement by permeation grouting of the weak and disturbed ground to
improve its stability
providing an alternative drain for the water in the perched water table down into the
chalk aquifer below the tunnel
controlling the water ingress by channelling it through drains in the lining and into
the existing tunnel drainage, or into an alternative dedicated drainage route
permeation grouting to reduce the permeability of the disturbed ground around the
tunnel and reduce water flows into it.
The feasibility of each of these options was assessed and given in an options study report,
including the potential advantages and disadvantages of each and estimated
implementation costs. This report provided the basis for discussions with Network Rail
and the final selection of preferred options for further development.
A do nothing option was not considered feasible, because in the short-term the ground
surface needed to be stabilised, and in the long-term the severe water ingress through the
tunnel lining was likely to lead to localised instability. Network Rail’s most urgent priority
was to reopen the fenced-off area of land to the public. It was initially agreed to develop a
single preferred option to mitigate the risk of surface subsidence only, allowing further
consideration of options to deal with the more complex problems associated with the
water ingress and the gradual deterioration of the tunnel lining.
2.7
PREFERRED OPTION FOR STABILISING THE GROUND
SURFACE
The preferred solution for addressing surface subsidence was the development and
construction of a geotextile ground reinforcement system. This involved stripping back
the ground over the area of subsidence and replacing it with a designed capping system of
engineered materials reinforced by horizontal tensile geogrid material, anchored at its
perimeter. The intention was that this construction would minimise the surface expression
of any subsidence occurring within its effective area, and mitigate the hazards associated
with rapid localised subsidence. This would ensure the safety of the area and allow it to be
reopened to the public.
272
The main advantages of this option were:
relatively economical treatment with predictable cost
rail possessions and tunnel closure not required
straightforward construction methods with quantifiable items
allowed the locality to be reopened to the public
future ground movements could be monitored periodically
relatively straightforward to remove if future ground treatment required it
would preserve ground drainage and allow development of vegetative growth.
The main disadvantages of this option were that it did not actually improve ground
stability at depth but rather covered it up, and that it could not mitigate unforeseen
ground movements outside its effective design area. These were considered acceptable
risks because it was only intended to be a temporary solution.
Construction of a concrete cap was considered, but this entailed extra serviceability
problems and would have been more difficult and costly to remove if future treatment of
the ground below the slab were required. Also, it would alter the local drainage and
present problems for landscaping the area.
A major advantage of the selected scheme was that it could be done rapidly without the
significant expense, disruption to services and long lead-in times associated with obtaining
rail possessions and closure of the tunnel. Detailed design and construction could proceed
without delay, before the development of a long-term solution to the remaining problems
of unstable ground at depth and water ingress into the tunnel.
2.8
PREFERRED OPTION FOR LONG-TERM GROUND
STABILISATION AND CONTROL OF WATER INFLOW
As the design of the geogrid capping option progressed, discussions between Mott
MacDonald and Network Rail resulted in a selection of a preferred remedial option for
mitigating the water ingress and its detrimental effect on the long-term performance of
the tunnel lining. The chosen option included local back-grouting of the permeable
gravels immediately behind the lining and for some distance beyond. The aim was to
significantly reduce water ingress into the tunnel by providing a low-permeability physical
barrier and sealing flow-paths in the funnel of permeable ground between the perched
water table and the tunnel lining. This would also consolidate and strengthen the weak
and loose disturbed ground adjacent to the weakened lining.
Due to the potential changes in the stress state of the lining during the works, it was
proposed to install a monitoring system to allow continuous evaluation of the lining’s
deformation response.
This option had several advantages over others considered, but particularly that it
represented a long-term solution to both the water ingress and unstable ground, and
included no requirement for maintenance beyond construction. However it would require
careful design, planning and control during construction with the most sensitive elements
of the work carried out in available rail possessions designated for other engineering
works on this section of the line. The area to be treated was clearly defined and of limited
volume, and of much higher permeability than the surrounding ground, which made it
suitable for the grouting solution.
CIRIA C671 • Tunnels 2009
273
2.9
DESIGN DEVELOPMENT AND CONSTRUCTION
Network Rail awarded the contract for the design and construction works to its
construction partnership framework contractor, Sir Edmund Nuttall Ltd, who involved
Mott MacDonald Ltd to further develop their outline designs to detail design stage.
Subcontractors were involved in the specialist surveying, groundworks and tunnel
monitoring tasks.
2.9.1
The geogrid capping layer
The geogrid capping was designed with factors of safety greater than 1.3, to bridge a void
of 2.5 m × 2.5 m plan dimension, or 3 m circular diameter, occurring within the central 5
m × 5 m area of the designed section (the effective design area). The road, gas and water
mains and public footpaths were close. This imposed constraints on the design and its
effective area, which was centred on the weakest ground above the tunnel, and included
the locations where subsidence had previously occurred at the ground surface. A void in
excess of the design size located within the effective design area might result in sliding
failure of the gabion baskets, with settlement potentially in excess of 200 mm, but even in
such circumstances it would not fail in a catastrophic way.
The capping layer, illustrated in Figure A1.22, required the construction of a 150 mm base
layer of granular self-compacting fill, with two 300 mm layers of similar fill above, each
with a layer of geogrid material at its base. The geogrid layers were anchored around the
perimeter of the design area by wrapping them around a double layer of gabion baskets.
The capping layer was constructed in a 1650 mm excavation and covered with 300 mm of
topsoil before re-landscaping in accordance with the local authority requirements.
Ground levels were monitored in the area before, during and after the construction works
to identify any changes in ground level that might indicate further settlement. Dipping
tubes were included to pass through the geotextile layers and allow inspection of the
ground immediately below the base layer to allow monitoring of any voids that might
develop below the cap.
To make provision for potential future drilling to allow permeation grouting of the
ground below the cap, an array of plastic collars were inserted through the apertures in
the geogrid material during the construction so that drilling and grout injection could be
carried out without risking damaging the geogrid and compromising its performance.
274
Figure A1.22
Design for the geogrid capping layer
Figure A1.23
Construction of the capping layer using geotextile and engineering fill. Note the
red plastic collars/tubes placed at regular intervals to allow access for injection
of grout in the ground below the cap in a later phase of work
CIRIA C671 • Tunnels 2009
275
2.9.2
Grouting the permeable ground around the tunnel
The design for the grouting scheme involved two main phases of grouting: the first to be
carried out from ground surface level and the second from within the tunnel itself. The
first phase would involve injection of grout through an array of tubes à manchette (TAMs)
into a zone beyond 3 m from the tunnel extrados so as to generally reduce water flow
through the body of the funnel while minimising the effect on the lining itself. This would
restore the low-permeability layer between the perched water table and the tunnel, which
elsewhere was provided by the clays of the Lambeth Group. Once this was done, and the
water flow into the tunnel largely attenuated, the second phase of grouting would involve
injection from within the tunnel in the area directly behind the tunnel lining, within the 3
m offset zone, to tighten up this area and seal off any remaining flow paths. Allowances
were made for further stages of grout injection within the tunnel and from the ground
surface, using a more closely spaced array of TAMs, should it be required. Any necessary
repairs to the masonry lining could then be carried out in an improved environment.
To limit potential generation of stress in the tunnel lining and avoid damaging it during
the works, limits were set on the pressure and volume of grout injections and on any
movement of the lining during the works, based on the results of structural modelling.
Structure movement monitoring system
The design for the monitoring system was based upon a robotic total station that would
continuously survey an array of target prisms installed around the lining intrados in the
area that might be affected by the treatment. The total station instrument was installed on
a bracket in a refuge (a recess in the lining) near to the area to be treated, with line of
sight to the prisms, and outside the potential zone of influence of the works. The system
was capable of making a full round of measurements, including all prisms in the work area
and extra reference prisms located well outside the potential zone of influence, at a
maximum frequency of about 20 minutes, although lower frequencies of measurement
were used during non-critical periods. The system was capable of a measurement accuracy
of better than ±1.5 mm and transmitted data to the offices of the monitoring contractor
and to the site offices for assessment. It was necessary to specify that the system was
capable of being remotely reset so that, if its operation were disrupted by train movements
while monitoring outside rail possession periods, it would not be necessary to re-enter the
tunnel to make it operational again.
Structural analysis and constraints on the works
Finite element (FE) analysis of the tunnel lining and the surrounding ground was carried
out to help evaluate the potential stresses generated in the tunnel lining and to agree
acceptable limits on lining deformation that might result from the works (Figure A1.24).
The FE analysis used information gained in the course of the site investigation
supplemented by a laser scanning survey of the tunnel lining in the area of interest that
gave an accurate 3D plan of the lining intrados, as well as a useful virtual image (Figure
A1.25).
276
Figure A1.24
Finite element structural modelling results for tunnel lining subjected to
full ground loading, hydrostatic water and grout pressures during injection
Figure A1.25
Section of the 3D laser-scanning survey results showing one side of the tunnel intrados folded
flat as a 2D image. This is a virtual image and great detail is visible, right down to the joints in
the brickwork. The repaired core sample holes (some of which were the result of previous
investigation of this area) and the wet area of the tunnel wall are also clearly visible (the
cores are ringed in red, the wet area is white)
CIRIA C671 • Tunnels 2009
277
Based on the analysis, two trigger levels were established to help control the grouting
works. A red trigger level was set at 15 mm movement of any monitored prism from the
baseline reading established before the works. This movement corresponded to a level of
stress in the lining at which tunnel serviceability might be affected, with a suitable factor of
safety, and if reached this would require all grouting work to be stopped at least until
further assessment and verification of the lining’s stability and condition could be carried
out. Also an amber trigger level was specified, required grouting work to be suspended
while an on-site assessment was made of the acceptability of continuing and the most
appropriate method of doing so. This would require an engineering judgement based on
how the tunnel was reacting to the works being carried out (eg the location and rate of
deformation, what was likely to have caused it, whether movement ceased immediately on
suspension of the work). The amber trigger level was set at a value two thirds of the red
trigger level (10 mm) to act as a holding point and to provide an extra element of control
when working near to it.
Also, based on the results of the structural analysis, upper limits were specified on the
grout injection pressures and injected volumes that could be used in areas close to the
tunnel lining.
Implementation
The design produced by Mott MacDonald was used in a performance specification issued
to the contractor, Sir Edmund Nuttall Ltd, who employed specialist subcontractors to
carry out the groundworks and monitoring elements respectively.
In the period of several months between finalising the design and planning the works
there had been some changes in circumstances that needed to be considered and required
a review of proposals. First, although still severe, the rate of water ingress into the tunnel
in the area to be treated appeared to have reduced considerably from its normally stable
high rate. Second, the two 52-hour rail possessions that were previously planned were
reduced to 36-hour possessions, which considerably reduced the working time available to
carry out the work.
When considering the design and altered circumstances, changes to some elements of the
groundworks specification were proposed by the groundworks subcontractor, System
Geotechnique Ltd, which were accepted by the contractor and designer. These resulted in
a modified approach whereby the use of an accurate downhole surveying system (the
Tigor system) improved control of the drilling operation and allowed the 3.0 m exclusion
zone (the red zone) around the tunnel to be reduced to 1.5 m. The amber zone between
1.5 m and 3.0 m was subject to strict controls on drilling for TAM installation and grouting
operations (work carried out only during rail possessions and reduced grout injection
pressure and volumes). This allowed a greater proportion of the total volume of grout
injection required to be carried out during the above ground phase of the operation,
which was more efficient and economic, and reduced the requirement for grout injection
within the tunnel.
Given the previous instability of the ground in the area of treatment and the limited
capacity of the geogrid capping layer installed, all works from the ground surface required
the use of load-spreading mats to minimise ground pressures imposed by the drilling rigs
and other plant. Also, ground surface levels were monitored during and for some time
after the works to identify any movements.
278
Figure A1.26
Design of the grout injection scheme, showing amber zone (1.5 m to 3 m offset from tunnel
extrados) where strict controls on drilling and injection were adopted. No drilling or grouting
was carried out within the red zone (less than 1.5 m offset from tunnel extrados)
Drilling the boreholes for TAM installation required the use of cased holes to ensure that
flushings were returned to the surface and minimise disturbance to the local ground. The
very high permeability of the gravels forming the funnel feature made this difficult and
over the course of the drilling some grout was added to the holes under hydrostatic
pressure. This was to initiate ground consolidation and reduce its permeability, which
would benefit later stages of the work. Where this was carried out in the holes directly
above the tunnel crown it had a double effect: first, the water inflow into the tunnel
reduced considerably, and second the monitoring system detected a downward movement
of up to 7 mm in monitoring prisms at the tunnel crown, which took place gradually over
the course of a single day. This indicated that the tunnel was very sensitive to work in this
area of ground above its crown, and also suggested that the grout added had found its
way into the main flow paths between the perched water table and the tunnel lining and
was already having a beneficial effect on water ingress. From this it was inferred that the
planned grouting scheme had a high likelihood of success in achieving its main aim. A
careful approach would be required for injecting grout into the ground close to the tunnel
to complete the works without lining deformation reaching the red trigger level when
work would have to stop.
Before proceeding with the critical phase of grout injection within the amber zone,
nearest to the tunnel lining, Mott MacDonald carried out a review of their structural
analytical model using the newly available deflection data. This reconfirmed the previous
results and trigger levels.
CIRIA C671 • Tunnels 2009
279
The critical phase of grout injection within the amber zone, in sensitive areas of ground
close to the tunnel, was undertaken during a single 36-hour rail possession, with the
automated deflection monitoring system augmented by staff within the tunnel who were
in direct two-way communication with the team carrying out the works above ground and
constantly monitored the lining to detect any visible evidence of distress. The sequence of
grouting was carefully planned to minimise extra stresses on the lining, particularly
around the sensitive crown area. It was important to clearly define the lines of
communication and responsibilities for control of the works and decision making. A clear
decision tree for working procedures on-site and actions to be taken in the event of
reaching trigger levels was agreed between the parties involved.
This critical phase of grouting was carried out successfully during the first possession,
using a cautious approach of initially low grouting pressures (2-bar above line loss) and
small volumes of grout (50 lt) per injection port, and increasing these (to a maximum of 4bar and 100 lt) only where it was found necessary. Grout injection outside the amber zone,
further from the tunnel, continued for some days beyond the end of the possession with
the monitoring system still in operation. At the end of these works the tunnel lining had
stabilised with a maximum deflection of 13.5 mm at its crown. The revised grouting
scheme had resulted in a high proportion of the total planned grout volume for the whole
of the works being injected in the first phase, and water ingress within the tunnel had
been dramatically reduced from its previous levels. On this basis it was agreed between
Network Rail and the contractor that the second phase of grouting, from within the
tunnel, would no longer be necessary.
At the end of the project the ground above the tunnel was re-landscaped and reopened to
the public. The ground at depth has been consolidated and stabilised by the grouting
works, and the water ingress within the tunnel, previously estimated at up to 5 lt/s via
general seepage and several pressurised water spouts, was reduced to occasional slow
dripping with the majority of the intrados starting to dry out. The performance of this
scheme will continue to be monitored and consideration will be given to carrying out
repairs to the deteriorated masonry lining in a much improved situation for working.
280
Case study 3: Strengthening Brunel’s
Thames Tunnel
Adapted by Martin Roach of Metronet Rail (Roach and Brunel, 1998)
3.1
INTRODUCTION
The 365 m long Thames Tunnel carries the East London Line (ELL) beneath the river
between Wapping and Rotherhithe. It is known as the world’s first shield-driven and
major sub-aqueous tunnel, built under the supervision of Sir Marc Isambard Brunel
between 1825 and 1843. The tunnel became part of the Wapping & Shadwell to New
Cross Line in 1869.
The tunnel consists of two bores with 64 intermediate arches forming cross-passages. The
lining of each consisted of rings of structural brickwork, with further bands of nonstructural (dentition) brickwork forming drainage channels. These were faced with layers
of roof tiles and render of variable thickness to provide a uniform finished profile.
3.2
SCHEME DEVELOPMENT
No major problems arose in service until a condition survey in 1994 revealed that the
tunnel finishes were significantly deteriorating in several areas, with numerous areas of
seepage. Investigatory work took place to cut out small areas of secondary lining, take core
samples and assimilate more information concerning the tunnel’s condition, and 30 of the
cross-passages were also bricked up to temporarily stiffen the spine wall.
Later analysis concluded that the tunnel had a factor of safety (FoS) of unity under certain
conditions, confirming it as a structure at risk. Various strengthening schemes were
investigated and sprayed concrete emerged as the favoured option. Its principal
advantages were the close construction tolerances achievable, plus it offered means of
providing early support. Trials were conducted and a visit to the Washington DC Metro
was organised to inspect older sprayed concrete tunnels, to provide assurances concerning
its application and durability.
The final recommended solution was that a 200 mm thick sprayed reinforced concrete
lining should be provided, backed by a PVC waterproof membrane. Also, the remaining
open cross-passages would be blocked up, to produce separate tunnel bores. As major
works elsewhere on the ELL were also planned, it was decided that closing the line for a
period of seven months would be a more cost-effective method than working in
possessions.
It was envisaged that the lining works would be completed within five months, with two
months allowed for re-commissioning.
CIRIA C671 • Tunnels 2009
281
3.3
HERITAGE CONSIDERATIONS
LUL has long recognised the tunnel’s historical significance. Good relations have existed
for many years with the Brunel Exhibition at Rotherhithe and the Trustees were aware of
LUL’s plans at an early stage.
To ensure that details of the original construction were not lost, plans were made to
record on videotape the details as work progressed and the Royal Commission for the
Historic Monuments of England was also invited to take photographs for national
archives. As the tunnel portals and parts of Wapping Station were listed structures,
English Heritage (EH) were also informed of the proposals.
282
Figure A1.27
The original tunnel after removal of services, track and ballast
Figure A1.28
Completed lining, including architectural features
The advance publicity campaign, launched in December 1994, immediately prompted
public controversy over the proposed use of sprayed concrete. On 24 March 1995, the
tunnel was Grade II* listed by the Heritage Secretary. This postponed refurbishment on
the tunnel apart from track removal and preparation works (Figure A1.27).
LUL immediately met with EH and established a working group comprising members of
LUL’s senior engineering management and representatives of the interested parties with
the aim of agreeing how the tunnel might be strengthened to meet the various
requirements and obtain listed building consent.
3.4
THE FINAL SCHEME
As discussions continued, it became clear that maintaining Brunel’s original vista of the
tunnel was important to EH. This would require preserving all the cross-passages,
together with their decorative features, plus the bands of brickwork in the tunnel bores.
Along with these architectural aspects, EH also required that any new lining would have to
be drained. EH also stipulated that the amount of break-out of the original structural
brickwork must be minimised. The design profile of the tunnel had to be reduced to the
absolute minimum possible to accommodate the kinematic envelope of the rolling stock.
By October 1995, LUL had developed an acceptable solution that incorporated all these
requirements, yet still provided the 200 mm lining (Figure A1.28).
3.4.1
Design considerations
The employer’s requirements stipulated that the new lining should consist of a 200 mm
minimum thickness of steel fibre reinforced concrete. The fire resistance period for
structural members was one hour.
Analysis of the tunnel lining was carried out using the finite element program CRISP.
Seven loading cases were considered, investigating combinations of several variables.
Back-of-lining drainage was incorporated to pick up water permeating through the
brickwork to be channelled into the low point sumps in the tunnels and be pumped out
through a rising main to the portals. Despite EH’s requirement to provide this drainage,
the lining was designed to resist full hydrostatic loading, lest the system should become
blocked in the future. The lining was also required to be watertight, so a fully welded PVC
waterproof membrane, with geotextile fleece backing, was specified between the existing
brickwork and the new lining.
3.4.2
Construction
Once the suspension was lifted, construction of the tunnel invert immediately started.
Because of the requirement to minimise brickwork cut-out, the new invert slab thickness
had to be reduced. To compensate for this, the design strength of the invert concrete was
increased from C40 to C60.
The invert was generally cast in 5.5 m bays, corresponding to the 18 ft centres of Brunel’s
cross-passages. Its brickwork was first trimmed to profile, the fleece and membrane then
laid. Reinforcement was fixed, along with two 150 mm diameter drainage pipes, installed
to provide extra storage capacity (Figure A1.29). Purpose built shutters were used to form
the top surface, incorporating half-round drainage channels at the edges, and sidewall
kickers. Concrete was mixed in the station batching plant and transported by dumper to
CIRIA C671 • Tunnels 2009
283
the work location. Construction of the invert took place as a continuous activity and
provided access for further works.
One of the original cross-passages, which were in the style of a Greek Doric arch, was used
as a template for new replicas to be cast in situ using four steel shutters.
284
Figure A1.29
Invert construction in progress
Figure A1.30
Temporary propping to tunnel during cross-passage reconstruction
Figure A1.31
Fixing waterproofing membrane to new tunnel lining
As with the main shell, the new cross-passage walls were required to be at least 200 mm
thick. Break-out of the existing walls necessitated substantial temporary propping of the
tunnel by braced colliery arches (Figure A1.30). To afford early warning of any increase in
loading, strain gauges and load cells were fitted to the arches. In the event, no significant
increase in load was noted.
Following completion of break-out and trimming, the drainage mat, fleece, membrane
and reinforcement were fixed before shuttering and concreting. The shutters were
progressively moved through the tunnel in a planned sequence to ensure that no enlarged
openings were closer together than 20 m. There were 56 cross-passages reconstructed: the
four northernmost passages were left bricked up, the four southernmost were preserved.
Once cross-passage reconstruction was well advanced, work on the new lining started.
Four collapsible shutters were fabricated to the required profile, two for each bore,
purpose built to allow the passage of plant beneath.
The sequence of construction for each bay was as follows:
1
Structural brickwork trimmed to provide the required minimum 200 mm concrete
thickness.
2
Flexible half-round perforated drainage pipes clipped to the brickwork
circumferentially, at about 700 mm centres.
3
Stainless steel collector hoppers fixed to the ends of the pipes above the junction with
the invert.
4
Drainage mat fixed all around the surface between the drainage pipes.
5
Drainage system overlain with the geotextile fleece and PVC membrane (Figure A1.31).
6
Mesh reinforcement and weep pipes fixed.
7
Shutter hydraulically manoeuvred into position, checked for line and level, stop ends
fixed.
CIRIA C671 • Tunnels 2009
285
8
Concrete pumped through openings installed in the shutter, externally vibrated.
9
Shutter struck, cleaned down and transported to its next location.
Generally, the new lining was also constructed in 5.5 m long bays. However, the earlier
coring exercises had indicated that the structural brickwork was thinner in some areas of
the tunnel and so it was decided to proceed in shorter lengths in these locations.
Electronically monitored frames were again used for temporary support during break-out
of structural brickwork. Minimum limits were also placed on separation distances between
areas of break-out in the same and adjacent tunnel bores.
The architectural features were formed later by rendering over strips of expanded metal
fixed to the finished concrete surface (Figure A1.28).
3.4.3
Sprayed concrete section at North End
The new profile is straight over most of its 365 m length, except for a section at the
northern end. Wapping Station is built on a curve, with the track transition curve starting
well within the tunnel. Such a profile did not easily lend itself to a shuttered solution, so
sprayed concrete was used in this area to form the lining.
The construction sequence as previously described. The lining concrete was then sprayed
(using wet-mix process) to form a minimum 175 mm thick base layer, including the steel
fibres. A 25 mm thick finishing layer incorporating polypropylene fibres was then applied.
Because greater amounts of structural brickwork required cutting out to accommodate the
increased size of lining, the bay lengths were reduced in size to about 1 m. Propping and
monitoring was again provided. About 47 m of lining at the northern end of the tunnel
was constructed in sprayed concrete, with a further 6.5 m length of sprayed lining also
placed at the southern end. This formed a transition from the formed concrete lining
(280 m) to the preserved section (31.5 m long).
3.4.4
Preserved section
As mentioned, the walls of the southernmost section of the tunnel were preserved intact as
part of the final scheme. EH only required basic cleaning and minor repairs to be carried
out, not full restoration. A new concrete invert slab was provided, as elsewhere in the
tunnel but this section of tunnel was strengthened by ground treatment from above,
carried out under a separate contract in November 1996.
3.4.5
Ground treatment works
LUL commissioned a specialist consultant to investigate means of strengthening the
southern tunnel section. They recommended that the Thames Gravel overlying the tunnel
should be injected with cement grout, forming a slab of about 4 m thick, 20 m wide and 40
m long, overlapping in plan with the fully strengthened section of tunnel. The grout slab
would reduce the potential water flows in the event of a breach.
The grout was injected via tubes à manchette, drilled into the gravels from three working
sites at Rotherhithe: one at surface level, immediately adjacent to Brunel’s Engine House
and the other two from 4.5 m deep shafts sunk in the area. Injection was carried out in
two stages: an initial bentonite-cement mixture, followed by a specially developed
microfine cement grout.
286
Apart from the Engine House, there were other properties nearby and ground heave
during the grout injection was a major concern, so extensive monitoring was put in place
to carefully measure this. This concern was justified, considerable heave actually taking
place, with the rate and volume of grouting then modified to minimise it. Electrolevels
were also installed within the preserved section to monitor movement, though this
eventually proved to be minimal.
3.4.6
Track work installation
The design for the new invert slab could not structurally accommodate sleepers, even
before its thickness was reduced. Direct fastenings to the slab were necessary. Slab track
was extended throughout the platform area of Wapping Station to provide greater lateral
stability. The listed tunnel portals and the presence of a large column supporting the
station lift shaft steelwork posed further constraints, resulting in the new alignment design
being a best-fit exercise.
3.4.7
Drainage
The Thames Tunnel is the low point on the ELL. Occasional pump failures elsewhere led
to more water flowing downstream than the tunnel’s pumps could handle, resulting in
flooding. So disruption of the train service had not been uncommon in recent years.
The final design of a drained lining meant that future water ingress needed to be
accommodated by enlarging the deep mid-point sump and casting in two 150 mm
diameter pipes into each invert slab. The rising main was renewed and duplicated to offer
a standby route and new pumps were later installed.
3.4.8
Programme
The final design scheme was reprogrammed to take into account its greater complexity
and a restart date of 29 April 1996 was agreed with a revised duration of fifty weeks. The
project was actually completed a week before the planned date of 13 April 1997, at a cost
of £23.2m.
CIRIA C671 • Tunnels 2009
287
Case study 4: Standedge north railway
tunnel, investigations and design of
major remedial works
By Robert Hills, Donaldson Associates, Peter Harris,
Donaldson Associates and Ian Wilson, Network Rail
4.1
SUMMARY
The Standedge tunnels form the first and longest of the major trans-Pennine
transportation routes. A total of four parallel tunnels were constructed, starting with the
canal tunnel at the turn of the 18th century followed by two single bores in the mid 19th
century and finishing with the twin tracked north tunnel at the end of the 19th century.
All of the tunnels are over three miles long. Major repairs to all three railway tunnels have
been undertaken over their lifetime. This case study describes the investigation and
monitoring, feasibility and design of ongoing major remedial works to three sections of
the twin tracked tunnel. Particular reference is made to the interrelationship of all four
tunnels and the back analyses of the effects of the construction of each tunnel on the
others. Remedial works entail the reconstruction of the tunnel lining and the construction
of a structural invert beneath the live tracks. The remedial works have been designed to
allow construction over a period of several years due to the strategic importance of the
trans-Pennine route, which only allows works to be undertaken during a short period each
year.
4.2
BRIEF HISTORY
4.2.1
Canal Tunnel
The first tunnel constructed in the Standedge area was Standedge canal tunnel, a section
of the Huddersfield Narrow Canal. The proposed length and width of the tunnel was
5025 m and 2.1 m respectively. Construction began in 1794 and finished in 1811. The
canal tunnel was 4988 m long, shorter than initially planned, but at the time of
construction was the longest tunnel in England only surpassed by the construction of the
Severn Tunnel in 1886. The tunnel was primarily unlined but historically sections have
later been faced with stone and brickwork and recently with sprayed concrete. The canal
tunnel was closed in 1944, under the LMS Canals Act. In the 1990s the canal and tunnel
were restored and Standedge canal tunnel remains open today.
4.2.2
Standedge centre tunnel
In 1840, the Manchester and Leeds Railway company constructed a railway line between
Leeds and Manchester through Hebden Bridge, well to the north of Huddersfield
through a gap in the Pennines. It was much quicker for goods traffic to travel from Leeds
to Manchester via this railway line than by travelling via the canal through the Pennines.
So in 1844 the Huddersfield Canal Company allowed the sale of the canal to the
Huddersfield and Manchester Railway Company, which was then renamed as the
Huddersfield & Manchester Railway and Canal Company (HMRCC).
288
The HMRCC line included a new tunnel beside the Standedge canal tunnel at Standedge.
Due to financial constraints, it was decided that one single track tunnel would be sufficient
to carry the anticipated traffic.
Construction of the tunnel was complete by January 1849, with a final length of 4885 m (3
miles 62 yds). The majority of the tunnel was masonry lined, with only 300 m (329 yds)
unlined. The line and tunnel were inspected on the 2 and 6 July 1849 and later opened
on the 13 July 1849.
Figure A1.32
Cross-section through the tunnels
4.2.3
Standedge south tunnel
By 1868, construction of a second railway tunnel was underway, parallel to and south-east
of Standedge centre tunnel. Construction progressed well and was complete six months
ahead of schedule in October 1870, taking just under two years to construct. The tunnel
officially opened on the 12 February 1871.
4.2.4
Standedge north tunnel
Due to expansion of the railways in the mid to late 1800s, a new double track line was
proposed to the north of the existing railway and canal in the 1880s by the LNWR (which
had taken over the HMRCC). Land was purchased for the storage of materials and spoil
in 1888, shortly after Royal Assent was received. Construction of a third twin track railway
tunnel at Standedge north of the existing two single track railway and canal tunnels began
in 1890.
Due to insufficient room at both ends of the proposed new tunnel, it was necessary to carry
the new tunnel over the canal tunnel. The canal tunnel was later extended by 203 m (221
yds) at the Diggle (western) Portal.
The constructed tunnel was 4884 m (3.66 miles) long, a mere four yards longer than
Standedge centre and south tunnels. The north tunnel opened on the 5 August 1894 and
had taken just over four years to construct.
Effects of tunnel construction
There are numerous records of the effects of each new tunnel construction causing
movement within the older tunnels.
CIRIA C671 • Tunnels 2009
289
Recent developments
In October 1967 the centre and south railway tunnels were closed to traffic, moving all
traffic to Standedge north tunnel. Since the reopening of the canal tunnel the centre
tunnel is used as the emergency access route to the tunnel.
4.3
TUNNEL LININGS
The centre and south tunnels have vertical sidewalls and a semi-circular arch. Standedge
north tunnel is of horseshoe profile with curved inclined sidewalls and compound curved
arch. The centre tunnel is constructed from masonry sidewalls (1 block thick ~ 500 mm)
and a brickwork arch (4 rings thick). Where the arch is constructed from masonry, the
blocks are typically 150 mm thick. Standedge south tunnel arch and sidewalls are
constructed from four rings of brickwork. The bricks used to construct the tunnel are soft
red clay bricks that are susceptible to freeze/thaw due to their high water content. The
facing brickwork lining was constructed from blue Staffordshire bricks. Standedge north
tunnel is entirely constructed from brick varying from 3 to 7 rings thick, set in hydraulic
lime. The lining is 3 rings thick where the surrounding ground is millstone grit and 7
rings thick where the surrounding ground is shale. See Table A1.2 for further details of
varying thicknesses of the lining.
All location references within the tunnels are referred to by tablet numbers (+ feet where
unspecified) as marked in the north (railway) tunnel. Tablet markers are placed at
intervals of 15.2 m (50 ft), Tablet 0 being at the Diggle (western) portal and Tablet 320
towards Marsden.
Table A1.2
4.4
North tunnel lining details
From
Tablet
To Tablet
Number of
brick rings
74
81
7
Cement Concrete invert Tablets 80 to 88
81
94
6
Brunn Clough Shaft Tablet 85
94
105
5
149
150
4
175
176
4
198
222
3
233
234
5
Flint shaft
253
254
4
Slight bulging in crown of arch. Few crushed bricks in crown of arch
254
279
4
Lime concrete floor Tablets 270 to 301
279
281
5
281
288
4
5 rings at Adit 12 Intersection
288
290
5
Lime concrete floor Tablets 270 to 301
Comments, repair details
Horizontal crack in masonry height 17ft length 25ft. Slight bulging in arch
GEOLOGY
Standedge tunnels are constructed through strata belonging to the Millstone Grit Group
deposited during the Namurian Epoch of the Upper Carboniferous Period. The Millstone
Grit Group regionally comprises a sequence of alternating thick beds of sandstone or
gritstone and mudstones or shale with subordinate marine bands, coal seams and
seatearths. The formation of the millstone grit strata results from rhythmic sedimentation
290
of fluvio-deltaic deposits, leading to depositional features such as channels and slumps.
The sequence has been subject to folding and faulting, as a result of the uplift of the
Pennine anticline. The regional strata dip about 5° towards the north-east, locally
steepening adjacent to faults. The dip also varies locally in the vicinity of folds and
flexures. Due to the dip of the strata and the sub-horizontal alignment of the tunnels, the
geological succession encountered is younger towards the Marsden Portal and the northeast. The geological succession encountered within the tunnels has been summarised in
Table A1.3, together with tablet references and typical lithological descriptions.
Several faults are shown to cross the line of the tunnel at ground surface. The locations of
the faults at the horizon of the tunnels have been established from observations made by
the BGS, from investigations carried out by Donaldson Associates Limited (DAL) and from
historical information and are shown in Table A1.4, which summarises the characteristics
of the faults. The downthrow of fault nos 1, 2 and 4 are large and in excess of 50 m. The
adjacent geology was probably disturbed during movements along this fault plane, which
has created shear zones, perhaps up to 5 m thick either side of the fault. Movement along
the faults and within the adjacent shear zones is probably ongoing.
Table A1.3
Geological succession within Standedge tunnels
Tablet reference
Diggle portals to
Tablet 20/21
Approx. Tablets
20/21 to 31/32,
103/104 to
147/148 and
166/167 to
184/185
Formation
Typical description
Geological
observations during
construction
Shale Grit
(120 to 150 m
thick)
Very thickly bedded
massive coarse
sandstone with thin
mudstone beds
NW-SE trending fault at
Tablet 20 to 21, 70°
dip to NE
Bedding dips from 23°
to 43° to NE
Grindslow Shale
(100 m thick)
Interbedded massive
silty or sandy
mudstone and silty or
carbonaceous
sandstones
NE-SW trending fault at
Tablet 103 to 104, few
metres downthrow to
SE.
Bedding dips from 3°
to 22° generally to
E/NE.
Steep joints noted.
Bedding dips from 4°
to 26° generally to
N/NE
N-S trending fault at
Tablet 196, few metres
downthrow to E
NW-SE trending fault at
Tablet 255, 75° dip,
45 m downthrow to NE
Thin coal seam of 200
mm thickness, (coal
seam no 4), between
Tablets 225 and 233,
resting on seatearth
and then sandstone/
gritstone.
Fault reported during
construction from
Tablet 303+49 to
306+18, 3.65 m wide,
“full of shale, softer
earth and coal”.
Coal seam no 1
between Tablets 260
and 273.
“Grey shale” noted to
Tablet 306+18.
Limestone bed and
nodules noted between
Tablets 299+40 and
306+18
Approx. Tablets
31/32 to 103/104,
147/148 to
166/167 and 184
to 255
Lower Kinderscout
Grit
(100 m thick)
Coarse sandstone
interbedded with
subordinate mudstone
and siltstone
Approx. Tablets 255
to 260
Undifferentiated
Millstone Grit
(15 m to 20 m
thick)
Medium to dark grey
mudstone and
siltstone with one
marine band
Approx. Tablets 260
to 311
Upper Kinderscout
Grit
(15 m thick)
Medium bedded
coarse sandstone
Approx. Tablet 311
to the Marsden
Portals
Undifferentiated
millstone grit
(20 m to 30 m
thick)
Dark grey mudstone
with siltstone and
sandstone beds, and
two marine bands
CIRIA C671 • Tunnels 2009
Structural
information
Sandstone noted at
fault location
Coal seam no 2
between Tablets 311
and Marsden Portals
291
Table A1.4
4.5
Fault locations
Fault No
Tablet reference
Trend
Dip (°)
Downthrow (m)
1
20 to 21
NW–SE
70
Large, NE
2
103 to 104
NE–SW
60 m, SE
3
196 to 197
N–S
Small, E
4
255 to 256
NW–SE
5
303 + 49 to 306 + 18
NW–SE
75
45 m, NE
Small, NE
TUNNEL MONITORING
In December 1995 convergence measurement arrays were installed at twelve locations,
from Tablet 1+05 to Tablet 288+30, within Standedge north tunnel. Monitoring by tape
extensometer has been carried out on a monthly basis and is ongoing. In 1998 further
arrays were installed in both the centre and south tunnels. The locations of the arrays
were determined by review of the examination reports and the known location of bulging
and/or cracking of the tunnel lining. Movements detected by the extensometer
monitoring prompted the installation of an array of five electro-level sensors at Tablet
262+20 and at Tablet 262+30 in December 1997. Between December 1995 and October
2000 the up and down sidewalls at Tablet 262 show a total convergence of 7.96 mm and
the up sidewall and down haunch converged by 8.47 mm. Convergence at similar rates is
continuing.
4.6
TUNNEL INVESTIGATIONS
DAL was commissioned to investigate the condition of the tunnel lining in the north
tunnel near to Tablet 262 in 1998. Initially the investigation comprised a total of 10 rotary
cored holes through the tunnel lining into the bedrock. These holes found that the lining
thickness of the brickwork varied between 310 mm to 650 mm and that the brickwork
recovered was generally of good condition. The material recovered from behind the
brickwork lining was typically a weak dark grey mudstone, initially highly fractured and
frequently recovered as gravel size fragments, with associated loss of flush and corehole
collapse. Pockets of clay and evidence of brecciation were encountered locally, with some
yellow staining and mineralisation. More competent mudstone was found at between 2 m
and 3.4 m depth.
292
Figure A1.33
Repair histories of tunnels
A review of major repairs in all the tunnels was made and Figure A1.34 shows the results
for the section at Tablet 262. As part of the investigation to establish the failure mechanism
causing distress to the lining two 40 m long rotary cored holes were drilled from the
centre tunnel to about 5 m above the north tunnel. These holes revealed the rock pillar to
comprise generally moderately strong to strong, fresh, thinly laminated mudstone.
Discontinuities are typically close to medium spaced and present in three sets: at 15° to
20°, 45° and 50° to 75° to the core axis. The discontinuities are generally planar and
smooth, with occasional 1 mm to 2 mm infill of clay or calcite. Stronger siltstone was
present at the base of one of the holes. There was evidence of ground movement in the
form of discrete shear zones and areas of brecciation and preferential weathering,
particularly in corehole no 2 where such zones occurred between 10 m to 36 m into the
corehole. To obtain geotechnical design parameters, laboratory testing for uniaxial
compressive strength (UCS), triaxial strength and Young’s Modulus was carried out on
core samples of mudstone. Other investigations carried out included trial pitting in the
cess to determine whether there was evidence of a structural invert, which was not found.
4.7
GROUND MODEL
The ground model was developed based on the results of the desk study, geological
review, visual inspections and data assessments.
4.7.1
Geological model
The geological review demonstrated the presence of a north-west to south-east trending
fault at 1100 m (Tablet 255) from the Marsden Portal, with a dip of 75°, and a throw of 45
m to the north-east. This fault was encountered during construction and it is recognised
by the BGS as one of the major structures intersecting the tunnel alignment. Rotary
coring at Tablet 262+20 to 262+30 has demonstrated the presence of many discrete
shears within the mudstone over a width of about 15 m, ie a shear zone that is trending in
a north-west to south-east direction, parallel to the known major fault. The extent of
information was such that it was not possible to be definitive regarding the dip and dip
CIRIA C671 • Tunnels 2009
293
direction of this shear zone. However, reasonable geological judgement would suggest that
the shear zone should have a very similar orientation to that of the fault. It was on this
basis that the geological model was developed. Given a true dip of 75° and a dip direction
of 075°, stereographic projection of this plane on to a section perpendicular to the tunnel
(trending to 335°) gives an apparent dip on the section of 65°. At Tablet 262+20, the
shear zone is located between the north and centre tunnels, and at Tablet 262+30, it is
located through and to the north of the north tunnel alone.
4.7.2
Geotechnical model
The geotechnical model involved a great number of variables, most were due to the
heterogeneous nature of rock masses, and could never be defined with any certainty. So it
was important that the geotechnical model incorporated not only the best estimate for a
given parameter, but also the likely range that the best estimate may fall. The following is
an indication of the number of variables involved in defining the geotechnical model for
this project:
Rock mass quality
The primary source of reliable and auditable geotechnical data was contained within
coreholes and associated laboratory testing. An assessment of the rock mass quality of the
rock from these coreholes was made, according to the updated “Q” system proposed by
Barton & Grimstad (1993). The results of the assessment are summarised in Table A1.5.
Table A1.5
Summary of rock mass quality assessment
Q′
Q
Material
Lower
Mean
Upper
Lower
Mean
Upper
Mudstone
0.08
1.7
3.9
0.21
3.54
7.9
Sheared zone
0.01
0.03
0.13
0.02
0.08
0.33
Rock mass strength
Based on the rock mass quality assessments correlations were made between Q′ and HoekBrown constants. Estimates of both peak and residual constants were made and the results
summarised in Table A1.6. Linear c′ and φ′ values for the sheared mudstone were derived
from a literature search for published values for shears within coal measures mudstones,
these being the most frequently studied materials most similar to those under
consideration. Several sources were referenced suggesting that fault gouge and shear
zones within mudstones (other than intra-formational shear zones parallel to bedding)
could be expected to have a shear strength of c′ = 0 and φ′ = from 8° to 25°, typically φ′ =
12°. However, it was noted that the majority of these shear strength values were derived
from low-stress, surface excavation, scenarios. The effect of confinement at depth would
be to increase the value of c’ and probably reduce the value of φ′. This was borne out by
tangents to Hoek-Brown envelopes over the relevant stress range, which suggested shear
strengths of c′ = 10 to 30 kN/m² and φ′ = 12° to 20°. A range of these parameters was
used where the sheared zone was modelled as a linear Mohr-Coulomb material to
highlight the uncertainty.
294
Table A1.6
Summary of range of adopted Hoek-Brown Constants
Peak
σc (MPa)
Material model
Mudstone
Sheared zone
Residual
mp
sp
mr
sr
Lower
20
0.33
0.00042
0.027
0
Mean
35
0.80
0.0067
0.16
0.00055
Upper
50
1.07
0.016
0.284
0.002
Lower
1.25
0.16
0
00.16
0
Mean
1.25
0.24
0
0.24
0
Upper
1.25
0.38
0.00065
0.38
0.00065
Notes
(σc) represents the uniaxial compressive strength of the intact rock
m and s are material constants that depend upon the properties of the rock and the extent to which it has been broken
s = 1 for intact rock
m is determined from tables.
Rock mass stiffness
In the absence of large-scale in situ deformation testing, the most reliable means of
estimating the stiffness of a rock mass is by correlation with careful assessment of rock
mass quality. The range in value of Q was correlated with rock mass deformation modulus
(Em′) using several published relationships and summarised in Table A1.7.
Table A1.7
Summary of deformation modulus correlations
Deformation Modulus (GN/m²)
Rock mass quality
Bieniawski 1978
Serafim &
Pereira 1983
Barton 1996
Hoek &
Brown 1997
Q = 0.08
0.81–1.47
1.88
2.52
0.84
Q = 1.70
1.67–3.04
8.91
8.41
5.27
Q = 3.9
1.96–3.57
14.12
12.49
9.98
Table A4.6 shows that there was a wide range in predicted rock mass deformation
modulus for a given rock mass quality. For example, the predicted modulus for the mean
rock mass quality ranges from 1.67 to 8.41 GN/m². This range was however restricted by
consideration of the known intact rock modulus determined from laboratory tests. These
results indicated an intact rock modulus of between 5.7 and 10.5 GN/m². This suggested
that Bieniawski’s 1978 correlation (which uses the intact modulus) may provide the most
satisfactory and reliable correlation. This relationship formed the basis for a sensitivity
study based on rock mass stiffness.
rock mass behaviour (non-linear vs. linear, dilatent vs. non-dilatent)
post-peak strength (perfectly plastic vs. strength reduction)
in situ stress (magnitude and orientation).
The Standedge tunnels are situated on the flanks of the Pennine Anticline within a series
of predominantly north-west to south-east trending faults. A subordinate series of northeast to south-west trending faults are also present. So it was expected that the major
principal horizontal stress would indeed be orientated towards the north-west to southeast. However, this may vary in proximity to the faults and shears that intersect the tunnel.
CIRIA C671 • Tunnels 2009
295
So, for modelling purposes, K0 from 1 to 2 was used together with a direction of the major
horizontal principal stress both in-plane and out-of-plane.
4.8
ANALYSIS
4.8.1
Modelling of ground movement and modelling philosophy
The modelling of ground movement was carried out using the computer program Phase2
produced by Rocscience Ltd in Canada. This program is a 2D finite element program for
calculating stresses and estimating support around underground excavations. The
purpose of this modelling was to simulate the broad trends and magnitudes of
known/suspected ground movements. This involved running a series of models, each
using a different combination of the many variables identified. Ranking of the results for
the best fit(s) was then used to identify the model(s) that most accurately replicated
historical and current trends in movement. Once a model, or series of models, had been
created that reasonably matched historical and current trends in movement, the design of
stabilisation works could then be progressed with a certain degree of confidence. The
computer modelling was progressed with the aim of replicating, first, the broad trends of
historical and current movements, and second, the approximate magnitude of historical
and current movements. Four basic excavation stages were included to keep the model as
simple as possible without oversimplifying matters. The stages were as follows:
4.8.2
Stage 1
Excavate canal tunnel (presumed unlined at Tablet 262).
Stage 2
Excavate centre tunnel and lining to sidewalls and crown (no lining delay).
Stage 3
Excavate south tunnel and lining to sidewalls and crown (no lining delay).
Stage 4
Excavate north tunnel and lining to sidewalls and crown (no lining delay).
Results
The modelling identified the following broad behavioural trends:
296
total displacements increase surrounding all four tunnels as the quality of the
mudstone become progressively worse
at the north tunnel, for the lower-bound quality mudstone, the maximum
displacement occurs at the invert. As the mudstone becomes progressively stronger,
the point of maximum displacement rotates such that for the upper-bound mudstone,
it occurs in the upper sidewall/lower haunch area
at the north tunnel, for the lower-bound quality mudstone, surrounding displacement
is roughly symmetrical and the influence of the sheared zone increases progressively
as the mudstone becomes stronger. For the upper-bound mudstone, the displacement
profile is strongly asymmetrical and skewed towards the shears with virtually no
displacement away from the shears
at the centre tunnel, total displacement profiles are aligned roughly parallel to the
sheared zone for the lower-bound quality mudstone. The profiles rotate as the
mudstone becomes stronger to become almost perpendicular to the sheared zone
at the centre tunnel, for the lower-bound quality mudstone, surrounding
displacement is roughly symmetrical and the influence of the sheared zone increases
progressively as the mudstone becomes stronger. For the upper-bound quality
mudstone, the displacement profile is strongly asymmetrical and skewed towards the
shears with virtually no displacement away from the shears
for the lower-bound quality mudstone properties, the major principal stress plot
shows a large zone of very low stress surrounding the north, centre and south tunnels.
These “de-stressed” zones decrease in size as the mudstone becomes stronger and
virtually non-existent for the strongest mudstone
stress concentrations within the sheared zone and especially surrounding the canal
tunnel are highest for the lowest quality mudstone and reduce progressively as the
quality of the mudstone increases
strength factors away from the sheared zone increase as the quality of the mudstone
increases. Strength factors within the sheared zone remain low irrespective of the
quality of the mudstone.
The sensitivity study allowed important conclusions to be drawn about what was, and what
was not, likely to be required for a successful geotechnical model to be established. A
ranking matrix approach was used to assess the results and determine the best fit(s). The
results of the ranking assessment found that three models would be used to design the
major stabilisation works to the north tunnel.
4.8.3
Structural analysis
Following this the brick masonry tunnel lining was analysed using PFRAME by CSC a
linear elastic analysis program using a nominal load case based on the ground
deformation around the tunnel lining. The analysis results were post processed using an
EXCEL spreadsheet to check the magnitude of the compressive stresses in the lining. The
loading applied in the PFRAME model was factored up until a pin “formed” in the lining
and the facility in the program to allow the model to take up the deflected profile was
used and a pin inserted in the model. The modified model was then re-analysed and the
loading incremented. During this iterative process the deflection of the model was
checked and, as the tunnel deflected further to the north into the rock mass, springs were
added to provide passive support. Eventually the model developed so many pins that the
model became a mechanism. This back analysis enabled the magnitude of the rock loading
to be assessed with a degree of confidence.
The analysis found that the arch could support only a nominal vertical load and as a result
has negligible capacity to support asymmetric loading due to ground movements. This
indicated that the lining as built (which in some instances was constructed after originally
designing the tunnel as unlined) was intended for durability and, structurally, to do no
more than prevent small, localised rockfalls. The analysis confirmed observations made
with respect to the lining thickness in comparison with many other tunnels previously
examined, where the linings have been considerably thicker and known to have been built
as true structural linings to support ground loading. The lack of structural capacity in the
lining was not unexpected as the majority of the Standedge tunnels are within competent
sandstones and gritstones, which is evident from the length of the unlined section in the
canal tunnel. Further corroboration was provided by the large amount of relining and
construction of inverts in all the tunnels required during the construction phases when
adverse ground conditions were encountered.
4.9
DESIGN OF REMEDIATION WORKS
The development of the design of the remedial works began with an inspection of the
defects. At Tablet 262 the movement was evident in the form of cracking of the lining, a
horizontal crack had opened up on the north (down) haunch at about 3 m above rail level
and a section of the south (up) sidewall in the haunch was bulging into the tunnel and also
cracking.
CIRIA C671 • Tunnels 2009
297
The next stage was to look at various options for remedial works. These had to be
constructible in the railway environment within several long railway possessions. The work
area at Tablet 262 is 890 m from the Marsden portal of the tunnel.
In view of the current level of cracking and ongoing continuing deformation it was
considered that the existing lining was close to its load carrying capacity over a length of
about 10 m. While temporary measures such as stitching the horizontal crack in the down
sidewall and cross pinning and grouting up the bulged and cracked area of the upside
haunch could be implemented carried out, these would not significantly improve the
lining’s long-term strength. It was decided that the long-term objective would be to
replace a 10 m length of the tunnel arch lining with a new significantly stronger structure.
Due to clearance considerations it was not feasible to internally line the existing brick arch.
The option chosen was to cut the lining out in transverse strips and insert steel lattice
girders and infill these with sprayed concrete. Similar work to this had been successfully
used for the remedial works to Conisbrough Tunnel designed by DAL and carried out by
Amalgamated Construction in 2000.
From the initial analysis it was apparent that provision of a new invert to the tunnel would
be structurally beneficial. There was concern that the required depth of excavation to
allow construction of a new concrete invert and also provide a minimum of 300 mm of
ballast below the sleepers could potentially undermine or significantly affect the lateral
stability of the sidewalls. So, it was decided to construct a new ground beam in the cess,
which would be founded on piles. The ground beam would be in the form of pre-cast
concrete trough units (see Figure A1.34). The piles would help to prevent lateral
movement of the sidewall footing when the ground beneath the track was excavated to
allow installation of the new invert.
Figure A1.34
Pre-cast concrete cess
trough and CHS pile
298
These works would be done over a 30 m length of tunnel centred on the proposed 10 m
replacement of the main arch (see Figure A1.35). Various options for constructing the new
invert were considered:
Figure A1.35
“hit and miss” in situ reinforced concrete across the full width of the tunnel
pre-cast units in panels under each track.
Plan of remedial works showing piles
The problems with these forms of construction are related to access. At least one track
must be kept operational to bring in materials and plant. The in situ option was rejected
due to the difficulty of working with both tracks in place while excavating under it and
fixing reinforcement. It was decided to take up one track at a time and lay half an invert
then repeat the operation on the other track in a separate possession.
At this time the objective was to use the largest possible invert panels that could be
handled by two road/rail cranes and to lay them in a “hit and miss” pattern in two passes
to minimise the lengths of open excavation. Attention was then focused on how to provide
support to the ballast and formation of the running track while excavating and running
road/rail vehicles on it to lift in the pre-cast invert slabs. A pre-cast concrete trough unit
similar to those used in the cess (but deeper) was proposed. This could be installed
between the sleeper ends and would support the running track and allow lapping
reinforcement to be fed through slots in the side faces to tie the two halves on the invert
together. This may have been feasible but the required tolerances would have been very
difficult to achieve. So a solid pre-cast concrete block base unit was designed with couplers
in the side faces to enable connection to the invert panels to be made and with a
detachable steel box to provide ballast retention sat eccentrically on top. This system
allowed the pre-cast concrete invert units to be rationalised in size as the block could be
positioned exactly on the centreline between the cess units (see Figure A1.36).
CIRIA C671 • Tunnels 2009
299
Figure A1.36
Pre-cast block in the six foot with ballast retention box
It soon became apparent that the works would have to be phased over several years and
the following phases were proposed:
install piles along each sidewall and connect with a ground beam and carry out
stitching works to the cracked and bulged areas
install a new concrete invert slab
replace the most severely damaged 10 m long section of the lining with a new sprayed
concrete lattice girder reinforced lining.
4.10
REMEDIATION WORKS
4.10.1
Mini-piling
The first phase of the works was installed in October to December 2001. Pre-cast concrete
trough units 3 m long were installed in each cess close up to the sidewalls and the gap
between them and the sidewalls filled with concrete. 225 mm diameter piles were then
bored through the holes in the trough units to a depth of about 6 m. A reinforcement cage
was fitted in the troughs and this was partly filled with concrete. This work was carried out
in eight no 19 hour possessions. The alliance contractor was May Gurney and the
specialist piling subcontractor was Systems Geotechnique.
4.10.2
Invert construction
The second phase of the works was programmed for November to December 2004.
Discussions were held with May Gurney and their subcontractor WA Developments to
refine the design details to ensure maximum efficiency on site, in particular with the size
of the invert units and the lengths of excavation that could be open at any one time. It was
decided to use only one road/rail crane and reduce the invert panel size. Eventually the
practical considerations of excavation and storage of material and installation of the units
within the tunnel resulted in the decision to reject “hit and miss” installation and a change
to working from one end in a continuous process of excavation and installation. This
change resulted in the lengths of open excavation exceeding the capacity of the
reinforcement in the cess troughs installed in 2001 and secondary support of the sidewalls
by RMD shores was introduced.
300
Figure A1.37
Invert construction
CIRIA C671 • Tunnels 2009
301
Ballast retention units were installed in the six foot, these had a permanent reinforced
concrete block base with rows of Erico Lenton™ couplers in each side face and a
detachable steel top unit to retain ballast. The sleeper ends were grout-packed tight to the
ballast retention units and the pre-cast troughs installed in phase 1. Following these
preparatory works there were two long 48 hour possessions. In the first the up line was
removed and the ballast and sub ballast progressively excavated from the east end with
the temporary shores being inserted between the ballast retention units and the cess
trough units to prevent inwards movement of the sidewalls. A total of 14 pre-cast invert
panels each weighing 4.2 tonnes were then brought in on the Down line and installed as
excavation proceeded. The invert panels were nosed under the cess troughs and the void
between them grouted up through grout injection tubes cast in the panels. Couplers in
the ballast retention units enabled lap bars to be fixed into recesses in the invert panels.
The design of the pre-cast invert panels minimised the quantity of in situ concrete
required and the use of rapid hardening cement and steel plates placed over the concrete
after initial set enabled the ballast to be replaced over the invert panels as soon as possible.
The shores were removed when the ballast was replaced and compacted. Once all the
invert panels were in and backfilled the track was replaced and packed and fixed where
cut with temporary fishplates.
In the second 48 hour possession the whole operation was repeated for the down line. In
later shorter possessions the track was re-welded and the top steel box sections of the
ballast retention units were unbolted and taken out and the ballast packed.
At the end of this phase further site investigation was carried out and it was identified that
the inner leaf of brickwork in this section of the tunnel had already been rebuilt at some
time in the past and was not bonded to the outer brickwork.
4.10.3
Tunnel relining
The third phase of works was started in the period October to December 2006. It was
proposed to cut out the existing tunnel lining in 0.6 m wide slices and install lattice
girders and infill with sprayed concrete, this was to be done on a “hit and miss” basis
during two 48 hour possessions until the whole 10 m length was replaced. The new
arched lining will be connected to the reinforced concrete tie beam already partially cast
inside the cess troughs. In the crown of the tunnel where clearances to trains was not a
problem the existing brick lining is to remain and be sprayed over.
Figure A1.38
302
Relining
Temporary works were installed to ensure the stability of the lining as it was partly cut out.
These consisted of:
stitching the brick rings together with grp bars in the areas where part of the
brickwork was cut out
2.5 m long rock bolts in the crown, which was to be over-sprayed to provide
temporary support. The analysis model was modified to incorporate these ties and the
forces in them determined. The load capacity of the rock bolts was confirmed on site
by a trial installation and load testing
temporary thin steel straps each side of the sections of brickwork to be cut out to
prevent the brickwork unravelling.
It was also necessary to break out, as designed, the back faces of the concrete troughs in
the cess installed in 2001 and locally form an opening in the brickwork and cast a footing
for the lattice girder arch with reinforcement connecting this to the troughs. This work
was carried out in short railway possessions before the main works and the opening
temporarily in-filled with concrete blockwork.
At present all the temporary works are completed and the final installation of the arches
and new lining has been carried out in late 2009.
CIRIA C671 • Tunnels 2009
303
Case study 5: Geophysical surveying to
identify hidden shafts
By Simon Brightwell, Aperio Ltd
As part of the Network Rail Tunnel management strategy, a three stage investigation to
locate hidden shafts was conducted throughout the Western Territory. The work involved
50 masonry tunnels in the territory and comprised:
Stage 1
Desk study and walkover.
Stage 2
Reconnaissance geophysical survey.
Stage 3
Targeted intrusive investigation.
To establish the number and location of likely hidden shafts a desktop study was
conducted by a dedicated researcher who accessed Network Rail records and external
archive sources such as public records offices and local newspapers dating from the time
of construction. This exercise proved successful in providing a wealth of information.
However, many of the positional references from the old records proved to be confusing
and contradictory when compared to the chainage systems within the tunnels today.
Walkover surveys were conducted at each tunnel to look for evidence of shafts, such as
spoil heaps or depressions in the ground, as well as determining land-use above the
tunnels. Photographs and notes taken were useful in places where the surface geophysical
surveys were later conducted. Walkthrough visual surveys within the tunnels were also
conducted where possible.
After considering various investigation options it was decided that the main geophysical
reconnaissance surveys would take the form of ground penetrating radar surveys (GPR)
from within the tunnels where possible, as opposed to surface geophysics from above the
tunnels. Key factors in this decision making process were:
the imprecise location of the tunnel line and suspected shafts on the surface. Without
carrying out detailed topographical surveying within the tunnel it is not possible to
transfer the positions of the tunnel centre line and possible shaft locations onto the
surface accurately, meaning that large areas would have to be surveyed to ensure
coverage of the shafts. Surveying within the tunnel gave a high degree of confidence
that the areas were covered
the ability to optimise resources by combining most of the GPR surveys with preplanned track possessions for the annual programme of tunnel assessments
the absence of overhead electrification in Western Territory, enabling unobstructed
access to the tunnel crown
the quicker progress and lower cost of the GPR method due to the factors described
here.
The GPR surveys were conducted from a scaffold tower erected on hand pushed trolleys.
304
This approach did not provide the quality of access afforded by a powered road/rail
vehicle with an adjustable platform, but was quicker and simpler to set up and allowed
cost savings. At least three longitudinal GPR profiles were collected from the crown area of
most tunnels, with more data being collected where time permitted. The position of all the
data collected was cross-referenced to tunnel portals and chainage markers. Shorter
tunnels such as Highertown (72 yards) were surveyed in as little as two hours, but longer
tunnels such as Chipping Sodbury (4444 yards) required two teams working for six hours.
In some cases, where there was a short distance between tunnels on the same line, such as
the Central Wales Line it was possible to survey three tunnels in one night.
Each GPR profile provided a continuous longitudinal section through the lining and the
materials to the rear. The results were processed and the data were analysed by
geophysical specialists with knowledge of Victorian tunnel construction practice and
considerable experience in interpreting tunnel GPR data. Typical shaft eye characteristics
that can be resolved by GPR include localised thickening of the lining, embedded timbers
or voids in the brickwork and the overbreak, and sometimes the walls and cavity of the
shaft itself.
Results were reported in a standardised format and a confidence rating was applied to
every shaft location reported. The confidence rating was based on the number of GPR
profiles that the shaft could be identified in, the clarity of the data (which can vary due to
soot, moisture, brick type etc), and the correlation with the recorded position derived
from the desktop study or other visual indicators such as spoil heaps.
Three tunnels were surveyed from the surface only because of factors such as difficulties in
gaining track access, or the presence of steel reinforced shotcrete on the lining.
In a very small number of cases the Stage 1 work (desktop study) led directly to Stage 3
(intrusive investigation). An example of this type of high priority site was at Colwall Old
Tunnel near Great Malvern, where records suggested the presence of a hidden shaft in
the garden of a domestic property. A detailed surface geophysical survey was conducted
using microgravity to locate changes in ground density, and GPR to map electrical
changes. Both methods pointed to an anomaly below the resident’s driveway, which was
later excavated to reveal a shallow concrete cap over an open shaft about 80 m deep.
Figure A1.39
Colwall Old Tunnel: concrete shaft cap exposed after targeting by geophysical survey
Stage 3 investigations in Western Territory are continuing on a targeted basis.
CIRIA C671 • Tunnels 2009
305
Case study 6: Relining of Blisworth
Tunnel
Adapted from a presentation given at the British Rail Works Conference 1983 by
R Garrett, British Waterways Board. Photos courtesy of Chris Reynard, British Waterways
Canal tunnels are the oldest tunnel structures in the country, and pre-date the railways by
at least 50 years, the major sewers of London by 80 years and the London tube tunnels by
about 100 years. Work at Blisworth on the Grand Union Canal near Northampton started
in 1794, but due to problems with groundwater and other difficulties was abandoned in
January 1796. Work re-started on a different line in May 1796 and was eventually
completed and the tunnel opened to traffic in March 1805.
Blisworth was, until recently, the longest navigable canal tunnel in the country at 2812 m
long and 5 m wide by 5 m high. The tunnel is almost entirely brick lined using bricks
made on site from locally dug clay. Now the longest is Standedge tunnel (5029 m) through
the Pennines, which was reopened in 2001.
In common with many waterway structures, there has been a general deterioration due to
the age of the tunnel. However, Blisworth also has a history of major repairs carried out
during its life. From as early as 1820 small areas of the tunnel have failed and required
repairs to include such items as replacing the brick invert, totally relining sections and
patch repairs to parts of the brick lining.
The major problem at Blisworth is related to the geology of the ground through which
the tunnel is driven and associated groundwater. The second tunnel attempt incorporated
an extensive network of drainage headings to control groundwater at the interface of the
Lias clay and the water bearing Blisworth sandstone.
In considering the problems at Blisworth the tunnel can be divided into about three equal
lengths. The northern third from chainage 0 m to 950 m is generally in reasonable
condition. As would be expected in a structure of this age there is a requirement for local
repairs to deteriorating brickwork, replacement of pointing etc. The only significant
problems in this length are associated with construction shafts, which were also drains
allowing the groundwater to flow from the sandstone down to the Lias clay.
Similarly the southern third from 1875 m to 2812 m, while not in such good condition as
the northern third, should be adequate for many years of further life given adequate
routine maintenance.
The major problems are concentrated in the middle third and most of the past failures,
including those that closed the tunnel in 1977 and again in 1979, have taken place in this
length. These failures have been closely associated with the clay/sandstone interface, which
dips from the north portal, where the tunnel is constructed entirely within the Lias clay,
and approaches the tunnel crown at the third point. Failures of the tunnel lining have
generally been initiated by heave in the invert and bulging of the lower part of the side
walls.
306
The 1979 failure involved a serious bulge in the lining leading to spalling and crushing of
brickwork. Temporary timber supports were installed to prevent any further movement
taking place. While a local repair could have been carried out, it was felt that with two
failures occurring in a short period a more detailed examination was required. It would
acceptable to deal with one failure and reopen with the possibility of further failures
taking place.
The first stage of investigation before adopting a method of carrying out remedial works
was the comprehensive survey of the tunnel by the Board’s mining engineer. The
assessment of the tunnel lining was carried out by a procedure used previously and now
adopted as a standard. The whole of the tunnel lining above water level was marked out
by a survey team in square metres. Each square metre was then categorised according to
the type and severity of deterioration (Figure A1.40). Also, other features such as openings
to side headings and shafts, water inflows and previous repairs were recorded on the
survey sheets.
Figure A1.40
Tunnel intrados marked out in 1 m squares to allow condition mapping –
this area exhibits some spalling of brickwork at the crown
A further feature of the comprehensive survey was a collation of all available information
concerning the tunnel. The sources of information included historical records, geological
records and boreholes. Following the preparation of this report in 1980, consideration was
given to the various options for remedial works. It was agreed that the existing structure
gauge should be maintained, permitting two-way traffic to continue. Strengthening the
tunnel by some form of internal lining, for example, reinforced sprayed concrete, was
excluded.
The final agreed solution adopted by the Board in 1981 was a pre-cast concrete segmental
lining to replace the brick lining, over a length of about 1000 m. Further work had to
await the provision of finance in 1982–1983.
In April 1982 a period of intensive investigation work started. This involved close cooperation between the British Waterways Board’s staff, soil investigation contractor
Geotechnical Engineering and consulting engineers, Mott, Hay and Anderson (now Mott
MacDonald Ltd). The work included extra boreholes, both from ground level and radial
drilling from inside the tunnel, dewatering of the tunnel to inspect the invert,
investigation of drainage headings, inspection of ventilation shafts and investigation to
locate backfilled construction shafts.
CIRIA C671 • Tunnels 2009
307
To start the remedial works quickly, two separate contracts were let. The main contract to
reline the middle third of the tunnel was inevitably delayed, both by its technical and
contractual complexity and by the necessity to analyse the results of the site investigation
and design the permanent lining. However, certain items of preparatory work could be
easily identified and were incorporated in a preliminary contract.
The preliminary contract, won by John Mowlem and Company, included the construction
of cofferdams and dewatering of the northern two thirds of the tunnel, construction of
access roads and local patch repairs and pointing to brickwork in the outer thirds (Figures
1.41 and 1.42).
Figure A1.41
Patch repairs underway supported off steel centering
Figure A1.42
Completed patch repair
When the main contract (also won by John Mowlem) began access was immediately
available to the tunnel to start work. An enlarged chamber 8 m long and 7.6 m diameter
308
was excavated by hand 950 m from the north portal, and lined with pre-cast concrete
rings. This enlargement provided a workshop in which the tunnel shield was erected
before starting the main drive (Figure A1.43).
The shield gave protection to the work force and support to the ground as it is
progressively jacked forward to complete the relining. All the equipment required moves
forward with the shield. This includes a hydraulic excavator, a conveyor to remove the
excavated material and erecting equipment to construct the new concrete rings. Any voids
behind the completed rings are filled with cement grout (Figure A1.44).
Figure A1.43
Construction of concrete segmental lining within the tunnel shield
Figure A1.44
Grouting behind the tunnel lining to stabilise and help to waterproof it
The work is inevitably more complex than constructing a new tunnel. The ground has
been disturbed and in some cases weakened, both at the original construction stage and
during the succeeding 180 years. Connections to ventilation shafts and the existing tunnel
drainage system have to be made as the work proceeds. Special problems may occur at the
location of the old backfilled construction shafts.
CIRIA C671 • Tunnels 2009
309
Case study 7: Leak sealing and
rehabilitation of sewer tunnels
By Chris W Rees, consultant to May Gurney Ltd (and formerly of Insituform Technologies)
7.1
OVERVIEW
The quality of the permanent works of the main sewer infrastructure created in the UK
during the late 1800s and early 1900s is a tribute to the skills and high standards of
workmanship of those times. The builders were however unable to benefit from the void
and annulus grouting techniques now standard in below ground construction as these
were developed only in c1900 and not in common use until the 1930s. Also, Portland
cement was not commonly used in the UK until the 1920s and in certain circumstances
the lime-based mortars used at the time of construction have deteriorated over time.
This case study considers the original construction techniques and outlines specific
examples where using present day materials and techniques has allowed the structures to
be returned to a condition equal to or surpassing that existing immediately post build.
7.2
INTRODUCTION
Although the great majority of man-entry sewer renovation contracts carried out over the
past 40 years have been directed at structural upgrading of the existing sewer, these works
also serve to dramatically reduce or stop any leakage previously existing within and
through the original structure.
The case studies in this guide are restricted to those that, in the opinion of the author,
may be most suitably applied to larger diameter tunnel structures used in transport.
7.3
COMPARING THE CONSTRUCTION OF SEWERS AND
TUNNELS
Sewers are constructed in trench, batter, heading or tunnel, with the choice of
construction method dependent on the topography and flow requirement of the section
under construction. Those constructed in heading or tunnel may be divided as:
Category 1
Constructed in unsupported excavation, ie through rock, marl, stiff clay or
chalk.
Category 2
Constructed within supported excavation, ie supports in timber, cast iron
segments or concrete segments.
In both categories the excavation works, whether supported or unsupported, were cut to
the tightest profile commensurate with allowance for working space for the bricklayers
and masons constructing the finished permanent works. Whereas the structure below mid
axis was supported by the backfill, introduced formally by packing in layers as the works
310
proceeded – and informally by broken bricks, mortar droppings and any other material
debris placed by gravity, the external void above springer height was filled only by hand
packing and was not subjected to any compaction.
Our engineering forefathers worked to very high standards and we must assume that
leakage was not apparent immediately following construction, so its occurrence is
considered to be time-related.
The integrity of the permanent works of the underground structures built at the turn of
the 19 century suffered from two main limitations in terms of materials and materials
technology:
1
The use of lime mortars before the widespread use of Portland cement based mortar
(c1920), pure or weakly hydraulic lime mortars were not ideally suited for use in
habitually wet conditions, although more strongly hydraulic limes (including natural
cements and Roman cement) typically exhibited better performance and improved
durability.
2
Cement-based grouting of the annulus, to refusal, was not common until 1930 – in
effect the void above springer level could only be loosely packed.
In many cases inspection after a century of use shows decay of the original lime mortar to
be clearly evident. This can lead to movement in the structure and formation of leakage
paths, exacerbated in some instances by the inadequate side support of any unfilled
excavated void.
The methods that we use now to stop leakage and to increase the structural integrity of
the original permanent works may be considered as remedying the limitations imposed by
materials and materials application technology at time of construction. In particular we
are now able to use grouts (generally of cementitious base but with polymer-improved
properties) that can penetrate and fill all voids resulting from the temporary works
associated with the original construction.
7.4
TYPICAL CONSTRUCTION DETAIL AND BREAKDOWN
SEQUENCE
To achieve the most effective result from treatment of the existing (leaking) structure it is
important to determine the sequence and procedures adopted at the time of original
construction. This allows us to assess the simplest way to address any perceived
weaknesses and to produce the strongest composite structure.
Figure A1.45 shows the construction of a two-ring brick sewer in timber heading at a stage
where the brickwork is part laid and awaiting construction of the arch above springing.
The relatively well compacted state of the void between horizontal excavated surface and
permanent invert (base) of the brick sewer is clearly shown.
Note the fact that the support timbers will generally be left in place.
CIRIA C671 • Tunnels 2009
311
Figure A1.45
Partly-constructed two-ring oval profile sewer with timber
heading, Piccadilly Circus c1928
Note also that the brick arch required to complete the structure will be built on a former
ring and that for ease of construction the bricklayers’ building procedure will apply the
mortar in layers as near horizontal as dimensions and working space allow. As the
completed arch-work approaches the arc running from 10.30 to 1.30 on the clock face the
angle of mortar face approaches the vertical before changing towards the horizontal again
for the final closure of the arch.
Should the completed structure have been built in cohesionless soil with a fluctuating
water table, groundwater can create significant paths through the sewer surround and
(with mortar missing and/or decayed) the walls of the sewer itself. We have learnt that the
preferred path for water ingress is generally in the lower part of the quadrants above the
horizontal axis – at or just above the boundary between the naturally compacted material
and the arch construction. An example of deterioration coinciding with these flow planes
(and sources of structural weakness) is illustrated in Figure A1.46.
The grouting process accompanying tunnel construction was not common until 1920 and
grouting as an adjunct to construction in timber heading was not in general use before
1960. So the outline above, although referring to an egg shaped sewer constructed in
timber heading, is common to brick ring structures in headings and tunnels excavated by
hand before the advent of segmental tunnelling and grouting in the early 1900s.
312
Figure A1.46
Deterioration of lower part of arch corresponding with typical location of water inflow
7.5
INTERSTICES AND EXTRADOS GROUTING AS A REMEDIAL
WORKS TECHNIQUE
The technique that has been found to be consistently effective in restoring masonry sewer
structures to a state where leakages are stopped consists of:
1
Creating an impermeable skin to the inner surface of the structure.
2
Drilling through the lining to relieve water pressure.
3
Injecting grout to refusal to the extrados and interstices of the structure.
The strength increase resulting from the use of this system is produced directly by the
rebuilding of the compressive ring strength by using modern high strength grouts to fill
all interstitial and extrados voids, this void fill also stops leakage.
The detailed sequence of operations is defined in the publication Brick sewer renovation
(Underwood & Rees, 1985), which gives information on the development of the
technique. In particular reference can be made to Water Research Centre Report 107E
(Procter & Fillingham, 1983). This WRC test report recorded an increase in load bearing
capacity, resulting directly from interstitial grouting of the fabric of the oval-profile brick
sewer behind the impermeable inner surface skin from 170 kn/m² (control length) to 430
kn/m².
It is worth considering that these techniques, which have been successful in achieving
strengthening and leak-sealing in sewer tunnels, are not necessarily suitable for use in
larger diameter tunnels without careful consideration of the potential influence on their
structural behaviour and long-term performance. In particular, where linings rely on their
inherent ductility to accommodate small ground movements the use of inflexible high
strength cementitious grouts around the tunnel annulus might result in a fundamental
change to the lining’s structural action, and possibly concentration of stresses and
structural damage. This is particularly a concern with larger-diameter masonry lined
tunnels, where the structural action of the lining is potentially quite different. The use of
weaker grouting materials with greater flexibility might be more appropriate in such
circumstances. In any case, the successes achieved with leak-sealing small diameter tunnels
does provide a useful basis for further development of these grouting techniques and
materials for use in other situations.
CIRIA C671 • Tunnels 2009
313
7.5.1
Materials
The surface render coat consists of high strength waterproof styrene acrylate polymer
reinforced mortar giving high mechanical strengths to the in situ surface as follows:
compressive strength at 28 days
50 N/mm²
flexural strength at 28 days
11 N/mm²
adhesion strength (sand blasted concrete)
19.2 kg/cm²
The grout properties are determined by the particular requirement for each contract.
Where the objective is to eliminate infiltration and exfiltration, filling of all voids and
interstices with high strength water displacement grout is essential and typically the grout
would be a polymer-enhanced cementitious material. To produce initial set times of about
thirty minutes this grout will contain an accelerator. The final requirement is that the
grout should be non-shrink, a characteristic achieved by the inclusion of a low-expansion
agent in the factory mix. As discussed previously, weaker more flexible grouting materials
might be more appropriate for other types of tunnel and the engineering implications of
grouting on the structure and the most appropriate grout characteristics need to be
carefully considered.
In sections subject to tidal conditions the final formulation of grout mix will generally be
modified in the light of actual site results. In all cases the final decision on mix and drill
hole spacing is taken only after completion of sealing of a test section.
7.5.2
Practical considerations
This approach, first used in 1979, has been used in the structural refurbishment and leaksealing of numerous brick sewer structures built in locations affected by tidal or high water
table conditions. In all cases the effect of the refurbishment work has satisfactorily sealed
all leaks in the structure.
The earlier works were carried out from 1980 in oval-profile brick sewers in Weston super
Mare and Bristol where the typical structure was some 900 mm to 1800 mm in internal
height and of single or twin ring brick construction. In some cases the bricks to crown
were wedge shaped and in extreme case the crown remained in place only via brick-tobrick contact. Ground conditions varied from stiff clays and marls in the Bristol sewers to
wet sands in Weston super Mare.
As experience grew over the years the system was used to seal and strengthen culverts and
locks in the canal waterways. Perhaps the most extreme example was sealing of leaks to
Diglis Lock on the River Severn, carried out in 1986. The water head within the
submerged structure was about 8 m and diving inspection had determined that total
mortar loss, and significant brickwork loss, had occurred following its construction in 1810
(Figure A1.47).
At this stage of development the grout used for filling interstices was mixed on site. The
water ingress was substantial and the site approach was as follows:
314
apply polymer reinforced render (with high wet adhesion properties) to surface of
area to be treated
release water ingress by drilling at points determined by preferred flow paths already
existing in the structure
inject (to refusal) bentonite/cement/flyash grout at drilled points, starting at lowest
level and working in upward sequence.
The thixotropic effect of the bentonite is crucial to the success of the leak sealing and it is
important to use the correct grout mixing and pumping equipment to produce the water
displacement property critical to success. Bentonite requires a high shear mixer for
successful incorporation in a grout mix and it is also critical to effective thixotropy that the
bentonite is mixed first in the tank before introducing the cement and fly ash. The
proportions of these latter constituents determine the set time and strength of the mix.
In the author’s experience there is no substitute for mix and performance testing on site –
the materials are cheap relative to overall cost, and time spent on experimentation on a
closely monitored trial section is well rewarded.
Before starting any works of this nature it is recommended that investigative work
(perhaps limited to obtaining as built information) be carried out to determine:
ground conditions
form of temporary works – heading, trench, batter
outside dimension of permanent works – is structure single, twin ring or greater
From such information an assessment can be made of the likely volume of void existing in
the structure and between the lining extrados and the excavated ground surface.
If the length to be treated is significant it is recommended that closely monitored trials are
undertaken to verify the refurbishment assumptions and also to arrive at the most suitable
design mix for the grout.
CIRIA C671 • Tunnels 2009
315
Case study 8: Management of a disused
and deteriorated rail tunnel
Based on work carried out by Mott MacDonald consulting engineers for BRB (Residuary) Ltd
8.1
INTRODUCTION AND BACKGROUND TO THE STUDY
This case study outlines the assessment of long-term performance of a decommissioned
tunnel situated on a disused railway line in the north of England. The tunnel, opened to
rail traffic in the late 19th century, is just under 2 km in length with about an 8 m
diameter circular cross-section. Structurally it comprises a brick arch supported by natural
chalk sidewalls. It includes five open construction/ventilation shafts along its length, with
depths varying between 25 m and 60 m. These are brick-lined, with shaft tops covered by
substantial metal grillages.
The tunnel was closed to rail traffic in the 1950s and has been inspected on a regular basis
after this date. Based on the results of an inspection in 1991, the progressive deterioration
of the brick tunnel arch lining and exposed chalk sidewalls had reached the point where
the structure was declared unsafe for manned entry.
A study was required to determine the condition of the tunnel and allow its future longterm performance to be assessed. The assessment was based on a ground investigation,
remote internal structural inspection and land-use/hydrology survey. The results of these
three elements of investigation were considered in combination to provide an evaluation
of the potential rate of deterioration of the tunnel and its possible results in terms of
effects on adjacent land and reinstatement/compensation liabilities.
In assessing possible long-term deterioration scenarios, the following aspects were
considered and discussed in detail within the report:
8.2
effect of collapse on existing statutory undertakers
effect on public highways
effect on adjacent property
effect on surrounding land-use and hydrology
possible remediation measures
cost implications
SCOPE OF INVESTIGATIONS
The investigation work consisted of four main elements:
316
8.2.1
Desk study review of existing information
No comprehensive as-built records of the tunnel or other contemporaneous records were
available. A desk study and review of existing information was carried out, including:
8.2.2
dimensional and conditional information gathered from existing inspection reports
and assessment records
a comprehensive search of existing services
enquiries of the Land Registry to obtain details of land ownership next to and
surrounding the tunnel.
Ground investigation
A ground investigation was carried out to confirm the geology and groundwater regime
above the tunnel and to determine the depth of weathering zone and degree of fracturing
of the rock above the tunnel.
Three rotary cored holes of 30 m, 50 m and 60 m were sunk next to the line of the tunnel,
at a distance of about 25 m away from its centre line. These confirmed the detailed
geology of the area, in combination with published data from the British Geological
Society.
8.2.3
Remote internal tunnel inspection
Because the structural condition of the tunnel precluded manned entry, a remote internal
inspection was carried out to provide information on the conditions within the tunnel and
shafts to assist in evaluation and assessment of the likely nature, extent and times of future
deterioration. This made use of the specialist camera and lighting equipment that was
lowered down each of the five ventilation shafts.
All cameras provided real time CCTV via a monitor located at ground level to allow the
engineer to investigate features of structural significance. Three no. low light cameras
operating on wide-angle facility recorded a 360° field of vision of the ventilation shafts.
Also, a low light camera with a zoom facility was also used for the inspection. The results
were recorded in VHS video format.
The following condition information was acquired:
identification of any displaced, fallen or otherwise visibly unstable areas of masonry to
the arch soffit lining
identification of any displaced, slipped, or visibly unstable areas of brick lining to
sidewalls
where visible to estimate the thicknesses of the brick linings
identification and estimation of any significant moisture ingress
the size and nature of bedding within the chalk
condition of the former trackbed
extent of brick lining to arch soffit and sidewalls
identification of any significant roof falls.
These criteria enabled a condition assessment to be undertaken on the tunnel to establish
long-term maintenance requirements.
CIRIA C671 • Tunnels 2009
317
8.2.4
Hydrology and land-use survey
A hydrology and land-use survey was undertaken to quantify the effects of potential
tunnel collapse on the surrounding landscape. A detailed survey was undertaken to
record all important features at a tunnel collapse.
The survey took the form of a visual inspection of the area surrounding the tunnel up to
500 m either side of the tunnel. Notes were taken on the following:
location and type of vegetation and crops
location and size of drainage channels and ditches
slope
areas of existing lying water/springs
other features of relevance to the survey.
The survey included a full annotated photographic record of the land within the
boundary defined here.
The information from the three stages of the options study was collated to determine the
effects of future deterioration of the tunnel.
8.3
RESULTS AND ASSESSMENT OF TUNNEL CONDITION
In assessing the long-term performance of the tunnel, two main elements were considered
in detail:
8.3.1
1
Condition of the tunnel.
2
Nature of overlying strata.
Condition of the tunnel
Visual inspection was carried out on those parts of the tunnel that could be accessed safely,
ie the portal areas and the tops of shafts at ground level. The tunnel portals were in a
moderate condition with spalling of brickwork up to one brick depth and open joints. The
metal grill was missing from the top of one shaft.
Internally, the remote inspection confirmed the construction details of the tunnel and its
shafts, and allowed safe observation and recording of their condition with CCTV
equipment being controlled from ground level above the tunnel.
The ventilation shafts were brick-lined throughout and suffered from some spalling and
open joints, particularly at the shaft eyes where they met the tunnel lining and where
brick sidewalls continued for a distance of about 30 m either side of the ventilation shaft.
There was no evidence to suggest that the brick sidewalls support the brick arch, and
beyond these areas the brick arch was supported on natural chalk sidewalls. Throughout
the tunnel the arch soffit had suffered from extensive spalling and the chalk sidewalls,
which are naturally fissured, also exhibited considerable spalling. The face of the chalk
sidewall had crumbled resulting in chalk debris spalling onto the former trackbed, leaving
the arch unsupported over the area of severely spalled chalk sidewalls.
318
8.3.2
Nature of overlying strata
Data from three borehole locations was used to determine the nature of the geology and
groundwater regime above the tunnel and to determine the depth of weathering zone and
degree of fracturing of the rock above the tunnel.
The strata above the tunnel horizon were found to be a structureless, weathered chalk
(weak) up to 60 m depth below ground level. It was determined that this material would
not be capable of spanning over the width of the tunnel without the support of the brick
arch.
From the geotechnical information gathered, a factual and interpretative report was
produced. This information was used to assess the likely results of a do nothing scenario
where the tunnel was left to deteriorate further until eventual collapse. This included a
quantification of the likely effects of such a scenario on the surrounding landscape.
8.4
LONG-TERM TUNNEL PERFORMANCE IMPLICATIONS
Based on the information from the investigations, several potential collapse scenarios were
formulated. In arriving at each scenario, consideration has been given to the following:
condition of the tunnel
geological condition of the strata overlying the tunnel.
The propensity of a void created by failure of the masonry arch to propagate to the
ground surface depends on the nature of the overlying strata, the depth of the tunnel
below ground level, and the tunnel cross-section. Subsidence calculations were used to
determine the likely effect of tunnel collapse on the ground above it and, where possible,
to estimate the extent (zone of influence) and size of the resultant settlement. Several
different collapse scenarios were assessed based on the results of the condition survey and
variations in the apparent competence of the ground around the tunnel along its length.
The collapse scenarios considered were:
complete simultaneous tunnel collapse
partial tunnel collapse between shafts
partial tunnel collapse next to the portals.
Several subsidence calculations were undertaken for each of these scenarios. The effect
and cost implications of tunnel collapse were related to the land-use above the tunnel.
These ranged from significant local subsidence in areas of agricultural land-use, resulting
in damage to crops and requirements for reinstatement, to much smaller ground
movements that might have led to very minor distress to buildings in the affected areas.
The methods used were based on those included in the Subsidence engineers handbook
(NCB, 1975).
CIRIA C671 • Tunnels 2009
319
8.5
ASSESSMENT OF POTENTIAL REINSTATEMENT
REQUIREMENTS
An assessment was made of potential costs associated with each of the subsidence scenarios
assessed. These were based on visual inspection of the site, land-use, and on the level of
subsidence as determined by the method given in the Subsidence engineers handbook.
Reinstatement costs were determined assuming reinstatement of the ground to match that
of the existing landscape. They considered:
the different types of landscape at ground level
the quantity of all materials used in the reinstatement process
vegetation, such as the realignment of hedges, replanting of trees and seeding of
grassed areas
buildings, including minor repairs
carriageway repairs, such as the replacement of sections of cracked carriageway
services, running perpendicular to the tunnel below the public road
costs associated with all labour and materials involved in the reinstatement process.
Based on maximum predicted settlements, reinstatement requirements varied from the
placement of relatively low volumes of inert backfill material and topsoil with seeding of
grass and reinstatement of hedges, to the placement of much larger volumes of material
with the potential associated costs of repairs to affected highways, buildings and services.
The long-term compensation expenditure was determined, along with a likely profile of
compensation expenditure based on an assessment of the predicted rate of structural
deterioration of the tunnel. This information was useful to the tunnel owner because it
provided a baseline for considering alternative remedial options, including continuing
structural maintenance and filling in the tunnel, as well as allowing budgetary provisions
to be made against the potential future liabilities associated with owning the tunnel.
320
Case study 9: Reconstruction of an
underground line tunnel at Old Street
Adapted from Northern Line tunnel reconstruction at Old Street, London (Burgess et al, 2002)
SUMMARY
For many years, London Underground’s Northern Line tunnels just south of Old Street
station had suffered from attack by sulphuric acid. These tunnels were constructed
between 1899 and 1901, and were enlarged between 1922 and 1924. Acid began to seep
into the tunnels in 1945 and cracks appeared in the tunnel linings in 1960. Over the
years, to ensure safety, London Underground Limited (LUL) monitored the increasing
levels of distortion and cracking of the tunnel linings and installed temporary
strengthening until the need for innovative long-term solutions became evident. This
paper describes various investigations into the problem, the formation of a solution and
the following construction works, which had to be carried out with minimal disruption to
services within the complex operating environment of the Northern Line. The solution
was to replace the grey cast iron linings with larger diameter acid resistant linings made
from cast duplex stainless steel. The works were carried out at night using a special shield,
through which the trains passed during the day. The project took six years from initial
concept to completion, including more than four years of research and design, a precasting contract for the linings and finally nine months of installation works. In total, the
project cost about £15m.
9.1
INTRODUCTION
London Underground’s Northern Line between Moorgate and Angel was constructed as
an extension to the existing City and South London Railway between 1899 and 1901. Two
years later, tunnels were built just above the Northern Line, now the Great Northern
suburban line. The Northern Line tunnels were originally 10 ft 6 in (3.20 m) internal
diameter (six segments and a key of grey cast iron), but were enlarged between 1922 and
1924 to nominal 11 ft 8¼ in (3.56 m) internal diameter, reusing the original segments and
inserting five new key-sized castings.
In 1945, 80 m of the Northern Line twin tunnel just south of Old Street station were
found to be suffering from attack by sulphuric acid, which had begun to seep into the
tunnel. Also, cracks appeared in the tunnel linings in 1960. London Underground
undertook several site investigations from the 1960s onwards. These showed that the
geology of the area was unusual and particularly unfavourable. Below the London Clay,
but above the Woolwich and Reading Beds (clays and sands) was a lens of sand, which on
chemical analysis revealed the presence of iron pyrites. Water, seeping down through the
clay, perched in this sand. The tunnels pass through this lens, and the passage of trains
pumped small quantities of air through the tunnel linings and into the wet pyrites. These
three ingredients react to form, among other products, sulphuric acid. Also, the dished
shape of the sand lens, with the tunnel passing through the lowest part, prevented the
acid from flowing away and was effectively a sump (Figure A1.47). The acid primarily
CIRIA C671 • Tunnels 2009
321
attacked the cementitious grout surrounding the tunnel rings, causing it to expand and
create high pressure on the tunnel axis. In many locations, the acid had reached the
linings themselves and had corroded the cast iron. The British Rail (BR) tunnels located
in London Clay above the Northern Line tunnels were not affected by the problem.
a
b
Figure A1.47
Cross-section showing the relative location of the two tunnels (a) and a 3D representation of
sand lens (b)
In 1963, an initial attempt was made to deal with the problem by pumping more than
2500 gallons of sodium hydroxide solution into the ground around the affected tunnels to
neutralise the acid. However, there was far more acid than had been envisaged, and it was
estimated that only 10 per cent would have been neutralised by this exercise, even if the
fineness of the sand had not prevented the acid and alkali from mixing.
Also in 1963, a length of the southbound tunnel was strengthened from the inside using
steel beams strapped vertically to the linings at axis level, but at the time no strengthening
works were considered necessary to the northbound tunnel. London Underground
Limited (LUL) continued to inspect and closely monitor the tunnel. Boreholes were
drilled in 1989 to provide more information on the ground surrounding the tunnels and
a theoretical assessment of the integrity of the lining was undertaken in 1990. The
increase in the rate of crack formation by this time was significant and there was
considerable ingress of acid.
In late 1990, in the course of installing reference studs in the flanges of rings to be
included in an expanded monitoring programme, cracking occurred in the
circumferential flange of a lining segment being drilled and at similar locations in flanges
of adjacent segments. LUL decided, as a temporary measure, to extend the reinforcement
of the linings at axis with steel strapping throughout the affected areas in both tunnels.
322
This work was undertaken early in 1991 in non-traffic hours. The contract included the
removal of several segments, which had been badly damaged by acid attack, for
investigation and their replacement by fabricated steel segments.
While the exigent strapping work was in progress, a study was undertaken to examine the
feasibility of replacing the existing lining in the affected lengths of tunnel with a suitable
alternative lining. While it was apparent that access to encircle the affected rings could be
achieved from a conveniently located site at ground level, it was equally clear that
considerable research of possible materials was required, both for an appropriate lining
and for a suitable contact grout.
A team was set up to manage the project, incorporating LUL, Charles Haswell and
Partners (lead consultant) and the Geotechnical Consulting Group. Several other
subconsultants were employed, including testing houses and materials experts.
Preliminary design of the works required to replace affected lengths of linings was
undertaken in conjunction with a programme of research and development to establish
the most suitable materials. A finite-element analysis, modelling the lining replacement
was undertaken using the ICFEP (Imperial College finite element package) programme to
establish likely ground loading on the lining and to predict ground movements that would
affect the twin BR tunnels on the same alignment and immediately above the lining to be
replaced. The analysis modelled one of the lining options, as well as the sequence and
timing of construction of all four tunnels. A site investigation contract was also
commissioned to confirm the extent of the lengths of tunnel that would need to be relined
and to supplement existing site investigation data required for design of the relining and
planning of the construction work. Following a condition survey of the existing BR
tunnels, reference points for monitoring movements and distortions were established in
the BR tunnels and base readings were taken.
9.3
LINING REPLACEMENT METHODOLOGY
In view of the nature of the ground around the lining to be replaced, the location of the
BR tunnels above and the need for the tunnels to remain operational, a unique shield was
designed to:
provide support to the track
allow free passage of trains through it
support the excavated face
allow for continuous grouting during shoving.
Room to construct shield chambers was an important consideration. Access to the working
site was planned as a temporary shaft from the surface to a level suitable for an adit tunnel
to be driven beneath the invert of the northbound LUL station tunnel, turning through
90° onto the alignment of, and directly beneath, the disused tunnel. Twin shafts were
intended to provide access for workers spoil, plant and materials between this blind adit
and the disused passage above. Traditional steel picture frames set in mass concrete were
detailed to form the openings between the disused passage and the shield chambers. The
replacement lining internal diameter was chosen to allow sufficient room for hand
excavation of ground around the existing lining while keeping the diameter of the shield
chamber lining to a practical limit with a view to minimising settlement effects on the
tunnels above.
CIRIA C671 • Tunnels 2009
323
In deciding the best use of working and possession hours, and how they might affect the
operation of the railway, several working method options for the construction were
planned, costs and programmes defined ranging from engineering hours only (ie no
appreciable affect on operations) to full closure of both tunnels. Extended engineering
hours, weekend closures and single tunnel closures were also considered. The major
factors proved to be the length of the construction programme with its associated
supervision costs, the use of a tunnelling shield, costs of cable diversions, replacement bus
services and loss of revenue due to the disruption of service. After consideration, the no
closure option was adopted.
Achieving the replacement of lining to an operational rail tunnel is clearly a complicated
matter. The time that work may proceed is strictly controlled by the need to maintain safe
unimpaired operation of the railway in normal traffic hours, albeit with a speed
restriction. This implies devising a method of working that allows quick, safe cessation at
any time of all tasks in the work cycle for lining replacement that could affect the
operation of the railway. It also demands sure and timely provision of all temporary
support and the removal to a safe location of all plant, materials and equipment that
would otherwise compromise normal railway working. Essential tunnel services for
signalling, power provision, lighting and communications needed to be relocated
temporarily, while the tunnel lining supporting them was renewed. Practically, this meant
their diversion from the length for renewal of the first tunnel to the adjacent tunnel,
followed by diversion of all services back from the adjacent tunnel to the renewed tunnel
length until relining of the second tunnel was complete. Diverted services were
accommodated between the rails in cable troughs laid in sections across the sleepers.
9.4
MATERIALS SELECTION
Given the nature of the damage to the existing lining, it was important that the
replacement lining should be resistant to corrosive damage and also have high flexural
resistance in view of possible future high distortions. Also, one of the design criteria for
the lining was a 400 year design life. Various materials were considered as possibilities for
the replacement lining including:
silica fume concrete with a protective coating
spheroidal graphite cast iron with a protective coating
glass reinforced plastic
cast stainless steel.
After consideration of all the materials the stainless steel was the only appropriate
alternative for use in the particular circumstances known to exist at Old Street. The high
chromium duplex stainless steel selected has excellent strength and corrosion
characteristics.
Finding a suitable inert grout proved problematical. An organic grout, although inert to
acid attack, would require the use of light organic solvent to transport it, possibly
encouraging the growth of anaerobic bacteria in the ground. Instead, a high bulk sand
grout, with bentonite and low cement, was chosen as the safest option, the properties of
the new linings being such that they could withstand distortion due to any acid attack on
the grout.
324
9.5
DESIGN
The replacement lining adopted comprised 4.68 m internal diameter stainless steel rings
of 12 segments and a key. The number of segments was chosen to provide a flexible lining
of segment length and weight suitable for easy handling and erection. The opening
supports to the disused passage and to the shield chambers were designed as traditional
picture frame steelwork surrounded by concrete. Particular design features of the project
were the track bridge incorporated in the shield and the openings from the shield
chamber to the connecting passages with the disused tunnel. These were considered
essential to the safe operation of the railway and were included in the original design
commission. The track bridge essentially supported the track through the length of the
shield. The system had to provide stable support to the track during traffic hours yet
during working hours allow free passage of the track along the bridge when the shield was
shoved in the ring construction cycle. The required adjustments to the support system at
the beginning and end of each work shift had to be achieved quickly and be effective.
9.6
LINING PROCUREMENT
The working shaft was a 32 m deep, 4-5 m internal diameter shaft, formed from standard
pre-cast concrete bolted linings. The initial rings were jacked down through made ground
and gravel into the London Clay, the remainder being constructed using the
underpinning method. On completion of this shaft, a 3.05 m internal diameter access adit
was driven beneath the southbound running tunnel to align with the centre of the disused
tunnel. The adit was hand dug initially on a 13 m radius and was formed from standard
pre-cast concrete bolted linings.
Simultaneously, cables from the northbound tunnel were being diverted to the
southbound tunnel and dewatering of the length to be relined began. For access and
egress the construction of one larger square shaft formed from steel sections lined with
steel plate with a traffic light system was proposed. This proposal was accepted and a pilot
chimney formed from the access adit up into the disused tunnel. From the disused tunnel
a 3 m square shaft was then constructed top-down into the access adit.
The next task was to form the shield chambers and tunnel openings. The construction of
the shield chambers was a complex operation involving working around live tunnels while
trains were running. The internal diameter of the existing running tunnels is about 3.6 m
and that of the shield chambers is 5.75 m, so it is not difficult to imagine the relatively
confined working space available to install the necessary and substantial temporary
ground support works before installing the new SGI shield chamber linings. The
contractor suggested efficient alternative openings in the shield chambers for access to the
adjacent disused tunnel. Instead of having to install very heavy lintels and sills, the
contractor proposed the installation of special steel segments at lintel and sill positions
bolted together with high strength friction grip bolts. In this way, the SGI segments
forming the lining between these fabricated segmental beams could simply be removed on
completion of the chambers. This proved very successful with the existing northbound
running tunnel, now safely cradled in the new shield chamber.
Actual relining works started in December 1995. Six weekend possessions and five Sunday
possessions had been booked from January through to March along with night
(engineering hours) working where the track had to be reinstated in time for the first
train. This is where the temporary track bridge came into play. The concrete supporting
the track had to be formed in two stages and attain a minimum strength of 25 N/mm²
before taking the weight of trains. The track bridge was designed so that it spanned the
fresh concrete as relining progressed and, by the time the end of the bridge had passed,
CIRIA C671 • Tunnels 2009
325
the concrete was capable of taking the full track loading. An early high strength concrete
was designed for this purpose.
Unfortunately, the access works took longer than expected and it became apparent that
they would not be ready in time. After careful consideration of available alternatives it was
decided to bring all the shield components in on a works train during the second booked
possession. This saved a considerable amount of time when compared to the original plan.
The first weekend possession would be used solely for track works (breaking out existing
concrete and installing temporary supports), with the break into the shield chamber
combined with the installation of the shield.
Every night, break-out of the track bed continued to a maximum of 15 m in front of the
shield and over this length the track was temporarily supported from the existing lining.
The track would be freed up from the temporary track bridge, linings would be installed
and the track behind the shield was carried on temporary timbers supported off the new
stage track bed concrete. On completion of the relining works, the temporary supports
were removed and second stage track bed concrete was placed.
In all, 135 new rings were installed in the northbound tunnel, the last ring being installed
at the beginning of May 1996. It is worth noting that during these works abandoned skins
from two old shields used for previous tunnelling works were encountered and had to be
cut up and removed as lining progressed. Also, some old timbering was found in places at
the crown of the old tunnel and extra back grouting was required once the timbers had
been removed. On completion of the last new ring, the hydraulic and electrical equipment
was stripped from the shield, the track bridge was removed and old tunnel segments were
rebuilt through the shield. The headwall was then constructed and the annulus between
the shield and old tunnel segments was grouted up. At this stage, new cable brackets were
installed in the tunnel using specially designed insulated connectors to avoid electrolytic
action between the stainless steel linings and steel strapping supporting the cast iron cable
brackets. The cables in the southbound tunnel were then all diverted into the completed
northbound tunnel.
In view of the success in installing the tunnelling shield in the northbound running tunnel
using a works train, in June 1996 the new tunnelling shield was installed in the
southbound shield chamber during one weekend possession adopting the same method.
9.7
SETTLEMENT CONSIDERATIONS
It was important that the effects of settlement because of the relining would not cause
operational problems for the BR tunnels above. For this reason, before work started on
the connecting passages from the disused passage to the shield-chambers, careful and
substantial arrangements of propping to the disused tunnel were provided as was full
timbering to the excavation faces. Movements of the BR tunnels were monitored very
carefully throughout construction and were well within acceptable limits (longitudinal
settlement profiles are shown in Figure A1.48). The second encirclement (of the
southbound tunnel) resulted in much greater settlement than the first (of the
northbound). This is a result of the close location of the two tunnels and the second
encirclement resulting in settlement of ground already disturbed.
326
Figure A1.48
Settlement profiles
9.8
CONCLUSIONS
The complete replacement of 160 m of tunnel lining was carried out successfully with
minimal disruption to trains. The total cost, including all research, design, procurement,
fabrication, installation and operating costs, was £15.3m. Even with the progressive
distortion and cracking of the linings and major ingress of acid, the tunnel had survived
50 years after the first discovery of seepage, before replacement became necessary. To
date, no other running tunnel in the London Underground has shown similar symptoms.
CIRIA C671 • Tunnels 2009
327
Case study 10: Inspection and
maintenance of a raw water tunnel
Adapted from Hieatt, Ellis & Locke (2005). By M Hieatt (Black & Veatch
Ltd), J Ellis (Veolia Water Partnership), K Locke (Three Valleys Water plc)
SUMMARY
A 7 km wedge block tunnel feeds river water to Iver Water Treatment Works, Slough. The
tunnel was completed in 1973 and has been in almost continuous use, permanently
flooded. There have been five maintenance inspections, a major tunnel cleaning exercise
and the construction of a second tunnel to provide a standby supply. Operational
demands and increasingly rigorous health and safety requirements have made the
planning and procedures for gaining access for inspection of the tunnel a major exercise.
This case study briefly describes the tunnel system, the condition of the tunnel as found
during the inspections, the planning and procedures for the 2005 inspection and provides
some practical suggestions for those planning similar exercises. Risk assessments enabled
the planning team to identify practical solutions acceptable to all parties for gaining safe
access within the constraints of the system and meeting health and safety demands.
10.1
THE IVER WATER TREATMENT WORKS RIVER WATER
TUNNEL SYSTEM
The Iver Water Treatment Works is fed with raw water from the River Thames via a 7
km long 100 inch (2.54 m) diameter wedge block tunnel, which was completed in 1973. It
lies below river level and is permanently flooded. River water flows into the tunnel by
gravity through intake trash racks, band screens and a 5 m diameter 17 m deep intake
shaft (shaft 1), passes through a dewatering wet shaft at about midway (shaft 3) and is
pumped out from a 8.6 m diameter wet shaft by suspended pumps. There is an initial
receiving shaft at the treatment works (shaft 5) where the tunnel diameter increases from
2.54 m to 3.66 m to reduce the velocity of the water before entering the pumping shaft.
The receiving shaft also accepts various return flows and overflows from the treatment
works. Typical present flows through the tunnel are 180 m litres per day, which is
equivalent to a water velocity in the tunnel of about 0.4 m/s.
The wedge block tunnel was constructed by open faced mechanised shield of Priestley
design, excavating through the over-consolidated London Clay without any major
incident. The design follows those established by Tattersall et al (1955) and used
extensively for tunnels in and around London. It runs at between 16 and 25 m below
ground. The tunnel lining is built up of rings each 27 inches (685 mm) long. Each ring
comprises 11 segments with radial joints and a key wedge at the crown. The segments are
pre-cast unreinforced concrete 140 mm thick, internal diameter 2.54 m with a specified
concrete strength of 50 N/mm².
328
10.2
MAINTENANCE INSPECTIONS
After the tunnel was completed in 1973 there have been five maintenance inspections, one
major tunnel cleaning exercise and, as mentioned, the construction of a second tunnel to
provide an alternative supply for emergency use. Increased operational demands on the
water treatment works coupled with the development of health and safety requirements
over the years have made the planning and procedures for gaining access for inspection of
the tunnel a major exercise. The tunnel was fully drained for the inspections in 1979,
1986, 1994 and most recently in January 2005. In 1982 the tunnel was partially drained to
check the extent of silt build-up in the intake shaft and first 100 m or so of tunnel.
Figure A1.49
General condition of lining and silt deposits in first leg of tunnel
1986 inspection. Note the absence of any significant biological growth
10.3
JANUARY 2005 INSPECTION
The following gives an overview of the preparation and the lead-time that was required
for the inspection carried out in January 2005.
In June 2003, a statement of need was raised for inspection of the River Water Tunnel
including: an assessment of the inspection costs including all before safety considerations,
draining down requirements, and the determination of a recommended frequency of
inspection of the river tunnel.
The eventual budget for the overall exercise including planning, capital costs and
operational costs was over £400 000, giving an indication of the scale of the exercise.
In 2004, the River Tunnel Inspection 2005 project objective was established as “to
successfully complete the inspection of the River Water Tunnel at Iver Water Treatment Works”.
CIRIA C671 • Tunnels 2009
329
The success criteria with highest weightings were established as:
sufficient information provided to assess future maintenance requirements
no accidents, injuries or near misses
no environmental issues regarding discharges, pollution, noise or flooding
condition of the tunnel immediately below gravel extraction location established
risks to the production results at Iver managed and reduced to a minimum
project completed to capital budget and time.
The dewatering contractors and the tunnel inspecting engineer were contacted and
advised of the impending project. In October, dates for dewatering of the River Tunnel
were agreed to start the week of 17 January 2005. Confined space consultants, the
Environment Agency and land agents were contacted and negotiations began for land
access and consents.
In November and December 2004 Three Valleys Water undertook a series of trial runs to
transfer the source of raw water entering the Iver Works from the River Water Tunnel to
the Reservoir Tunnel. Detailed planning meetings were held with key parties involved
with the inspection during November, December and early January 2005. Risk
assessments were carried out and detailed method statements and emergency plans
prepared and agreed between the various parties involved covering both the operational
aspects and health and safety.
10.3.1
330
Some practical points for consideration
confined space entry: a specialist company was employed to provide confined space
access equipment, emergency services co-ordination, attendance on the tunnel
inspection team and to draw up the method statement for the confined space access
and traverse of the tunnel system. A member of the specialist company accompanied
the inspection party during the traverses of the two main legs of the tunnel. Basic
procedures for safe entry into confined spaces are well established in the industry. In
the case of entry into relatively long tunnels some of the following may be of assistance
in planning such exercises. For the main lengths of the tunnel, the inspection party
comprised four people, with an extra pair accompanying the party to provide an
intermediate communication relay post. There was a comprehensive permit to enter
regime including interfaces with the treatment works operation and confirmation of
valve lock-offs and similar aspects
gas monitoring: low level explosive gas alarms from portable monitors had occurred
on at least two previous inspections and caused those particular entries to be
abandoned. Gas detectors were lowered into the shafts before any man entry and
were then carried by the inspection parties at all times
ventilation: with shaft hatches removed and the tunnel drained there is normally a
steady draught through the tunnels induced by the atmospheric pressure gradient at
the surface. The direction of the draught will change depending on the wind
direction and it is appropriate to change the traverse direction to suit the draft. It is
preferable to traverse the tunnel in the same direction as the draught
intrinsically safe equipment: for the purposes of the inspection, where equipment was
essential for use during ingress and egress and for communication it was required to
be intrinsically safe (explosion proof), for example, cap lamps, torches, gas monitors,
radios. However for equipment that was not essential for access and for which use
could be controlled, it was acceptable to use standard equipment, eg ordinary digital
cameras with flash. Their use was subject to demonstration that the atmosphere at any
particular time was below the acceptable lower limit for explosive gases. A similar
approach was taken for use of sampling tools
access facilities: the tunnel shafts are fitted with aluminium ladders, pre-cast concrete
intermediate platforms and aluminium handrail. The accesses had been found to be
serviceable on previous occasions but it was known that there was some pitting of the
aluminium and no assurance that the access facilities were still safe (Figure A1.50a). So
once the shaft had been tested for gas, the first person (or the bottom man) was
lowered in a bosun’s chair from the open surface at shaft top to shaft bottom on a
man-riding safety winch from a tripod (Figure A1.50b). Once it was established that
access platform and ladders appeared usable the second entry team member was
secured to a safety line and went down the ladders leaving the safety line running
through the ladder openings. Other members used the ladders but were safeguarded
by a fall arrestor attached to the safety line while on them. Safety harnesses were worn
permanently by all staff entering the shafts and tunnels for the inspection to aid
retrieval in case of an emergency
a
Figure A1.50
b
View of intake shaft access with ladders that were, in the absence of contrary information,
assumed to be unsafe so that alternative safe access methods were required (a) and entry to
dewatering shaft using a safety winch and tripod (b)
communications: both main lengths of the tunnel are each about 3.5 km long.
Experience from previous inspections found that radios were effective from the
surface to shaft bottom but were not effective in the tunnels. Air horns carry a
considerable way along the tunnel and a signal system was established to maintain
positive contact throughout. However previous experience had found that
communication with air horns started to become uncertain at about 1.5 km. To ensure
reliable communication two extra staff accompanied the inspection party until the
routine air horn signals became faint, which then became a communication relay until
the inspection party establish communication with the shaft to which they were
travelling
emergency breathing equipment: the traverse time along each main leg of the tunnel
is several hours and subject to the conditions found. Compressed air cylinders to give
each member sufficient air time to escape to the starting point in the event of a gas
alert would be impracticable. The solution was for each member to carry two escape
re-breather sets that would give reasonable time to return to the entry point and were
practicable to carry
CIRIA C671 • Tunnels 2009
331
lighting: for maintenance inspections to date, lighting in the tunnel has been by
standard mining cap lamps supplemented by intrinsically safe torches. If tunnel walls
are clean and there has been the opportunity for the concrete segments to dry out in
the natural draught there is a reasonable amount of diffused reflected light However
once a layer of biological growth has established and if the walls remain wet then the
field of view is restricted more or less to the beam of the cap lamp and inspection is
more difficult. During the construction of the sister tunnel, final quality inspections
were made using two car spot-lamps mounted on a sack trolley with 12v car batteries
carried on the trolley pushed along by the inspecting engineer. The arrangement lit
up the whole circumference of the tunnel for some distance ahead and proved very
effective. A modern intrinsically safe equivalent would be worth considering for future
maintenance inspections
photography: as with any inspection, photographs provide an invaluable record
(Figure A1.51). For the last inspection, a relatively modest specification digital camera
with a good low light capability gave good results but it is worth having a separate
powerful flash gun available rather than relying on the typical built-in flash
provisions. On the last inspection, digital photo improvements were effective in
revealing detail not immediately apparent. Consider:
for security, have two cameras each carried by a different member of the party
and take parallel sets of record photographs: the extra cost is minimal
compared to the value of the record if one camera fails or is damaged or lost
avoid taking flash photographs of people wearing reflective safety clothing: the
reflective materials are very effective and the resultant glare masks everything
else
if there is a draught, either take photographs facing into the draught or hold
your breath when taking flash photographs facing downwind: the
condensation cloud from breathing can be highlighted by the flash and form a
very effective smokescreen that might be visible only after the inspection is
over.
Figure A1.51
Image from the 2005 inspection
showing 3.66 m dia. tunnel with
persistent old longitudinal cracks
in cast in situ concrete lining at
crown and shoulder positions
made visible by the use of lowangle lighting
332
10.3.2
measurement: chains along the tunnel were measured with a measuring wheel (and
confirmed on the way out), diameters were measured using an infrared one-ended
measuring device. Measurement locations were marked with cable ties on the disused
chlorine dosing pipes. A few cable ties left on the pipes from work in the tunnel in
1985 were still in good condition and showed that this system of marking would stand
the test of time
PPE: normal Personal Protective Equipment such as overalls, hard hats, harness,
gloves, goggles etc and also waders rather than boots as mud can spread. Some
members also wore inflatable flotation aids (water sports type life jackets with a
manually activated gas cylinder)
compliance with before safety obligations: when an inspection involves staff from
several different companies it is important to ensure early liaison on health and safety
issues to ensure that each company is comfortable they are meeting health and safety
obligations in relation to their own employees. This will entail the circulation of
method statements for review, comment and agreement and can take some time to
complete. Risk assessments proved a useful tool in this process and in reaching a
consensus. Staff carrying out the inspection should undergo appropriate confined
space training including emergency procedures and to be physically fit
carrying equipment: have a bag or other means of carrying equipment so that both
hands are kept free
recording information: plan ahead. Whatever you use needs to be resistant to water
and mud and easily managed in the dark with gloves
taking and labelling samples: using pre-numbered sample bags and containers makes
the recording in the wet and mud easier and more reliable.
Risk issues
Formal risk workshops were held and a project risk register was developed as part of
planning meetings in the run up to the inspection. Risk assessments on health and safety
issues were carried out as part of the development of method statements. Where possible,
risk mitigation and reduction measures were incorporated into the budget for the
inspection.
10.3.3
Inspection programme results
The detailed planning allowed the inspection to be completed within programme and
with significant budget savings. The tunnel was confirmed as being in sound condition.
No accidents, injuries or near misses occurred during the project and there were no
environmental issues caused relating to discharges, pollution, noise or flooding.
10.4
CONCLUSIONS
the structures of the wedge block lined tunnel and associated shafts are in sound
condition after thirty years and have required minimal maintenance
use solid section fittings, such as ladders and hand railing, so that the extent of pitting
and corrosion are evident. Using stainless steel fixtures and fittings pays dividends
over time
start planning early and consider both capital and operational costs. Over time
savings may be significant in extra operational costs
circumstances change with time: check the condition and serviceability of systems and
equipment before the inspection
CIRIA C671 • Tunnels 2009
333
334
allow plenty of time for the actual inspections
risk assessments provide useful means of identifying mitigating risks and are valuable
tools in reaching a consensus between the different parties involved
detailed consideration of simple practical measures will help make life easy for the
inspection team and improve the quality of the inspection
when carrying out the inspection and analysing the results be aware that dewatering
changes the conditions within the tunnel, eg it may reverse the normal operating
head differential
testing standby equipment and the standby pumping station before helped establish
operational method statements and identification of any plant maintenance required
to enable the successful completion of the inspection
resources for the project were made available early and this helped the success of the
inspection. A considerable resource was required from Three Valleys Water at preplanning stages and in operation of the plant and liaison with contractors to help with
the overall inspection.
Case study 11: Investigation and
construction joint mapping of Haymarket
tunnels
By Simon Brightwell of Aperio Ltd and Jack Knight (formerly
of Scott Wilson and Charles Haswell and Partners)
As part of the site investigation for the proposed commercial development of the former
Morrison Street Goods Yard, Edinburgh, information about the condition and
construction arrangement of the two existing railway tunnels located beneath one corner
of the site was required. The tunnels, known as Haymarket north and south tunnels, were
opened in 1846 and 1896 respectively and are both about 920 m (46 chains) long. The
areas of interest were limited to the first 200 m length of tunnel from the Haymarket
Station end, eastwards towards Waverley Station. These tunnels carry the main Edinburgh
to Glasgow lines and the main Edinburgh to North of Scotland lines via the Forth Rail
Bridge.
The survey methods selected by the client included detailed visual inspections and ground
penetrating radar (GPR) from within the tunnels.
The visual inspection involved a walk-through survey noting the condition of the present
lining, including areas of spalling and water ingress and also, where possible, the location
and spacing of any construction joints (joint mapping). While the present surface
condition of the tunnel lining was important in deciding the possible need for a
strengthening/repair plan, interpretation of construction joint patterns in both length and
distribution, could be a helpful indicator of the method of construction and the location of
hidden shafts. It was thought that the direct overlaying of the joint locations onto the GPR
surveys could reveal much more about the inner lining than the use of isolated,
destructive coring.
GPR surveys were conducted over 200 m in both tunnels using road/rail access vehicles.
Although this was originally programmed for completion within a single night possession
by two survey teams, the work required three possessions to complete because of
scheduling problems.
11.1
NORTH TUNNEL
The north tunnel was opened in 1846 and has two tracks, with no overhead electrification.
Records indicate that it was built by a combination of cut-and-cover and tunnelling
methods. The cut-and-cover section was the length under investigation.
After opening, extensive and severe spalling has occurred in the tunnel lining and
substantial areas of the brickwork have been re-cased with two rings of engineering brick.
The suspected cause of spalling was identified as being sulfate attack because of the
sulphate-rich filling used to create the goods yard above the tunnel.
While a survey of the original construction joints in this tunnel was prevented by the recasing works, the visual survey identified areas where the original brick work had not
CIRIA C671 • Tunnels 2009
335
been re-cased and required repair, and also areas of the re-casing that had suffered
further degradation from sulfates. The main concern in this tunnel was delamination or
gaps between the old outer brickwork rings and the newer re-casing rings of brick. The
GPR survey objective was to locate and map any de-lamination or voiding between these
brick rings.
As it was considered that the difficulty in effecting a bond between the re-cased brickwork
and the original outer rings of brick would be greatest in the arch, the GPR survey was
confined to the crown and haunches. No GPR survey work was undertaken on the
sidewalls. Thirteen, 200 m long survey profiles collected from the arch between the crown
and the haunches. Most data were collected using a medium frequency 900 MHz antenna
with a range of up to 1.5 m, extra data were collected using a higher frequency 1500 MHz
antenna to help calibrate the velocity of radio signals through the structure.
Despite the presence of highly conductive engineering brick, the GPR data quality was
adequate to plot delamination and voiding at three depths within the lining. One plot
mapped shallow defects (within 300 mm of the surface), another mapped defects in the
rear part of the lining (from 340 mm to 700 mm), and a third mapped voids and other
variations in the materials behind the lining (from 800 mm to 1500 mm). Three main
areas of delamination were also found at shallow depth to some smaller and less significant
areas of delamination deeper in the lining.
Figure A1.52
Severe spalling to original brick lining
11.2
SOUTH TUNNEL
The south tunnel opened in 1896, has two tracks and overhead electrification. Records
indicate that it was built by hand tunnelling methods using access shafts and headings. It
is lined throughout in red brick.
Although the south tunnel was visibly in much better condition than the north tunnel,
concerns had recently been raised regarding possible variations in lining thickness, and
also the ability of this tunnel to absorb changes in loading conditions because of the
proposed development over the tunnel. Some variations in lining thickness had been
uncovered in some adjacent railway tunnels, which were identified as because of poor
workmanship during construction. The GPR survey objectives in this tunnel were to
determine the lining thickness and to map any construction variations.
336
Twelve, 200 m long profiles were collected from the arch between the crown and
haunches using a lower frequency 400 MHz antenna to achieve depth penetration to 2 m.
Again no profiles were taken along the sidewalls as the variations in the construction
depth were thought most likely to occur in the arch.
The collection of the GPR profiles from the south tunnel was slightly hindered by the
presence of overhead electrification gantries but it was still possible to determine that the
lining thickness was typically seven or eight brick rings throughout this section of tunnel
with only minor variations in the depth of brickwork. Differences in depth were identified
by joint mapping, as being within discrete construction lengths and were possibly formed
as break-ups where individual lengths of tunnel had been constructed from the temporary
headings. These break-ups had to be particularly strong and well built, as they had to
stand and take the full weight of the full tunnel construction either side. These lengths
were seen particularly within the GPR survey results as being clear sections of tunnel
without traces of water ingress or voiding.
The joint mapping survey when viewed in conjunction with the GPR survey also strongly
indicated the position of a hidden shaft within the goods yard area where cover to the
tunnel was particularly low and convenient for a shallow depth shaft to expedite
construction. The pattern of break-ups at this suspected shaft location is consistent with
the presence of a hidden shaft.
Figure A1.53
Haymarket south tunnel – GPR with joint mapping
11.3
CONCLUSIONS
GPR surveys combined with tunnel joint mapping can be mutually beneficial in
establishing the construction pattern of a tunnel, including the positioning of hidden
shafts, break-up, junction and shaft lengths and the location of bad ground or faulting
encountered during construction. The combined use of GPR and joint mapping can also
lead to more efficient repair: GPR surveys can identify defects and repair locations can be
specified to target the lengths between joints rather than arbitrary areas. This can lead to
more effective repair and less waste.
CIRIA C671 • Tunnels 2009
337
Case study 12: Relining of Sugar Loaf
tunnel
By Jack Knight, formerly of Scott Wilson and Charles
Haswell and Partners, courtesy Network Rail
Sugar Loaf Tunnel lies on the Central Wales Line between Craven Arms and Llanelli and
carries a single railway track. The tunnel was opened to passenger traffic in 1886 and is
915 m (1000 yards) long with a central ventilation shaft. There are a further three
reported backfilled shafts in the northern section of the tunnel.
The tunnel is curved in plan throughout its length and is 5 m wide by about 5 m high with
vertical sidewalls and a semicircular arched roof. Originally the tunnel was partially
unlined but is now fully brick lined from the southern portal for 300 m and for the
remainder of the tunnel to the north portal has vertical stone clad sidewalls and brick arch
roof.
The tunnel has a history of severe water ingress and mortar deterioration particularly in
the central and northern sections where the tunnel has low ground cover and lies directly
along an existing fault line. Wedging and wedge tightening in the brickwork of the crown
were regular features of repair works. A collapse in 1947 saw extensive brickwork relining
in the tunnel.
Since then further deterioration of the lining including mortar loss, bulging of the lining
and bricks falling on to the track particularly in the area of low cover and faulting where
backfilled shafts were known to exist, increased the need for urgent repair work. In 1992
a complete tunnel closure enabled 160 m of tunnel arch to be relined with a reinforced
concrete lining from the northern portal. A smaller section of lining, closer to the centre of
the tunnel, was also completed during the same closure using a Hungarian sprayed
concrete lining system.
In 1997 as part of the structures renewal programme, Sugar Loaf Tunnel was again
inspected and it was found that the majority of water ingress related defects occurred in
the northern section of the tunnel from the end of the concrete lining at chainage eight,
up to the air shaft at chainage 26. The inspection reported sections of lining loose and
ready to fall.
A complete tunnel shut down, as used before, was not now acceptable to the tunnel
operators and a system of supporting the lining had to be devised without disruption to
traffic. The lining would have to be strong enough to be self supporting in only a few
hours after placing, should have good adhesion under wet conditions and could be placed
quickly and efficiently to give the necessary production during the limited possessions
between booked services.
338
Figure A1.54
Sugar Loaf Tunnel after relining with sprayed concrete
Tunnel lining strips were devised that were of sufficiently small volume to allow spraying
in a four hour period within a possession by one machine. More machines meant that
more strips could be completed. However, as a precaution stainless steel anchors were
drilled and epoxy-resin applied to the lining to support any partially completed strip,
should a breakdown occur. Mesh reinforcement was rejected as being a potential risk to
passing trains if erected and not covered with sprayed concrete. Steel fibres were chosen
to form the strip reinforcement. The main item in the dry sprayed concrete was natural
cement, which came pre-bagged with the steel fibres. Adhesion under wet conditions was
excellent and immediate rapid strength gain meant that the strips could easily resist full
speed trains passing within two hours. This rapid gain in strength and low rebound
percentage meant that spraying and clean-up could be left almost until the end of the
possession.
The strips were completed by a toe-anchor block support system using fibreglass bolts to
finish the arch, and a water management system (plastic dimpled sheets) set behind the
strips to conduct any water that found its way to the inside of the brickwork down to the
cess drains.
CIRIA C671 • Tunnels 2009
339
Case study 13: Structural monitoring
strategy for the Channel Tunnel
Adapted from Choquet, Schwenzfeier and Lamont (2005)
13.1
BACKGROUND
The construction and operation of the Channel Tunnel (formally known as the Fixed
Link) was agreed by Treaty between the French and UK Governments in 1986. This
allowed the construction and the operation, by private concessionaires, of a fixed twin
bore tunnel rail link, with associated service tunnel, under the Channel between Cheriton
in Kent and Fréthun in the Pas-de-Calais Region. The Concession, awarded to
Eurotunnel, indicated the general characteristics of the Fixed Link and the rules to be
applied during its construction and later operation. The tunnels were driven by TBM
from each side of the English Channel between 1987 and 1991, and the commercial
operation of the tunnel began in 1994.
13.2
THE SAFETY AUTHORITY AND ITS DUTIES
Through the Treaty and the Concession Agreement, the Governments acquired the power
to monitor the construction and operating conditions of the Tunnel. An
Intergovernmental Commission was established to supervise safety on behalf of the two
Governments, and a bi-national Safety Authority was established to advise it. The main
duties of the Safety Authority are to ensure that national and international safety law is
enforced in the Tunnel, to examine reports concerning any incident affecting safety in the
Tunnel, and to carry out any necessary investigations and report to the Intergovernmental
Commission.
The Safety Authority established five permanent and specialist working groups to provide
detailed technical analysis based on their advice to the Intergovernmental Commission. Of
these, the Civil Engineering Working Group (CEWG) is responsible for the infrastructure
of the tunnels and terminals. A major aspect of its work is to ensure the structural integrity
of the tunnels and that there is a proper asset management system in place for this to be
achieved. In all its activities, CEWG has worked towards ensuring the asset management
system treats the infrastructure as a single tunnel complex. This has involved the
integration of inspection and reporting procedures through the development of the
comprehensive tunnel lining monitoring strategy described here.
13.3
DETAILS OF THE TUNNEL STRUCTURE
The Eurotunnel system comprises two running tunnels on either side of a service tunnel.
The tunnels are about 50 km long, of which about 37 km are under the English Channel.
The running tunnels contain the track for the movement of trains on a closed circuit of
track between the two terminals used for Eurotunnel’s shuttle trains. Running tunnel
north normally handles traffic from the UK to France and running tunnel south from
340
France to the UK. The service tunnel provides ventilation, a safe area for evacuation, and
a method of access for emergency teams.
The two running tunnels and the service tunnel are lined with pre-cast reinforced
concrete segments or, in some places, cast iron segments. The nominal diameter of the
running tunnels is 7.6 m, and of the service tunnel 4.8 m. Each running tunnel contains a
walkway on the service tunnel side for the evacuation of passengers and crew in the event
of an incident. On the opposite side, there is a walkway for maintenance purposes or for
the inspection of trains that have broken down.
The running tunnels are connected to the service tunnel by cross-passages at intervals of
about 375 m. At each cross-passage and recessed into the side of the running tunnels, are
cross-passage doors that separate the running tunnels from the safe haven of the service
tunnel. These doors are fire resistant and are normally closed.
Piston relief ducts with dampers, some 2 m in diameter, are found at intervals of about
250 m, and connect the two running tunnels to relieve the build-up of air pressure caused
by the passage of trains and to reduce aerodynamic resistance.
There are two undersea crossovers, and two land crossovers to enable trains to pass from
one running tunnel to the other, when part of either tunnel is closed for maintenance.
The undersea crossovers are equipped with massive sliding fire-resistant doors, controlled
from the rail control centre. When closed, these maintain the separation between the two
running tunnels and allow them to be separated into three 17 km length sections.
13.4
DEVELOPMENT OF THE MONITORING STRATEGY
From when the tunnel opened in 1994 until 1997, the regulator mainly depended on
annual reports of measurements of the 22 sections of lining fitted with instrumentation, to
check the adequacy of condition monitoring of the linings. From 1997, the CEWG
requested its inspectors and civil engineering experts to analyse these reports and to make
comments, which would enable the monitoring policy to be improved. This was done
through meetings between Eurotunnel, its consultants (Mott MacDonald, SETEC), and
CEWG experts. At the same time, Eurotunnel started to formalise its monitoring strategy
in a single document, which became the Civil engineering maintenance strategy for the
tunnels (CEWG, unpublished).
The monitoring strategy was developed to satisfy the objectives of both the tunnel
operator and the regulator:
1
Tunnel operator: the requirement is to maintain the structure in a functionally
operational state, to ensure that the tunnel can be used with minimum disruption to
traffic. The operating and monitoring system is specifically developed to deal with all
structural faults, which could disrupt traffic, as quickly as possible. For example,
everything possible is done to deal expeditiously with minor leakage through joints in
the lining. Even small amounts of water dripping onto sensitive equipment such as
signalling equipment or the catenary can cause corrosion or short-circuiting resulting
in disruption to traffic operations. Dust is also a major problem
2
Regulator: the objective is to check that the operator’s monitoring policy guarantees
safety in terms of the integrity of the lining and any changes to it. Information about
any part of the structure must be easily accessible, and should enable any change in
the condition of the structure to be identified, through successive reports on its
condition.
CIRIA C671 • Tunnels 2009
341
The development was carried out within a framework, agreed with CEWG, which
required that it should:
cover the periodic routine monitoring of all tunnel lining and deliver results of the
monitoring in a report to CEWG every two years. The report should be interpretative
as well as providing factual data, should give opinion as to the likely future safety of
the structures and should be in the same format for all the tunnel complex
identify parts of the tunnel complex with special monitoring requirements
permit the identification of locations within the tunnels where further monitoring
would be required because of unexpected corrosion, significant leakage or movement
cover the provision and maintenance of instrumentation as necessary to give
continuity of monitoring data
define both relative and absolute alarm thresholds
require appropriate testing of materials where necessary for the monitoring
programme.
As described in the following sections the two principal components of the monitoring
strategy are:
13.5
1
Instrumentation of the structure
2
Visual inspections
INSTRUMENTATION OF THE STRUCTURES
The instrumentation was originally installed in sections of the tunnel to verify the validity
of the assumptions made during the design and construction phase, and to check whether
behaviour of the structures was as predicted. The locations of these sections were
proposed by the designer and constructor of the Tunnel, TransManche Link, taking
account of the geological properties of the ground and the strata profile, and approved by
Eurotunnel. Factors considered when locating instrumented sections included depth of
overburden, chalk interface profile, proximity of the Gault clay and of the service tunnel
at the crossovers etc.
Having verified the validity of the design assumptions, the in situ instrumentation now
provides information, which is proving to be extremely valuable for long-term structural
monitoring purposes. Because the design assumptions for the tunnel lining differed
between the UK and French sides, the instrumentation is also slightly different. On the
UK side, the lining is assumed to be drained, whereas on the French side, the lining is
fully gasketed.
13.5.1
Instrumentation used
The following is a brief description of the instrumentation used in the tunnels:
Normal sections of the tunnels: a normal section generally covers the three tunnels.
During construction, eight sections on the French side and six sections on the UK side
were fitted with instrumentation divided between the two running tunnels and the service
tunnel. Each instrumented section has:
342
24 to 30 vibrating wire strain gauges, to measure strain (Figure A1.55)
one piezometer
six to eight monitoring points for measuring convergence by invar wire and optical
techniques (Figure A1.56).
Figure A1.55
Figure A1.56
Instrumentation box for piezometric and
vibrating wire strain gauges (VWSG), installed
in cross-passage for permanent access
Convergence measurement on the upper part
of the tunnel by invar line, with the help of the
hydraulic access platform
Tunnel crossovers are fitted with:
extensometers of 1 m, 4.5 m, and 9 m in length.
piezometers
monitoring points for measuring horizontal convergence
levelling markers
rock pressure cells
monitoring points across the joints between the central wall and the vault
monitoring points on the headwalls.
Fire damaged section of running tunnel south:
monitoring points for measuring convergence by invar wire and optical techniques
vibrating wire strain gauge to measure strain
extensometers.
CIRIA C671 • Tunnels 2009
343
13.5.2
Frequency of monitoring
The frequency of monitoring varies from six months (for the piezometers and
extensometers) to two years for optical convergence, to six years for the highly accurate
but time-consuming invar wire convergence monitoring.The final section of tunnel before
the UK portal passes for some 500 m through an ancient landslip on the seaward slope of
Castle Hill. Instrumentation consists of boreholes fitted with inclinometers and
piezometers. Regular monitoring of instrumentation is carried out from the hillside, as
well as checks on the flow of water from drainage galleries within the body of the slip.
13.5.3
Reporting and interpretation
Monitoring reports for the tunnels and Castle Hill are prepared annually to an agreed
schedule. The reporting period ends in the spring and should cover the wettest half of the
year (autumn to spring) within a single reporting period. The reporting schedule is as
follows: reports on the results for a year are submitted to the Safety Authority no more
than six months after the last measurement, and presented orally to the experts three
months later. Clarification or extra measurements may be requested by CEWG experts to
address any concerns. The experts report to the CEWG, who may then propose
inspections, or further investigations, to verify the behaviour of sections where the
measurements appear to be abnormal. Finally, the CEWG co-chairmen report the results
to the Safety Authority.
13.6
VISUAL INSPECTIONS
Visual inspections form the second part of the structural monitoring regime. They are
based on French regulations for the monitoring of civil engineering structures forming
part of the national road network, but they are implemented in different ways because of
the specific features of the Tunnel, as it is of a different size than conventional civil
engineering structures. There are two types of visual inspections:
1
Primary inspections, made every six years, consist of a visual examination of all
surfaces accessible from the walkways. They are supplemented by an inspection of the
vault using an access platform where there is a need to monitor changes to specific
defects, or to deal with them (for example, by injecting grout to stop seepage dripping
on to the catenary).
2
General inspections of the lining are carried out annually, chiefly to identify the need
for operational cleaning, surveying of obvious major defects, the development of fresh
defects and the functioning of the drainage networks.
In areas known to be problematical, particularly due to the influx of water, frequent
monitoring may be required.
Finally, if routine monitoring has indicated the presence of a serious defect, a detailed
expert appraisal is organised immediately. That appraisal may lead to repairs or to other
actions such as increased frequency of monitoring or the installation of instrumentation
and an inspection by the CEWG experts.
344
Figure A1.57
Digital plotter with data-logger
13.7
CONCLUSIONS
The monitoring and maintenance strategy for the tunnel lining has been established by
Eurotunnel through a development programme, taking into account the requirements of
the regulator. The strategy is based in part on the monitoring defined at the time of the
design and construction of the Tunnel, but equally it takes into account changes in the
structure over time, as evidenced by the monitoring results. For sections of the tunnels
where lining defects are identified, the response must take into account the history of their
development along with anticipated future changes and all in sufficient time to limit the
adverse effects on tunnel operation.
The strategy is described in a live document and is subject to change should new
information arise from the monitoring and maintenance operations.
CIRIA C671 • Tunnels 2009
345
Case study 14: Invert reconstruction and
other structural repairs to Netherton
Canal tunnel
Adapted from Haider and Richards (1987)
Netherton Tunnel was opened to traffic in 1858 and was the last major canal tunnel built
in Britain. The tunnel was built by the Birmingham Canal Navigation Company to relieve
the congested Dudley Tunnel, which was carrying over 40 000 boats a year at the time on
a one-way system. Now, about 1000 boats a year pass through the tunnel. When it was
built, Netherton Tunnel embodied the skills and experience gained from almost 100 years
of canal construction. It was completed in just 31 months at a cost of £155 000. Walker, the
engineer for the original construction, provided a detailed record of the execution of the
works (see Walker, 1860).
The dimensions are impressive for a canal tunnel, being 2777 m long, 7.4 m high and 8.2
m wide. The waterway width of 4.6 m is designed for two-way narrow boat traffic and has
towpaths on either side as shown in Figure A1.58. The thickness of the brick lining varies
but the majority of the tunnel has 560 mm thick brickwork in the sidewalls and crown and
340 mm thick brickwork in the invert. Localised sections of lining were increased to 675
mm in the sidewalls and arch and 560 mm in the invert where ground conditions were
found to be bad. The original contract drawings indicate that the engineer considered that
certain sections of the tunnel could be constructed without a structural invert. In the
event, a structural invert was provided for the full length of the tunnel, presumably
because ground conditions were less favourable than anticipated. The lower part of the
tunnel is provided with a puddle clay lining off which the towpaths walls are built. The use
of such a lining was unusual but was intended to prevent the canal water from draining
into the adjacent coal mines. Apart from its size, Netherton is notable for the exceptional
quality of its construction.
Figure A1.58
Section through the
original canal tunnel
346
The strata through which the tunnel was driven consist mainly of mudstone, sandstone
and coal measures. The coal seams on either side of the tunnel are mostly worked out but
pillars of unworked coal have generally been left close to the tunnel. These pillars were
stabilised by the canal proprietors purchasing areas of adjacent mines as mining damage
threatened the stability of the canal and its structures.
The tunnel was subjected to damage from mining subsidence during construction and up
to the period of the 1930s when mining in the area ceased. The invert heaved in several
places during construction and Walker attributed this to swelling of the underlying marl.
The damaged invert was generally cut out and rebuilt with thicker brickwork. In some
areas, the rebuilt invert was also given greater curvature, possibly because of more adverse
ground conditions. In 1895, the invert failed over a 100 m length with a maximum heave
of about 800 mm. Temporary supports were installed to keep the tunnel open until 1902
when the failed invert was replaced with thicker brickwork with a greater curvature.
Following incidents of boats grounding on the invert, a comprehensive survey of the
tunnel was undertaken in 1976. This indicated three areas of heave totalling about 270 m.
Further deterioration of conditions caused the tunnel to be closed to public navigation in
late 1979 until an extensive programme of repair work could be completed. As an interim
measure, an array of rock bolts was installed in the sidewall of the failed area. These
succeeded in arresting the continued convergence of the sidewalls until 1983 when the
tunnel was dewatered and a detailed study of the invert could be made before
undertaking repairs.
The investigation comprised:
survey and dimensional checks on the lining
stress measurements in the lining using photoelastic devices
stress change measurements in the lining using vibrating wire gauges
site investigation of failed areas
deformation monitoring of tunnel lining.
Laboratory tests showed that the swelling pressures in the mudstone would not have been
sufficient to have been the prime cause of the failure (although this was thought by Walker
to have been the reason for the earliest invert heaves). There was no evidence from the
site investigation to suggest that mining subsidence or groundwater pressures were a
contributory factor in the failures. Analytical studies showed that the most likely cause of
heave was failure of weak mudstones under the relatively high rock stresses in the areas of
greatest cover (up to 100 m). The original design of the invert was purposely made
relatively flat to minimise the requirements for puddle clay. Unfortunately this proved to
be inadequate to withstand the higher ground pressures in areas where rock failure has
occurred.
The repair works were designed, on the basis of rock-support interaction methods, to
withstand horizontal and radial loads appropriate to the ground conditions. The repairs
consisted of a replacement concrete invert with a slightly greater curvature than the
original (Figure A1.59). An overslab detail was provided for strengthening the transitions
between replacement and existing inverts. Where sidewalls were in poor condition, the
towpath was designed as a structural element and was built of reinforced concrete.
CIRIA C671 • Tunnels 2009
347
Figure A1.59
Details of replacement invert
Very careful control was required to maintain stability of the tunnel during excavation of
the puddle clay and invert. Temporary supports were used that provided a comparable
pressure to the water and fill within the tunnel. Excavation of the failed invert was carried
out in 2 m maximum width transverse bays. Invert reconstruction had to be completed
and specified concrete strengths obtained in any bay before an adjacent bay could be
opened. Multiple bay excavation was permitted only at a maximum spacing of two tunnel
diameters.
Working in a canal tunnel environment required special and sometimes complicated
arrangements for excavation, removal of waste and construction. In particular, the
handling of materials was demanding. For the excavation of the towpath walls and soft
material, up to eight boats were used to transport the excavated material to the site
compound where it was unloaded by a crane and grab. It was finally taken to a tip by
lorry. In all, 750 m³ of brickwork and 2300 m³ of soft materials were removed in this way.
For excavation of the invert a conveyor system was used to transport waste from the
dewatered section of the tunnel to waiting boats.
For the concreting operations, aggregates and cement were brought in by boat and mixed
in the tunnel before pumping to the discharge point. Where the areas to be concreted
were close to a shaft, ready-mix concrete was delivered to site and pumped up to 250 m to
its destination, which increased the rate of progress by about 80 per cent. A total of 3100
m³ of concrete was required for the invert and towpath repairs.
Safety aspects were given a high priority and an engineer was allocated almost full-time to
tunnel monitoring duties. Gas monitoring sets were maintained at the work areas to check
the work atmosphere. Breathing apparatus and escape-sets were issued individually to all
staff with rescue sets maintained at the site office. Wherever possible, electrically powered
plant was used to minimise the risk of pollution and the need for forced ventilation in the
tunnel. A high degree of supervision was maintained.
An intensive programme of monitoring was used throughout the construction period to
check the stability of the tunnel and the adequacy of the temporary supports. The main
monitoring methods were borehole and tape extensometers to measure rock and lining
deformations and hydraulic load cells to measure temporary support loads. The
instrumentation and survey methods are shown schematically in Figure A1.59. Typical
graphs of tunnel convergence are shown in Figure A1.60. Following dewatering a steady
convergence of the sidewalls of up to 5 mm was noted until this was arrested by
installation of the temporary supports. At the same time, the load cells indicated modest
load increases from 10 kN to 40 kN.
348
Figure A1.60
Schematic diagram of instrumentation and survey methods
Figure A1.61
Graph of tunnel convergence during the construction process
As the work proceeded, areas of poor brickwork were revealed below the water-line.
Brickwork replacement of about 25 m length of sidewall and 30 m³ of patch repairs were
carried out in short bays with the adjacent lining being stabilised with rock bolts. Also,
about 200 m² of brickwork, mainly at the crown level, was strengthened with sprayed
concrete 50 150 mm thick. This was carried out from the canal using a specially adapted
boat with hydraulic platforms to provide the required operating height. The compressor
and gun were located in an adjacent boat, which was also used to bring in materials at the
end of each shift.
The works took about 16 months and cost £1.3m (at 1984 prices).
CIRIA C671 • Tunnels 2009
349
Case study 15: Piling adjacent to deep
and near-surface tunnels in London
By Ganga Prakhya, McAlpine Design Group
This case study summarises the risk assessment of driving piles (both bored and CFA) next
to LUL tunnels at a development in London. CFA piles of 750 mm diameter for the south
bridge and bored piles of 600 mm diameter for the works were used at about 3.6 m from
the tunnel face. The development required a pile cap of size 6 m × 18 m and 2 m deep.
The tunnel near the south bridge (Figure A1.62) of the development is 3.8 m OD, of
segmental construction and consists of six circumferential units of 20" (508 mm) width.
The 20" units are bolted together through the flanges of the plate. The plate (of grey cast
iron) is 7/8" (22 mm) thick and the bolts are 1¼" (32 mm) diameter. The tunnel has a soil
cover of 4 m above the crown and is founded in London Clay.
15.1
MOVEMENT PREDICTION AND MEASUREMENT
Worst credible movements in the tunnel (both in concrete and cast iron tunnels) were
predicted using the following methods and calculations performed for assessment. As far
as possible simple techniques were used for conservatism:
volume loss due to the bored or CFA pile
loss of ground pressure in temporary condition (elastic analysis)
non-linear FE analysis in FLAC (axisymmetric model for piling in the ground)
analysis of joints for prying action on bolted segments and friction in concrete
segments.
The model predictions were compared with the data presented at the CIRIA conference
on the response of buildings to excavation-induced ground movements (Jardine, 2001)
wherein 2.1 m diameter piles were bored within 1.5 m from the face of Post Office mailrunning tunnels to give confidence in the techniques and methods. They were also
compared with the data published in Ground Engineering (July 2002) for bored piles
next to Victoria Line Tunnels in London Clay.
The conclusions from the findings were:
350
the total expected transverse movements in the tunnel due to piling were likely to be
less than 1 mm under worst credible scenario (Figure A1.63) corresponding to CI
tunnels even with the interaction of adjacent piles
the joints in the axial direction of the tunnel appeared satisfactory for the expected
movements
the risk to the tunnels (both CI and concrete) because of this movement was low as the
structural integrity was not endangered for both concrete and bolted rings.
A condition survey of the tunnel was carried out before piling.
Monitoring of the ground movements was carried out using inclinometers in a borehole
adjacent to the tunnels as it was not possible to monitor inside the tunnels. A satisfactory
agreement was reached between the finite element model predictions and measured
values. Previous observations in similar circumstances have shown that the vibration
during installation of piles do not exceed 10 mm/s.
Figure A1.62
Deep tunnels: piling and
pile-cap construction
Figure A1.63
Deep tunnel – influence of piling from the face of the tunnel
CIRIA C671 • Tunnels 2009
351
15.2
PILING ADJACENT TO SUBSURFACE TUNNELS IN LONDON
At a development project in London, piling and pile cap construction close to subsurface
LUL masonry tunnels (built in 1890s) was carried out. The piles are 1.5 m to 1.8 m in
diameter and the pile cap is 11 m × 24 m × 1.8 m deep. A detailed analysis was carried out
to determine the risks to the tunnel from various construction activities near to the
tunnels.
Detailed finite element analysis revealed the following:
the movement of the construction machinery adjacent to tunnels could have a
significant effect on the stresses in the tunnels
construction machinery, such as excavators and pile rigs, was restricted to operate
within certain constraints imposed by the risk assessment analysis. These included
keeping a minimum distance from the tunnel face and restricting the loads imposed
by the plant at or above the thrust level of the tunnel
the construction for a pile cap that is 24 m × 11 m was carried in three steps to limit
the unsymmetrical loads on the tunnels to a minimum. In cases where the tunnel
section was weaker, the tunnel was suitably propped during the construction of the
pile cap (see Figures 1.64 and 1.66).
Monitoring of the tunnels was carried out using electrolevels during the entire period of
the development and beyond.
The observations recorded in the tunnel during demolition of the existing structure,
construction of piles, pile caps and transfer structure showed a good agreement with the
theoretical predictions incorporating soil structure interaction models.
352
Figure A1.64
Subsurface tunnel (seven ring masonry) – piling and pile cap construction close to the tunnel
walls
Figure A1.65
Subsurface tunnels (five ring masonry) – propping of tunnel during construction of a pile cap
Figure A1.66
Pile cap construction in steps
15.3
CONCLUSIONS
Not surprisingly, the stresses in the tunnels are very sensitive to the way the tunnel
experiences the loads. For example, the tunnel capacity for uniformly distributed loads on
the entire span is much higher than for non uniform or unbalanced loads applied on the
crown. The capacity of the tunnel was reduced when unbalanced loads were applied
because of highly unsymmetrical excavations on the sides of the tunnels. This effect is
more predominant in subsurface tunnels compared to deep tunnels. The analysis gave
insight into the type of loading that tunnels experience without exceeding the stress limits.
It also allowed the relaxation of undue conservatism in terms of the various construction
operations such as movement of plant, demolition of the existing building, and
construction relating the development etc. The analysis resulted in the safe operations of
the trains together with considerable economies in construction. It is important to assess
these effects during any operations near tunnels.
CIRIA C671 • Tunnels 2009
353
Case study 16: Predicting and monitoring
the effects of adjacent construction on
masonry-lined tunnels
By Ganga Prakhya, McAlpine Design Group
The case study, which is published in detail in CIRIA SP199 (Jardine, 2001) is of a
development in Birmingham, above 19 century brick railway tunnels that are still in full
use. To understand the behaviour of the tunnels during the various stages of construction
and to set trigger levels for movement and vibration monitoring, detailed finite element
analysis was undertaken. The importance of modelling the tunnels with soil structure
interaction, non-linearity of the materials, and the influence of imperfections was
demonstrated. The finite element predictions removed undue conservatism in the
traditional methods of analysis, particularly MEXE, and helped to provide realistic risk
assessments. The models also helped to devise safe constraints on working methods where
real risks were expected. The constraints formulated into the construction method
statements to allow safe operation of railway included limiting the type of machinery
above the tunnels and the length of unsymmetrical excavations.
16.1
BACKGROUND
The Bull Ring redevelopment in Birmingham involved replacing the existing 1960s
concrete shopping centre with a new one. Two parallel rail tunnels (210 m long), carrying
four lines east from New Street station, run under the site and bear on Upper Sandstone.
They are of different brick construction and were built in open cut by different companies
at different times between 1849 and 1863. Railtrack required real time monitoring
systems for deformation and vibration to be installed in the tunnels.
16.2
TUNNEL GEOMETRY
Laser measurements were taken at regular intervals to map the existing profiles of the
tunnels, which revealed imperfections of three to five per cent (difference between
measured and design radii/design radius).
Core-hole sampling and probe-hole investigations were carried out at selected locations to
determine the properties of the brick for assessment.
354
Figure A1.67
Typical sections showing the proposed development near tunnels
16.3
FINITE ELEMENT MODELLING
Analysis objectives
The objectives of the analysis were to generate envelopes of movement and vibration
levels for safe operation of the railway, and to propose limitations on type of plant
operation above the tunnels and size of excavation around the tunnels.
FE model
Classical models (Pippard’s analysis) and methods such as plastic methods (Heyman’s) do
not take into account the effect of soil flexibility in modelling the arches. Also, it is not
possible to take into account uneven imperfections in classical methods. So realistic
movements can only be obtained with a detailed FE model that incorporates soil structure
interaction, non-linearity of the materials such as brick and soil, interaction between the
tunnels because of location, imperfections in the tunnels, and flexibility of the vertical
walls. Finite element models were developed in ANSYS (commercially available software).
The results from ANSYS were calibrated against the results from FLAC (commercially
available program for soil structure interaction).
Calibration using trial test rig
The movement from the sensors for both tunnels were recorded when a trial test rig
weighing 700 kN was located on the crown of the arches. The existing fill at the location of
the test was 2.5 m. The results showed that the non-linearity of the fill material influenced
the results to about 20 per cent. A satisfactory agreement between model predictions and
sensor measurements was obtained, which improved confidence both in the model and
analysis predictions.
Figure A1.68
FE model of the tunnels and soil
CIRIA C671 • Tunnels 2009
355
Construction operations and sensitivity analyses
The loading conditions that could cause highly unsymmetrical loads on the arches and
collapse conditions at various stages of construction work (Figure A1.69) were analysed.
Sensitivity analyses were carried out with respect to soil properties (reduction of about 30
per cent from test results), brickwork properties, and imperfect arch profiles.
a
Figure A1.69
b
Typical construction operations, excavation on south side (a) and excavation between tunnels (b)
Figure A1.70 shows the size of excavation and its effect on the tunnels. From the table, the
size of the excavation larger than 20 m could pose a risk by creating higher stresses than
the in situ condition.
Figure A1.70
Effect of unsymmetrical excavation vs predicted in situ stresses
In Figure A1.71 at location five, the capacity factor reduces from 1.1 to 0.9 because of the
effect of imperfections.
356
a
b
Figure A1.71
Capacity factor vs. imperfection (a) and construction of transfer beams over tunnels (b)
16.4
MONITORING SYSTEMS AND METHODS
The two tunnels each have 13 transverse electrolevel arrays, each consisting of six
beams/sensors as shown in Figure A1.72. Also, each array has anchor points for tape
extensometer measurements. Intermittent sections of tape extensometer points are
provided between the electrolevel sections. Six vibration monitoring geophones and four
tilt meters are also located in the tunnels and on the retaining walls near the station
portal. Individual trigger levels were set with reference to BRE recommendations for
structures (BRE, 1995).
Figure A1.72
Monitoring stations
16.5
RESULTS OF MONITORING
Large excavation on south side
Table A1.8 shows a typical comparison of predicted movement and the measured
movement during the excavation on the south side of the south tunnel. A large
unsymmetrical excavation including a surcharge load due to construction equipment on
the tunnel crown with a fill depth of about 2 m resulted in both longitudinal and
transverse strains and stresses. The results fall within the envelopes of the sensitivity study.
CIRIA C671 • Tunnels 2009
357
Table A1.8
Predicted and observed movements during excavation on south
Displacements in mm
South tunnel
16.7
Crown
Springing
Predicted using lower bound soil properties
5.3
3.6
Predicted using upper bound soil properties
3.8
2.5
Measured
3.2
2.1
CONCLUSIONS
The use of FE models included soil structure interaction and based on measured material
properties and geometries gives less conservative predictions of movement and vibration
than traditional methods. This allows economies to be made in the construction process.
Monitoring confirms that predictions are good and are still safe.
358
Case study 17: A feasibility-based risk
matrix for option selection
By Danny Swannell and and David Jarvis, Owen Williams Railways
SUMMARY
This case study introduces a simple feasibility design and risk matrix as used on two
Network Rail tunnel feasibility schemes. The matrix is an example of simple innovation
and good practice that delivers an accountable, justifiable optioneering solution based on
several categories and weighting. Each option is scored and ranked giving the user and
client a clear tool that aids both budgetary planning and technical approval. Whole-life
cost issues are described in some detail and the study concludes with a worked example
detailing the effectiveness of the tool and how it might be used in the future to benefit
clients for initial and detailed optioneering schemes for any type of asset.
17.1
INTRODUCTION
The risk matrix outlined here provides an example of a simple, effective and efficient tool
to establish the most appropriate solution for proposed works. It was first used for the
optioneering study of Box Tunnel lining works for Network Rail and is innovative in its
simplicity and accountability. It is based on pre-agreed categories with weightings to
emphasise importance, and has the potential to be incorporated into good practice
procedures. So far it has been used on two tunnel optioneering studies, as well as a station
platform refurbishment scheme with favourable results and feedback. It has proved useful
during technical approval and illustrates potential risk in an easy to manage format.
Note that the methodology described here is provided as an example of one possible
approach and is not necessarily recommended for unmodified use on other projects
without careful consideration of the specific situation and its individual requirements.
17.2
WHY USE A MATRIX?
A matrix has been used for many reasons. Primarily it allows informed decisions to be
made that are both justifiable and traceable. Often in feasibility studies the result can be
subjective or not obvious. This system instead allows for a preferred option to be
promoted with a greater degree of confidence and client agreement.
The matrix is also flexible. Categories are selected in accordance with the client
requirements such as cost and duration. Associated with each category is a weighting, and
these weightings can be easily adjusted to reflect the importance of the categories.
The matrix is simple. The tool is a simple spreadsheet that can be updated live. It also
provides the ideal extract and answer (in a design statement) for justifying why certain
options have been considered and rejected. It can be easily adapted to meet changing
client emphasis, and is an objective tool that is simple to grasp and use.
CIRIA C671 • Tunnels 2009
359
17.3
HOW TO USE THE MATRIX – A TWO-STAGE APPROACH
The matrix should reflect the stage of the feasibility study. An initial brainstorming stage
needs to consider a range of options at a relatively high level with various options
following that can be discarded due to their low scores. The remaining options then follow
through to the detailed stage.
The initial phase only considers four categories, namely buildability, construction cost,
construction duration, and whole-life cost. The detailed phase considers these four
categories together with health, safety and environmental assessment, design cost,
possession over-run risk and residual risk. These categories have worked well with
Network Rail remits and requirements.
Having defined the categories, weightings must be applied. These weightings must be 100
per cent but otherwise there are no further constraints. Cost categories clearly demand
the greater weighting, and for the initial stage the cost categories total 60 per cent and for
the detailed stage the construction cost plus whole-life cost categories total 40 per cent.
The weightings can be adjusted as part of a sensitivity analysis to establish if, for example,
cost is made to be the dominant weighting how much would the overall scores and
ranking of each option being considered vary.
17.4
HOW TO USE THE MATRIX – SCORING
Scoring is the most subjective element of the process however each score must be justified
with an explanatory comment. There are two simple rules for the scoring:
1
The scores must range between zero and ten, with zero representing a fundamental
failure of the brief and ten representing the most advantage to the client
2
The scores relate to the range of options, ie a score of eight does not represent an
option twice as advantageous as an option that only scores a four for the same
category. This generates some distance between options, which is then modified
further by the weightings.
The scores represent the best advantage to the client. An option scoring ten for one
category means it is the perfect option for the client based on that category. Normally the
best and worst options for each can be identified, and a suitable score given accordingly.
The remaining options can be scored knowing how they compare to the best and worst, as
well as to each other.
17.5
EMPHASIS ON COSTING
A lack of useful costing information can be overcome by appointing experienced
contractors to advise on initial and maintenance costs together with buildability issues. On
the Box Tunnel feasibility study (Section A17.6) and Rhosferig Tunnel feasibility study the
contractor provided detailed construction and maintenance costs for each of the proposed
options to ensure accuracy and reliability. These costs are based on informed possession
regimes for the particular tunnel, and include all temporary and permanent works
required for the various options.
The whole-life cost of each option is calculated on the design life specified in the remit,
typically 120 years. This is the present worth of the future costs, which is the sum of the
inspection costs (not necessarily the same for each option) and associated routine
maintenance costs. These costs depend on the option, in particular the use of steel
360
elements that might need repainting, brickwork repairs that might remain after the
solution is installed and so on.
These costs are discounted using the procedure outlined in the DMRB (HA, 1992).
However, the DMRB value of internal rate of return (eight per cent) is somewhat out of
date. At Owen Williams Railways four per cent was used for two reasons. First the
National Audit Office had recently reduced its six per cent value to 2.5 per cent. Second,
the lower the internal rate of return, the more cost and emphasis it places on the
maintenance value. For example, £25 000 of steel repainting in 25 years time has a
discounted value of £3650 at eight per cent, but £9378 at four per cent. The internal rate
of return can be easily changed to undertake a sensitivity analysis on the discount rate.
17.6
WORKED EXAMPLE – BOX TUNNEL
In 2003, Owen Williams Railways were commissioned to carry out a feasibility study for
the design of a new lining in sections of the 1¾ mile long Box Tunnel in Wiltshire. In
some areas of the tunnel, the oolitic limestone strata forming the unlined areas had
suffered deterioration and recent minor rockfalls. At these locations, the tunnel profile has
been formed in caverns around 9 m wide and up to 12.5 m high.
Because of the risk of a rockfall causing damage to infrastructure and passengers,
Network Rail was keen to explore a means of protecting the track area from any further
potential damage.
The optioneering design process involved a series of brainstorming meetings involving
industry expertise in the fields of geology, tunnel lining design and underground
construction techniques in a rail environment.
The matrix was a key tool that was used during the design process and presented to the
client in conjunction with the feasibility report. It was populated during brainstorming
meetings throughout the optioneering process.
17.6.1
Feasibility matrix
The feasibility matrix (Table A1.9) shows the initial stage for the optioneering/
brainstorming process, scored under the four feasibility categories as previously
mentioned. The weightings adopted were buildability (20 per cent), construction cost (50
per cent), construction duration (20 per cent) and whole-life cost (10 per cent).
The options included:
1
Arched lining forms, such as brickwork, concrete and steel.
2
Horizontal lining forms, such as crash decks or tensioned cable netting.
3
Surface lining forms applied against the rock surface to prevent any loose rock from
falling.
Much of the optioneering involved practical discussion on methodology, resources,
materials and access to site. A comments column is included in each category to record the
advantages or disadvantages applicable and to assist the reader in understanding the score
given. For example, Option 3b has large steel panels, which require consideration of
handling and storage of materials, and site access issues relating to delivery of pumped
concrete. Access is especially important at Box where the nearest rail access and storage
facilities are one mile from the portal.
CIRIA C671 • Tunnels 2009
361
The construction costs were estimated by the specialist contractor and were entered
directly and scored manually. There were more than 10 options at this stage and to keep
scores between zero and 10, costs close together were grouped with the same score, eg
Option 2a (£6.6m) and Option 6a (£6m) were scored at 5.
It was noted that the do nothing option attained top score because of zero costing and
durations, and the option failed to meet the remit so was given a zero rating and not taken
further.
Possession durations were estimated in accordance with methodology and possession
availability and scored accordingly. Most options were considered on the basis of a series of
30 hr possessions except where it was deemed necessary to close the line for longer
periods, for example, the brick arch and the concrete segment options.
The whole-life costs were calculated for the next 60 years. The calculation included the
initial construction cost, replacement of design expired items and maintenance costs. For
example, the asset multiplate steel arch (3a, b and c) needs to be repainted every 20 years
and the corrugated steel panels replaced every 60 years. It was considered unnecessary to
calculate beyond this time despite the 120 year design life, because the net present worth
values were insignificant beyond 60 years.
The total score column revealed the top six options excluding the do nothing options and
these were taken through to the detailed stage.
17.6.2
Detailed matrix
The detailed matrix is shown as Table A1.10. Four further categories were added at this
stage and the redistributed weightings were as follows (new categories in bold):
CDM assessment (10 per cent)
design cost (five per cent)
buildability (15 per cent)
possession risks (10 per cent)
construction cost (25 per cent)
construction duration (15 per cent)
whole-life cost (10 per cent)
residual risk of lining failure (10 per cent).
Further detail was applied to the drawings at this stage and the risk scoring re-assessed
following further information received at optioneering meetings.
The contract remit required the decision making process to involve the client, so the live
matrix was displayed at the final presentation. The matrix is adaptable to meet changing
client objectives and this proved useful at the meetings. The client proposed an extra
option, which was a combination of two existing options in conjunction with increased
inspection and maintenance of the rock surface. This option was then re-scored at the
meeting and became the preferred option. The adaptability of the matrix enabled a
prompt decision to be made by the client.
362
17.7
CONCLUSION
This exercise demonstrated the importance of involving engineering specialists at the
start. This not only adds to the scope of options that are raised in brainstorming sessions
but also provides sound principles for the future, or to discard impractical solutions. It
also reduces the chance of surprises emerging in the later stages, especially technical
approval when a specialist may become involved at too late a stage to make major
alterations to the scheme.
One of the advantages of the matrix is the simplicity and logical format. This enables the
client to focus on the practicalities of procurement and cost for the installation phase. It is
also adaptable and flexible enabling its use on many types of construction optioneering
projects, simply by changing the risk categories and weightings to suit the project goals.
The two stage approach was the preferred way of narrowing a wide range of ideas in these
projects. It has already proved successful in existing feasibility studies for Owen Williams
Railways on a variety of optioneering projects and should prove a useful tool for other
organisations in the management, maintenance and repair of infrastructure tunnels.
CIRIA C671 • Tunnels 2009
363
Table A1.9
Feasibility matrix for initial assessment of options for box tunnel relining
Feasibility category and weighting expressed as a percentage
Six year
examination.
Re-point
£50 000/6 years
£15 170 538
1
120
£6 600 000
5
1 year blockade
2
6 year
examination.
£5000/20 years
£6 603 799
7
440
£9 600 000
2
14 years, based
on 7 no 30 hour
possessions per
year
2
6 year
examination.
£5000/20 years
£9 603 799
3
250
7
6 year
examination.
Minor repairs.
Repainting 20
years.
Replacement
after 60 years.
£100 000/
20 years
£7 567 631
5
490
7
6 year
examination.
Minor repairs.
Repainting 20
years.
Replacement
after 60 years.
£100 000/
20 years
£8 115 162
5
510
8
6 year
examination.
Minor repairs.
Repainting 20
years.
Replacement
after 60 years.
£100 000/
20 years
£7 512 878
5
550
6
6
6 year
examination.
Minor repairs.
Repainting 20
years. £20 000/
20 years
£3 365 195
9
720
3
2
£15 000
000
Large but heavy
interlocking modular
segments, heavy
specialist plant.
Adaptable timings.
Special design.
4
Large but heavy
interlocking modular
segments, heavy
specialist plant.
Adaptable timings.
Special design.
4
2b Concrete segments
(possessions)
3a Asset multiplate
(from haunch)
3bi) Asset supercor
(from haunch)
3bii) Asset supercor (from cess)
Large heavy
modular sections,
Awkward handling
of preformed
sheets. In situ
concrete required.
Bracing required
before backing fill.
Bolt on site. Storage
limited. Transport to
worksite.
Large heavy
modular sections,
Awkward handling
of preformed
sheets. In situ
concrete required.
Bracing required
before backing fill.
Bolt on site. Storage
limited. Transport to
worksite.
Modular fixing. Precast concrete
hollow slab decking.
Heavy lifting rock
and boring
requirements
5
6
7
8
Ranking
1
Score/10
36 years, based
on 7 no. 30 hour
possessions per
year
1 Brick arch
1
High labour costs,
well known practice.
Standard materials
and equipment/
labour
Large sections,
Awkward handling
of preformed
sheets. In situ
concrete required.
Flimsy sections,
Bracing required
before backing fill.
Bolt on site. Storage
limited. Transport to
worksite.
Total
Comments
Comments
Comments
Whole-life cost
Score /1000
Construction
duration
Score/10
Construction
cost
100%
Score/10
Buildability
10%
Score/10
20%
Number
4 Horizontal
crash deck
364
50%
2a Concrete
segments
(blockade)
Option
20%
£6 850 000
£7 350 000
£6 800 000
£3 350 000
4
4
3 years, based
on 52 no. 10
hour possessions
per year
3 years, based
on 7 no. 30 hour
possessions per
year
4
2 years, based
on 7 no. 30 hour
possessions per
year
7
6 years, based
on 7 no. 30 hour
possessions per
year
Comments
ff
5a Hollybank Arches
(from haunch)
5b Hollybank Arches
(from cess)
Standard arch
sections with variable
intermediate
sections. Core/
anchor supports. RRV
lifting plant. Foam
concrete backfill
6a Tensioned cable
and netting
Labour intensive,
multi rock anchor
drilling. Single plane
mesh. Significant
volume of rock cut.
6b Rock bolt and
surface mesh
Labour intensive,
Cherry picker access,
Multi core drilling.
Standards bolts and
surface covering
mesh.
7a Sprayed concrete
(protective layer)
No construction
requirements, labour
intensive rockface
preparation.
Specialist operator.
Flexible installation
times. Waste of
rebound material.
No construction
requirements, labour
intensive rockface
preparation. Specialist
operator. Flexible
installation times.
Waste of rebound
material. Rock cutting
for profile
No construction
requirements, labour
intensive rockface
preparation. Specialist
operator. Flexible
installation times.
Waste of rebound
material. Rock cutting
for profile
8 Aluminium
canopy
Standard arch
sections with variable
intermediate
sections. Core/
anchor supports. RRV
lifting plant. Foam
concrete backfill
7b Sprayed concrete
(lining from Haunch)
Feasibility matrix for initial assessment of options for box tunnel relining (contd)
7c Sprayed concrete
(lining from cess)
Table A1.9
Large light sections.
Backfill required
CIRIA C671 • Tunnels 2009
6
6
£7 100 000
£5 350 000
5
480
9
730
2
6
2 year
examination.
Major repairs
£150 000/
£5 045 725
25 years and
replacement at 50
years
8
670
4
4
6 year
examination.
Minor repairs
£6 692 162
£25 000/20 years
and replacement
at 50 years
7
500
3
6 year
examination.
Minor repairs
£25 000/20 years
and replacement
at 50 years
£8 973 587
4
330
4
6 year
examination.
Metalwork
£8 008 068
protection
£20 000/30 years
5
420
4 years, based
on 7 no. 30 hour
possessions per
year
7
5 years, based
on 7 no. 30 hour
possessions per
year
5
7 years, based
on 7 no. 30 hour
possessions per
year
3
9 years, based
on 7 no. 30 hour
possessions per
year
3
8 years, based
on 7 no. 30 hour
possessions per
year
£8 000 000
620
2 year
examination.
£2 925 336
Mesh replacement
every 30 years
8
7
8
7
£1 800 000
£7 850 000
£5 365 195
7
5
4
6
6 year
examination.
Minor repairs.
Repainting 20
years. £20 000/
20 years
2 year
examination. Cable
re-tensioning.
£8 052 650
6 years. Cable/
mesh replacement
every 60 years
5
£5 850 000
500
6
£6 000 000
5
6
5 years, based
on 7 no. 30 hour
possessions per
year
2
£4 350 000
£7 115 195
4
4 years, based
on 7 no. 30 hour
possessions per
year
6
6
6 year
examination.
Minor repairs.
Repainting 20
years. £20 000/
20 years
11 years, based
on 7 no. 30 hour
possessions per
year
5
365
9 Hollow section
crash protection
Feasibility matrix for initial assessment of options for box tunnel relining (contd)
Unknown
manufacturing
methods
3
£7 000 000
10 Do nothing
Table A1.9
None required
10
£0
4
8 years, based on
7 no. 30 hour
possessions per
year
10
None required
4
6 year examination.
Minor repairs.
Metalwork repainting
£100 000/20 years
£7 075 974
6
400
10
1 year examination.
£5000/year
£113 117
10
1000
1
Notes
1
2
3
4
366
All scores are relative to the other options and represent the advantage to Network Rail, a higher score indicating a better
advantage to the client
If any option category scores zero, this indicates that it fails to meet the project remit and so the option total score is set to
zero.
Construction costs exclude possession management and alternative transport and disruption costs.
This matrix considers options generically over the whole tunnel length (ie not area specific)
Temporary works
to install haunch
supports and
multiple heavy
lifting/limited
access bolting up
of lining sheets
Multiple heavy
lifting/limited
access bolting up
of lining sheets
Temporary works
to install haunch
supports and
multiple heavy
lifting of pre-cast
slabs
Stability bracing,
limited heavy
lifting of
steelwork and
multiple light
lifting of plank
shuttering
5
Footing
design and
multiple arch
geometric
setting out
5
Simple
modular
design
6
Arch design
and multiple
arch
geometric
setting out
5
Large light
sections. In situ
concrete
required
6
Large light
sections. In situ
concrete
required
7
Standard
equipment,
modular fixing.
In situ concrete.
Heavy lifting
rock and boring
requirements
5
Standard Arch
sections with
variable
intermediate
sections. Heavy
lifting. Backfill
6
6
8
7
High risk
High risk
Medium risk
Medium risk
4
4
7
5
£7 350 000
£6 800 000
£3 350 000
£5 350 000
4
4
7
5
6 years
3 years
6 years
5 years
6
6 year
examination. Minor
repairs. Repainting
20 years.
Replacement after
60 years.
£100 000/20 years
8
6 year
examination. Minor
repairs. Repainting
20 years.
Replacement after
60 years.
£100 000/20 years
6
6 year
examination.
Minor repairs.
Repainting 20
years. £20 000/
20 years
6
6 year
examination.
Minor repairs.
Repainting 20
years. £20 000/
20 years
£8 115 162
£7 512 878
£3 365 195
£5 365 195
Ranking
4
Haunch
support
design and
multiple arch
geometric
setting out
Score /1000
Total
Score/10
Residual risk (of
lining failure)
Comments
Whole-life cost
Score/10
Construction
duration
Comments
Construction cost
Score/10
Possession risks
(over-run)
Comments
Buildability
Score/10
Design cost
Comments
CDM assessment
Score/10
100%
Comments
10%
Score/10
10%
Comments
15%
Score/10
25%
Comments
10%
Comments
15%
Score/10
3bi Asset supercor
(from haunch)
3bii Asset supercor
(from cess)
4 Horizontal crash
deck
367
5b Hollybank arches
(from cess)
5%
5
Limited risk of
unforseen rock
anchor
overload
7
525
5
Very limited
residual risk of
unforseen
backfilled arch
overload due to
high factor of
safety and load
spreading
capability
8
600
5
8
Limited
residual risk of
unforseen rock
anchor/deck
overload
6
670
2
7
Limited
residual risk of
unforseen
backfilled arch
overload due to
high factor of
safety
7
605
4
Detailed matrix for further assessment of shortlisted options
Number
Option
10%
Table A1.10
CIRIA C671 • Tunnels 2009
Feasibility category and weighting expressed as a percentage
7a Sprayed concrete
(protective layer)
10a Do nothing
10b Periodical
de-scaling
No design
risk to assess
Simple design
risk to assess
7
Limited
design
Specification
only
10
No design
cost
9
Minimal
design cost
8
9
No construction
requirements,
labour intensive
rockface
preparation.
Flexible
installation
times. Waste
10
9
9
6
10
Labour
intensive, waste
9
Low risk
Low risk
No risk
Low risk
9
£1 800 000
8
£4 350 000
10
No
construction
cost
9
£400 000
8
6
10
9
4 years
5 years
0 years
2 years
Notes
1
All scores are relative to the other options and represent the advantage to Network Rail, a higher score indicating a better advantage to the client.
2
If any option category scores zero, this indicates that it fails to meet the project remit and so the option total score is set to zero.
3
Construction costs exclude possession management and alternative transport and disruption costs.
4
This matrix considers options generically over the whole tunnel length (ie not area specific).
7
6
2 year
examination.
Minor repairs
£50 000/
20 years
10
8.5
None
None
£2 925 336
£5 045 725
£0
£778 988
9
9
840
1
7
Residual risk of
water pressure
build-up behind
sprayed
concrete lining
causing
premature
failure
4
635
3
10
Unacceptably
high residual
risk of rockfall
hitting train so
fails to meet
project remit
10
0
7
10
Significant
residual risk of
rockfall hitting
train within 20
years
2
833
1
Detailed matrix for further assessment of shortlisted options (contd)
Health risk
from cement
reaction
requiring high
levels of
containment
and PPE
9
2 year
examination.
Mesh
replacement
every 30
years
Little or no
residual risk
because rock
face stability is
improved by the
grid of rock
anchors and
effect load
potential is
removed
Table A1.10
6b Rock bolt and surface
mesh
368
Simple
construction
installing rock
bolts using
light plant
Labour
intensive, Multi
core drilling.
Standards bolts
and surface
covering mesh.
Flexible working
times
Case study 18: Tunnel fires, collapses
and other serious incidents
This section includes data on, and several short summaries of, a variety of serious
accidents and incidents involving tunnels, including fires, collapses and environmental
pollution. Although for many reasons the sudden catastrophic structural failure of tunnels
has become increasingly rare, recently there have been several major tunnel fires
worldwide that in aggregate have resulted in hundreds of fatalities. It is important for
those involved with tunnel design, management, maintenance, repair and operation to
understand and learn from such past experience, and with the benefit of hindsight, many
of the incidents described here were avoidable.
18.1
OVERVIEW OF FIRES AND OTHER SERIOUS ACCIDENTS IN
RAIL AND METRO TUNNELS
Adapted from a paper by T Andersen and B J Paaske, Det Norske Veritas (DNV)
Consultants, Norway (Andersen and Paaske, 2002)
Andersen and Paaske have identified 26 serious accidents in rail and metro tunnels during
the period 1940 to 2001, collected from a variety of reference sources. Thirteen of the
accidents involved fires in passenger trains. There are no uniform criteria for selection of
these accidents, except that they occurred in tunnels or in subsurface spaces of metro
systems. It is reasonable to believe that the most serious fires during this period are
included.
A study of known tunnel and underground metro accidents provides an important
contribution to understanding the conditions that caused them, and may indicate how
preventative actions or emergency responses might have reduced their effects, or even
allowed them to be completely avoided.
The following provides a summary of their findings and conclusions.
18.1.1
Examples of rail passenger vehicle fire incidents in tunnels
The most serious accident occurred in the Armi tunnel, Italy, in 1944 where between 400
and 500 people were killed by carbon monoxide poisoning caused by smoke from the two
steam locomotives hauling the train that, lacking power to climb the gradient, became
stuck in the tunnel. Finally, the train reversed out but by this time most of the passengers
had died. This happened during WWII and it seems that the passengers may have been
transported on open flat cars. Although not really a fire accident and not important in
today’s rail infrastructure and operations, the combustion intensity and smoke production
from the two steam engines (6 15 MW thermal effect) may have been comparable with
what can be expected in a relatively severe fire in a single passenger car in today’s train or
metro systems.
CIRIA C671 • Tunnels 2009
369
Among other serious accidents, two that are also particularly notable due to the high
number of fatalities caused are the fire in the metro of Baku in 1995 (289 people killed)
and the fire in the tunnel on the cableway to Kitzsteinhorn, Austria, in 2000 (155 people
killed). Both tunnels had relatively small cross-sectional area (Kitzsteinhorn 10 m² and
Baku Metro 28 m²). There are reasons to believe that the narrow cross-sectional area of
these tunnels contributed significantly to the severity of the accidents because most of the
people who died did not manage to get out of the train, or got out very late. In the Baku
fire about 245 casualties were found in the train and only about 40 were found outside it
in the tunnel. In both incidents, there were problems with opening the doors, but rapid
development of the fire and smoke accumulation also made a considerable contribution to
the casualties. It has been speculated that if these tunnels had a larger cross-sectional area,
there may have been more time for evacuation before the heat and smoke concentration
became unbearable. In these cases it is unlikely that improved escape ways from the
tunnels would have significantly reduced the effects, but may have saved a few people.
It should be mentioned that in the accident in the Baku metro many of the people killed
were trampled and crushed to death in the panic to escape.
Another serious accident occurred in 1972 in the double tracked Hokuriku tunnel (13.9
km) when a fire started in a restaurant wagon in a night train. The train stopped halfway
in the tunnel to disconnect the actual wagon, but was not able to continue. The train
carried more than 700 passengers and 30 of these were killed. The tunnel was not
sufficiently equipped with provisions for ventilation and lightning and this was heavily
criticised after the accident.
There are also examples of serious fires in trains that have stopped inside a tunnel and
where the passengers have rescued themselves by escaping through the tunnel.
The accident on the BART metro in San Francisco in 1979 shows that a twin-bored tunnel
concept with frequent intervening cross cuts is no guarantee of safety in a case of fire and
does not necessarily lead to sufficient working conditions for the rescue team. This tunnel
also had a service tunnel and two single tracked tubes, but still one person from the rescue
team was killed and several were injured in this fire.
18.1.2
Brief analysis of accidents
In total about 1400 people have been killed in the 26 identified accidents, of which 1000
were killed in the 13 accidents classified as fires, with associated asphyxiation/toxicity from
fumes. The remaining accidents are primarily related to collisions, tunnel collapses and
station overcrowding. The great majority of the casualties (90 per cent) were found
onboard the trains or within station areas. Only a small proportion of the victims were
suffocated in the tunnels outside the train. Also, in fire accidents, most of the people were
killed inside or near to the train. So it seems equally important to ensure the possibility of
evacuating a train and escaping from the immediate scene of the fire as it does to ensure
safe escape from the tunnel.
In four of the 13 fire accidents the train stopped in the tunnel following a technical failure
that also started the fire, or the fire resulted in a technical failure that caused the train to
stop. The Baku and the Kitzsteinhorn accidents, as well as many other severe accidents,
were of this type. In two other fire accidents (Hirschen-graben and BART) the train was
forced to stop in the tunnel due to application of the emergency brake, after which it was
not able to continue out of the tunnel and had to be evacuated inside the tunnel. These
situations could potentially have been avoided if it had been possible for the driver to
override the emergency brake, or if passengers were instructed not to use emergency
brakes in tunnels. In the six fire accidents where the train came to an unwanted stop, the
370
stop was made at an arbitrary point along the tunnel’s length. To ensure safe evacuation in
such cases it should be possible to carry out rapid evacuation from the train at all locations
in the tunnel, and the interval between two cross-connections to a second tube/escape way
should be quite short or the tunnel should have a large cross-sectional area to provide
adequate evacuation time.
In four scenarios (Eurotunnel, Hamburg, Hokuriko and Simplon) the train was
deliberately stopped in the tunnel or at an underground station to evacuate people,
decouple carriages on fire, and/or fight the fire. Apart from the Hamburg U-Bahn
incident where the train remained at a station platform, the other events happened in
relatively long tunnels (14 km or longer). In the Eurotunnel fire the train stopped next to
an emergency exit but the concentration of smoke and gases following the train made it
difficult to use the exit. A bubble of fresh air at overpressure was injected into the tunnel
through the emergency exit. In Simplon the carriage on fire was decoupled and the diesel
powered train moved out of the tunnel with most of the passengers. In the Hokuriko
accident the train was not able to move after the effort of decoupling the carriage on fire,
and 30 of the 700 passengers lost their lives before the remaining passengers were saved
by trains on the neighbouring track.
The remaining three accidents include events with a train on fire driving through the
tunnel without stopping (Salerno in 1999), co-toxification from steam locomotives not able
to pass through the tunnel (Armi, 1944), and collision between trains with fire (Batignolles
in 1921). In two of the events (Eurotunnel in 1996, BART in 1979) the passengers escaped
into parallel tunnels. For the other accidents it is doubtful whether the tunnel concept
made such an escape route viable. Tunnel fires with the highest number of fatalities have
all occurred in tunnels with narrow profiles with a single track bore either on a single
track line or as part of a double tube tunnel concept.
18.2
CLIFTON HALL TUNNEL DISASTER
Clifton Hall tunnel in Manchester was a brick-lined rail tunnel opened in 1850. The
original records of its construction were badly damaged in a fire caused by wartime
bombing in 1940, and the remainder destroyed by a fire in the records office in 1950. On
13 April 1953, fifteen days before the collapse, some brickwork fell from the roof of the
tunnel where an old and unknown construction shaft had been filled in. Immediate steps
were taken to stop rail traffic and arrangements were made to strengthen the tunnel, but
before the protective work had been completed, the roof collapsed early in the morning of
28 April. The contents of the shaft poured into the tunnel forming a crater at the ground
surface, causing a pair of semi-detached houses to collapse suddenly and violently, and the
outer wall of a third house to fall outwards. The five occupants of the first two houses lost
their lives but fortunately two others in the third house were rescued, suffering from
shock and minor injuries. A government inquiry was instigated to investigate the causes of
the collapse and how in future such incidents might be avoided, and its results were
published in 1954. It exonerated the individual rail engineers dealing with the situation,
but pointed out that the situation could have been avoided if adequate records of the
tunnel’s construction had been available at the time.
CIRIA C671 • Tunnels 2009
371
Figure A1.73
Plan and section of the tunnel after the accident showing timber bulkheads and backfill
Figure A1.74
Scraper being used to level off the ingressed sand
Drawings of the tunnel’s cross-section (see Figure 2.16) reveal that the brickwork in the
crown was constructed entirely from headers, and this could have been a significant
influence on the collapse.
The lessons to be learned from this disaster are summarised in the inquiry report’s
conclusions, included here in abridged form:
“The roof of this tunnel caved in at a point directly underneath an old brick-lined
construction shaft, the contents of which must have fallen in one mass into the space
below. The surrounding soil, which was a mixture of sand and clay, poured into the
hole and formed such a large cavity underneath the foundations that the two houses
collapsed without warning to the unfortunate inhabitants.
372
The failure of the roof was in no way attributable to mining subsidence, but was due
to an inherent weakness in the construction of the tunnel…Prompt steps were taken
to protect rail traffic as soon as the defect was noticed…and I am satisfied that (the
staff) were fully alive to the need for urgent action, but none knew of the existence
of the construction shaft and so they did not appreciate the very dangerous
conditions, which were set up when the initial brickwork fell.
…the loss of the tunnel records contributed materially to this accident, and the
events leading up to it have shown all too clearly the danger which arises when vital
knowledge is not readily available. The maintenance staff should know of the
existence of old shafts and other features which may cause weaknesses but in many
cases the only records are the original construction drawings which, with the growth
in the number of documents to be preserved in the engineers’ offices, may possibly
be overlooked, though in this case they were burnt in circumstances over which the
staff had no control. I recommend, therefore, that all tunnel records be reviewed
and any special features brought to the attention of the maintenance and examining
staff. It is also desirable that the position of disused shafts should be permanently
marked on the tunnels themselves, so that these places may be particularly watched.”
18.3
TRAIN DERAILMENT AND FIRE AT SUMMIT TUNNEL
Based on information published in a report on the derailment and fire that occurred
on 20 December 1984 at Summit Tunnel (Department of Transport, 1986)
Summit Tunnel is a double-tracked rail tunnel, 2885 yards long, built in 1841 and located
about 15 miles north of Manchester on the line to Leeds. It was driven through a variety
of ground including shales, gritstones, coal measures and broken ground, with a
horseshoe-shaped lining of brick masonry varying between five and 10 rings in thickness
depending on the geological conditions encountered during construction. The tunnel
included 14 construction shafts, two of which had been closed by the time of the incident.
At 5.50 am on 20 December 1984 a freight train, comprising a locomotive hauling 13
loaded bogie tank wagons conveying 835 tonnes of petroleum spirit, was passing through
the tunnel at about 40 mph when an axle bearing on the front bogie in car four
overheated, failed and caused the car to derail. The automatic air-brake pipe ruptured
resulting in emergency application of the train’s braking system, bringing it to a halt about
midway through the tunnel. Cars six and 10 had overturned and landed on the adjacent
track. Fire broke out almost at once and the train’s crew rapidly escaped from the tunnel
on foot and raised the alarm from a signal telephone box just outside the tunnel portal.
The fire brigade entered the tunnel and made fire-fighting attempts in the first hours
after the accident. The train’s crew re-entered the tunnel under fire brigade supervision
and succeeded in uncoupling the train between the third and fourth cars and pulling
these first three cars out of the tunnel with the locomotive. At this point the fire seemed
under control but by 9.40 am it developed rapidly requiring a complete emergency
evacuation. The fuel supply to the fire was so rich that some of the combustibles were
unable to find oxygen inside the tunnel to burn: they were instead ejected from the two
nearest open shafts as superheated, fuel-rich gases that burst into flame the moment they
encountered oxygen in the air outside the tunnel, with flames reaching a height of 50 m
above ground level (Figure A18.3). This set local vegetation on fire and caused the closure
of the A6033 road. To gain control of the fire, water and foam was pumped down the
shafts but it was not until 3 am the following night that the flames from the shafts
decreased. Early attempts to inspect the tunnel were abandoned after fire crews reported
daunting noises emanating, presumably, from the contracting metal, brick and rock.
CIRIA C671 • Tunnels 2009
373
As well as the risk from fire, one problem affecting the down-gradient (Manchester) end of
the tunnel was the risk of pollution from petroleum spirit entering the drains and
watercourses. This led to a decision to evacuate members of the public in the surrounding
area, including a partial evacuation of the nearby towns of Todmorden and Summit.
The fire was not considered to be under control until the evening of 24 December, and it
was sealed off until 27 December. Later inspections found that the great heat generated
and sustained over such a long period by the fire had caused some spalling of brickwork
up to two rings in depth, and vitrification and partial melting of some brickwork surfaces.
Later analysis indicated that the mean temperature in the region of the fire was around
1300°C, with evidence that up to 1530°C was achieved locally (tests indicated that the
melting point of the bricks was about 1250°C).
Fortunately, the effects of the fire were confined to the train, tunnel and associated rail
infrastructure, and there were no injuries or fatalities.
Considerable problems faced the teams tasked with rehabilitating the line. Tank wagons
had been badly damaged in the incident and the confines of the tunnel hampered the
salvage crews as did the emergence of petroleum vapours from the ballast. The last vehicle
was removed from the tunnel on 1 March 1985 and the line was not reopened to traffic
until 19 August 1985, 34 weeks after its closure. Amazingly, the damage done to the
tunnel was relatively minimal and localised with about half a mile of track had to be
replaced, as did all the electrical services and signalling. The bases of the two nearest
construction shafts were shored up and filled with inert foam. One of the most remarkable
observations was how well the brick lining had stood up to the fire.
Figure A1.75
374
Superheated, fuel-rich gases combust as oxygen becomes available at the top of the shafts –
the plumes of fire reached 50 m above ground level and caused closure of the local A-road
Figure A1.76
Investigators stand amid the twisted wreckage of one of the wagons
18.4
DUCKMANTON TUNNEL COLLAPSE
Duckmanton Tunnel was a masonry-lined rail tunnel located on the closed Great Central
line between Chesterfield and Lincoln, in an area subject to the effects of mining
subsidence such that the rock mass in certain areas is fragmented and weak. In 1959 a
volume of rock burst through the masonry crown of the tunnel and fell to the trackless
tunnel floor. A sizeable annular void was visible behind the tunnel crown, with support
timbering still in place. No one was hurt due to the collapse and the tunnel has now been
filled in. The surrounding strata were extensively fractured, so rockfall was probably
precipitated by the effects of mining subsidence.
Figure A1.77
Rubble fills the tunnel below the collapsed area of the tunnel crown, revealing a void
behind the lining with construction timbers still in place
CIRIA C671 • Tunnels 2009
375
18.5
PENMANSHIEL TUNNEL COLLAPSE
The lining of Penmanshiel Tunnel collapsed on 17 March 1979 resulting in the loss of
three lives. Penmanshiel Tunnel was located on the East Coast Main Line 18 miles northwest of Berwick-on-Tweed and was 244 m, containing a double line of railway. The single
bore tunnel was driven in the 1830s through a formation of sedimentary rock. The lining
was a brick arch consisting of six rings of brickwork springing from brickwork sidewalls
that extended to the original track level. The local geology was vertically bedded
greywacke. The arch ring was not in contact with the overlying rock and many voids
existed with one cavity, at the northern end, with a clearance of 9 m from the lining
extrados to the face of the rock. Brick piers were present supporting the lining against the
cut tunnel. The tunnel was generally dry except near the portals.
The collapse occurred during major works to lower the track bed to increase the effective
headroom. The rock floor was being lowered by between 1000 mm to 1700 mm, and the
track and ballast was being replaced by paved concrete track. The two tracks were dealt
with separately, the work on the up line had been completed and lowering of the rock
floor on the down side had been undertaken in readiness for track installation. At the time
of the accident, work was underway trimming the rock floor with hydraulic picks and
other plant machinery. At 3.45 am the railway works inspector noticed pieces of rock
bursting off the vertical face 200 mm below the springing of the brickwork. As he made his
way to make arrangements to shore up the tunnel, it then collapsed without warning.
About 30 m of the tunnel, 90 m from the south portal, had completely filled with rock,
engulfing a dumper and a JCB and both their operators. It was estimated that 20 m of the
arch ring had collapsed. Each face of the fall displayed a scree of rock, the largest size
being 150 mm.
The previous annual examinations of the tunnel, carried out according to British Railways
normal practice, noted no major defects with the brick lining apart from surface spalling
on the innermost ring and some loss of mortar. There had been no serious deterioration
in the rock walls and no serious distortion or movement in the arch ring. All preparation
and preliminary investigation that took place before the lowering of the tunnel floor
revealed nothing untoward and was completed to the required standard. Similarly the
work that had been carried out on the floor lowering had gone according to plan before
the collapse.
A comprehensive site investigation was prevented as it was deemed unsafe to remove the
debris as this may have resulted in a complete collapse of the tunnel. However an
investigation of the brickwork revealed several inconsistencies in the construction of the
tunnel. In many places the arch ring was found to taper down to the side walls, varying in
thickness between six and 34 inches. The annular void was found to be full of loose rock.
Large cavities were found behind the lining crown, originating from the tunnel’s
construction where tree trunks had been placed between the back of the brickwork arch
and the excavated face of the vertically bedded rock. Over the years the timber had rotted
away and the rock had settled onto the top of the arch. The nature of the rock was such
that it unravelled upwards, putting a surcharge on the top of the brick arch. Broken rock
was discovered overlying the lining to a depth of eight feet in places. Later it was found
that there were narrow cavities/voids in the side walls as well.
Following the collapse, engineers were able to get a good appreciation of the geological
bedding in the hill above the tunnel, which showed a complex anticlinal structure
comprising sharply folded beds of rock with a central core of shattered and sheared rock.
This feature appeared to intersect the line of the tunnel at the collapse site. So the area of
collapse would have vertical bedding planes hanging over the arch ring, which seemed to
have failed causing the lining to collapse.
376
Although the collapse was foreseeable, and one BR employee had expressed deep concern
over the effect of the works, no evaluation had been carried out following the engineering
excavation on the existing tunnel structure. British Rail pleaded guilty in the High Court
in Edinburgh to a case brought under Health and Safety at Work Act Sections 2 and 3. A
plea of not guilty by the company involved in construction was accepted by the Lord
Advocate, appearing in court on behalf of the Crown. The old tunnel was consecrated as a
burial ground and abandoned. The line was then re-routed through a cutting 40 m from
the tunnel itself.
The incident led to a review of the then BR practice of inspecting tunnels.
18.6
BLACK BOY TUNNEL COLLAPSE
On 25 November 1865 a section of Black Boy Tunnel near Exeter collapsed on to the line,
and a passenger train later ran in to. The tunnel is 200 m in length and situated one mile
east of Exeter Central station. It was dug through a geology consisting of red brick earth,
marl and red sandstone with little water present. It is lined throughout with five rings of
brickwork. The tunnel was opened to traffic in 1860.
The first warning of an incident was when the 7.15 am Yeovil to Exeter train entered the
tunnel at a moderate pace and ran into a mass of earth, rock and bricks lying on the down
side rails. There were no major injuries and the passengers disembarked the derailed
service. Shortly after another section of the tunnel gave way and crushed one of the
carriages. The original fall occurred 90 m from the east end of the tunnel over a section
21 m long.
Immediate action was taken to prop up the tunnel at the site of the opening in the roof.
The opening was boarded up to prevent more material falling on to the track. The
engineers discovered that a cavern, about six metres in height from the crown, had
opened up in the roof from the where the collapse occurred.
A large piece of rock was found at the base of the infilled material. After the original
construction of the tunnel the workmen had failed to fill in the space created above the
arch when the draw logs were removed. It was deduced that the lump of rock had fallen
on to the arch ring, from a height created by the timber support removal, affecting the
lining on the extrados and causing the collapse. There appeared to be no water providing
lubrication for the ground movements and the brickwork was in sound condition.
18.7
BOROUGH STATION TUNNEL COLLAPSE
Since August 1922 progress has been made to widen the tunnels at Borough station on
the Northern line of the London Underground, which was to accommodate updated
rolling stock. The work consisted of removing the existing cast iron segmental lining,
excavating the surrounding clay and constructing the new lining to form the enlarged
tunnel. The work was being carried at night around running services. At the end of each
shift there would be a temporary joint constructed between the old and new linings as the
work progressed down the tunnel. This joint was secured with timbering. It was the
failure of one of these joints that caused the collapse and the later subsidence that
occurred above ground. The event occurred about 120 m south of Borough station. A
passing locomotive struck some dislodged timbering resulting in 650 tonnes of coarse
gravel and sand entering the tunnel. This influx of material caused a cavity 14 m diameter
and 4.5 m deep below Newington Causeway, which is 12 m above the roof of the tunnel.
The creation of this cavity caused severe fracturing of services underlying the roadway,
specifically a gas main that filled the cavity with gas. This was ignited causing a violent
CIRIA C671 • Tunnels 2009
377
explosion. However, shortly afterwards a fractured water main flooded the cavity and
extinguished the fire. The surface subsidence occurred 10.5 m away from a bridge
carrying four lines of the Southern railway over the road. No apparent damage to the
bridge was noticed.
The tunnel lining was composed of rings 432 mm long built of six segments and a key
piece. Each piece was 305 mm thick. The nature of the work required that two rings had
to be removed before one could be replaced as the original rings were closed up tightly
under pressure. So there was a two ring gap almost constantly during the renewal. This
space had to be secured by timbering before traffic could pass. At the time of the
subsidence the gap was 990 mm for the length covered by the two top segments and the
key piece. In this case timber had been placed end to end to form a continuous bearing.
Previous exploratory borings (taken in 1917) had found that the surrounding material
was generally good, dry clay. The exception to this was between ring numbers 1013 and
1015 where water had been seen pouring in at the crown. However later borings did not
discover anything like the same amount of water.
Preliminary evidence from the site of the collapse suggested that water was present and
that immediately after the fall it was noticed flowing in considerable quantity. A further
examination approximately half an hour later found that the tunnel had been completely
filled by ballast, sand and half-buried timbers.
From the beginning of the works the material at the face had always been good, dry clay.
However the previous night water was seen trickling through the joints at the accident
site. All evidence seems to suggest that the water was being bled from a stone found
lodged at the site as the quantity was so small and the phenomenon was dismissed. It was
the general conclusion that the collapse occurred due to insufficient clay cover under
water bearing gravel, this being aggravated by the water leaking through and the
vibration of the passing trains. A contributing factor was the insufficient thickness and
bearing of the temporary poling boards at the gap between the iron rings.
18.8
COFTON TUNNEL COLLAPSE
On 11 May 1928 a section of Cofton Tunnel, on the Birmingham to Gloucester section of
the London Midland Scottish Railway, which was in the process of being removed to form
a cutting, collapsed resulting in three injuries and the loss of four lives. The tunnel was
situated seven miles south of Birmingham and was constructed in 1838 1841. It was 402 m
long and was built of red brick and varied in thickness between four to six rings. At the
time of the works the brickwork was in good condition although most of the inner ring
had been replaced by blue brickwork in the course of maintenance. The tunnel was damp
and after the incident it was discovered that some of the inner brickwork was in
deteriorated condition because of this. To convert the tunnel into a cutting the overlying
material was removed. Windows were then cut into the brickwork using pneumatic
hammers. The final demolition would then take place using explosives to blow out the
remaining material in between the windows. The workers at number eight section were
completing the cutting of the windows when the collapse happened.
It appears that this collapse can be attributed to the structure of the tunnel being less
sound in places than was originally thought and it was possible that the brickwork was not
bearing uniformly on the springings. It was also excessively weakened by the cutting of
the windows. It was decided that this was a case of bad practice even though the method
of weakening before demolition had been used previously.
378
18.9
EFFECTS OF THE COLLAPSE OF A DISUSED TUNNEL ON
THE CANTERBURY WHITSTABLE RAILWAY
This account is based on a 1975 report by Harris and Sutherland (consulting civil engineers)
and Prof A W Bishop of Imperial University London for the University of Kent, which is not
available in the public domain. We are grateful to the University of Kent for its co-operation
in allowing us access to the information for the purpose of this publication.
Summary
In July 1974 the collapse of a railway tunnel on the disused Canterbury Whitstable railway
line resulted in rapid subsidence, severe structural damage and the near-collapse of a
recently constructed two-storey building, part of the University of Kent, located directly
above it. This incident illustrates the importance of adequately maintaining old tunnels
whether in use or disused, how quickly serious structural problems can develop, and the
potential effects of collapse at ground level, even where the tunnel is located at significant
depth.
18.9.1
History of the tunnel
Construction of the historic six mile Canterbury Whitstable line, one of the first passenger
railways in the world, took several years of arduous digging and preparation before it was
opened in May 1827. In particular, excavation of the 770 m Tyler Hill Tunnel proved
challenging and lengthy. The line carried passengers for just over 100 years until the last
passenger train ran on 1 January 1931. It continued to be used for goods traffic until its
final closure on 1 December 1952, and the track was removed. In 1963 it was sold by
British Railways (Southern Region) to the University of Kent.
The tunnel passed through London Clay at a depth of about 20 m below the University’s
campus buildings. Its masonry lining had a horseshoe profile and comprised four courses
of brickwork (18” or 457 mm). Inspection records indicate that the bonding patterns
varied erratically from place to place, and that the quality of workmanship and materials
was highly variable.
While in the University of Kent’s ownership the tunnel continued to be inspected at
regular intervals by their own staff and by engineering consultants, with nothing unduly
disturbing being noted until 1973.
18.9.2
Tunnel collapse and subsidence
In July 1973 cracks were noted in a link bridge between two campus buildings and
settlement identified as a potential cause. In November of the same year, a detailed
inspection was carried out by the University of Kent’s consultants, who identified more
severe deterioration (cracking and scaling of the brickwork) developing in the tunnel
directly beneath the affected building (the Cornwallis building). It was recognised that the
situation could not be left to develop further and required early attention. However, while
a contract to repair and strengthen the lining using sprayed concrete was being
negotiated, the situation developed rapidly towards a dramatic conclusion.
In late April 1974 a further inspection of the tunnel identified deformation of the
brickwork lining beneath the Cornwallis buildings. Thereafter, throughout May and June,
the tunnel lining continued to deform. Accompanied by several small localised falls of
brickwork, the sidewalls began to move progressively inward and eventually breached,
with some clay entering the tunnel. Heave of the invert was noted. Meanwhile the
cracking in the university buildings continued to worsen over a period of a few weeks until
CIRIA C671 • Tunnels 2009
379
on 11 July the sudden complete collapse of a length of the tunnel directly below the
buildings resulted in a final dramatic settlement of almost 700 mm occurring practically
overnight. The buildings, which had already been evacuated and valuable equipment
removed, were severely damaged. One block was in danger of collapse and had to be
disconnected from adjacent buildings and demolished. Others had their foundations
supported on jacks until emergency ground stabilisation works could be carried out.
Several services, including water mains, electricity and a foul sewer, were seriously affected
over a wide area and required support or re-routing. In the course of the remedial
measures a significant proportion of the remaining tunnel was filled with grout, working
from the ground surface.
A 30 m length of the tunnel was blocked by fallen material, which comprised a voided
mixture of brick, clay and sand. Boreholes from the surface found that the subsided clay
above the tunnel was substantially intact with fissures and shear planes but no actual voids.
At the ground surface the subsidence created a bowl-shaped area of about 750 m², with a
maximum settlement of 700 mm and an estimated volume of 154 m³. The original volume
of the collapsed area of the tunnel was about 348 m³.
18.9.3
Mechanism of subsidence
The recorded sequence of events in both the Cornwallis building and the tunnel below
indicated that the subsidence was because of, and followed, the failure of a section of
lining from chainage 240 to 270. The Harris and Sutherland report identified the
subsidence as resulting from breaches in the lining that allowed the ingress of clay into the
tunnel and development of a cavity above it. When the size of the cavity reached a critical
value the clay above failed as a plug in undrained shear, as illustrated in Figure A1.78.
Figure A1.78
380
Postulated subsidence mechanism showing clay plug failing in undrained
shear and relative locations of the tunnel and the Cornwallis building
There was no evidence from the soils investigation to suggest that the subsidence was
caused by a cavity progressively caving in and rising to the surface. Calculations indicated
that failure of the clay as a single plug in undrained shear was possible at a critical
condition equivalent to a 12 m length of collapsed tunnel.
18.9.4
Mechanism of failure of the tunnel lining
The Harris and Sutherland report concluded that the tunnel failure occurred because of
the breakdown of the arch action in the walls leading to failure of the brickwork in
bending and breaches. The available evidence indicated that the breakdown of the arch
action in the walls was probably caused by failure in the invert. The postulated mechanism
for the failure is shown in Figure A1.79
Figure A1.79
Possible collapse mechanism of tunnel
The most likely mechanism is: first the invert failed and this resulted in removal of the
effective haunch to the sidewall arches of the tunnel, which then deformed under the load
of the clay behind and finally failed in bending. This led to ingress of clay into the tunnel
and the formation and growth of cavities, which precipitated the subsidence of the clay
plug.
CIRIA C671 • Tunnels 2009
381
It was not possible to establish clear reasons for the initial failure of the invert. The most
likely explanation is that its bricks were imperceptibly weakened because of their location
beneath the invert fill, creating a weak link in the arch system of the lining.
Several factors were considered but rejected:
18.10
scaling of the brickwork in the sidewalls and arch, although serious in places, was not
thought to have been a direct cause of the lining failure because there was clear
evidence that the sidewalls failed in bending. However, the reduction in the crosssectional area of the brickwork may have caused local weakness
a theory that pre-existing backs in the clay moved under the load of the building
above was thought difficult to prove or disprove. The principal argument against this
theory was that if a previous failure surface or back had been reactivated by the
weight of the Cornwallis building and this had caused the brick lining to fail, then it
would be expected that this movement would accelerate as the resisting force from
the lining reduced. In fact after the first breach of the tunnel lining, a cavity is known
to have existed for eight days before the major subsidence occurred. So it was
considered that this theory could not be substantiated
other factors, including a reported WWII bomb and leaking service pipes above the
tunnel, were not thought to have contributed to the failure. Likewise tectonic
movements and unfilled cavities in the clay surrounding the tunnel were discounted.
GROUTING AND GROUNDWATER POLLUTION AT
HALLANDSÅS TUNNEL, SWEDEN AND ROMERIKSPORTEN,
NORWAY
Based on information in the report produced by Risk & Policy Analysts Ltd for the
Department of Environment, Transport and Regions (RPA, 2000). Reproduced here by their
kind permission
The need to consider risks to the environment when carrying out works in tunnels
(particularly grouting works using potentially hazardous materials) is demonstrated by
two incidents that occurred in the mid-1990s during the construction of Hallandsås
Tunnel in Sweden and at Romeriksporten in Norway. The incident represents the misuse,
rather than normal use, of grouts, and what can go wrong when insufficient consideration
is given to the potential environmental hazards associated with tunnel works, particularly
grouting.
18.10.1
Hallandsås Tunnel, Sweden
The tunnel through the Hallandsåsen ridge in the south west of Sweden was
commissioned in 1991 to form part of the west-coast rail link between Gothenburg and
Malmö. The project was intended to improve transport between Oslo and Germany. The
ridge forms an important aquifer with considerable use made of the water for vegetable
growing. The area is also of national interest, with high biological diversity and is
protected by law to prevent harm to the natural or historic environment. Tunnel
construction began in 1992 and continued until October 1997.
It had become necessary to seal the tunnel due to the high rates of water flow into it
during construction. Initially, the Rail Administration had requested permission to drain
33 litres per second from the entire tunnel and seal the cracks by injecting concrete into
the tunnel walls. However, this method did not work despite the use of large volumes of
concrete. Indeed, the volumes of water being drained from the tunnel approached 70
382
litres per second. An alternative method of sealing the leak was required and a chemical
grout that contained acrylamide and n-methylolacrylamide (NMA) was chosen. These are
potentially hazardous and carcinogenic constituents, but in normal use when combined
with other components in properly set grout, they do not represent a risk to the
environment.
When only one-third of the tunnel had been grouted it was discovered that water seeping
out of the tunnel had high acrylamide and n-methylolacrylamide (NMA) content. Cows
had become paralysed from drinking the contaminated water and fish had been killed in a
breeding facility on one of the affected water courses. The source of the contamination
was the chemical grout that had been used to seal the tunnel. More than 1400 tonnes of
the grout had been used, which corresponded to an estimated 140 tonnes of acrylamide
and NMA. Acrylamide and NMA were also found in groundwater.
Acrylamide and NMA leaked from the tunnel because it did not set correctly. This was
because:
the groundwater leakage rates were so high that they caused dilution of the
components, which meant that polymerisation was slowed or could not occur
the temperature in the tunnel, less than 9°C, was too low and lay outside of the
application range for the grout
return flow from the process of injecting the grout into the rocks added to the
concentration of acrylamide and NMA in water discharging directly from the tunnel.
Following the discovery that acrylamide and NMA were present in water discharging from
the tunnel, a municipal emergency plan was put into operation. This involved
recommendations as to the risks of drinking or using water supplies and in eating meat
and vegetables produced in the region. There were considerable effects on commercial
activities within the area, particularly on farmers. A total of 370 animals were slaughtered.
Milk suppliers were also affected with 330 000 kg of milk disposed of and nine milk
suppliers excluded from milk collection services. Vegetable and root crops were also
destroyed. Water was analysed from 310 wells in the area, with 29 having detectable levels
of acrylamide and/or NMA. Water was delivered by tanker to households within the risk
zone. The total value of claims paid out because of the incident was SEK 26m (£1.9m or
€3m).
The methods used to apply the grout meant that workers were exposed to acrylamide and
NMA. In total, 20 workers were found to be suffering from effects on the nervous system,
many of which persisted over a period of several years.
The final report of the Tunnel Commission (Tunnelkommissionen, 1998) summarised:
“The problems caused by the use of Rhoca-Gil were the result of a chain of errors
and mistakes. There was no adequate analysis of the risks, and the relevant expertise
was lacking. The information provided by the supplier was incorrect and was not
checked. The safety organisation, company health care services and trade union
representatives were not sufficiently involved. If the risks had been evaluated in
advance and then followed up during the course of the project, these problems
could have been avoided.”
One of the major problems for the contractors was meeting the allowable limits for
drainage of water from the tunnel as set by the Water Rights Court. The levels were set at
such a limit that it is unlikely that grouting could ever have reduced leakage rates
sufficiently. There are also questions as to the adequacy of the environmental impact
CIRIA C671 • Tunnels 2009
383
assessment. Rather than use the exercise to minimise effects on the environment, the
National Rail Administration described the problems as challenges and used the study to
justify the project. No alternative approaches were considered.
Following the decision to use the chemical grout, the County Administrative Board and
the Municipality (the supervisory bodies for the project) decided that there was a need to
consider the risks in greater detail. However, they did not contact the National Chemicals
Inspectorate for another three weeks. Before any decisions could be made as to whether
less hazardous materials should be used, streams, wells and groundwater had all been
polluted with acrylamide and NMA. Also, the National Rail Administration had not
contacted the supervisory bodies until large-scale experiments with the grout had already
begun.
18.10.2
Romeriksporten, Norway
In 1995 similar problems occurred at Romeriksporten in Norway, with tunnelling
contractors attempting to reduce water leakage rates to meet pre-set guidelines. The
problems with groundwater leakage were potentially more serious in Norway because
there were several lakes that were undergoing dramatic decreases in water levels. The
tunnelling operation was behind schedule following disagreements over the value of the
contract to the parties involved with its construction. The high leakage rates put the
tunnelling project even further behind. Grouting of the tunnel was started in an attempt
to prevent loss of water in the lakes. The grouting material used at Romeriksporten was
the same as that used at Hallandsås. Tests were then undertaken to assess whether the
lakes had been contaminated with acrylamide, and these found that there was a
concentration equivalent to 560 times the limit value set by the EU. Contaminated
drinking water was also found.
One of the major concerns in Norway was that cancer producing chemicals could be used
in large volumes without the knowledge of the authorities or local inhabitants contrary to
national legislation. It was not known that the grout contained acrylamide until August
1997 as the information provided by Rhone-Poulenc (the producers) had not stated it. As
the contractors had not previously known that the grout contained acrylamide, workers
had been exposed to it. A statement by one worker indicated that respirators were not
used and that the vapours inside the tunnel were so strong that they suffered from
headaches. One of the workers was also not wearing a hood and water from the boreholes
(probably including the grout) ran down the back of his neck. This is somewhat anecdotal
but highlights that the safety instructions given with the grout were not followed.
Health checks and blood tests were taken from 23 tunnel workers who had been involved
with the injection of the grout plus a further 20 tunnel workers who may have had some
contact with it. Many exhibited symptoms including nausea, dizziness and eczema/skin
irritation, although none of the workers who had been exposed were found to have lasting
health effects when examined two years after the incident. Blood tests taken shortly after
the incident found that three workers did have elevated levels of acrylamide. These
workers had been recently exposed.
Following the incidents at both Hallandsås and Romeriksporten, the manufacturer
stopped production of the chemical grout involved.
Because of the problems at Hallandsås, several major construction companies have
introduced environmental management systems for site operations. These systems should
ensure an increased level of environmental expertise and should clarify the various levels
of responsibility.
384
A2
Sources of existing information
This section contains further information on potential sources of existing information on
tunnels and contact details for some of the principal sources.
A2.1
SOURCES OF HISTORICAL INFORMATION
A2.1.1
Infrastructure-specific sources
Primary sources of information for the main transportation infrastructure owners are
given in Table A2.1.
Table A2.1
Primary sources of infrastructure-specific sources of tunnel information
Enquiries for information on those structures, including tunnels still in use as part of the operational railway
should be directed to Network Rail at regional offices listed in telephone directories.
Railway structures
For tunnels and land forming part of disused parts of branch lines, BRB (Residuary) Ltd is responsible for
management and disposal of assets which were once owned by British Rail and were designated as surplus to
the needs of the national mainline railways, so not acquired by Railtrack (and subsequently Network Rail)
under the 1993 Railways Act. They retain ownership of land, buildings and other structures (including many
tunnels) and hold information relating to these:
BRB (Residuary) Limited
Hudson House, Toft Green, York, YO1 6HP
Tel: 020 7904 5100 (London office)
<http://www.brb.gov.uk/>
A series of Gazeteers, authored by Lawrence Popplewell and published by Medgellen Press, give
comprehensive information on railway line opening dates, contractors, engineers and are an ideal starting
point for data searches at bodies outside Network Rail.
For older structures, the National Railway Museum has a limited amount of infrastructure related information:
National Railway Museum
Leeman Road, York, YO2 4XJ
Tel: 08448 153139
Fax: 01904 611112
<http://www.nrm.org.uk>
Advice for carrying out research is included in the Guide to railway research and sources for local railway
history (Kay, 1990). Enquiries may also be made to local record offices.
Information about private railways and tramways may be available from the same sources or from their own
archived records.
Highway structures
Canal and waterways
structures
Many historical records are held by the National Waterways Museum in Gloucester:
National Waterways Museum
Gloucester Docks
Llanthony Warehouse, The Docks, Gloucester, GL1 2EH
Tel: 01452 318200
Fax: 01452 318202
<http://www.nwm.org.uk/>
Enquiries may also be made to British Waterways regional offices (listed in telephone directories), and to local
record offices.
The construction of most UK road tunnels has been documented in ICE proceedings.
The British Tunnelling Society has a database of UK tunnels that includes basic information on road tunnels.
The database can be accessed from:
<http://www.britishtunnelling.org>
An international directory of road tunnels can be found in the reference section of
<http://www.tunnelbuilder.com/>. This page also includes a tunnel history section that has information on
various UK road tunnels.
CIRIA C671 • Tunnels 2009
385
A2.1.2
General sources
The potential sources of information on the historic development of UK infrastructure
and, specifically, the construction and operation of tunnels, are many and diverse. An
experienced researcher will be familiar with many of the principal general sources, for
example, the National Archives and the British Library, and many engineers will be
familiar with others such as the ICE library. The following is a brief summary of some of
the principal sources with comments on their potential usefulness to the researcher trying
to obtain historic information on tunnels.
Researchers should also refer to Appendix 1 of the IStructE publication Appraisal of existing
structures (IStructE, 1996), which includes a comprehensive list of sources for construction
and engineering related information.
The National Archives and the Catalogue
The National Archives, which covers the United Kingdom, was formed in 2003 by
bringing together the Public Record Office and the Historical Manuscripts Commission. It
is responsible for managing the records of central government and the courts of law, and
ensuring they are accessible to everyone. The collection is one of the largest in the world
and spans an unbroken period from the 11th century to the present day. The catalogue
(formerly PROCAT) is often a useful source. Catalogue reference MT contains records
created or inherited by the transport departments and of related bodies, and of the
London Passenger Transport Board. This contains several sub-divisions of records
inherited and created by the Ministry of Transport and successors, Railways, Inland
Waterways, and Ports and Harbours Divisions. For example, MT29/1 Railway Inspectorate:
inspectors reports contains the Board of Trade reports of inspections carried out between
1840 and 1910 as well as details of slips, accidents etc.
The catalogue is searchable online <http://www.nationalarchives.gov.uk> and documents
can be reserved for viewing in advance, at whichever location they are currently held. A
reader’s ticket is required, this is issued free of charge but requires a personal visit with
suitable identification.
Records are also held at the Public Records Office of Northern Ireland (PRONI) and the
National Archives of Scotland (details in Table A2.2).
Libraries and map collections
386
the British Library Newspaper Collection is an excellent source of information,
particularly the Colindale Newspaper library in London – see newspaper records
the British Library (in particular its map library and its newspaper library at
Colindale, see below) and the ICE Library and archives are useful for specific
references. The ICE Proceedings are available online (for a fee) and they include
papers on specific tunnels. Non-members of the ICE can be introduced by members,
and then use the library at will
also at the ICE, the Panel for Historical Engineering Works (PHEW) has also
generated documents called HEWs (Historic Engineering Works), which include
information on certain tunnels
university libraries are a good source of information relating to local engineers, for
example, the Brunel collection held at Bristol University, and the Goldsmith
Collection and the Rastrick Collection at Senate House Library (University of
London)
museum libraries are also worth a mention, in particular the Science Museum Library.
The Science Museum also houses the transactions of the Newcomen Society, which
include many potentially relevant articles (available to visitors by arrangement or
online for a fee). The Bodlean library (Oxford University) has one of the most
extensive collections of Ordnance Survey maps available, but in the past has not
proven to be a good source for other types of information
there are a variety of picture and map collections that can yield useful information.
For further details, including equivalent sources of information in other parts of the UK,
see Table A2.2.
Newspaper records
The British Library Newspaper Library at Colindale holds the national archive collection
of British and overseas newspapers. It is part of the national library of the UK and is, as
such, a research library not a public reference library. The library’s collection of
newspapers and periodicals is held in closed access storage and is not available for
browsing. It is necessary to identify the titles, months and years of the items that are to be
consulted.
The vast majority of newspapers have no subject index, which makes specific searches
difficult. There is an electronic database of publications available which uses a keyword
search facility. It is possible to search by region, town or using specific subjects. This can
produce a long list of publications that can be simplified using search criteria such as
opening date, or date of the start of construction that may be up to two years, and
sometimes more, before opening.
As there is no subject index, searching the newspaper archives is a very time consuming
process, but events connected with tunnel construction, such as opening ceremonies, start
of construction and accidents, were often reported in great detail.
The Internet
The internet is the most useful access, as most libraries have online catalogues for
advanced ordering. Also groups such as caving groups, local industrial archaeology
groups, and others have archives which can often be viewed and downloaded. Particular
caution should be taken to verify uncontrolled information from the public domain.
Note: Many of these and other potential sources of historical information are included in
Table A2.2, along with their contact details.
A2.2
SOURCES OF GEOLOGICAL AND HYDROGEOLOGICAL
INFORMATION
Potential sources of geological, hydrogeological and other information are included in
Table A2.3, along with their contact details.
A review of sources of information for geological site investigation is included in Perry and
West (1996).
The main sources for geological information are the British Geological Society (BGS) and
the Geological Society (GeolSoc) geological maps. Geological maps provide an instant
source for identifying the geology of an area associated with a tunnel, including drift
(superficial) and solid formations, large structural features and the dip of the strata. The
CIRIA C671 • Tunnels 2009
387
most useful maps are at a scale of 1:10 000 and 1:10 560, depending on the date of
mapping, as these often show geological information associated with tunnels and shafts.
Some large-scale maps are available for certain areas. Note that at the scales mentioned
above many smaller scale features can be missed, so it is important, from a geological and
geomorphological point of view to carry out a site walkover survey (see Section A3.5).
The GeolSoc is located in Piccadilly, London and holds a large selection of geological
maps, books and periodicals. The collection is available to Fellows of the Geological
Society, although non-members can use the library by prior arrangement but a charge for
using the library may be made.
Hydrogeological maps may be purchased through the BGS website or from their libraries
at Keyworth or Edinburgh. The BGS holds many public borehole records that have been
drilled either on behalf of the survey or obtained from site investigations carried out by or
for other public bodies and local government, and these can be ordered over the internet.
Enquiries relating to geophysical and geochemical information in specific areas can also be
made online.
Understanding the groundwater regime associated with the tunnel is very important,
because this can affect the tunnel in many ways and is often a primary influence on
deterioration and structural distress. The two principal sources of information relating to
groundwater are:
1
Groundwater vulnerability maps: these are used to identify whether the site of
interest is located above a major, minor or unclassified aquifer and how vulnerable the
soils above the aquifer are to leaching (transferring contaminants from above). These
are published at a scale of 1:100 000 and at present only cover England and Wales.
2
Hydrogeological maps: these provide information associated with geology, surface
water, groundwater and artificial features (such as wells and pumping stations). They
can also provide information on groundwater levels (potentiometric surface), relief
and average annual rainfalls for the region and hydrochemistry. These maps are only
produced at a scale of 1:125 000 and are again produced mainly for England and
Wales. A large scale 1:625 000 map of England and Wales is available although details
provided are for generally large scale features only.
The Environment Agency (EA) provides hydrological maps. On their website details such
as areas prone to flooding and other relevant information can be identified for England
and Wales. If the information required is not available on the internet then the regional
office covering the area of interest will need to be contacted, a charge may be levied
depending on the information required. In Scotland the Scottish Environmental Protection
Agency (SEPA) should be contacted to obtain similar information.
Information on historic mine works is available from The Mining Records Office
(telephone number 01623 637 000), which is sited at the Coal Authority headquarters in
Mansfield. Within the Mining Records Office the Authority currently holds three main
sets of records, these being:
388
1
Coal abandonment plans.
2
The Coal Holdings Register and associated records (dealing with the transfer in
ownership of coal before the nationalisation of the coal industry on the 1 January
1947).
3
Licence Register (information on all current licences).
The coal abandonment plans are most likely to be of use for tunnel researches. These have
been required to be deposited since 1872, and the collection numbers in excess of 100 000
plans. All plans have been microfilmed, and while copying of the original plans is
prohibited, due to possible damage to the plans, prints can be supplied from the
microfilms for a modest fee. Inspection of the coal abandonment plans requires an
appointment to be made in advance but is free of charge subject to a maximum visit of half
a day per working week. Visits in excess of this incur a charge.
A2.3
AERIAL PHOTOGRAPHS
Aerial photographs are a useful source of information and are able to provide historical
information on the location and construction of a tunnel and associated features (ie shafts,
portals). An important use of these photographs is their contribution to identifying
features of engineering significance such as soil type, drainage conditions and marshy
areas, unstable ground, shaft locations, subsidence etc. For best results the photographs
should be examined alongside the relevant Ordnance Survey, geological, and other maps
of the same period. Aerial photographs can be taken either vertically or obliquely (at an
angle).
Recently geographical information systems (GIS) have been developed which permit the
merging of mapped and photographed information. This technique is becoming more
popular as this information can be web-based allowing people to access the information
from any location if required.
As well as several private companies that offer aerial photographic record services, The
National Monuments Record Centre, located in Swindon, holds over 600 000 oblique
aerial photographs and three million vertical aerial photographs ranging in date of origin
from between 1940 to 1984, and allows visitors to inspect these free of charge.
A2.4
UTILITIES AND SERVICES
Identifying the location of utilities (gas, water, electricity and telecommunications) close to
the tunnels can prove invaluable and should be carried out at the earliest opportunity.
Utilities can affect the type of investigation, remedial or inspection work carried out and
be identified as potential areas of concern eg leaking water mains, fire hazard for gas.
Unfortunately it can sometimes be very difficult to locate and identify services associated
with a site, particularly from plans provided by the utility companies themselves. It may be
possible to gain a better idea of the exact location of certain utilities by visiting the site in
question and looking for physical evidence such as drain covers and monitoring points.
There are several specialist companies that can search for buried utilities using nondestructive techniques such as radar if knowledge of precise location is critical.
Alternatively, careful excavation may be carried out.
A2.5
WALKOVER SURVEY
Once all the documentary information regarding the tunnel’s history and associated
geology has been gathered a site walkover is recommended to see if features identified
from the data-gathering exercise can be confirmed.
A walkover survey involves an inspection of the area surrounding the tunnel in
conjunction with the examination of local records concerning the tunnel. Access to the
internal parts of the tunnel may require special arrangements, for example, suspension of
CIRIA C671 • Tunnels 2009
389
traffic or drainage of water/sewer service tunnels, or may be programmed to take place in
periods when disruption can be avoided or minimised. It is particularly important to
determine if there are any external geological/geotechnical constraints that could affect
the stability of the tunnel. It is essential that all the information concerning the site is
studied thoroughly before carrying out the walkover survey.
Features that should be identified during a walkover survey include:
1
Land-use: an understanding of land-use above the tunnel is very important. Asset
owners may have their own policies on this, for example, Network Rail Standards
require that a list containing details of all the landowners and tenants within the zone
of influence of the tunnel has to be held by Network Rail. If a shaft is identified then
the relevant landowner(s) should be informed. The land registry can be approached
to assist with identifying land ownership.
2
Trees: tree roots can add extra loads to tunnel linings and also provide pathways for
water ingress if the lining is breached. If the tunnel is located in a clay deposit,
desiccation could affect the tunnel lining. The location and types of trees present
should be recorded. Reference to the National House Building Council standards
(NHBC, 2003) should be made to determine if the trees are high water demand
(leading to desiccation of the soil).
3
Buildings: buildings can show evidence of movement by the appearance of cracks in
the structure, indicating possible ground settlement.
4
Sources of contamination: the location of any possible sources of contamination
should also be identified such as flammable liquid storage tanks, which could pose a
threat to the use of the tunnel.
5
Standing water/marshy ground/springs can indicate groundwater close to surface
and/or poor drainage.
6
Unstable ground: evidence of scree slopes or slips above the tunnel or in any
surrounding cuttings could indicate that material is moving in towards the tunnel.
7
Joint sets, faults and swallow-holes: if the tunnel is in an area where rock outcrops at
surface a preliminary rock mass rating (RMR) can be determined from these
exposures. An indication of the internal angle of shear resistance can be obtained
from tilt tests and point load tests can give an indication of strength. This will give an
indication on dominant joint sets present in the area, which depending on orientation
could affect the stability of the tunnel. If the tunnel is in an area known to suffer from
natural subsidence, the location of swallow holes and fissures should be identified.
8
Mine workings: location of any mining or quarrying activity around the tunnel
should be noted so that issues such as mining-related subsidence and potential
damage to the lining due to blast vibrations can be addressed.
9
Construction work: construction work around the tunnel may affect the tunnel itself
depending on the intensity of the works. The location of these works should be
identified so that an assessment of the likely effects can be carried out. New works can
affect the groundwater regime both during and after the construction period and
they may, as a result, affect the integrity of the existing tunnel.
10 Accessibility: if the potential for further work is identified then all access routes to the
area above the tunnel should be noted.
390
Table A2.2
Sources of historical information – contact details
<http://www.google.co.uk>
<http://www.ask.co.uk>
<http://www.lycos.co.uk>
Search engine websites:
Archives of historical information
The British Library (map collection)
St Pancras
96 Euston Road
London
NW1 2DB
Tel:
Fax:
Email:
Website:
+44 (0)20 7412 7702
+44 (0)20 7412 7780
<maps@bl.uk>
<http://www.bl.uk>
Bodleian Library
Broad Street
Oxford
OX1 3BG
Tel:
Fax:
Email:
Website:
+44 (0)1865 277013
+44 (0)1865 277139
<maps@bodley.ox.ac.uk>
<http://www.bodley.ox.ac.uk>
Pictures and Maps Collection
National Library of Wales
Aberystwyth
Ceredigion
SY23 3BU
Wales, UK
Tel:
+44 (0)1970 632800
Fax:
+44 (0)1970 615709
Address your fax for the attention of the Pictures and Maps
Collection
Email:
<holi@llgc.org.uk>
Website: <http://www.llgc.org.uk>
Map Library
National Library of Scotland
Causewayside Building
33 Salisbury Place
Edinburgh
EH9 1SL
Scotland, UK
Tel:
Fax:
Email:
Website:
+44 (0)131 623 3970
+44 (0)131 623 3971
<maps@nls.uk>
<http://www.nls.uk>
Landmark Information Group Ltd
5-7 Abbey Court
Eagle Way
Sowton Industrial Estate
Exeter, Devon
EX2 7HY
Tel:
Fax:
Email:
Website:
+44 (0)1392 441700
+44 (0)1392 441709
<sales@promap.co.uk>
<http://www.landmarkinfo.co.uk>
The Institution of Civil Engineers
One Great George Street
Westminster,
London
SW1P 3AA
Tel:
Fax:
Email:
Website:
+44 (0)20 7222 7722
+44 (0)20 7222 7500
<library@ice.org.uk>
<http://www.ice.org.uk>
The National Archives
Kew, Richmond, Surrey
TW9 4DU
Tel:
Fax:
Email:
Website:
+44 (0)20 8876 3444
+44 (0)20 8392 5286
<enquiry@nationalarchives.gov.uk>
<http://www.nationalarchives.gov.uk>
National Archives of Scotland
Historical Search Room
H M General Register House
2 Princes Street
Edinburgh
EH1 3YY
Scotland, UK
Tel:
Fax:
Email:
Website:
+44 (0)131 535 1334
+44 (0)131 535 1328
<enquiries@nas.gov.uk>
<http://www.nas.gov.uk>
Panel for Historic Engineering Works (PHEW)
One Great George Street
London
SW1P 3AA
Tel:
Fax:
Email:
Website:
+44 (0)20 7665 2250
+44 (0)20 7976 7610
<library@ice.org.uk>
<http://www.ice.org.uk>
Public Records Office of Northern Ireland (PRONI)
66 Balmoral Avenue
Belfast
BT9 6NY
Northern Ireland
Tel:
Fax:
Email:
Website:
+44 (0)28 9025 5905
+44 (0)28 9025 5999
<proni@dcalni.gov.uk>
<http://www.proni.gov.uk>
CIRIA C671 • Tunnels 2009
391
Table A2.3
392
Geological and other sources of information – contact details
BGS Library – Keyworth
British Geological Survey
Keyworth
Nottingham
NG12 5GG
Tel:
Fax:
Email:
Website:
+44 (0)115 936 3205
+44 (0)115-936-3200
<libuser@bgs.ac.uk>
<http://www.bgs.ac.uk>
BGS Library – Edinburgh
British Geological Survey
Murchison House
West Mains Road
Edinburgh
EH9 3LA
Scotland, UK
Tel:
Fax:
Email:
Website:
+44 (0)131 667 1000
+44 (0)131 668 2683
<mhlib@bgs.ac.uk>
<http://www.bgs.ac.uk>
The Geological Society
Burlington House
Piccadilly, London
W1J 0BG
Tel:
Fax:
Email:
Website:
+44 (0)20 7434 9944
+44 (0)20 7439 8975
<enquiries@geolsoc.org.uk>
<http://www.geolsoc.org.uk>
Environment Agency (EA)
Contact regions for addresses. Check website to
obtain details.
Tel:
Email:
Website:
+44 (0)8708 506 506
<enquiries@environment-agency.gov.uk>
<http://www.environment-agency.gov.uk>
Scottish Environmental Protection Agency (SEPA)
SEPA Corporate Office
Erskine Court
Castle Business Park
STIRLING
FK9 4TR
Contact regions for addresses. Check website to obtain details
Tel:
+44 (0)1786 457700
Fax:
+44 (0)1786 446885
Email:
Enquiries form available on website
Website: <http://www.sepa.org.uk
National Soil Resources Institute
Cranfield University
Silsoe, Bedfordshire
MK45 4DT
Tel:
Fax:
Email:
Website:
+44 (0)1525 863000
N/A
<nsri@cranfield.ac.uk>
<http://www.cranfield.ac.uk>
National Monuments Record Centre
English Heritage
Customer Services Department
Freepost WD214
PO Box 569
Swindon, Wiltshire
SN2 2UR
Tel:
Fax:
Email:
Website:
+44 (0)870 333 1181
+44 (0)1793 414926
<customers@english-heritage.org.uk>
<http://www.english-heritage.org.uk>
Customer Contact Centre
Ordnance Survey
Romsey Road
Southampton
SO16 4GU
Tel:
Fax:
Email:
Website:
Fax:
Email
+44 (0)23 8030 5030
+44 (0)23 8079 2615
<customerservices@ordnancesurvey.co.uk>
<http://www.ordnancesurvey.co.uk>
+44 (0)20 7215 0936
Via website: <http://www.bnsc.gov.uk>
The United Kingdom Hydrographic Office
Admiralty Way
Taunton, Somerset
TA1 2DN
Tel:
Fax:
Email:
Website:
+44 (0)1823 337900
+44 (0)1823 284077
<helpdesk@ukho.gov.uk>
<http://www.ukho.gov.uk>
The National Meteorological Library and The
National Meteorological Archive
Fitzroy Road
Exeter, Devon
EX1 3PB
Tel:
Fax:
Email:
Website:
+44 (0)870 900 0100
+44 (0)870 900 5050
<enquiries@metoffice.gov.uk>
<http://www.metoffice.gov.uk>
The Mining Records Office
The Coal Authority
200 Lichfield Lane
Mansfield, Notts
NG18 4RG
Tel:
Fax:
Email:
Website:
+44 (0)1623 637 000
+44 (0)1623 629100
<thecoalauthority@coal.gov.uk>
<http://www.coal.gov.uk/>
A3
Visual inspection procedures and
observations
A3.1
PREPARATION
The inspection should be carried out strictly in accordance with an approved inspection
plan, which should contain all the necessary procedures to allow the inspection to be
carried out efficiently and safely. It should follow a risk assessment, preparation of a
method statement and, where appropriate, a reconnaissance visit. Tunnel inspections will
normally involve several people, both engineers and operatives, and a team leader should
be nominated to direct and control all aspects of the team’s activities (Swannell, 2003).
Advice on preparations for visual inspections, including health and safety and
environmental considerations, is included in Section 4.6.
Inspection techniques are discussed in Section A4.1.
A3.2
OBSERVATION AND RECORDING
The inspection should be carried out systematically and be adequately recorded. Proforma are usually considered to be the most appropriate and efficient means of recording
the results of inspections, although some infrastructure owners, for instance the operators
of the Channel Tunnel Rail Link (CTRL), have successfully adopted the use of hand-held
data-loggers for this purpose.
Visual inspection of the tunnel begins with the portals, where these are present, or the
transition between the tunnel and some other structure (eg stations). Sometimes the
sections of tunnel adjacent to portals are constructed differently from other sections of the
tunnel, for example, they may be cut-and-cover whereas the rest of the tunnel is bored.
There may also be other variations in tunnel construction along its length, for example,
changes from lined to unlined sections, changes in lining type, in intrados profile etc. It
may be necessary to adapt the inspection procedure to look for different types of features
and phenomena in different sections. Frequently changes in tunnel construction mark a
transition in the ground conditions encountered while tunnelling and it is important to
look for evidence of differential settlement at such transitions.
Visual inspection should be carried out on all exposed surfaces of structural elements.
Where defects or changes in condition are found, their location should be noted and their
details recorded in an objective manner, preferably using a pre-defined and standardised
system of classification. Descriptions should preferably be augmented by measurement
and recording of feature dimensions (eg crack length and width, depth and extent of
spalling) and by photography or video recording. Records of the previous inspection
(including photographs, measurements etc) should be carried so that changes in condition
can be discerned while on-site.
There are simple non-destructive or slightly destructive testing techniques that can be
used by inspectors with a minimum of training and specialist equipment. The most
commonly used is hammer tapping to identify and determine the extent of hollow and
CIRIA C671 • Tunnels 2009
393
delaminated areas, but other simple methods are available and might be used to
supplement visual observation where this would be advantageous – either to investigate
features that are not visible to the naked eye or to provide a more objective measure of a
specific parameter such as material condition (see Section A4.1).
A suitable system of spatial reference is essential so that the position of features within the
tunnel can be accurately recorded and easily located during later inspections or
maintenance/repair works (see Section 3.5.3).
The following sections describe the principal features of interest during visual inspections,
relating to defects and changes in the condition and performance of a tunnel’s structural
fabric, and include recommendations for simply classifying and recording them, and for
further investigation, where appropriate.
A3.2.1
Lining type and features
The following is a list of lining features worthy of note in the course of observations. It is
not intended to be exhaustive and other features may be of particular interest in certain
tunnels.
Masonry linings:
the cross-sectional shape and intrados dimensions, and especially any changes in these
the presence of foundations or an invert where this can be ascertained visually
the colour and bond pattern of brickwork should be recorded, along with its apparent
condition and that of the mortar
the spacing of construction joints in brick masonry (see Section A4.1.4), particularly
where there are changes in spacing, can provide important clues as to the hidden
structure of the tunnel, its construction sequence and changes in ground conditions
behind the lining
changes in the pattern of bonding at the tunnel crown can be an indication of an
original construction shaft which has been sealed off at the lining intrados
changes in construction materials, and the presence of repairs, can frequently be
assessed by noting variations in the colour and appearance of bricks, stonework and
mortar.
Segmental concrete linings:
394
cross-sectional shape and intrados dimensions and any changes in these
details of segment dimensions and arrangements, number of segments forming the
complete ring, presence of an intrados and whether this is visible
details of the key between segments and bolting arrangements, particularly missing,
loose or heavily corroded bolts
details of the joints between segments, including the presence and nature of packing
materials, caulking grooves and caulking materials, as well as the condition of
caulking and areas where it is missing or clearly defective (caution should be exercised
because such materials may comprise asbestos materials)
visible changes in the appearance, dimensions or shape of segments or their jointing
arrangements
the presence of any obvious repairs, coatings or finishes.
Metal linings:
A3.2.2
lining material type, ie grey cast iron, expanded grey (flexible) cast iron, spheroidal
graphite iron (SGI), steel linings etc
details of segment and key arrangements, including the number of segments forming
the complete ring and the position of the key relative to the segments within the ring.
The use of special segments should also be recorded
details of the formation of the invert and whether segments in the invert are buried
the extent of circumferential radial bolting noting in particular missing or loose bolts.
This is generally a greater concern around tunnel openings where, as a result of an
engineering assessment, the structural integrity of the lining may depend on the
capacity provided by the bolts
changes in diameter of the linings should be recorded and the materials used for the
formation of the infill or headwalls between varying diameters
whether or not a caulking groove is present and if so, the type and condition of the
caulking. In some cast iron lining designs the caulking groove is allowed to rust. In
other designs an apparent groove would be visible on the intrados but in reality if
formed by the separation of individual segments by packing material
the presence and condition of any packing used between segments including the
packing material. Packing is often used to deflect segments for alignment purposes or
to make up gaps that may form from inconsistent ring building
the presence and condition of any coatings or other protective or insulating materials
presence of repairs, eg filler repairs to cast iron linings, which were carried out in the
19th century using cosmetic filler material largely composed of iron filings and which
can be detected using a magnet.
Cracking
A crack is a linear fracture caused by tensile forces exceeding the tensile strength of the
construction material (whether rock, brick or stone masonry, metal or concrete). Cracks
can occur for a variety of reasons, and it is important that they are properly recorded so
that their potential cause and consequences can be assessed, and further investigation or
remedial action carried out if necessary, particularly where they are associated with
structural distress of the tunnel.
Where they are wide, occur on flat, light coloured and untextured surfaces, or are marked
by corrosion or deposits from water ingress, cracks can be readily visible, but conversely
they can be very difficult to discern where they occur on dark, dirty and irregular
surfaces, for example, on old, soot-covered masonry. Fine cracks on metallic surfaces can
be discerned using liquid dye penetrant testing.
It may be possible to distinguish between recent and longstanding cracking from, for
example, the presence and thickness of waterborne surface deposits and, in metal linings,
of corrosion products, or, in concrete or masonry, the presence of dirt and carbonation
along cracks (requiring on-site testing or preferably petrographic examination of core
samples, see Section A4.2.2).
It is important to record the main features of cracking:
location
type (single or multiple, linear, irregular or interconnected)
CIRIA C671 • Tunnels 2009
395
orientation (longitudinal, transverse or diagonal to tunnel’s axis, horizontal in
sidewalls, or random/meandering)
relationship with other cracks (if present) and with tunnel features (eg services,
changes in section or construction type)
length (distance from beginning to end)
width (the widest opening at the surface, variation along its length and, where
possible to discern, its depth)
displacement (whether there is any offset, indicating relative movement, across the
two sides of the crack)
description (for example, associated features such as spalling, water ingress, surface
deposits, corrosion etc in masonry, whether it follows the mortar joints or passes
through the masonry units)
any visible evidence of previous repair attempts.
Photography is a useful tool for objectively recording information about cracks and
allowing subsequent comparison. Where crack patterns are of interest, but the cracks do
not show up well in photographs taken from the required distance, their positions can be
marked for instance using chalk, wax crayons or spray paint. Crack widths can be
estimated by comparison with an Avonguard-type crack-width gauge (a small card printed
with lines of varying specified thicknesses) or by insertion of the varying thickness blades
of a feeler-gauge, which may also give some idea of what a crack does below its surface.
Note that cracks may appear significantly wider at the intrados surface than they are just a
few millimeters below this, so inexperienced inspectors may overestimate the severity of
cracking, but can be fully investigated by examination of core samples. Simple devices may
be attached to the tunnel intrados to monitor crack movements between inspections (see
Section A4.4).
New cracking in rock-lined tunnels can be difficult to detect since most bodies of in situ
rock already have a system of fractures or joints, which are likely to vary in spacing and
orientation along the tunnel’s length. Also, rock surfaces may be highly irregular making
cracking difficult to discern. Where such ready-made cracks exist, they may react to
changes in tunnel stress state by visible contraction (which may extrude joint filling
material) or dilation (widening joints which may result in cracks that were once closed
becoming open). New cracking may also be evident where it results in spalling and
instability of rock at the tunnel intrados. It is very important to properly record and assess
the significance of such phenomena.
In some tunnels defects may be covered by deposits of dirt or soot. If excessive, this may
need to be removed carefully before inspection to visually inspect and assess, for example,
the conditions of discontinuities.
A3.2.3
Spalling and delamination
Cracking can also occur within the plane of the tunnel’s structural material, ie parallel to
the face of the lining, in which case its presence and extent may not be readily apparent to
visual examination. This delamination cracking is most common in brick linings, where it
causes detachment between adjacent rings of brick, or in reinforced concrete, which can
delaminate in the plane of corroding reinforcement. Delamination can be difficult to
detect, although it may be found using a hammer survey or by observation of
irregularities and bulges in the intrados. Where delamination is found or suspected, the
perimeter of the delaminated area can be marked on the tunnel wall and its extent
recorded. It may be necessary to carry out intrusive coring through the intrados,
accompanied by endoscope inspection, to investigate delamination. Non-destructive
396
methods, such as radar surveys and ultrasonic techniques, may also give useful
information.
Spalling occurs where subsurface delamination cracks cause separation and removal of
parts of the intrados surface. It can result either in crater-like depressions over an
irregular area where the delaminated material has fallen away, or elongated depressions
where the loss of material is along a linear feature such as a crack or joint. The location,
extent and depth of spalling should be recorded, along with any features with which it is
associated (eg cracks, wet patches) or which might give a clue as to its cause. In concrete,
spalling is often associated with exposure of corroded reinforcement. In masonry or in situ
rock, the freshly spalled surfaces are often clean (suggesting frost-attack if near to portals
or shafts) or may exhibit white deposits of sulfates (suggesting possible sulfate attack).
Where the spalled surface is discoloured, this suggests water percolation through a crack
that may have been present for some time. It may be possible to assess the causes of
cracking and spalling by petrographic assessment of material samples (see Section A4.2.2)
As with cracking, spalling and delamination is important as it can indicate structural
distress, and is potentially hazardous because it can reduce the structural capacity of the
lining, or present a falling hazard if it occurs overhead. So it is necessary to accurately
record its details (by description, measurement and photography) and assess the likely
cause and significance of recent delamination and fresh spalling.
A3.2.4
Bulging and distortion
Minor changes and irregularities in intrados shape are not uncommon in old tunnels,
particularly those constructed from lime mortar masonry which tends to be ductile and,
over long periods of time, may adjust from the shape of its original centering to one that is
better in equilibrium with ground pressures. As a result, it is not uncommon for the lining
of old masonry tunnels to exhibit undulations and minor bulges that are longstanding and
not progressive, frequently without any further evidence of distress such as cracking and
spalling. However, recent or ongoing changes in intrados profile are clearly a cause for
concern and may indicate changes in ground pressure and serious structural problems,
possibly leading to the development of structural instability and, ultimately, collapse.
Because of this it is vitally important to monitor bulges and distortions for signs of
progressive movement and the development of associated deterioration such as cracking
and spalling. Where bulges and distortions occur, other phenomena, such as local water
ingress and changes in its pattern or quantity, should also be monitored.
Whereas old masonry with lime mortar can frequently accommodate an appreciable
degree of distortion without any significant damage, in less ductile materials, such as
concrete and in particular grey cast iron, distortion is likely to result in clearly visible
cracking. A complete loss of tensile strength, resulting in failure before any appreciable
degree of distortion can occur.
The visual identification and assessment of bulges and distortions in their early stages may
be problematic, and the subjective nature of visual assessment makes it difficult to
determine whether they have worsened between inspections. Modern techniques such as
laser-profiling provide a solution here, because they can be used to rapidly capture an
accurate profile of the whole tunnel intrados and identify areas which require special
attention during inspections. Also, repeat profiles may be compared to identify apparent
changes. Discrete bulges may also monitored by accurate surveying or the installation of
suitable instrumentation.
CIRIA C671 • Tunnels 2009
397
A3.2.5
Material condition and degradation
Deterioration of metals
Corrosion is the principal mode of deterioration in metallic materials. Corroded iron or
steel varies in colour from dark brown to red. Initially it is fine grained and confined to
the metal surface, but becomes flaky or scaly in character as it progresses deeper and
causes pitting and loss of section. The location, extent, appearance and depth of corrosion
should be recorded during visual inspection, and may be discerned using a simple
classification scheme such as:
Minor
Light surface pitting and discolouration, confined to the metal surface and
making it slightly rough and irregular.
Moderate
A definitely discernible area of loose corrosion product with the formation
of scaling and flaking.
Severe
A heavy, stratified corrosion or scaling and pitting of the surface resulting in
loss of section.
When assessing loss of section in metallic linings, it is important to appreciate that
corrosion products may expand to between five and 10 times the original volume of the
metal, growing outwards from the corroding surface, so that a considerable amount of
rust can be generated from a relatively modest loss of section. For an accurate assessment,
corrosion products should be removed, for example, by mechanical cleaning methods
such as grit or shot-blasting (Bussell, 1997).
Factors including the type, hardness and residual thickness of metallic linings can be
assessed on-site using a variety of non-destructive testing methods including scratch tests,
radiography, ultrasonics and metallographic examination. Cracking can be discerned by
the use of dye penetrant testing or by the presence of obvious water ingress or associated
surface deposits.
Where graphitisation corrosion (described in Section 2.6.2.2) has occurred there may not
be any obvious metal wastage. A sharp metal scraper can be used to help locate any areas
of graphitic attack although it will be necessary to blast clean surfaces to determine the full
extent of any attack and accompanying reduction in wall thickness.
Deterioration of masonry
The most common form of deterioration in masonry materials is softening and loss of
material from the mortar joints. Where loss of mortar results in deep joints this reduces
the structurally effective cross-section of the lining and can cause bricks to loosen
(particularly dangerous in the crown area). In the course of a visual inspection, areas with
deep joints (mortar loss greater than 20 mm) should be recorded, along with an estimate
of the area affected and the typical and maximum depth of joint loss, and particularly the
location and extent of any potentially loose masonry.
Cracking and spalling of the masonry units themselves has already been discussed in
Sections A3.2.2 and A3.2.3 respectively.
The presence of surface deposits, encrustations and efflorescences, such as the typical
hard whitish deposits of carbonate which are commonly found on masonry surfaces which
have been subject to the passage of water, should also be recorded.
398
Figure A3.1
Extensive whitish surface encrustations of carbonate-minerals (typically calcite)
on a masonry tunnel lining – these have been leached out of the mortar by water
seepage and gradually deposited on the lining surface
Stone, brick and mortar hardness can be estimated by scratching or chipping using a
sharp metal object, which gives clues about the type of materials present (eg soft brick or
hard engineering brick, soft lime mortar or hard cement mortar) and their condition.
Semi-quantitative evaluations may be made (see Ferro, 1980) and can be useful in making
relative assessments between different areas, monitoring changes in condition over time,
or in specifying repairs.
Deterioration of concrete
The causes of concrete deterioration are discussed in Section 2.6.2.3. The most common
cause of deterioration in reinforced concrete is reinforcement corrosion, the visible
symptoms of which are typically cracking and spalling (see Sections A3.2.2 and A3.2.3)
which may be accompanied by reddish staining from corrosion products. Spalling may
also be caused by freeze-thaw damage of saturated concrete or by structural distress. The
presence and location of any such features should be recorded as well as a description of
features such as their shape and extent (eg crack width, orientation and length, and depth
of spalling) and any associated features such as water ingress, to allow later assessment of
their potential cause and significance.
Softening and disintegration of the concrete may also be seen, sometimes accompanied by
whitish deposits, which suggest the possibility of sulfate attack and again the presence of
any such phenomena should be recorded.
Guidance on assessing fire damage to concrete structures is given in Assessment and
repair of fire-damaged concrete (Concrete Society, 1990).
A3.2.6
Tunnel wetness and water ingress
Since tunnel wetness and water ingress are important causes of deterioration and distress,
it is necessary to record the wetness condition of tunnel materials and the presence,
location, extent and severity of any water ingress. The severity of wetness and water
ingress is difficult to describe objectively in the course of a visual inspection, however a
classification system that has been used successfully in rail tunnels is given in Table A3.1.
CIRIA C671 • Tunnels 2009
399
Table A3.1
Example of descriptive wetness index system
Wetness
index
Descriptors
0
Surface dry (no visible sign of moisture and no wetness to the touch)
1
Surface damp (detectable moisture in material – some darkening/discolouration seen in porous
materials and may feel damp to the touch)
2
Surface wet (thin film or beads of water visible on surface, definite wetness to the touch with moisture
transferred to fingers)
3
Light surface flow (water is flowing down walls in a light trickle or thin film, or is dripping from
overhead areas)
4
Rapid surface flow (water is flowing down walls in a continuous trickle or thick film, or is pouring
continuously from overhead areas)
5
Local high-pressure flow (water spurting or pouring from walls or overhead areas under pressure)
These values have been assigned to each cell of a tunnel survey grid, based on the worst
water ingress present, to indicate zones where water ingress is a particular problem and to
assist with distinguishing patterns of ingress. The semi-quantitative measurements can also
be used to provide a comparative measure of wetness between surveys, either locally or
over the whole tunnel, by comparing numerical averages of wetness measurements.
Figure A3.2
Typical appearance of surface-wet masonry (wetness index of 2 or 3 in accordance
with the classification given in Table A3.1) near to a tunnel portal that is gradually
spalling due to freeze/thaw cycling
Figure A4.1 shows an example of how oblique lighting can be used to clearly delineate
surface-wet areas of tunnel lining, and used with photography for recording purposes.
Wherever possible the relationship of the water with tunnel features should be identified
and recorded. For example:
400
general permeation through tunnel fabric (eg large areas of weeping)
associated with local damage or deterioration (eg missing mortar, missing bolts/
caulking, tree-root penetrations, cracked, spalled or corroded areas, damaged drainage)
associated with features of tunnel construction (joints in rock or construction
materials, geological faults and changes in ground or lining type).
The presence and appearance of water-borne deposits should also be recorded, since
these may provide useful information on the cause and effect of the inflow, how long it has
been occurring and even on the water chemistry and local ground conditions. In tunnels
with brick or stone masonry or concrete linings, hard whitish build-ups of minerals are
common, often calcium carbonate from the groundwater or, in concrete and masonry
linings, leached out of the lining materials themselves, but it can also be gypsum (calcium
sulfate) where there is a source of sulfates. Bright red ferruginous deposits are indicative
of corrosion of metallic elements or may be from areas of ground rich in iron minerals.
Deposits of clay and silt may also be present, indicating loss of fines and washout of the
ground behind the lining – this can be significant over long periods where there is the
potential to reduce ground stability or create voids behind the lining. Over time build-up
of waterborne deposits can cause problems with clearance for traffic, or obscure the lining
intrados and hamper visual inspection.
In the course of a tunnel inspection, samples of ingressing water can be collected for
laboratory testing to determine its chemical characteristics, and help to assess the response
needed. The results can also help to determine the likelihood or cause of materials
deterioration, for example, where metals are subject to rapid corrosion or sulphate attack
is suspected in masonry or concrete. It may also help to identify the source of water and
whether it presents any biological hazard, for example, whether it is likely to be mains
supply water, sewage/wastewater, and groundwater or surface/rain water. In built-up areas,
water in tunnels is often a mixture of water from more than one source, so it may prove
difficult to discern the source with adequate confidence.
This information may be used in combination with details of the ground conditions
behind any lining present and historical records (original construction records, the results
of ground investigations, nearby borehole records and past inspection reports) to
determine the likely source of the water, its pathway to and into the tunnel, the
requirement for remedial action, and the measures that are most likely to meet with
success in preventing or controlling the ingress, or mitigating its undesirable effects.
A3.2.7
Other observations
Depending on the circumstances of the tunnel and its environment, there are many other
observations that may be important or relevant in a visual inspection, for example:
the local weather conditions at the time of the inspection and those of the previous
few days (in particular rainfall)
the condition of the portal areas and any evidence of instability or ground movements
any new construction or change in land-use that might affect the tunnel
the presence of vegetation at portals and within the tunnel (particularly any tree roots
that might have penetrated it).
Inspectors should be vigilant concerning other phenomena or situations that are relevant
under their duty of care and responsibilities, for example:
clearly damaged, loose or deteriorated elements of tunnel equipment and services
where these might present a hazard to tunnel users and normal operations
damage to boundary fences, evidence of any unauthorised entry to the tunnel (eg
graffiti), fires or other evidence of vandalism
CIRIA C671 • Tunnels 2009
401
the presence of flammable, toxic or otherwise hazardous items and materials,
including litter, vermin, needles and syringes etc.
Where there are inspection requirements that are specific to a particular tunnel, for
example, special attention to be paid to the condition of a certain element, measurement
of a certain parameter or examination of known feature such as the lining at the base of a
shaft, these should be clearly identified as a requirement in any description of scope for
the work or pre-inspection checklist.
402
A4
Inspection, investigation and monitoring
techniques
This appendix includes information on those techniques that can be used in the course of
visual inspections, in situ investigations and monitoring, and in the laboratory analysis of
samples. The range of available techniques for different types of investigation and
materials evaluation is enormous and a comprehensive review of them all is beyond the
scope of this document. However some of the most commonly used methods are
discussed, along with more sophisticated state-of-the-art techniques that have had limited
use to date but are considered to have potential for further development. The range of
techniques considered in this appendix is listed in Table A4.1.
Table A4.1
General, specialist, testing and monitoring techniques for tunnel investigation
Inspection, mapping and
simple on-site tests
Sampling and testing
techniques
(Section A4.1)
(Section A4.2)
Specialist nondestructive
investigation
techniques
Monitoring techniques
(Section A4.4)
(Section A4.3)
visual inspection
photography and
videography
hammer-tapping
spatial measurement
remote visual inspection
joint mapping
simple on-site
assessment techniques
spatial measurement
and surveying
techniques.
core drilling and core
samples
laboratory
investigation
techniques
geotechnical
investigation
techniques
measurement of in
situ stress.
infrared thermography
gravimetric survey
magnetic survey
ground resistivity
survey
conductivity survey
seismic survey
3D seismic
tomography
ground penetrating
radar
transient
electromagnetic
broadband
electromagnetic
ultrasonic pulse
velocity (UPV).
crack monitoring
strain monitoring
stress monitoring
displacement
monitoring
corrosion monitoring.
It is important to note that any invasive investigation of a structure that is subject to
statutory protection (for example, listed building or Scheduled Ancient Monument status,
see Section 3.6.2), or that might cause environmental damage or disturbance or damage
to protected species (see Section 3.6.3) may require consent from relevant authorities in
advance. Also, the health and safety aspects of investigations and techniques employed
should be considered and compliance with statutory requirements and any relevant
policies of the asset owner (see Section 3.6.1). There should be consideration of the
potential effects on the performance and safety of the tunnel structure and associated
equipment and systems, and of the requirements to repair any damage, particularly where
intrusive or destructive investigation is planned.
CIRIA C671 • Tunnels 2009
403
A4.1
INSPECTION, MAPPING AND SIMPLE ON-SITE TESTS
A4.1.1
Visual inspection
Visual observation is used as the first and most efficient method of obtaining basic
information on a tunnel and determining and monitoring its condition. In many respects
it is also the most effective however it does have shortcomings, as discussed in Section
4.3.1. These can be overcome by supplementing it with extra simple and rapid techniques
such as photography, dimensional measurement, hammer tapping and other simple onsite actions. These can be applied in the course of an inspection where extra information
obtained would be beneficial, as discussed in the following sections.
A4.1.2
Photography and videography
Photography is a basic surveying tool and should be used routinely when recording the
condition of tunnels, in particular the appearance of any features of interest such as
defects and evidence of deterioration. Likewise, moving video images are used
increasingly for recording site conditions and features. Visual images are invaluable for
communicating information in a relatively objective manner to others, supplementing
written observations. Their objectivity also benefits the continuity of inspections, because
they can be less subjective than the observations of a single inspector and useful for
comparing condition over time.
Conditions in tunnels are often not conducive to taking good photographs. Cameras with
integral flash units may be adequate for close-up work within a range of a few metres, but
photographing larger areas may require the use of higher-power flash units, and possibly
wide-angle lenses to give adequate field of view if space is constricted. Both traditional film
and modern digital cameras are potentially suitable, however digital imaging has
additional benefits for inspection because image quality can be checked while still on-site,
and are more easily incorporated into reports and shared between interested parties.
It is recommended that all inspectors are equipped with cameras and any necessary
ancillary equipment (tripods, flash units) necessary for obtaining good photographs, and
should be trained to an adequate standard in their use. Where photographs may be used
for later comparisons, consistency should be ensured in the images captured to facilitate
this (ie in the area photographed, angle and position of view, lighting conditions and
inclusion of scale measures).
404
Figure A4.1
Photography combined with oblique backlighting can be a very useful aid to recording
areas of surface-wetness on tunnel linings, because these are highly reflective
A4.1.3
Hammer-tapping surveys
Hammer tapping is a simple technique used to give a quick appreciation of the condition
of concrete and brick and stone masonry. It is useful for identifying voids, loose and
spalling material and, in multi-ring brick arches, delamination between rings. When the
intrados surface is tapped with a hammer or sounding pole, the integrity of the material
can be discerned from the tone and resonance of the sound generated. An experienced
operative may be able to achieve good results although the subjectivity of the assessment
means that they are not always repeatable.
Although this is a relatively quick and easy technique, with no requirements for specialist
equipment or training, direct access is required to the surface under investigation, and it
may be included in touching-distance surveys. Hammer surveys can detect delamination
near to the surface of an element, but are of no use when it occurs at depth. The
technique is not suitable for use on frozen concrete or brickwork.
A4.1.4
Construction joint mapping
During the construction of brick or stone tunnel linings, construction joints were dogtoothed to provide bonding to the next construction length. However, they were hard to
build exactly to match-in. The lack of mortar and badly twisted bricks forming the start of
the new length has left a clear indication of where the joint is in the tunnel lining. Some
joints are clearer than others but mistakes in identification can usually be quickly rectified
and double-checked. A description of construction joints in brickwork, including
photographs, is given in Section 2.3.5.
The spacing of the joints relates directly to the story of the tunnel construction and in
good ground construction lengths were 14 feet (4.6 m) long. When ground turned softer
and more support was needed to keep the tunnel shape, excavated lengths were
shortened to 10 feet (3 m) or even six feet (1.8 m) and the brickwork of course had to
follow suit. The reduction in joint spacing could represent worsening of ground
conditions. The known practice of paying piecework however meant that there was no
lack of willingness to go for longer lengths after the poorer ground conditions had passed.
CIRIA C671 • Tunnels 2009
405
Shafts that were used for construction were generally not bricked across until after
construction had finished and at nine feet (2.7 m) diameter would represent in the lining
a shorter length of brickwork. These lengths would join up the two shaft side lengths built
either side of the working shaft to carry the weight of the shaft above. At this point either
the shaft was left open for ventilation or was infilled.
After completing the lining across the shaft, infilling the shaft started. This would require
backfill material, which would probably be low grade material, un-compacted and loose,
attracting water as a drainage point directly above the tunnel. It would be likely that the
lining in the tunnel crown would show, over a short length between joints, twisted
brickwork as the coursing from the two separate side lengths was matched up across the
opening, and the area would most likely be wet.
Other features that might be evident in the lining joints could be break-up lengths where
full sized sections of tunnel were erected in between shafts using a break-up from the
bottom heading. These short sections of tunnel were formed in a constructed slot, into
which a 10 feet (3 m) long section of tunnel was built. This stand-alone section of tunnel
was then used to construct sections of tunnel both sides of this length. These break-ups
allowed many more gangs to be employed along the length of the tunnel and the work to
be completed much quicker.
Gangs working towards each other may meet at a junction length, which could be much
shorter than a standard length due to either miscalculation or one gang taking the lion’s
share leaving little or nothing for the other gang on their piecework scheme.
There are several features within the pattern of tunnel joints that can reveal more about
the history of tunnels construction.
Joint-mapping is probably most useful when combined with a geophysical survey of the
tunnel lining, such as ground probing radar (see Figure A4.2) in which the joint lengths
can isolate areas of water ingress, poor workmanship, lack of mortar etc. All of these
defects can be confined to joint lengths making repair and remedial works more focused.
During inspections areas that have been identified with sudden problems such as bulging
or distortion can be located within the joint-mapped tunnel and the pattern of joints in
this area may give an indication if there was a blind shaft, junction length, break-up length
or just an area of bad ground at that location.
Without having a fully joint-mapped tunnel, mapping of the joints for 50 m to 100 m
either side of a sudden problem such as bulging could help identify the known
characteristics, from previous tunnel mapping experience and results, of bad ground or
an infilled shaft.
Case study A1.11 gives an example of how joint mapping, combined with a geophysical
survey, has been used in practice to identify construction details and locate hidden shafts.
406
Combined Joint Mapping and GPR Survey
Figure A4.2
Results of combined joint mapping and ground penetrating radar (GPR) survey
to identify construction features and defects within a masonry-lined rail tunnel
(courtesy Jack Knight and Aperio Ltd)
A4.1.5
Remote visual inspection
Remote visual inspection techniques have a key advantage over manned visual inspection
in potentially hazardous situations: they are conducive to the achievement of a safe system
of work because they allow inspection without man-entry to confined spaces or unstable
structures. Also, they provide a permanent visual record of the inspection, which is
objective and can easily be shared, reviewed and compared to previous inspection records.
Previously records have been made on analogue video-tape but newer systems provide
digital records which can be stored on hard-drives, CD-Roms and DVDs, as continuous
video images or as screen grabs, and this date can be incorporated in databases and
geographical information systems (GIS).
CCTV inspection systems are commonly used for pipework inspection and sewer/drainage
tunnels. Miniature cameras can provide high quality colour video images and be adapted
for use in a wide variety of situations, for example, by the addition of remote control
robotic trolleys, lighting units, pan-and-tilt systems and zoom functions. CCTV inspection
of shafts using a camera suspended under a remotely controlled helium balloon has been
trialled by LUL Ltd with encouraging results (Chew and Roberts, 2005) though there is
currently some debate about the acceptability of remote inspection techniques in tunnels
and shafts, and they are not favoured by some asset owners.
CIRIA C671 • Tunnels 2009
407
Although requiring man-access to the inspected areas, another remote inspection
technique involved the use of rigid endoscopes or flexible image-scopes, both of which
provide still and video records and can be used to examine hidden parts of tunnels. Access
can be through existing holes/cracks, core-holes taken for other purposes, or small holes
(typically 10 20 mm) drilled into the structure specifically for this purpose.
A4.1.6
Simple in situ assessment techniques
Where appropriate, it may be advantageous to incorporate other quick and simple
methods of qualitative, semi-quantitative and quantitative assessment in the course of
visual inspections. For example:
use of a metal spike as a probe to assess the condition of materials including concrete,
stone, brick and mortar: this can either be used to allow an inspector simply to gain a
feel for relative condition and to identify weak and deteriorating areas of tunnel
fabric, or in a semi-quantitative manner by using a scale such as the Russack scale for
assessing the hardness of mortar and bricks (Ferro, 1980). Such simple methods may
be useful, for example, for assessing the extent of areas requiring repair
dimensional measurements: these can be made using tape measures or hand-held
laser-measurement devices, which are accurate to within a few millimetres and could
be used as a technique for checking for relative movement (eg whether lining bulges
are progressive) between surveys
checking for local, relative movements (eg across cracks): this can be achieved by the
use of simple surface-mounted displacement gauges and tell-tales (described in
Section A4.4.2). These are susceptible to vandalism or accidental damage when
attached to other types of structure, but this should be less likely in most tunnels.
There are a variety of relatively quick and simple in situ characterisation techniques that
can be used to determine material properties and condition, and could potentially be used
in the course of a detailed inspection. These often provide quantitative or semi-qualitative
measures, but some can also be calibrated to provide quantitative estimates of
characteristics such as strength, within a certain margin of error.
Examples of such techniques include:
scratch tests and hardness tests (eg the Brinell test for metals)
pull-out tests (eg the screw pull out test for mortar strength described in BRE digest
421, or similar techniques for other materials such as brick, stone or concrete)
rebound and impact tests (eg the Schmidt hammer and the Windsor Probe, which can
be used to evaluate concrete quality and estimate its compressive strength)
measurement of drilling energy (for comparative evaluation of hardness/strength,
potentially for use in concrete, brick and mortar)
Although such techniques often have limitations in terms of accuracy and repeatability,
they can provide a relatively straightforward and inexpensive method of evaluating
materials properties and some are even suitable for incorporation in routine visual
surveying exercises. The main limitation of these tests is that they characterise only the
material at the surface of an element, which for various reasons (not least exposure to
weathering and other deteriorative mechanisms) may not represent the whole material.
This is not a problem where it is the condition and deterioration of surface material that is
of interest. For example, they could be useful in monitoring deterioration or making
relative comparisons between the condition of materials in different areas.
408
A4.1.7
Spatial measurement and surveying
Recent developments in instrumentation technology and data processing capability have
greatly increased the power and flexibility of survey techniques. In all but the smallest
applications, modern automated and semi-automated equipment with onboard recording
and processing capabilities and PC-connectivity has almost completely replaced manual
instrumentation, providing benefits in terms of the accuracy and dependability of data, as
well as the safety and productivity of survey teams. Through minimisation of manpower
requirements and survey times, the use of modern automated survey instruments
provides considerable potential for reducing the overall cost of surveying work.
The monitoring of structural movement using electro-mechanical instruments, such as
linear variable differential transformers (LVDTs) and vibrating-wire strain gauges, is
relatively well-established. However, in this field, as with surveying, there have been
significant advances in recent years:
the development of precise and reliable electro-mechanical sensors (such as
electrolevels) and the increasing experience of their use
the introduction of fibre optics and advanced light measurement techniques
the application of modern survey instruments and methods
advances in the data processing power of computers and the development of
increasingly sophisticated software applications.
In many situations it is important to undertake an accurate survey of the tunnel intrados,
or to determine the size and shape of features and their relationship to other parts of the
tunnel. Traditional surveying techniques using modern equipment can achieve accuracies
to within a few millimetres and rapidly obtain information from several individual points.
Also, accuracy can be achieved by using reflective targets (of the disposable stick-on type or
better quality glass prisms). The theoretical accuracy of high precision system is sub-arcsecond and sub-millimetre, but there are many systematic and random factors that can
degrade this accuracy.
More recent specialist surveying techniques, such as laser-scanning and photogrammetry,
have improved in recent years and now offer the possibility of determining the 3D shape
of elements:
laser scanning encompasses both 2D profiling and the 3D laser scanning which can
create a complete image of the structure. Both systems use a time-of-flight method to
measure how long it takes for each laser pulse to hit a surface and return to the
scanner. Incident angles and ranges are recorded for each laser pulse. The laser
scanning uses a laser cloud principle to collect scans containing millions of points
which are then converted by specialist software into 3D images to be viewed in a CAD
environment. Scanning takes place from several known locations to ensure complete
coverage of the site
photogrammetry involves taking multiple digital images using a specially designed
camera to dimensionally model a structure. The images taken from different locations
are compared and the differences used to retrieve 3D information from the image.
Specialist software is used to identify correlations between the two images and to make
measurements. This is often done by digitising common locations on both images.
Visible features can be used to generate the corrected dimensional data.
Both techniques can be used to create a virtual 3D model of a structure or element which
can be viewed and measured in a variety of software applications such as CAD. A major
CIRIA C671 • Tunnels 2009
409
benefit of these techniques is their ability to capture very comprehensive data, avoiding
the need for repeat surveying visits if further data is required. Another benefit is that the
equipment required is now relatively light and portable and can be mounted on a vehicle
or trolley before passing through the tunnel, at a steady walking speed. A detailed scan of
an entire tunnel intrados can be produced very quickly and relatively cheaply, to an
accuracy of within a few millimetres. This makes these methods particularly useful where
access is limited in duration. The geometric information obtained forms a 3D tunnel
profile which may be useful for more advanced methods of structural analysis or for
monitoring changes in structural condition, or even as a general asset record.
This type of information can be used for (Network Rail, 2004b):
1
Evaluating tunnel intrados geometry and dimensions and assessment of clearances.
2
Checking alignments of ducts and lift guide rails.
3
Monitoring changes which can indicate problems of deformation.
4
Compilation of inventories and as-built drawings.
5
Determining volumes of excavation or lining materials.
6
Indication of structural failure.
7
Collection of information for refurbishment.
8
Checking the driving of tunnels.
9
Monitoring progress of projects.
Laser scanning techniques have been used for quantity surveying purposes, for example,
to accurately measure the volume of material removed and replaced in lining repair
contracts, rather than relying on contractor’s estimates, with claims of considerable time
and cost savings (Trimble, 2005).
A4.2
SAMPLING AND TESTING TECHNIQUES
These include a range of techniques for on-site measurement and testing and obtaining
samples of material for subsequent laboratory investigation.
A4.2.1
Core drilling and core samples
Cores are usually cut perpendicular to a surface using a rotary hydraulically jacked
cylindrical core barrel to produce parallel-sided core samples. Coring rigs come in a
variety of sizes, from portable rigs which need to be bolted to the tunnel surface, to larger
and more powerful vehicle or trolley-mounted units which can be more productive. Cores
are normally taken from concrete, rock and brick or stone masonry to provide samples for
establishing material properties, however where full-depth cores are taken through a
tunnel lining the hole itself is very useful because it allows direct observation of features
that are normally hidden behind the intrados. Examination of the core-hole can provide
confirmation of construction details such as lining thickness, identification of features such
as voiding and delamination within it, and information about the nature of the ground, its
wetness and its contact with the lining extrados. Examination down core-holes is often
facilitated by using an endoscope.
Cores can be taken to considerable depths however the disruption caused by the drilling
process means that weak material, such as old and deteriorated brickwork, often
fragments and there may be significant material loss from the resulting core sample. This
can be minimised by good drilling practice, equipment and technique. Operatives should
410
understand that the objective is to produce a good quality and substantially intact core
sample, not simply making a hole in the structure.
Figure A4.3
Taking a 100 mm core through a brickwork tunnel lining. The lightweight rig is bolted to
the wall. Progress can be slow in hard masonry materials and in the investigation of rail
tunnels use is often made of heavier and more powerful coring equipment mounted on track
trolleys
The core diameter is an important influence on the quality of the extracted core and its
usefulness. For most applications a 100 mm core diameter is advisable to enable recovery
of broken material and provide samples suitable for a range of laboratory tests. Sometimes
larger diameter cores (up to 300 mm) may be necessary for certain purposes, and smaller
diameter cores (as small as 25 mm) may be useful, depending on the aims of the
investigation. The type of material to be cored and the quality required of the core sample
will dictate the type of core bit and barrel used. Water-flush and diamond-tipped core
barrels are advisable where the recovery of good quality cores is paramount, but other
flushing fluids are sometimes employed and tungsten-tipped core barrels are also
available but may not be suitable for harder materials. A double tube core barrel can be
employed in which an inner tube supports the core as it is cut, reducing the erosion of the
core by the flushing medium. However a recent development is the use of triple tube core
barrels which yield the best core recovery.
To prevent disturbance, cores should be withdrawn from the barrel horizontally, ideally by
employing a hydraulic extruder. To increase the quality of core recovered, the inner
barrel can be lined with a plastic sleeve before drilling, ensuring that the samples are as
undisturbed as possible.
Where a core-hole is intended to penetrate the full thickness of a tunnel lining, or where
there is a risk of doing so, the potential for encountering water under pressure behind the
tunnel lining should be considered. Where necessary, measures can be taken to control
the ingress of water (potentially bringing with it any loose ground), for example, by using
a stuffing box for the coring rig. Reinstatement of core-holes is important and should be
undertaken using appropriate methods and materials.
The retrieved cores should be clearly labelled with their location and orientation within
the tunnel, logged and measured, with the materials and visible features recorded and a
scaled colour photograph taken. They should be wrapped in cling-film to preserve their
CIRIA C671 • Tunnels 2009
411
condition. Careful handling and packing for transportation from the site, normally to a
laboratory for examination and testing, is very important to avoid damage.
If the cores are used to determine strength, it should be remembered that they are usually
not taken in the direction of principal stress and so should be interpreted accordingly.
Cores are also taken through the ground, either from the ground surface or from within a
tunnel, to investigate geological conditions and recover samples for examination and
laboratory testing. They can also give excellent information on the occurrence and
condition of shafts. Close supervision during the drilling to monitor and record changes
in the rate of penetration, changes in the colour of the flush returns and the loss of
flushing medium into cavities within the rock can all help to identify abandoned workings.
Detailed geological logging of the cores obtained allows correlation between different
boreholes. The core logs should identify features such as rock type, degree of weathering,
strength, number of fractures, dip of strata and amount of solid recovery. Careful
assessment of these logs will enable an understanding of the local geology to be obtained.
A4.2.2
Techniques for laboratory investigation
The range of tests and analyses performed on-site during a site investigation are limited,
so samples of materials (including lining materials, local ground, chemical deposits, water
etc) are often obtained and sent to a laboratory where are subjected to a range of other
techniques to gather information including:
identification of the materials present
their physical and chemical characteristics
their current condition
the presence, cause and extent of deterioration.
The aim of the laboratory-based investigation is to provide additional information on the
materials present in the samples and on the in situ materials in the tunnel. It is important
that the samples selected are representative of the materials, features or conditions under
investigation, and that the relationship of the sample to the in situ material is understood.
For example, samples may be taken from typical best or worst areas, or those that exhibit
a particular feature of interest (as discussed in Section 5.4.2). As with on-site testing, the
wider cost of obtaining and analysing each sample should be justified in terms of the
potential benefit from the results and unnecessary sampling and testing should be avoided.
Laboratory based techniques commonly used for materials investigation include:
microscopic/petrographic assessment of metals, concrete and masonry materials to
evaluate material types and characteristics, their quality, variability and condition, and
to identify and assess deterioration and investigate its causes
physical testing to provide qualitative measures of material parameters such as
compressive strength, hardness, modulus, porosity and permeability (on lump or core
samples)
chemical characterisation using a range of analytical techniques such as wet chemistry,
X-ray diffraction (XRD), differential thermal analysis (DTA), scanning electron
microscopy (SEM) with electron dispersive analysis (EDAX) etc.
Consideration should be given to the sample types required and the minimum and
optimum sample sizes for the intended methods of analysis, and the likely accuracy and
repeatability of results.
412
Figure A4.4
Compressive strength testing of a 300 mm diameter concrete
core while simultaneously measuring strain (courtesy CERAM)
A detailed description of the range of laboratory investigation techniques is beyond the
scope of this document, but the following publications include more detail:
for brickwork and masonry in general – Testing of ceramics in construction (Edgell et al,
2005)
for metallic materials – Appraisal of existing iron and steel structures (Bussell, 1997)
for concrete – Analysis of hardened concrete: A guide to test procedures and interpretation of
results (Concrete Society, 1989b), Guide to testing and monitoring the durability of concrete
structures (CBDG, 2002), Corrosion of steel in concrete, part 2: investigation and assessment
(BRE, 2000).
CIRIA C671 • Tunnels 2009
413
A4.2.3
Techniques for geotechnical investigation
Geotechnical investigations aim to establish the type and characteristics of the ground
above and around a tunnel. Full details of such investigations are outside the scope of the
publication, however a brief description of some of the more commonly used investigation
methods is given. For full details, refer to BS 5930 (BSI, 1999).
As described in BS 5930, for investigations into existing structures observations and
measurements of the structure are first needed to determine whether the ground
conditions are partly or wholly responsible for any defects seen. If this is the case,
geotechnical investigation is required to ascertain the condition of the strata and the
groundwater conditions, both as they exist in situ and as they existed before the works
were constructed. Indications of the probable cause of a defect often result in detailed
attention being directed to a particular aspect or stratum of soil.
A considerable variety of methods of ground investigation exist, consisting of those in
which samples are retrieved from the ground for description and testing and those in
which the properties of the soil are described or measured in situ. Sampling inevitably
introduces sample disturbance. This is reduced in the case of in situ testing which can give
more representative results.
It may be easier to ensure that the sample orientation is correctly related to the proposed
loading conditions in the laboratory than with an in situ test. Long-term tests are better
handled in the laboratory. However, it is possible to test a more representative volume of
soil in situ than in the laboratory (eg in a direct load plate test). Any factor that influences
the soil properties should be carefully considered when deciding the mode of testing.
Trial pit investigation
Shallow trial pits provide an economical method of examining in situ conditions.
Exploration depths are usually between two and four metres, and require temporary
support if personnel are to enter them. Investigation is normally limited to levels above
groundwater in non-cohesive soils.
A close examination of the ground is carried out using a systematic scheme for the
description of the soils. Colour photographs of side faces should be taken with a
prominent scale marker and located on a plan. Trial pits can be used to locate statutory
undertakers’ installations and any other buried equipment. Sampling can be carried out
from either disturbed samples obtained by excavation, or open tubes driven into the base
or side face. High quality block samples may be cut from a bench formed in a trial pit. In
all cases, the exact location and orientation of the sample relative to the pit and walls
should be recorded.
Within rail tunnels it may be necessary to excavate trial pits down through ballast to
determine the presence and condition of any invert, or to inspect tunnel foundations.
Asset owners are likely to have their own requirements for limiting the methodology and
extent of excavations below, and for the replacement and re-compaction of ballast below
rail supports (see Figure A4.5).
414
Figure A4.5
Trial-pit through ballast at base of tunnel sidewall to prove invert depth
Borehole investigation
Exploration by means of boring or drilling, and the recovery of samples, is an established
technique, and may be used exclusively or to supplement trial pits. It is unlikely to be used
unless there are real concerns about the competence of the foundations or refurbishment
includes new-build and/or change of loading regime.
Many of the most frequently used sampling and in situ testing methods can be carried out
with a wide variety of boring rigs, so that the accessibility or labour costs are often
controlling factors regarding rig selection. For rough terrain or inaccessible locations, light
rigs are advantageous. Environmental issues need to be strictly observed, in particular
with respect to the use of water and the disposal of material.
Figure A4.6
Intrusive investigation through metallic segmental tunnel lining
CIRIA C671 • Tunnels 2009
415
Other common techniques
Other common techniques are presented in Table A4.2, adapted from CIRIA SP25,
(Weltham and Head, 1983).
Table A4.2
Techniques used in the investigation of ground around tunnels
Method
100 mm diameter
open-tube sampler
(U100)
Standard
Penetration Test
Application
Firm to stiff clays, insensitive
or stony clays, clayey silts,
some weathered rocks.
Advantages
Simple robust equipment,
usually dynamically driven.
Provides a reasonably large
sample. Inexpensive. Rapid.
Widely accepted and used.
Derivation of a standardised
blow count from dynamic
penetration in granular soils
(silts, sands, gravels) and often
other materials such as weak
rock or clays containing
Simple, robust equipment.
gravels which are not readily
Procedure is straightforward
sampled by other methods.
and permits frequent tests.
Convenient both above and
Inexpensive.
below the groundwater table.
Vane test
The results should be used in
conjunction with laboratoryderived values of cohesion and
measurement of plastic index
(PI) sor that an assessment of
the validity of the results may
be made.
2
Camkometer, self-boring
type.
3
Stressprobe, pressed into
the soil from the base of
a borehole.
Pressuremeter test
Plate bearing test
416
Horizontal load tests are
possible to determine backfill
capacity and deformational
characteristics for numerical
modelling
Results require interpretation.
Test insensitive in loose sands.
Causes little disturbance to
the soil.
There is some dependence on
the PI of the clay.
Can be used directly from the
surface, or from the base of a
borehole.
To be used in conjunction with
careful description and backed
up with high-quality sampling
and laboratory testing.
Results are direct and
immediate.
Tests can be rapid.
In situ low disturbance
measurement of important
soil and weak rock
parameters.
Less expensive than direct
bearing tests.
Rapid test procedure
Lateral stress and Ko
measurements are possible
For determination of elastic
modulus and bearing capacity
of soils and weak rock, with
minimum disturbance.
Simplicity of the equipment
belies its sensitivity to
operator techniques,
equipment malfunction and
poor boring practice.
The results are affected by
silty or sandy pockets, or
significant organic content in
the clay.
Three similar types of
pressuremeter are available:
Menard pressuremeter,
installed into a borehole.
Accurate control of sampler
penetration is difficult. Quality
depending on the care taken
by the driller.
Permits in situ measurement
of the undrained strength of
sensitive clays with cohesions
up to 100 kN/m². The
remoulded shear strength may
also be measured in situ.
Small hand-operated vane test
instruments are available for
use in sides or base of
excavations.
1
Produces disturbed samples.
Misleading results in fissured
clays.
The blow count (N value) may
be used directly in empirical
formulae to determine soil
strength parameters.
Measurement of undrained
shear strength of clays and
measurement of remoulded
strength.
Disadvantages
There is close control of
loading intensity, rate and
duration
More representative than
laboratory testing.
Results are in terms of total
stress only.
Specialist technicians
required.
In some soils and rocks, the
operation of the equipment
can be uncertain, particularly
coarse granular soils.
Drainage conditions have to
be assumed and the
consequential effect on the
porewater pressures.
Specialist technicians
necessary.
Expensive and time-consuming
test.
Scale effects should be
considered.
Specialist technician required.
Can be carried out in pits or
boreholes.
Ground conditions changed by
test.
Lateral loading is possible
Pore-water changes are
difficult to take account of
A4.2.4
Measurement of in situ stress
Knowledge of the magnitude and direction of in situ stresses in the rock mass around a
tunnel, or in a tunnel lining itself, is highly desirable, because it can be used for assessing
structural performance, adequacy and safety, for verifying design assumptions, and to
identify structurally sensitive areas. The results of stress measurements can be used as
input for structural models, for example, using finite element techniques.
Measurements of in situ stress is a highly specialist task, most commonly used in the
mining and tunnelling industries but the techniques have also been used in the evaluation
of existing civil engineering structures. Suggested procedures for evaluating stress
measurements are described in the International Society for Rock Mechanics (ISRM)
publication Suggested methods for rock stress determinations (ISRM, 1987) but these are for
guidance only, based on papers published by researchers in this field, and do not by any
means constitute approved norms or standards. Development is underway to allow the
practical use of stress measurement as a tool for a range of civil engineering, however
their routine use in such applications is still some way off.
The most common methods are based on instrumented stress-relief techniques (involving
either over-coring or slot-cutting with pressure compensation), hydrofracturing in rock,
or, most recently, acoustic emission measurements from rock cores. A brief overview of the
principal techniques is given here.
Overcoring techniques
A commonly used method that involves releasing stress in concrete or rock elements by
making a saw-cut or a drilling a core-hole and accurately measuring the strain response
(using either a directional stress-cell within the hole or an array of vibrating-wire strain
gauges previously installed on the surface adjacent to the hole, sometimes in combination
with Demec studs). The technique can be applied in minimum core lengths of 200 mm to
300 mm.
For example, using in-hole instrumentation, the basic procedures involve diamonddrilling a hole to the desired depth before removing the core and preparing the hole to
receive a stress measurement cell containing a strain-gauge rosette, which is inserted and
fixed in place at the bottom of the hole. A baseline reading is taken, then a new core is
drilled beyond the base of the original hole and this is broken off at its base to produce a
discrete core, separated from the surrounding in situ material and released from stress.
The stress cell is used to take a second reading from the core in its relaxed state.
Comparison of these values provides a strain measurement which, along with the elastic
parameters of the material determined from laboratory testing, allows the in situ stress to
be estimated.
A full stress evaluation involves several measurement cycles, which is likely to require one
or two days of in situ testing, making this a relatively time-consuming and expensive
technique.
Hydraulic fracturing techniques
Hydraulic fracturing techniques can be used to determine in situ rock stresses in a plane
perpendicular to a borehole (a borehole drilled for stress measurement by overcoring, as
previously described, can be used). Tests can be carried out on vertical holes drilled from
the ground surface or vertical, inclined or horizontal holes drilled from within a tunnel.
CIRIA C671 • Tunnels 2009
417
A test section of the hole is isolated using packers before fluid (normally water) pressure is
applied until the rock fails in tension. The hydraulic pressures required to generate,
propagate, sustain and reopen tensile fractures are recorded as a function of time and can
be related to the magnitude of the existing stress-field. The orientation of the
hydraulically induced fracture, which is parallel to the orientation of the secondary
principal stress, is also measured.
As for the overcoring method, with which it can be combined, the hydraulic fracturing
technique for stress measurement is a highly specialist test and can be time-consuming
and expensive.
Acoustic emission (EM) techniques
Acoustic emission (AE) is a technique that relies on the phenomenon of the Kaiser effect,
which pertains to the stress memory of rock, related to its peak in situ stress, measured
under uniaxial and triaxial conditions. The Kaiser effect suggests that previously applied
maximum stress can be evaluated by stressing a rock specimen to the point at which
internal microfractures begin to propagate rapidly, correlating to a substantial increase in
AE activity that can be measured.
In practice, this requires core-holes to be drilled into rock and prepared for testing in a
laboratory. The test procedure involves cyclic uniaxial compressive loading at a constant
stress rate, while AE sensors attached to the specimen record its acoustic response.
This is a relatively recent methodology but has been used successfully in real mining
applications (Villaescusa et al, 2002). Experimental studies have suggested that the EM
technique gives similar results to those obtained by the overcoring method, while being
relatively cheaper, requiring less time in-the-field, and not relying on the use of
sophisticated equipment on-site.
Flat jack techniques
Flat jacks have been used to measure in situ stresses in masonry in one direction with some
success (Hughes and Pritchard, 1994a and 1994b, Abdunur, 1995). Most experience has
been gained with masonry arch bridges but the techniques have also been applied to
masonry tunnel linings (Hughes, pers comm.). Various types of jack have been tried from
simple rectangular jack (see Figure A4.7) through to a series of segmental jacks that are
incrementally introduced into successively deeper grooves. The principle is the same and
involves fixing a series of reference points on the surface of the masonry, cutting an
appropriate groove into the surface and installing the jack such that it is intimately in
contact with the sides of the groove. The jack is then pressurised until the reference points
are deemed to have returned to their original (as-constructed) position. This basic
procedure assumes uniform stress for the depth of the jack, which is a simplified
approach. More sophisticated techniques try to compensate for the variation in stress with
depth by progressively cutting deeper grooves and inserting larger jacks building up a
picture of the stress state.
418
Figure A4.7
Flat jack developed by Cardiff University
There are several sources of error one being the simplification of the procedure in
ignoring potential stress variations mentioned above. Note the material stiffness and
loading may vary through the depth of the lining, making the surface measurements
somewhat uncertain. Also, it can be difficult to achieve a close and even contact between
the jack and the material and the jack should be calibrated because of edge effects that can
cause the pressure exerted to be less than the internal pressure. Ongoing development of
flat jack systems seeks to overcome these potential drawbacks.
International standard recommendations for the use of flat jacks to measure stress and
elastic moduli in a compressive environment have been published by RILEM (1994).
Standard test methods include those from the ASTM (ASTM, 1991a and 1991b) and a
digest on flat jack testing has been published (BRE, 1995).
As noted, flat jacks can be used to determine the elastic modulus of the masonry. This can
be done by cutting two slots in the masonry at about the length of the jacks apart. The
slots should not be more than 1.5 times the length of the jacks apart. Best results are
achieved by cyclic loading at increasing stress levels. Constant load tests can be performed
to check the effects of creep.
A4.3
SPECIALIST NON-DESTRUCTIVE INVESTIGATION
TECHNIQUES
A brief description is given below for a range of specialist investigation techniques, which
are often non-destructive testing (NDT) techniques. From the outset many require the use
of sophisticated equipment, and a high level of expertise is necessary to interpret their
output. Consequently, they can be expensive procedures and need to be used carefully to
optimise their chances of success and the usefulness of data collected. Those specifying the
use of specialist methods of investigation should have a good understanding of the
potential benefits and the limitations of each technique, how these relate to the objectives
of the investigation and situation of the tunnel in question. A comprehensive overview is
presented in the Highways Agency Advice Note BA86 (HA, 2004), and further review of
geophysical methods for ground investigations is given in BS 5930 (BSI, 1999) with
recommendations as summarised in Table A4.3.
CIRIA C671 • Tunnels 2009
419
Because there is often uncertainty surrounding the interpretation of the output of the
techniques described here, it is recommended that sole reliance upon one technique
should be avoided where a high level of confidence is required in the results. Such
techniques should ideally be used in situations where more conventional and reliable
techniques cannot be used, are unsuitable, or their use should be minimised. It is often
advantageous to combine the use of specialist and conventional techniques in a
complimentary manner, making best use of their strengths and minimising their
weaknesses.
For example, if trying to determine the thickness and uniformity of a masonry or
unreinforced concrete tunnel lining along its length, a radar survey may be used to
rapidly cover a large area of lining intrados, and this can be supplemented by drilling a
limited number of small diameter cores through the full thickness of the lining and
directly measuring its thickness using the cores and (more reliably) the holes. In this case,
the benefits of the radar survey over using coring only are rapidity, good area coverage
and minimisation of damage to the structure and disruption to its normal use. The
potential weakness of the technique (lack of confidence in accuracy) is minimised by the
accurate spot results from the coring, which allow calibration of the radar results and
further verification. In such an investigation there is an opportunity to gather additional
information by examining the core-holes using an endoscope (to assess condition in situ
and inspect the area behind the lining) and to use the masonry cores for laboratory-based
examination and testing (to determine materials characteristics and condition). For a
similar investigation on other types of lining radar surveying might not be suitable for
example, the success of radar on reinforced concrete strongly depends on the amount and
spacing of reinforcement, and in metallic linings other techniques would be required.
The following sections include descriptions of some of the main non-destructive
investigation techniques and comments on their strengths and limitations for use in
characterising features of tunnel linings and the surrounding ground.
420
Table A4.3
Usefulness of engineering geophysical methods for geotechnical investigations (BSI, 1999)
Groundwater exploration
Water quality
Porosity
Permeability
Temperature
Flow rate/direction
Buried channels
Clay pockets in limestone
Sand and gravel
Basic igneous dykes
4
1
1
2
0
0
0
0
0
4
1
2
1
Reflection – land
2
2
2
1
2
0
0
0
2
1
2
0
0
0
0
0
1
0
0
1
Reflection – marine
4
4
2
2
4
0
0
1
0
2
0
0
2
0
0
0
4
0
0
0
Acoustic tomography
2
2
3
3
1
4
2
2
3
2
0
0
0
0
0
0
2
0
1
2
Cavity detection
2
Rippability
3
Density
4
Dynamic elastic modulus
4
Fault displacement
3
Fractured zones
4
Lithology
4
Stratigraphy
Refraction
Geophysical methods
Depth to bedrock
Buried artefacts
Applications
Seismic
Electrical
Resistivity sounding
4
3
3
2
2
0
0
1
2
1
4
4
3
1
0
0
3
0
3
0
Induced polarisation
2
2
3
1
0
0
0
0
0
0
3
1
3
2
0
0
2
1
1
1
Electromagnetic and
resistivity profiling
3
2
2
4
1
0
0
0
3
3
4
4
1
0
0
0
3
4
3
3
Electrical imaging
4
3
3
3
3
0
0
0
3
1
4
4
3
4
0
0
3
4
3
3
Other
Ground-probing radar
2
3
1
2
3
0
0
0
3
4
2
2
1
0
0
0
2
2
1
2
Gravity
2
0
0
0
2
0
2
0
2
1
1
0
0
0
0
0
2
1
1
2
Magnetic
1
0
0
0
2
0
0
0
2
3
0
0
0
0
0
0
1
3
0
4
Borehole Logging
Self-potential
2
4
4
1
1
0
0
0
1
1
4
2
0
0
0
0
0
0
0
0
Single point resistance
2
4
4
0
0
0
0
0
0
0
4
2
1
0
0
0
0
0
0
0
Long and short, normal
and lateral resistivity
2
4
4
0
0
0
0
0
0
0
4
2
4
0
0
0
0
0
0
0
Natural gamma
2
4
4
0
0
0
0
0
0
0
2*
2
1*
3*
0
0
0
0
0
0
Gamma- gamma
3
4
4
0
0
0
3*
0
0
0
2*
0
3*
2*
0
0
0
0
0
0
Neutron
2*
4
4
0
0
0
3*
0
0
0
3*
0
3*
2
0
0
0
0
0
0
Fluid conductivity
0
1
0
0
0
0
1
0
2
0
4
4
4
1
0
0
0
0
0
0
Fluid temperature
0
0
0
1
0
0
0
0
1
0
2
3
0
0
4
2
0
0
0
0
Sonic (velocity)
3
4
2
3
0
3
2
1
2
0
1
0
1
0
0
0
0
0
0
0
CIRIA C671 • Tunnels 2009
421
A4.3.1
Infrared thermography
Infrared thermography converts temperature differences into a thermal image. A typical
thermographic imaging system comprises a hand-held infrared camera which is self
powered and is operated in a similar fashion to a typical video camera. Data can be
recorded as both stills and video images. This technique is able to distinguish between
materials with different thermal properties, and has been applied successfully in tunnels to
identify hidden changes in construction and also to locate defects such as voids, indicate
areas of dampness and locations where water ingress was occurring (Mitchell and Jardine,
2002). However, a prior condition for the successful application of this technique is the
requirement for a temperature gradient, through the lining or within the ground, of at
least 2°C, making application within tunnels more difficult as they are often stable thermal
environments away from the portal areas.
Rapid tunnel survey method allowing tunnel surveying at speeds of several kilometres per hour.
Strengths
Rapid above-ground survey method allowing very high rates, particularly as an aerial tool for hidden
shaft identification
Requires a temperature gradient through the lining or within the ground of at least 2°C.
Limitations
A4.3.2
Survey image can be obscured or confused by, for example, areas of dense ground cover, moisture
build-up and other local heat sources.
Gravimetric survey
Gravimetric methods comprise the measurement of local variations in the Earth’s gravity
field caused by density changes relating to different subsurface rock and soil types and
their content.
High-resolution microgravity surveys are employed in engineering investigations and are
capable of a resolution to a height of less than 20 mm and position within 10 m. This
technique is generally slow as measurements take several minutes each while the sensitive
equipment is stabilised. The success of the technique directly depends on the relationship
between the depth and the size of the feature and the relative difference in density
between it and the surrounding geology or structure. It is commonly used to gather data
on a feature whose location is well known or to support results of more general overview
techniques.
A4.3.3
Strengths
A sensitive method with a track record in accurately locating features such as shallow voids and mine
shafts. Not adversely affected by interference from the electromagnetic sources or ferrous objects that
limit the use of many geophysical methods.
Limitations
Relatively slow rate of progress, most useful for discrete measurements at a single location.
Local movement, vibration or other site activity is likely to affect the success of a measurement.
Use within a tunnel may be complicated, as the ground overburden profile will affect the results. It may
be more appropriate for surface features eg hidden shafts.
Extensive data processing and expert interpretation of data is vital as a single gravity profile can have
a large number of geological solutions.
Magnetic survey
Magnetic survey techniques involve the measurement of variations in the Earth’s magnetic
field as subsurface materials with varying magnetic properties distort it. By using a
technique incorporating a digital data acquisition system and (above ground) a GPS
capability, the method can be continuous and rapid. The results of this method of
assessment for potential voids or hidden shafts can be produced on-site in the form of a
map.
422
This technique has been used for the location of buried mineshafts and buried metallic
objects such as pipes and storage tanks. It is also commonly used in archaeological surveys
and geological fault location.
A4.3.4
Strengths
Allows relatively rapid walkover survey and is easy to carry out in the field.
Microprocessor manipulation of the raw data has developed to increase the effectiveness of the
method.
Limitations
Difficult in areas of live electrical cables, for example, overhead electrification lines in the crown of
tunnels.
Results influenced by the presence of ferrous materials (present in blue engineering bricks and some
other types of brick) and carbon soot build-up.
Large survey area to distinguish deep anomalies from the background response data.
Ground resistivity surveying
Electrical resistivity measurements are undertaken by the insertion of probes into the
ground at set intervals, commonly along a pre-determined grid. In electrical resistivity
surveying, an electrical current is passed into the ground through two electrodes and the
voltage measured across a second pair of electrodes. Results are presented in terms of
apparent resistivity (AR) as the ground is assumed not to be fully homogeneous. This is
determined by calculating the resistance measured in respect of a constant that relates to
the spacing between the electrodes.
Three different survey techniques have been developed for different applications. These
are:
1
Constant separation traversing.
2
Vertical electrical sounding (VES).
3
Electrical imaging.
Resistivity techniques are well-established, and techniques capable of surveying to depths
of between 5 m and 20 m are used. Resolution is related to the degree of contrast between
the resistive properties of the features being surveyed. The technique is suitable for use in
most ground types and in solid materials such as brick or rock.
A4.3.5
Strengths
A relatively rapid walkover survey technique. Greater resolution is provided where used with fixed
electrodes pegged into ground. This is slower than the walkover reconnaissance method, but provides
greater penetration (12 m to 20 m). It is suitable for use in tunnel linings to locate hidden shafts
providing there is contact between the lining and the fill.
Limitations
Resolution of shaft heads that are buried at a depth greater than their diameter can be difficult.
Sensitive to materials within the ground that may be electrically conductive. So use within an urban
area with buried iron and steel services and abandoned articles could present difficulty.
Conductivity survey
Electromagnetic conductivity measurements are achieved by directing an electromagnetic
pulse into the ground and measuring the variations in the response based upon the
conductivity of the medium that the wave travels through.
The electrical conductivity of the ground is a function of several factors, including:
porosity
porosity geometry
CIRIA C671 • Tunnels 2009
423
variation of moisture
concentration of ionic salts
temperature.
Survey equipment ranges from a hand-held rod, with emitter and receiver aerial coils at
either end, for shallow investigations, to separate aerial units, positioned at measured
distances from each other, for deeper surveys. The technique does not require direct
coupling to the ground and is suited to rapid surveys. The smaller rod-type equipment
can be used for walkover surveys and all have the potential for road/rail vehicle mounted
surveys within a tunnel.
Figure A4.8
Resistivity survey traverse line laid out over a tunnel with suspected hidden shafts
The technique is used in civil engineering investigations to determine relatively shallow
features, particularly in assessing the moisture movement behind or within masonry or for
the identification of hidden composite construction features. The technique is commonly
used in association with other geophysical methods, particularly magnetic surveys, to
confirm the location of features.
A4.3.6
Strengths
Data download and assessment can be rapid, with the potential for on-site analysis.
Rapid survey method, without the need for contact with the ground or any other surface to be
surveyed. Can be used inside non-metallic tunnels to map moisture both within and behind the lining.
Limitations
Certain ground conditions can detrimentally effect the penetration depth of the method:
surface layers of clay soils can mask features at lower levels
moisture differences may need careful interpretation to avoid recording false subsurface ground
features.
The presence of ferrous metals or sources of electromagnetic interference such as high voltage cables
can mask local features if they are in close proximity to the feature of interest.
Seismic survey methods
Seismic surveys are well-established as a tool for geophysical investigations, particularly for
investigations of large features and rock strata. Major advances in this survey technique
have been in the area of signal enhancement using digital microprocessing to replace
previous analogue methods.
424
The method is based upon the recording of seismic waves created by a source, usually a
hammer or explosive charge. Waves travel through the ground and the time taken to be
recorded at the detector is measured. The wave returned to a detector geophone is either
by a route of reflection or refraction from a geological or man-made material layer
provided it has sufficient physical contrast from that of the upper surface ground layer.
This means that, for example, in searching for a hidden shaft, the surrounding ground
and the hidden shaft wall and shaft infill material will need to have a physical contrast for
this method to be successful.
Two distinct techniques have been developed that could be applied to surveys of geology
around tunnels and shafts, and for detection of hidden shafts, respectively based on the
reflection and refraction of seismic waves. The reflection method is often used for regional
engineering studies, when surveying ground geological structures in preparation for dam
or tunnel projects.
A4.3.7
Strengths
Potential to provide geological information for tunnels and shafts.
Reflection method allows the location of deep structures, although it is limited to examining features
less than 10 m deep.
Reflection methods are suited to providing survey information for cavities and shafts.
Limitations
Refraction method is only likely to be successful for the location of voids or hidden shafts up to a
maximum depth of 3 m.
Time consuming method, suited to the investigation of small areas rather than as an overall linear
survey technique.
Reflection methods are not suitable for identifying shallow features.
Requires specialist to interpret output of the survey, and not suitable where geology is complex.
3D seismic tomography
The technique of tomography is an improvement upon standard seismic and acoustic
techniques because it allows both direct and angular measurements of sound waves along
a large number of ray-paths that can be resolved into a 3D image of wave velocity
distribution by computer processing. Local variations in the velocity of the response wave
recorded relate to differences in continuity and materials within the ground where energy
waves will travel at different rates.
Tomography employs several transducers, normally positioned at a distance from the
seismic/acoustic source. It is important that sufficient data is measured to ensure that a full
image of the location of interest can be resolved. Data scatter between successive tests in
the same location is common and care in interpretation should be exercised to avoid the
identification of false anomalies.
This method is proven for the identification of voids, faults and hard inclusions within a
ground mass. The availability of more powerful computers has greatly influenced the
development by aiding the resolution of the responses from an array of receivers.
However, the use of multiple transducers and potentially several repeat measurements at
each location makes this a slower, more expensive alternative to the more simple seismic
techniques.
CIRIA C671 • Tunnels 2009
425
A4.3.8
Strengths
The technique is superior to standard seismic methods because it allows a local 3D image of the area
of interest to be resolved from an array of transducers
Modern computer techniques have allowed recent improvements in resolution of the responses and
computational data smoothing techniques.
Limitations
The technique employs a relatively large number of transducers to receive the response for each
measurement. Data smoothing techniques are likely to require several measurements to be made at a
single location, so:
each location will take an appreciable time to prepare for a test
technique may be more expensive than simple seismic methods.
Careful planning is required to ensure that the position and number of reading stations provides
sufficient results to map the test location.
Interference from local vibration sources, such as heavy plant working or being too close active road or
rail trafficking may affect the results.
Ground penetrating radar surveying
Ground penetrating radar (GPR) uses the reflection of pulses of electromagnetic energy to
detect interfaces in the surveyed medium whose components (or elements or constituents)
have contrasting physical properties.
The system comprises a transmitting radar antenna that introduces electromagnetic
energy in pulses into the ground. A corresponding receiving antenna detects the partially
reflected electromagnetic wave, and the identification of inconsistencies and features is
characterised by the strength of the reflected signal and the time it takes between
transmission and receipt. Depth of penetration is a function of the frequency of the energy
waves employed and resistivity of the substrate. Practical penetration depth, providing a
good resolution is up to four metres, but much greater depths are being achieved with
further developments in low frequency antenna, although the degree of resolution is not
as good.
The technique normally employs a linear traverse across the area of interest, with the
antenna coupled to the ground or structure surface. So progress is relatively rapid
(typically a steady walking pace) where a clear, uninterrupted access route is available.
Within a tunnel, multiple runs are often possible within a relatively short possession
period, allowing different areas to be examined or a change in antenna used to refine the
focal depth of the survey.
Figure A4.9
Radar survey from a track-mounted
cradle, with the aerial, mounted on the
end of a telescopic arm, swept over the
intrados at the location of a possible
hidden shaft (courtesy Aperio Ltd)
426
This method is becoming increasingly used for the investigation of civil engineering
studies and has been used successfully on tunnels, including identification of voids
between brick rings in masonry tunnel linings.
Strengths
Relatively rapid progress rates are possible. GPR surveys conducted from road/rail vehicles can enable
a survey team to progress at 4 8 km/h for a single profile, although multiple profiles may be
necessary for improved accuracy. GPR surveys within tunnels can help to determine lining construction
thickness and condition, variations in moisture content, overbreak characteristics, blind shaft location,
and invert characteristics.
Interpretation of objects with complex geometries, detected within the typical effective range of one
metre in high resolution mode and three metres in low resolution mode.
This method produces an output that requires detailed interpretation by an expert. Success is directly
related to the competency of the expert.
Features and their resolution can be hidden by radar reflective materials such as metals, which can
severely limit the effectiveness of the survey. Masking scenarios in tunnels include:
tunnels with metallic linings, reinforced concrete with multi-layered or dense reinforcement, or
areas of shotcrete where metal reinforcement is used
areas of engineering bricks (blue bricks), which have a high ferrous content
thick soot deposits, which are highly conductive
electrical services and wiring.
Limitations
Features may not be detectable if their electromagnetic response is similar to that of the surrounding
material, for example, where:
infill material in shafts is composed of a similar material to the adjacent ground and of the same
compaction
brick linings and shaft brick haunch/lining may not exhibit a clearly distinguishable boundary
where they are backed by a well-compacted clay soil.
Conductive (wet and particularly saline) materials can cause problems by severely attenuating
electromagnetic waves, and results may be poor in clays and saturated soils.
Ground radar systems are low-powered radio transmitting devices and the possibility of interference
with communications cables should be considered.
A4.3.9
Transient electromagnetic (TEM) scanning
The transient electromagnetic (TEM) technique employs current pulses instead of the
constant waves which are applied in the standard GPR method. This induces eddy
currents within conductive media, producing a secondary magnetic field that can be
detected via a receiver coil as a time dependant decaying electrical signal. The
conductivity and geometry of a ground feature can be determined on analysis of the
transient electrical decay response.
A TEM scanning technique is commonly employed to provide a rapid survey of substrates
from a few metres in depth to several hundred metres deep.
A4.3.10
Strengths
Rapid survey method.
Capability for deep penetration at a test location, in excess of 100 m.
Limitations
The technique is affected by electrical and magnetic sources, eg by the proximity of live power sources,
including electrification or service cables.
Urban areas with a large incidence of buried services and objects are likely to obscure the response.
Broadband electromagnetic scanning
Broadband electromagnetic (BEM) scanning is a non-destructive technique developed to
identify anomalies and to undertake thickness measurements of ferrous materials, such as
cast iron segments. In simplistic terms, when a ferrous metal is subjected to an
electromagnetic field it generates an electrical current that induces a secondary magnetic
field in the material. A series of such fields can be generated at different frequencies by
transducers and can be measured in a broadband spectrum to provide an estimate of
material thickness. When this information is interpreted, thickness contour plots can be
produced that indicate material anomalies, such as cracks and casting defects within the
CIRIA C671 • Tunnels 2009
427
sample. This is a relatively new technique which has been trialled by LUL and although
further development is required, encouraging results were obtained (Chew and Roberts,
2005).
A4.3.11
Strengths
Can potentially measure cast iron lining thickness through secondary linings.
Provides a large amount of data suitable for statistical interpretation to highlight areas of concern, ie
areas of reduced section thickness, cracks, hidden blowholes.
Limitations
Equipment needs further development before it is suited to routine use.
The technique is affected by electrical and magnetic sources, eg by the proximity of live power sources,
including electrification or service cables.
Ultrasonic Pulse Velocity (UPV)
Ultrasonic pulse velocity (UPV) is a non-destructive technique used in testing a wide range
of building materials to determine properties including strength and elastic modulus, and
to investigate defects such as the presence of delamination and depth of cracking.
Although its principal use is the thickness measurement of, and identification of flaws in,
metallic elements.
a
Figure A4.10
b
Surface preparation (a) and subsequent ultrasonic testing (b) of cast iron segmental lining
(courtesy Aperio Ltd)
The technique is based on the velocity of ultrasonic pulses travelling in a solid material
and depends on the density and elastic properties of the material. The quality of some
materials is related to their elastic stiffness so that measurement of ultrasonic pulse velocity
can often be used to evaluate this. It is also possible to detect voids and cracks within
otherwise solid materials, and even to map the extent and depth of voiding and
delamination.
428
Strengths
System is small, portable and self powered.
Accurate depth readings can be gained.
Confidence in accuracy is increased with multiple readings, eg where measurements are made along
a profile.
Routinely used method for assessing thickness of metallic elements.
Limitations
Needs intimate contact with the surface under investigation.
Some systems require the use of coupling gel to work.
Individual readings are subject to relatively high levels of variability.
A4.4
TECHNIQUES FOR MONITORING
Periodic inspection and manual measurement is the cheapest form of monitoring and is
generally very effective. However, it does have limitations and there are a variety of
circumstances where it is necessary to install systems of instrumentation to carry out
specific monitoring tasks, for example, where it would be unfeasible, unsafe or
uneconomic to visit the monitoring location to make measurements at the necessary times
and frequencies.
Considerations for using monitoring systems within tunnel environments (power, cabling,
data transmission, access, reliability etc) and system design are considered in Section 4.5.
A4.4.1
Crack monitoring
Crack monitoring can be carried out in several ways with relative variations in
sophistication, from the most basic crack width gauges (tell-tales, graduated plastic rules)
to data-logging systems continuously measuring movements at sub-millimetre accuracy, as
well as other parameters such as temperature.
Although cheap and easy to install and read, crack width gauges are prone to breaking off.
Also, they need to be read manually and have to be accessible, which may be a problem in
some tunnels. They also only give a snapshot of the movement that may reflect, for
example, seasonal temperature variations rather than a deteriorating structural condition.
It is clear that Avonguard type gauges or similar may not be appropriately placed across
some of the joints in older brick lined tunnels. Demec studs, which are installed on the
structure and can be measured using a dial gauge, have the benefits of being
straightforward and relatively simple to use, and give a good accuracy and repeatability.
It has been proposed that a thinner-than-normal sprayed concrete lining to a brickwork
tunnel would help to support it and could act as a very large tell-tale around the structure
by cracking when movement occurred, however the authors are not aware of this system
having being used in practice.
More sophisticated (and expensive) automated systems are available, usually incorporating
an extensometer which can be read remotely via a data-logger. A thermometer can be
incorporated to allow displacement and temperature to be logged simultaneously,
enabling thermal effects and cyclic temperature variations to be taken into account where
a high level of accuracy is desirable.
A4.4.2
Displacement monitoring
Displacements between points on a structure are usually measured using potentiometric,
linear variable displacement transformers (LVDTs) or occasionally manually-read dial
gauges. These transducers should be mounted directly or via invar wires to an
independent frame. Invar wires are used to minimise temperature effects and are kept
short to reduce wind induced oscillation and disturbance from passing vehicles. They
represent a relatively straightforward and robust technique with proven accuracy which
has been used successfully in a variety of monitoring situations, although the need to
physically connect the measurement points using invar wires can create obstructions,
making them unsuitable in some circumstances.
There are other ways of measuring displacements such as electro-levels, laser scanning
techniques and photogrammetry. Automated robotic laser theodolite systems can be
programmed to make repeated measurements to several reflective targets attached to the
CIRIA C671 • Tunnels 2009
429
structure, and can be used to determine their displacement under changing loads – either
relative to one another (identifying changes in shape) or relative to a reference datum
which can be positioned outside the potential zone of influence. For greatest accuracy, the
targets should be reflective prisms but cheap stick-on reflective targets are also available.
This technique has been used successfully for monitoring tunnel response to changes in
loading during grouting operations (see Case study A1.2). Laser surveying and
photogrammetric methods have also been used to identify and monitor changes in the
intrados profile of rail tunnels over time, making periodic readings referenced to a set of
datum points established in a stable area. Such systems are likely to have a time lag in
results being available, depending on the number of monitoring points included in each
round of measurement, which can affect their suitability for measuring rapid dynamic
changes. Their accuracy can be affected by atmospheric distortion where there are
extremes or significant variations in, for example, air temperature, moisture or airborne
dust.
For monitoring applications, techniques such as laser-scanning and photogrammetry/
videogrammetry may find increasing use in the periodic measurement of large and
complex areas, particularly where it is difficult to foresee the location of potential
movements and measurement points cannot be clearly defined. In such circumstances
conventional survey and instrumentation techniques, which are suitable for making
measurements to specific pre-defined locations, would be unsuitable. The production of a
comprehensive virtual archive of the measured elements provides the user with great
flexibility, allowing any points to be chosen and measured as required.
Tiltmeters (electrolevels) are used to detect and measure angular displacement of the
elements on which they are sited. They are small fluid based sensors with no internal
moving parts so are robust. The system is energised with an electric circuit and the output
depends on the amount and direction of tilt of the instrument. The sensors can be
mounted within beams that may be fixed mechanically to the structure. Relative
displacement between points on a tunnel intrados and changes in shape can be detected
with great accuracy using a string of several electrolevel beams connected in series,
mounted on pins so that they tilt as the structure moves. The Bassett convergence system
was developed in a London Underground Ltd running tunnel before its installation in
existing tunnels requiring deformation monitoring during construction of the Jubilee line
extension project (Bassett, 1999). The system can be configured as a closed loop or left
open-ended. With closed loops, a correction can be applied to minimise the effects of
errors. Open-ended systems are referenced to conventional surveys periodically unless
one end is known to be stable. Once the instruments are fixed in place they can be zeroed
and any subsequent movement logged at user defined intervals. Because they can be
difficult to calibrate, they can be used in conjunction with digital tape extensometers to
improve their reliability. A measurement system based on electrolevel sensors can produce
results which are very close to real time with very high accuracy and sensitivity (detecting
movements of tenths or even hundredths of millimetres) but they require expert
installation and a full series of beams can be expensive. An installed array of electrolevel
beams reduces available clearances in the tunnel, which may preclude their use in some
situations, although low-profile custom-shaped beams are available to minimise clearance
requirements.
The sensitivity of certain electrical transducers such as LVDTs to dust, vibration and
electromagnetic fields makes them unreliable in some applications, and electrolevels are
considered to be more robust (Basset et al, 1999). Strings of electrolevel beams are capable
of high angular resolution, with accuracy of measurement related to the number used and
their configuration. Individual three metre beams are capable of resolution to
approximately 0.01 mm over their length. Sub-millimetric measurement accuracy is
possible using both LVDTs and electrolevels and, with careful system design,
430
displacements in the order of 0.001 mm can be measured with a repeatability of
approximately 0.01 mm (Kimmance et al, 1999). The achievable accuracy of systems
employing surveying techniques and equipment may be less, a few millimetres, although
in suitable circumstances with careful system design automated systems can achieve
accuracies of +/- 2 mm or less which is adequate for many purposes. The case studies in
A1 involved the successful application of automated structural movement monitoring
systems based on both electrolevels and surveying equipment.
A4.4.3
In situ strain monitoring
Accurate strain measurements are often made using vibrating wire (VW) gauges. These
comprise a fine wire tensioned between mounting blocks and protected by a tube. The
wire is plucked by a coil that also reads the frequency at which the wire vibrates.
Temperature can also be monitored. A resolution of 0.5 micro-strain over a range of 3000
micro-strain is achievable for a gauge with a gauge length of 140 mm. Installation can be
difficult particularly when the surface is wet or cold. Masonry is often irregular resulting
in the necessity to provide make-up shims to ensure that the wire’s protective tube does
not make contact with the surface. Also, the adhesive should be selected to ensure rigid
attachment. If the surface layer of material is suspect then careful consideration to
installation is necessary and may result in the need for steel pins or bolts being installed to
which the VW gauge can be secured. The main disadvantage of VW gauges is that they
need a finite period of time to be read and so cannot be used in dynamic load situations.
They are also sensitive to electromagnetic fields, which can cause problems if they are
used in proximity to electrical cabling.
Other transducers, including linear variable displacement transducers (LVDTs), linear
potentiometers, foil (electrical resistance) strain gauges and also fibre-optic based
measurement systems are available, their suitability depending on specific monitoring
needs and the environment.
Although more commonly used for measuring crack movements, Demec (demountable
mechanical) gauges can be used for strain measurement. They comprise a spring-loaded
lever system operating a dial gauge. Pins protruding from the instrument are located in
pre-drilled studs that have been fixed to the structure. The dial gauge is read manually.
Several gauge lengths are available (100 mm, 200 mm and 250 mm are the most
common). It is important to use the same dial gauge for all readings. The accuracy of this
method may be affected by temperature but it is possible to compensate for this if readings
are properly referenced to the accompanying invar reference bar.
A4.4.4
In situ stress monitoring
As discussed in Section A4.2.4, there are various techniques which allow the measurement
of in situ stress in rock masses or lining materials. Making repeat measurements of stress
over a period of time could provide a method for the early detection of stress changes,
before their results (distortion and structural distress) become detectable by other
monitoring techniques such as visual surveying or repeated dimensional measurement.
This provides advanced warning of structural deterioration, allowing remedial measures
to be instigated where necessary. Although currently the authors are not aware of stress
monitoring being used in this way, in the future existing methods may be used to develop
this useful technique for carrying out periodic health checks on structurally sensitive areas
of lined and unlined tunnels.
CIRIA C671 • Tunnels 2009
431
A4.4.5
Corrosion monitoring
There are a wide variety of techniques available for measuring and monitoring corrosion
in metallic tunnel linings, differing greatly in their complexity and the associated cost.
Techniques can be based on mechanical, electrical or electrochemical devices that
determine the corrosive potential of the environment and rate of metal loss. The most
commonly used devices are corrosion coupons (weight loss measurements), electrical
resistance techniques and linear polarisation resistance techniques. Other techniques are
also used, for example, galvanic techniques, hydrogen penetration and microbial/biofilm
techniques, however these tend to be more specialised and may not be appropriate for
most tunnel monitoring.
Corrosion coupons are tag-like samples of metal exposed to the working environment for
a given duration of time which are then removed for analysis. The corrosion rate of the
metal is then determined from the sample weight loss over the exposure time. Typical
exposure times are about 90 days, thereby providing quarterly corrosion rates. The
technique is very simple to install and coupons can be fabricated from any commercially
available alloy. However this method has restrictions when trying to determine if localised
corrosion is present in an existing structure. Coupons are mostly of benefit in
environments where the corrosion rates do not change over significant periods of time, as
is often the case in tunnels.
Electrical resistance monitoring is an in situ technique that uses electrical resistance probes
to measure metal loss at any point in time. This method provides a read-out of the section
loss based on the changes in resistance of the metal with changes in section thickness.
Readings can be taken at any point in time and the probe remains in the environment.
This method also has the advantage of being able to incorporate alarms that can be
triggered by particular factors, such as when a section reaches a minimum thickness. They
also have a quick response time to variations in the environment.
Linear polarisation resistance monitoring is an electrochemical technique whereby the
corrosion rate is determined directly by application of a small voltage to electrode material
in solution. This method has the advantage of immediately providing the corrosion rate
without the need for further calculation, however it is sensitive to contamination and the
presence of oil will affect results. Linear polarisation techniques have also been used
successfully for estimating corrosion rates of reinforcing steel in concrete structures, and
apart from requiring a small breakout to expose the steel for connection of an electrode, it
is non-destructive.
In addition to these monitoring techniques there are non-destructive testing techniques
that can be used to identify corrosion, for example, radiography, or the likelihood of
corrosion, such as the well-established technique of electrical half-cell testing for concrete
reinforcement.
432
A5
Detection and location of hidden shafts
As discussed in Chapter 3, it is important that tunnel owners are aware of the location of
all tunnel construction shafts to ensure that they are safely managed. This may require
measures to identify the risk of any unknown and possibly hidden shafts which do not
appear on existing records and cannot be confidently detected from visual inspection of a
tunnel and the ground above it. Tunnel owners may have their own policies and
procedures on the location and management of shafts, but the following is an example of
how one major infrastructure owner is dealing with this issue.
Once a hidden shaft is located, it will be necessary to carry out a detailed examination to
evaluate its construction and condition, the level of risk it represents to the tunnel and the
ground above it, and to determine the best course of action in dealing with it. Details of
shaft management procedures are provided in Section 3.8.
A5.1
MULTI-PHASE APPROACH TO SHAFT LOCATION
Network Rail has adopted a multi-phase approach to ensure that the location of all hidden
tunnel shafts is identified (Network Rail, 2004b), as described below:
PHASE A: desk study which aims to find independent records of the existence of a
tunnel shaft and its location to a tolerance of ±10 m. A reconnaissance walkover
survey could be included in this Phase, if deemed necessary, but otherwise is included
in Phase B
PHASE B: non-intrusive investigation, including walkover and walk-through surveys,
and using non-destructive techniques from the ground above the tunnel (eg radar,
magnetic and seismic survey techniques) or from within it (eg radar and ground
resistivity surveying), to identify features which might be associated with a shaft
PHASE C: intrusive investigation of areas where Phases A and B suggest a strong
likelihood of the presence of a shaft (using techniques such as boreholes and ground
probing, trial pits, penetration tests) to confirm its presence and location.
This process provides a method of efficiently investigating potential shaft locations with
increasing confidence until a stage is reached at which confidence is sufficient to discount
the existence of a shaft or to confirm its presence and location.
A5.1.1
PHASE A
A study of existing historical records where available may indicate the presence and
possibly the location of unknown shafts, which will guide the following, confirmatory,
phases of investigation. However, it cannot be assumed that historic records will be
accurate, comprehensive or available in this respect and a general knowledge of early
tunnel construction methods is useful in assessing the likelihood of unrecorded shafts
being present.
Early site investigations or borings taken at the surface along a rudimentary alignment
given by a simple horse drawn rope, were common in the first canal tunnels and these
were soon turned into access points or shafts into the tunnel to accelerate progress. Shafts
were common on very long tunnels but also played a part in the shorter tunnels. Where
very long cuttings were to be formed either side of a tunnel, it was found worthwhile to
CIRIA C671 • Tunnels 2009
433
sink a shaft at a convenient point along the proposed tunnel, for example, near an access
road (for brick supplies), and then proceed with the tunnel construction at the same time
as the cuttings. Very short tunnels such as Porthkerry No 2, 73 yards long are thought to
have had shafts sunk in this situation. Distances between shafts were variable, depending
on the ground encountered, length of tunnel and programme of work. Simms (1844)
stated that: “in ordinary kinds of strata, if the work is required to be completed in twelve
months, the interval between the shafts should not exceed 100 yards, but I believe about
200 yards has been the distance more generally adopted”. The distance of 200 yards can
be seen on several tunnels and relates to an accident during the construction of the
Watford Tunnel where a shaft and a partially completed section of tunnel being
constructed at the bottom of the shaft collapsed. Learning from this accident it was
decided to go at least 50 yards from the shaft before constructing a break-up (see Gripper,
1879) for the first section of tunnel and then working back towards the shaft. This distance
of 50 yards was found to be an ideal length of tunnel to construct and a single break-up
between shafts would give shafts at 100 yards or 50 yards each way to the shafts and later a
break-up at 100 yards from the shaft would give 200 yards between shafts and four
sections of 50 yards of tunnelling. However with or without deep cuttings either side of
the tunnel it was found difficult to start-in at a semi-vertical face as full cover to the tunnel
was difficult to achieve. A shaft close (50 yards) to the portal face allowed the tunnel to be
constructed in relative safety from the inside working towards the portal. Bad ground or
shallow cover also made for more frequent shafts as seen at Sugar Loaf Tunnel. There is
thought to be no fixed distance for shaft spacings and it is likely that conditions such as
adjacent cuttings, bad ground, access roads and engineers preference all played a part in
the final layout of shafts.
This is applicable to older tunnels remaining in service from that period, as it pre-dates
the first use of dynamite (1866) that substantially aided the construction rate. By 1879
advice on the necessity for working shafts was that the possible advance per annum, from
a working shaft, including both working faces, was 300 yards (274 m) (Gripper, 1879). It
appears that shafts where good tunnelling conditions were present were equidistant and a
mid-point permanent ventilation shaft was often used where the tunnel was at least 700
yards (646 m) long. It is important to capture this level of construction detail in the
planning phase of a tunnel survey so that the correct assumptions are made at the time of
the investigation.
As part of the desk study, the earliest aerial photographs available should be used to trace
outlines of spoil heaps or even original shaft chimneys built above ground but now
demolished and backfilled.
Following a comprehensive desk study, Phase B of investigation (reconnaissance and
surveying) is undertaken or a purely intrusive approach (outlined in Phase C) employed
where more reliable location information exists.
A5.1.2
PHASE B
Investigations to identify and confirm the presence of unknown shafts can be considered
as three main activities:
1
Overview reconnaissance survey.
2
Detailed above-ground survey.
3
Internal tunnel survey (both reconnaissance and detailed).
Where ground conditions are relatively open and features clear, useful reconnaissance
information could be gathered by a visual walkover survey, particularly where it can be
434
combined with other, more detailed, methods. At locations where access to the ground
surface is limited or a strategy of investigating ground features for several tunnels is
planned, an aerial technique may be the most appropriate.
When carrying out reconnaissance surveys, there are some visual clues indicating the
presence of hidden shafts from within the tunnel, and can be seen during visual
inspections, particularly:
circular damp patches on the crown of the tunnel
variations in the brickwork pattern, including the evidence of keystones or more
recent repairs
timber identified in brickwork linings.
Also, there may be clues on the ground surface along the vertical alignment of the tunnel
circular spoil heaps (visible on aerial photographs, see Figure A5.5)
stones or some other type of marker post.
Although observation of such features is useful in suggesting the possibility of a hidden
shaft and its location, they do not provide conclusive evidence, and not all shafts will
present such clues. Spoil heaps can give clues during the walkover but can be some
distance form the exact location of the shaft. Visual observation should not be relied upon
as the sole means of hidden shaft location.
Figure A5.1
Chimney clearly marks the location of a shaft at ground level – not all are easily located
Once a reconnaissance survey of the tunnel has been carried out, potential shaft locations
may have been identified and further investigated at a more detailed level of survey.
Alternatively, where there is no visible evidence of hidden shafts but the presence of shafts
is suspected for other reasons, for example, historical records or supposition based on the
construction method of the tunnel and its length, a more extensive survey along the
tunnel’s whole length may be necessary. These surveys will normally involve the use of
geophysical ground investigation techniques (NDT techniques) to:
confirm the existence and location of suspected shaft sites highlighted by existing data
or documentary evidence
CIRIA C671 • Tunnels 2009
435
confirm the existence and location of suspected shaft sites highlighted by the
reconnaissance survey data or other survey methods where anomalies have been
discovered
discover the existence of a shaft where the presence is suspected but no clear evidence
of its existence or indication of its location has been found or produced.
The decision on the most appropriate type and method of survey largely depends on site
details and availability of access to the tunnel. However, as a general approach the survey
strategy would be expected to consist of two or three distinct phases, either within or
outside the tunnel, assuming that no documentary information identifies shaft locations
before the site survey work. These are:
1
Reconnaissance: these methods are the most rapid and provide a low level of detail
or resolution. Success in identifying a hidden shaft location is not expected to be
good, although this is offset by the rate of progress and low cost.
2
Detailed line survey: these methods provide a more detailed approach along the
entire tunnel and employ NDT techniques which are relatively rapid while providing
an intermediate level of detail.
3
Detailed survey of anomaly locations: this approach employs NDT techniques that
are capable of providing high resolution to investigate sites that have been identified
as probable locations of hidden shafts. However, these methods are not suitable for an
overall survey of the tunnel, unless in a low resolution mode.
An evaluation of the usefulness of a variety of geophysical and other NDT methods carried
out for Network Rail by Mott MacDonald Ltd. Consulting engineers concluded that the
following methods were most likely to be successful in the detection of hidden shafts:
Table A5.1
Survey techniques for the identification and location of hidden tunnel shafts
Survey type
Reconnaissance
Figure A5.2
436
Technique
Aerial survey
Photography (inc. stereo-pairs)
Thermal imaging
Walkover survey
Ground resistivity
Magnetics
Ground conductivity
Detailed investigation from ground surface
Ground conductivity
Ground resistivity
Microgravity
Detailed investigation from within tunnel
Ground penetrating radar
Ground resistivity
Ground conductivity
Results of a ground resistivity survey traverse, taken along the crown of a masonry tunnel
lining that clearly identify a potential location of a hidden shaft (courtesy Aperio Ltd)
Figure A5.3
Results of a ground penetrating radar (GPR) survey traverse along the crown of a masonry
tunnel lining that identify a potential shaft location of a hidden shaft (courtesy Jack Knight
and Aperio Ltd)
Table A4.2 includes further information on the potential suitability of engineering
geophysical techniques for the detection of cavities.
When planning a survey it is important to optimise the technique(s) selected based on
important information known about the site, including:
obstructions to survey, eg vegetation and buildings on the surface or electrification
equipment inside the tunnel
ground soil/rock type(s) present
water table data, moisture condition at the location
location of adjacent development
local power sources
known buried services or structures (redundant and commissioned)
power sources within the tunnel (for internal survey methods).
It is likely that the condition and status of a hidden shaft will be unknown and success will
be guided by the difference in specific physical characteristics between the shaft and
surrounding ground whether it is infilled or present as a void.
Section A3.1 includes more information on these survey techniques and considerations in
their use.
On completion of this phase of survey work, where evidence has been found to establish
the possible or probable locations of suspected or known shafts, their presence may be
confirmed by intrusive investigation in Phase C.
A5.1.3
PHASE C
Following completion of Phase A and Phase B (where appropriate), intrusive methods for
the confirmation and identification of shafts can be used. The recommendations in this
section are largely drawn from CIRIA Special Publication 32 Construction over
abandoned mine workings (Healy and Head, 1984, reprinted in 2002) together with the
unpublished Network Rail Good Practice Guide Location and treatment of tunnel shafts
(Network Rail, 2004b).
CIRIA C671 • Tunnels 2009
437
The Phase A investigations should indicate the possible presence of a shaft or shafts,
although records are often imprecise about locations. Occasionally, different records show
several locations for the same shaft. Such findings necessitate a more extensive area to be
covered by an intrusive investigation.
Intrusive shaft location methods include:
trial pits and trenches
probing and drilling
remote visual inspection.
Trial pits and trenches
Trial pits and trenches are a quick, cheap method for investigating the ground to shallow
depths above the groundwater level. The depth of pits can be up to four metres, but can
be extended to six metres. Temporary support should be supplied if personnel are to
enter the pits.
Excavation is the most effective methods of finding shafts, but these methods may not be
possible where the ground surface cannot be disturbed by excavation (for example, in
built-up or environmentally sensitive areas) or where there is a substantial depth of
overburden. Knowledge of the probable shaft diameters should be used when
determining the spacing of pits or trenches (eg shafts of 1.0 or 1.5 m width may be missed
if excavations are 2.0 m or more apart). A series of parallel trenches are excavated where
the distance between the near walls of adjacent trenches is no greater than about ¾ of the
expected shaft diameter or ¾ of the minimum width if shafts were rectangular.
Typical indications of the presence of backfilled shafts revealed by trial excavations
include:
a different colour, density or composition of soil from the surrounding ground
inclusions in the soil (eg timber, brick or other exotic materials) indicating that it is
disturbed ground
remains of the headworks, linings or capping materials.
Probing and drilling
Where trenching and pitting cannot be used for any of these reasons, probing and drilling
can be undertaken. This is carried out by making small diameter vertical open holes with
rotary drilling or dynamic probing equipment.
A shaft can be detected by the difference in resistance to penetration from that of the
surrounding ground and, if rotary drilling equipment is used, by the difference in colour
and nature of the arisings.
The most efficient exploration sequence is a spiral pattern (see Figure A5.4) on a grid of
dimensions appropriate to the probable shaft diameter. Typical investigations to locate a
shaft comprise holes spaced at about 1.5 m centres over an area of 30 m x 30 m centred
on the supposed position of the shaft.
438
Figure A5.4
Drill sequence to locate obscured shaft when position has been reasonably well-established
Each hole should be made to a depth that confirms it is either in stable ground or in a
shaft and often this will be 10 m or more. Drilling is expensive for general searching of a
site, as a large number of holes would need to be formed. In soft ground, dynamic
probing, in which a conical pointed bit is driven by percussion methods into the ground
and the blows taken counted, provides a cheaper method. However, no arisings are
produced so it is not appropriate where the shaft backfill may offer similar penetration
resistance to that of the surrounding ground. Figure A5.5 shows a dynamic probe rig
undertaking an investigation to locate a potential hidden shaft, which involved 20 holes to
a maximum depth of 16 m. An example of the output is given in Figure A5.6, which
shows an iso-surface of equal resistance (blow count) outlining an area of soft ground that
indicates a possible shaft (see Case study A1.2).
Figure A5.5
Dynamic probing at possible hidden shaft location
CIRIA C671 • Tunnels 2009
439
Figure A5.6
Iso-surface interpolated from dynamic probing
to locate backfilled shaft
To penetrate rock, rotary drilling is required. A rotary hole can be cored, where a
continuous sample of the material penetrated is produced, or open holed without
obtaining a core sample. The latter method is faster and the rate of advance of the hole
can be monitored to provide an indication of ground density. Rock fragments returned to
the surface in the flushing medium can allow a crude log of the strata penetrated to be
produced. The greater speed of open rotary drilling makes this method significantly
cheaper than rotary cored drilling and is more suited to shaft location investigations.
Rotary coring is of limited use for shaft location as it is slow and costly. However such
techniques should be used to determine the ground conditions surrounding a shaft and
can form part of an integrated site investigation to locate shafts and provide geotechnical
remedial design parameters. Triple tube coring using diamond bits with polymer mud
flush is recommended for such investigations to ensure optimum core recovery and
quality. Coring can also be employed to determine shaft lining type by using angled holes
to intersect a known shaft location. All cores retrieved should be logged and
photographed in accordance with BS 5930 and ISRM recommended methods for rock
logging. Consideration should be given to the use of down-hole geophysical methods to
detect possible shafts adjacent to the cored holes.
If probing and drilling methods fail to locate known shafts, it may be necessary to excavate
the entire area around the supposed shaft location.
Remote visual inspection
Remote investigation of land-surface features can be carried out using remote sensing
methods, which use the art and science of observing and measuring items on the Earth’s
surface from a distance. By this definition remote sensing encompasses the field of aerial
photography, which is a useful technique for observing characteristic topographic features
and other tell-tale signs of ground conditions and excavation such as changes in
vegetation. Aerial photography has many potentially useful features:
440
it offers an improved vantage point, revealing features that may not be visible in a
ground-based walkover survey
it can cover an enormous area of land-surface in a short period
it provides a permanent record that can be incorporated in GIS systems
it has broader spectral sensitivity than the human eye
it has better spatial resolution and geometric fidelity than many ground-based sensing
methods.
The frequency range of photographically recorded visible light can also be augmented by
thermal/infrared frequencies, which may reveal additional detail.
Stereo-pair photographs, taken from two cameras with slightly different viewpoints, can
be examined using simple viewer and allow the user to benefit from the effect of
perspective to emphasise variations in topographic relief.
Aerial photography becomes increasingly economical as larger areas of land are mapped,
so there is benefit in combining the investigation of the land above several tunnels where
this is feasible. Sources of existing aerial photograph resources are identified in Section
A2.3.
Figure A5.7 shows an interpretation of possible hidden construction shaft locations based
on an aerial photograph of the land surface above a tunnel, based on evidence of a series
of earth excavations aligned at regular intervals between the portal locations.
Figure A5.7
Aerial photograph from which tunnel alignment and possible location of construction shafts
can be discerned based on topography. Ideally such photographs can be viewed as stereo pairs
A5.1.3
Note on safety
Extreme care is required when carrying out such exploratory work, because shafts may
suddenly collapse when the fill or a support staging is disturbed. Safety precautions are
necessary to protect personnel and equipment. When investigating shaft fill by drilling,
the rig should be supported on beams which extend a safe distance beyond the potential
shaft diameter, and all persons working in the vicinity should wear a harness attached to a
safety line and picket. This process can be expensive and slow if several positions are to be
drilled.
Geotextile mesh can be pegged out over the site as an extra precaution. The cost is
moderate when compared with drilling, and it may prevent injury should collapse occur
(but could still result in loss of equipment). Such grids have the advantage of durability for
reuse.
CIRIA C671 • Tunnels 2009
441
A6
Investigation and assessment of unlined
tunnels and shafts
A staged approach to the assessment of unlined tunnels and shafts through rock is
recommended, comprising an initial desk study followed by a reconnaissance visit and
finally a detailed rock mass survey.
The objectives of the assessment are to:
determine the rock mass classification
establish the stability of the tunnel/shaft
determine all geotechnical factors where the excavation geometry may be sensitive
develop rock joint map for each tunnel including existing or potential instabilities.
Once the assessment is complete, remedial measures can be designed.
A6.1
DESK STUDY
As with lined tunnels, a desk study should first be undertaken (refer to Section 4.2).
A6.2
RECONNAISSANCE VISIT
The principal objectives of a reconnaissance visit are to record:
general rock mass condition
main geotechnical and geological features
particular stability or condition problems
constraints to certain types of data collection (which may preclude a particular rock
mass classification scheme)
evidence of tunnel construction techniques and any form of tunnel support
access/safety requirements for the detailed survey.
A reconnaissance visit would comprise a visual assessment, and would allow prioritisation
of further detailed surveys either between tunnels or sections within a tunnel.
A6.3
DETAILED SURVEY
Detailed surveys describe fully the geological features and geotechnical condition of a
tunnel or shaft. The overall condition and individual stability problems are defined and
possible remedial measures can be described. Where particularly acute stability problems
are found, further detailed surveys can be recommended. The principal methods used are
scan line surveys to record representative discontinuity characteristics and rock mass
mapping to identify and define zones of similar rock mass condition and highlight
particular instabilities.
442
When assessing unlined tunnels and shafts, note that the engineering behaviour of the
ground is not solely controlled by the strength of an intact piece of rock but that the rock
mass as a whole should be considered. The rock mass is a body of rock that contains
discontinuities such as joints, faults and bedding. A detailed survey should characterise the
properties of these discontinuities to the rock mass to be classified so provide quantitative
data for condition zoning and remedial design.
A6.4
SCAN LINE MAPPING
Scan line surveys are used to obtain quantitative data on the discontinuities within a rock
mass to carry out a rock mass classification. They may be carried out at waist height on
each side wall and vertically across the tunnel cross-section. The properties of every
discontinuity that crosses an imaginary line (or one marked by a tape measure or string
tacked to the excavation wall) along the length of the survey are recorded. Rock surfaces
may require cleaning before detailed surveys are undertaken as accumulation of soot and
debris can obscure discontinuity properties.
Discontinuity properties should be recorded in accordance with International Society for
Rock Mechanics (ISRM, 1978) recommendations include:
dip
dip direction
chainage
discontinuity type
planarity
roughness
aperture
infilling
termination
persistence
JRC (joint roughness coefficient)
wall strength
weathering
seepage.
A summary of these properties is given in BS 5930 (BSI, 1999). These properties can be
used to classify the rock mass, as detailed in Section A6.6.
An example of scan line mapping carried out at Abbotscliffe tunnel (near Folkestone) is
given in Figure A6.1.
CIRIA C671 • Tunnels 2009
443
infilling material
D/c termination (lower,
upper) (ISRM, 1978)
Persistence measurmenet
D/c wall strength
D/c wall weathering
seepage rating (ISRM, 1978)
0.7
Low
Unstained/
tunnel dust
ii
Open and
infilled
Silty sand
DB
0.85
View
2
Open and
infilled
Silty sand
DD
0.18
Low
Slight black
speckling
ii
Matt
1–2
Open
Clean
DX
0.32
5 mm
Slight black
speckling
ii
Matt
1–2
Open and
infilled
Silty sand
RD
0.64
3 mm
Tunnel dust
ii
D/c aperture observation
DB
D/c aperture measurement
Clean
Surface appearance
ii
Joint roughness coefficient
Unstained/
tunnel dust
Intermediate scale planarity
(ISRM, 1978)
5 mm
Small scale roughness
(ISRM, 1978)
0.53
D/c dip direction
DB
D/c dip
ii
D/c type
Unstained
D/c number
7 mm
D/e set reference number
0.9
Chainage to end
TB
Chainage to start
Silty sand
chalk
3.1
3.15
A
A1
F
59
006
Rough
Planar
7
Matt
<1
Open and
infilled
3.3
3.4
A
A2
F
64
012
Rough
Undulating
12
Matt
3.3
3.55
A
A3
F
65
358
Rough
Planar
5
Matt
2
Open
3.45
3.6
A
A4
F
54
051
Rough
Undulating
7
Matt
1
3.8
3.9
C
A3
F
83
269
Rough
Stepped
18
Matt
4
4.35
C
C2
F
89
295
Rough
Undulating
7
4.8
5
A
A5
F
64
044
Rough
Planar
7
5.15
5.55
B
B1
F
56
141
5.4
5.55
B
B2
F
81
192
Rough
5.45
5.8
A
A6
F
71
033
Damaged
6.1
6.3
D
D1
F
77
131
Rough
7.2
7.5
7.8
7.8
D
D2
F
67
133
Tunnel
dust
<1
Open and
infilled
8.4
8.5
B
B3
F
77
162
Maske
d
0
Tight
Tight
D/L wall softened and damaged while cleaning
Planar
Undulating
7
12
Matt
Matt
ii
DR
iv
1
Open and
infilled
Silty sand
DD
1.1
2–3
Open and
infilled
Orange
silt
XR
1.8
7
Open
Clean
BB
1.2
2–3
mm
RR
0.8
7 mm
iv
BX
0.4
5 mm
iv
ii
Comminuted
chalk
iv
Tunnel dust
ii
Zone of fractures in crown of adit spaced 10 mm – 20 mm
Figure A6.1
Example scan line discontinuity log
A6.5
ROCK MASS MAPPING
Silty sand
In addition to the scan line surveys, mapping is carried out to quantify joint orientations
for stability assessments. Individual discontinuities are plotted onto plans of the tunnel or
shaft and dip and dip direction measured. These orientations are plotted on stereographic
projections (Hoek, 2007) to determine the number of joint sets. Observations of
kinematically feasible unstable blocks throughout the tunnel are also made and plotted
onto the rock mass map. A structural analysis can be carried out to define potentially
unstable blocks/wedges using the data on specific discontinuity orientations and conditions
defining specific blocks/wedges.
The rock mass map produced allows zonation of similar ground conditions and similar
remediation methods. An example of rock mass mapping is given in Figure A6.2.
444
Figure A6.2
Example discontinuity map
A6.6
ROCK MASS CLASSIFICATION SYSTEMS
The detailed assessments provide information that allows the rock mass to be classified
(see Bieniawski, 1989 for a full description). Rock mass classification systems provide a
means of developing a quantitative description of rock mass for use in engineering design.
The classification systems developed by Barton et al (1974) and the rock mass rating
system of Bieniawski (1989) are the most commonly adopted. The Barton system allows
derivation of a tunnelling quality index, Q, based on assessment of the following
parameters:
rock quality designation (RQD), which provides a measure of discontinuity spacing
joint set number (jn) is a measure of the number of joint sets in the rock mass, which
when combined with RQD allows assessment of block size
joint roughness number (jr) and joint alteration number (ja) provide a measure of the
shear strength and dilational characteristics of discontinuities based on assessment of
discontinuity wall conditions, aperture and infilling
joint water reduction factor (jw) is a measure of water pressure and its effect on
discontinuity strength and behaviour
stress reduction factor (SRF) is a measure of the potential stress conditions acting on
the tunnel due to either loosening loads, ground stress or swelling.
Bieniawski’s geomechanics classification rock mass rating (RMR) system classifies the rock
mass by assigning ratings to six basic parameters:
1
Uniaxial compressive strength of intact rock material.
2
Rock quality designation (RQD).
3
Spacing of discontinuities.
4
Condition of discontinuities.
5
Groundwater conditions.
6
Orientation of discontinuities.
Generally the geomechanics classification uses parameters that are assessed from site
investigation drill hole cores and may provide the more reliable assessments at that stage.
However, the Barton Q system is more applicable to mapping data.
The systems can be adopted to allow assessment of general tunnel stability based on
precedent experience because they include sufficient information to provide realistic
assessment of the factors that influence the stability of underground excavations. Based on
the numerous correlations between practical support experiences, from which
relationships between these classifications and specific support systems have been
developed, preliminary assessments of support requirements for tunnels can be assessed.
CIRIA C671 • Tunnels 2009
445
These can be adopted for existing tunnels without further analysis.
Care needs to be taken in amending the systems to cater for specific conditions as this may
invalidate the use of precedent experience from the database correlations.
Rock mass classifications also provide a means of deriving rock mass strength and
deformation parameters for use in rock mechanics design by means of empirical
correlations, which have been shown by experiment and experience to replicate
engineering behaviour. Again amendments to the systems can invalidate the applicable
correlations, so care needs to be taken in making any system site specific.
In certain situations it may be necessary to carry out more detailed analyses to understand
behaviour. Input to these analyses would be provided by rock mass classification derived
parameters, lab testing if appropriate and the rock discontinuity data.
The structural analyses previously described are broad brush analytical techniques that
allow support requirements to be assessed separately for stress induced deformations and
discontinuity controlled failures. Where a tunnel has been constructed in strong intact
rock or very closely jointed rock that may be considered as a continuum in relation to the
size of the excavation, computer modelling techniques such as finite element and finite
difference methods can be used. These methods may also be used if the discontinuities in
the rock are tightly closed. For continuum analyses, programs such as the finite difference
program FLAC (Fast Langrangian Analysis of Continua) can be used. Such programs
allow inclusion of typical ground support elements such as rock bolts, dowels and sprayed
concrete in the assessment of ground support interaction. A 3D analysis can be considered
for complex intersections or geology.
Where the rock mass discontinuity system in relation to the size of the tunnel or shafts
means that continuum analyses may be inappropriate consideration should be given to
discontinuum analysis using programs such as UDEC (universal distinct element code) to
assess the effect of rock mass behaviour. This program is a distinct element – boundary
element code that treats the rock mass as an assemblage of blocks interacting through
deformable discontinuities and can model rock deformability, fluid flow and ground
support interaction. However, the level of geotechnical data required in respect of the
discontinuity regime is greater than for a continuum model.
446
A7
Guidance on structural assessment
When considering appropriate analysis methodologies, care should be taken to ensure
compliance with any requirements of asset owners and/or overseeing authorities unless prior
approval of any variation is obtained.
A7.1
LIMIT STATE ASSESSMENT
Limit states represent states beyond which the structure no longer fulfils the relevant
assessment (design) criteria. Limit states are divided between ultimate limit states (ULS)
and serviceability limit states (SLS).
The limit states that concern safety of people and/or the structure are classified as ultimate
limit states. Assessing the structure against ultimate limit states means proving, with a high
level of confidence, that in a given assessment situation, loss of equilibrium of part or all of
the structure, failure by excessive deformation or transformation of part or all of the
structure into a mechanism, does not occur.
The limit states that concern the functioning of the structure or structural elements under
normal use, comfort of people and the appearance of the asset are classified as
serviceability limit states.
An assessment in accordance with the limit state principles can be carried out by using the
partial safety factors method. Actions, effects and resistances should be interpreted
according to the definitions given in the glossary (compliant with BS EN 1990, BSI, 2002).
Ultimate limit states are only discussed as serviceability limit states and are not assessed.
From an operational point of view, according to this method, the assessment involves
verifying that in all relevant assessment situations, namely under all possible actions on the
structure, no relevant limit sate is exceeded when assessment values for actions or effects
of actions and resistances are used in the assessment model.
The verification is carried out using assessment values for actions, effects and resistances.
Assessment actions and resistances are derived from representative loads and
characteristic material properties by applying partial factors of safety to them.
The factors on the representative loads cover uncertainties on the deviations from
nominal values, modelling uncertainties and dimensional variations. Different partial
safety factors can be used in different action combinations.
Effects should be calculated under action combinations of the type:
∑γ G + ∑γ
g
f
F
1
In which G and F are respectively the permanent and the variable actions γf and γg and the
partial safety factors.
The factors applied to the characteristic properties of materials cover deviations of actual
properties, uncertainties in the resistance models and dimensional variations.
CIRIA C671 • Tunnels 2009
447
The appropriate factors for loads and material properties are given in the relevant codes
for the form of construction to be analysed.
A7.2
ASSESSMENT PRINCIPLES
As explained previously, to carry out an assessment it is necessary to identify the relevant
assessment actions on the structure, calculate (or simulate if physical models are used)
their effects by using an appropriate model and compare them with the associated
resistances. The following sections provide guidance on how to proceed through the
necessary steps to achieve this.
A7.2.1
Modelling of actions
Failure of the tunnel is associated with failure of its lining. However, the tunnel structure
comprises both the lining itself and the undefined volume of ground around it. So both
the ground and the lining should be included in the model of the structure. In this case
some of the actions are applied to the ground and some to the lining. The actions applied
to the ground are the self weight and the superficial loads, which are actions from
building foundations, or highway/railway loads. The actions applied to the lining are self
weight, loads due to vehicles (highway or railway loading) and/or the weight of the
equipment running along the tunnel (eg ventilation fans). Water pressure can be applied
either to the ground or directly to the tunnel depending on the model used for the
ground.
Loads applied to the lining should be clearly defined, consistent along the tunnel length
and do not need further discussion. However, there are issues related to loads applied to
ground that need to be addressed.
The main issue relates to superficial loads. In principle, these actions could be determined
in detail based on actions that are imposed onto buildings, highways and railways above
the tunnel. This has two main drawbacks:
1
Analysis of all the buildings above the tunnel would be required and the effects of
each specific action would be taken into account in models that probably have a
degree of accuracy below the degree of accuracy to which the action is defined.
2
The effort required for this analysis is likely to be considerable – this will be discussed
later.
Detailed analysis of superficial loads, starting from actions on buildings, can be required
for sections of tunnel, when the assessment of a shallow tunnel is required because a new
building is planned in the vicinity of the tunnel or an existing building is to be
considerably altered.
Detailed analysis of superficial loads is also required for shallow tunnels (usually cut-andcover construction) underneath highways or railways that act structurally more like
bridges than tunnels.
For deep tunnels, in which the local distribution of superficial loads is not important, it is
convenient to define a uniform surcharge (which can be split into permanent and variable
components) representing an averaging of the actual actions (loads) in the area crossed by
the tunnel. The surcharge is representative at the same time of highway/railway loading
and building loading. It can be defined by identifying the typical construction form above
the tunnel and smearing the resulting action onto its footprint. This is because building
loads (including the weight of the building) should be higher than highway ones.
448
Note that this approach may result in using 2D structural models as the variation of the
actions along the tunnel is not considered.
The determination of a threshold depth for considering a tunnel deep enough to use a
uniform surcharge at ground level is not easy and depends on the geology of the site, the
nature of the actual loads and the construction form.
Preliminary analysis can be used to decide the accuracy required for modelling the
superficial load. This can be a simple linear elastic analysis of a continuum semi-space with
a hole representing the tunnel, under different arrangements of superficial loads. The
superficial loads in the preliminary analysis need not be the assessment ones as only
sensitivity and not absolute values are to be investigated. A tunnel with a depth of three or
four times its major dimension can be regarded as a deep tunnel – for which a uniform
surcharge is appropriate.
Once the permanent and variable components of the surcharge have been determined the
following assessment situations should be considered:
1
All actions acting on the ground plus water pressure if directly applied to the lining in
the model adopted and if it is unfavourable (in some cases by increasing the axial
force in the lining it can increase the resistance against bending effects due to
distortional loads), plus self weight of the lining.
2
All actions including those directly applied to the tunnel lining due to use or weight of
equipment (highway/railway loading inside the tunnel and self weight of equipment
such as ventilation fans).
In equations the load cases to be considered are:
γ G,soilGsoil + γ G,surchargeGsurcharge + γ G,liningGlining + γ Q,surchargeQ surcharge
(2)
and:
γ G,soilGsoil + γ G,surchargeGsurcharge + γ G,liningGlining + γ Q,surchargeQ surcharge +
∑
+
(3)
γ Q,liningQ lining
Where:
Gsoil
is the permanent action on the ground including ground water action if
this is not included in Glining
Glining
is the permanent action on the lining, including the ground water action if
this is not included in Gsoil
Gsurcharge
is the permanent component of surcharge
Qsurcharge
is the variable component of surcharge
Qlining
is a variable action on the lining.
The partial factors of safety in Equations 2 and 3 should be obtained from the relevant
standards.
The values given for design in the Eurocodes and the British standards which can be used
for the assessment and are given in Table A7.1.
CIRIA C671 • Tunnels 2009
449
Table A7.1
Partial factors of safety for actions recommended by Eurocodes and British Standards
Eurocodes
British Standards*
γG,soil , γG,surcharge , γG,lining
1.35
1.4
γQ,surcharge , γQ,lining
1.5
1.6
Note:
* Values refer to combination dead + superimposed (other combinations are not applicable)
The Eurocode referred to in the table is BS EN 1990:2002 basis of design and the British Standards are those relative to the structural
materials (BS 8110, BS 5950 etc).
A7.2.2
Evaluation of effects
When the assessment situations have been defined the effects on the structures should be
evaluated and in theory can be done using physical or mathematical models. In practice it
is done using mathematical models for several reasons:
the level of generality, accuracy and reliability they have attained in recent times
cost and scaling problems associated with physical models
measurements on physical models can seldom be complete whereas in mathematical
models all results are readily available
few people have the appropriate experience in working with physical models and
access to sufficiently equipped laboratories.
In this section the characteristics that the mathematical models should possess are
discussed, without discussing formulation and solution details.
The physical components to be included into the model are (i) the ground, and (ii) the
tunnel lining. In most cases the structures above the tunnel can be conveniently
considered only through the surcharge. Often it is convenient to consider the interface
between ground and lining as a third component (when modelling of relative sliding is
desired).
To determine the effects of the assessment actions on these components, it is necessary to
use a geometrical model and a mechanical model. The geometrical model describes the
shape of the structural components and the way they can deform and interact, while the
mechanical model describes the material properties and the relationship between stresses
and strains.
Geometrically the ground should be represented by a continuum while for the lining a
shell or beam schematisation can be convenient. The latter is obvious because the
thickness of the lining is small compared to the length of the axis of its cross-section and its
longitudinal extension.
The geometry of the lining should be representative of the real situation, so joints and
construction details should be modelled realistically.
Note that in a continuum model the effects are quantified through stresses and strains at
each point while for a shell or beam through bending moments, axial and shear forces and
curvature, elongation and shear strains. The model used dictates the choice of relevant
effects and resistances.
In general the resistance model is implicit in the mathematical model used for the analysis
450
and, introducing the assessment action as an input, a pass or fail result is obtained directly.
For example, if the lining is modelled by a nonlinear material incorporating damage, the
latter is accumulated as the structure is loaded and the local stiffness and strength of the
material is automatically reduced until it is no longer possible to balance the applied load.
The load at which this occurs represents the assessment resistance of the structure and the
assessment effect coincides with the assessment load. However, such an approach
necessitates the use of advanced numerical analysis techniques so it is generally not the
preferred option for many engineers due to operative difficulty and reliability issues.
A convenient and general approach, consistent with the recommendations in the various
national standards, is to determine the effects in the lining in terms of axial forces and
bending moments (considering the lining linear elastic) and checking them against the
corresponding combined ultimate resistances. In a 2D model in which the lining is
modelled as a frame (or a ring) this amounts to check that the point representative of the
bending moment and axial force under the assessment action is within the interaction
domain at each section of the lining. Redistribution to limit the conservativeness of the
approach can be applied consistently with the code indications for the specific lining
material.
Shear is usually addressed in isolation (ie coupling with other effects is neglected).
According to this approach, the lining material is modelled as linear elastic.
Modelling of ground is much more complex, and use of nonlinear models may be
appropriate. In modelling the ground action on the lining the history of stresses and
strains in the ground should be observed.
Note that in a design scenario the excavation should be simulated by using a 3D model.
This is not always done and a wished in place assumption is generally made. By wished in
place it is implied that the tunnel has been built instantaneously without any disturbance
of the ground by excavation. The load taken by the ground inside the tunnel envelope
before construction is then distributed between lining and ground outside the tunnel
envelope according to their stiffness.
In an assessment situation taking into account the stress-strain history of the ground
around the tunnel is even more difficult because of time dependent effects (the assessment
is usually carried out some time after construction). So a wished in place assumption is
even more appropriate in this case. It is not good practice to resort to more accurate
modelling than is justified by consideration of the likely accuracy of available input data.
This is especially true when the better accuracy of a refined model cannot be easily validated.
Advice should be sought from a geotechnical engineer about the best model to be used for
the ground and the parameters characterising it.
Ground water pressure can be applied directly to the lining if an effective stress analysis is
carried out for ground. In this case the submerged specific weight should be used for
ground below the water table.
In summary the steps for the evaluation of the assessment effects and assessment against
ultimate limit state on the tunnel lining (and so on the tunnel) are as follows:
define a geometrical model for the ground
define a geometrical model for the lining
define a geometrical model for the interface between ground and lining if appropriate
CIRIA C671 • Tunnels 2009
451
A7.2.3
define a mechanical model for ground
model lining as linear elastic
apply assessment actions
calculate assessment bending moments, axial forces and shear forces at each crosssection of the lining
check that the cross-sections have enough capacity to sustain the assessment actions as
defined at the previous point (use the appropriate material factors in evaluation
capacities).
Evaluation of assessment resistances
To complete the assessment by carrying out the final step in the list, the assessment
resistances should be evaluated. According to the previously proposed procedure the
checking of lining resistances has been reduced to a standard check of a beam/column or a
slab/wall. Methods to determine assessment capacities are easily found in the codes of
practice relevant to the materials used in the tunnel lining (for reinforced concrete and
masonry). The exception to this is cast iron linings where there is lack of guidance for
determining the ultimate capacity.
This can be overcome by using a permissible stress approach and referring to old codes of
practice – a useful guide is the Historical structural steelwork handbook also known as the
Brown book (Bates, 1984). For recent linings material properties can also be obtained from
SCI publication 172 (SCI, 1996).
Alternatively the elastic limit of the section can be assumed as ultimate limit state and the
material factor used to obtain the permissible stress can be selected to provide the same
safety as the permissible stresses method applied in accordance with old codes of practice.
The factor can be calculated by calculating for each load case the ratio γ = , where E is
M
the effect obtained applying the characteristic loads and Ea is the effect obtained applying
the assessment load. This ratio varies with the load combination and with the effect being
considered. The material factor for the derivation of the permissible stress can be then
-1
conservatively assumed as γ M = min E Ea ×α where α is the conversion factor from the
characteristic strength to the permissible stress for cast iron ( α = σ f /σ permissible ) derived
from the codes of practice applicable at the time of construction of the lining.
{
}
The advantage of using this approach is that once the factor γM has been determined the
assessment of the cast iron lining is formally identical to the assessment of linings made of
any other material. This simplifies the assessment of tunnels where more materials are
coexistent. This approach is also consistent with the limit state philosophy. Further if a
factor γM for cast iron has been determined for several structures, it makes sense to use its
average value, or a characteristic value of some kind, for all other structures made of the
same type of cast iron.
Refer also to BS EN 1990:2002 for further information on the determination of an
appropriate γM for materials not covered by the codes.
A7.3
ASSESSMENT WORKED EXAMPLES
This section includes worked examples illustrating the assessment methods discussed.
There is one worked example for each of the principal tunnel lining types – cast iron,
concrete and masonry. These worked examples were provided by Guiseppe Simonelli
from Mott MacDonald.
452
Worked example 1: Cast iron
2
18
Assessment of a cast iron lined tunnel
4,07
3,85
4,07
110
600
Tunnel configuration
Cast iron section
Thickness: 25 mm
Area: 19250 mm^2
I: 13.4*10^6 mm^4
CIRIA C671 • Tunnels 2009
453
Tunnel external diameter
De := 4.07m
Depth of tunnel
h := 18m
( at axis)
wsoil := 14
Soil specific weight (dry)
kN
3
m
porosity e=50%
water table 2 m below ground level
Ko soil
Ko := 0.6
vsoil := 0.2
Esoil := 50MPa
Elining := 125000MPa
vlining := 0.25
Cast iron compressive permissible stress
σ := 150
N
2
mm
Cast iron tensile permissible stress
σt := 46
N
2
mm
surcharge
LL := 10
kN
2
m
We assume “wished in place” construction and consider the lining under a uniform load
corresponding to the stress state in the soil at axis level
This uniform load will have a horizontal component Po and a vertical component Pv
The wet weight of soil
wsoilwet := wsoil + 0.5 × 10
kN
3
m
wsoilwet = 19
kN
3
m
As for cast iron we use a permissible stress approach, nominal loads do not need to be factored
Pv := [ 2m × wsoil + ( h − 2m)wsoilwet ] + LL
Po := Ko × Pv
Po = 205.2
kN
2
m
454
Pv = 342
kN
2
m
Po
Diagram of actions on the tunnel lining
CIRIA C671 • Tunnels 2009
455
The formulas used for the derivation of the axial forces and bending moments in the lining
have been taken from Curtis (1974).
The notation used is as in the following figure
θ
R2
Rm
The dashed line
indicates the location
of the structural axis of
the lining
Rm :=
De
2
− 24mm
Rm = 2011 mm
Uniform component of pressure
Pu := 0.5 × ( Pv + Po )
Pu = 273.6
kN
2
m
Distortional component of pressure
Pd := 0.5 × ( Pv − Po)
Pd = 68.4
kN
2
m
hh :=
456
2
19250mm
600mm
hh = 32.083 mm
(hh is the equivalent thickness of a simple ring lining
having the same area as the ribbed actual one)
R2 :=
De
is the external radius of the lining
2
For a thin ring:
Q1 :=
Esoil
Elining
1 − vlining
×
2
1 + vsoil
×
Q1 = 0.02
Rm
hh
Uniform thrust in lining:
Nu :=
R2
1 + Q1
Nu = 546.08
Ilining :=
Q3 :=
Sn :=
St :=
× Pu
kN
m
13.4 × 10 mm
6
0.6m
4
(
(Iining is the moment of inertia of the lining section per metre of tunnel)
R2 × Esoil × 1 − vlining
3
2
)
12 × Elining × Ilining × ( 1 + vsoil)
1 − Q3
⎛ 3 − 2 × vsoil ⎞
1 + Q3 × ⎜
⎟
⎝ 3 − 4 × vsoil ⎠
1 + 2 × Q3
⎛ 3 − 2 × vsoil ⎞
1 + Q3 × ⎜
⎟
⎝ 3 − 4 × vsoil ⎠
Ntot ( θ ) :=
Md ( θ ) :=
− R2
3
2
Rm
6
Q3 = 9.827
× Pd
× Pd
× ( Sn + 2 × St) × cos( 2 × θ ) + Nu
× ( 2 × Sn + St) × cos( 2 × θ )
CIRIA C671 • Tunnels 2009
457
7 .10
5
6.5 .10
5
6 .10
5
Ntot( θ )5.5 .10
5
5 .10
5
4.5 .10
5
4 .10
5
0
0.5
1
1.5
θ
2
2.5
3
0.5
1
1.5
θ
2
2.5
3
7
1.5 .10
7
1 .10
6
5 .10
Md( θ )
0
5 .10
6
1 .10
7
7
1.5 .10
0
These diagrams are plots of the axial force and the bending moments over half of the tunnel lining.
The axial forces and bending moments over the remaining half can be derived by symmetry
458
Ntot ( 0) = 426.602
⎛π⎞
⎟
⎝2⎠
Ntot ⎜
⎛π⎞
⎟
⎝2⎠
kN
(axial at crown)
m
= 665.557
Md ( 0) = 10.965
Md ⎜
kN
kN
(axial at axis)
m
m
(bending moment at crown)
m
= − 10.965
CIRIA C671 • Tunnels 2009
kN
m
(bending moment at axis)
m
459
The interaction diagram for the lining to be assessed is plotted in the figure below:
Cast Iron Lining Interaction Dom ain
3500
3000
2500
Nu (kN )
2000
Elast ic Int er act ion
Cr own
Axis
1500
1000
500
0
- 40
-30
-20
-10
0
10
20
30
Mu (kN m )
The diagram demonstrates that the representative points of the stress state of the lining at crown
and axis are well within the permissible envelope. It is concluded that the tunnel is safe.
46 0
Worked example 2 : Concrete
2
18
Assessment of a concrete lined tunnel
4,07
3,77
4,07
Plain concrete lining
150mm THK
Tunnel configuration
CIRIA C671 • Tunnels 2009
46 1
Tunnel external diameter
THK := 150mm
Thickness of lining
Depth of tunnel
De := 4.07m
( at axis)
h := 18m
wsoil := 14
Soil specific weight (dry)
kN
3
m
porosity e=50 %
water table 2m below ground level
Ko soil
Ko := 0.6
Esoil := 50MPa
vsoil := 0.2
vlining := 0.18
Elining := 15000MPa
Concrete assessment strength
σ := 16
N
(is equal to 0.45 Rck)
2
mm
surcharge
LL := 10
kN
2
m
We assume “wished in place” construction and consider the lining under a
uniform load corresponding to the stress state in the soil at axis level
This uniform load will have a horizontal component Po and a vertical component Pv
The wet weight of soil
wsoilwet := wsoil + 0.5 × 10
kN
3
m
wsoilwet = 19
kN
3
m
So applying a factor of 1.4 to the permanent loads and 1.6 to the surcharge
Pv := 1.4 × [ 2m × wsoil + ( h − 2m)wsoilwet ] + 1.6 × LL
462
Pv = 480.8
kN
2
m
Po := Ko × Pv
Po = 288.48
kN
2
m
Po
Diagram of actions on the tunnel lining
CIRIA C671 • Tunnels 2009
463
The formulas used for the derivation of the axial forces and bending moments in
the lining have been taken from Curtis (1974).
The notation used is as in the following figure
θ
R2
Rm
The dashed line
indicates the location
of the structural axis of
the lining
Rm :=
De
2
−
THK
Rm = 1960 mm
2
Pu := 0.5 × ( Pv + Po)
Pu = 384.64
kN
2
m
Pd := 0.5 × ( Pv − Po)
Pd = 96.16
kN
2
m
hh := THK
464
hh = 150 mm
R2 :=
De
Q1 :=
Nu :=
2
Esoil
Elining
R2
1 + Q1
Nu = 756.185
Ilining :=
Q3 :=
Sn :=
St :=
1 − vlining
×
2
1 + vsoil
×
Q1 = 0.035
Rm
hh
× Pu
kN
m
THK × 1m
3
12m
(
Ilining = 281250 mm
R2 × Esoil × 1 − vlining
3
2
)
12 × Elining × Ilining × ( 1 + vsoil)
1 − Q3
⎛ 3 − 2 × vsoil ⎞
1 + Q3 × ⎜
⎟
⎝ 3 − 4 × vsoil ⎠
1 + 2 × Q3
⎛ 3 − 2 × vsoil ⎞
⎟
⎝ 3 − 4 × vsoil ⎠
1 + Q3 × ⎜
Ntot ( θ ) :=
Md ( θ ) :=
− R2
3
2
Rm
6
3
(Iining is the moment of inertia of the lining section
per metre of tunnel)
Q3 = 6.711
× Pd
× Pd
× ( Sn + 2 × St) × cos( 2 × θ ) + Nu
× ( 2 × Sn + St) × cos( 2 × θ )
CIRIA C671 • Tunnels 2009
465
5
9.5 .10
5
9 .10
5
8.5 .10
5
8 .10
5
Ntot ( θ )7.5 .10
5
7 .10
5
6.5 .10
5
6 .10
5
5.5 .10
3 .10
0
0.5
1
1.5
θ
2
2.5
3
1
1.5
θ
2
2.5
3
7
7
2 .10
1 .10
Md( θ )
7
0
1 .10
7
2 .10
7
3 .10
7
0
0.5
The above diagrams are plots of the axial force and the bending moments over
a half of the tunnel lining.
466
Ntot ( 0) = 587.234
⎛π⎞
⎟
⎝2⎠
Ntot ⎜
kN
= 925.136
Md ( 0) = 20.68
kN
(Axial at crown)
m
kN
(Axial at axis)
m
m
(Bending moment at crown)
m
⎛ π ⎞ = −20.68 kN m
⎟
m
⎝2⎠
Md ⎜
CIRIA C671 • Tunnels 2009
(Bending moment at axis)
467
The interaction diagram for the lining to be assessed in plotted in the figure below:
Concrete Lining Interaction Domain
2500.0
2000.0
1500.0
Nu (kN)
Interaction domain
Crown
Axis
1000.0
500.0
-50.0
-40.0
-30.0
-20.0
-10.0
0.0
0.0
10.0
20.0
30.0
40.0
50.0
Mu (kN m)
The diagram demonstrates that the representative points of the stress state of the lining at
crown and axis are within the permissible envelope. It is concluded that the tunnel is safe.
468
Worked example 3 : M asonry
Assessment of a circular masonry lined tunnel
Di := 6m
Tunnel internal diameter
h := 12m
( at axis)
wsoil := 14
Soil specific weight (dry)
2
Depth of tunnel
12m
De := 7.2m
Tunnel external diameter
kN
3
m
porosity e = 50 %
water table 2m below ground level
wlining := 18
Masonry specific weight
kN
6m
3
m
Ko := 0.8
Ko soil
7.2m
Fk := 11.5
Masonry compressive strength
N
2
mm
γm := 2.5
Tunnel configuration
ULS compressive strength
σ :=
surcharge
LL := 10
σ = 4.6 MPa
Fk
γm
kN
2
m
We assume “wished in place” construction and consider the lining under a
uniform load corresponding to the stress state in the soil at axis level
This uniform load will have a horizontal component Po and a vertical component Pv
The wet weight of soil
wsoilwet := wsoil + 0.5 × 10
kN
3
m
wsoilwet = 19
kN
3
m
Hence applying a factor of 1.4 to the permanent loads and 1.6 to the surcharge
Pv := 1.4 × [ 2m × wsoil + ( h − 2m)wsoilwet ] + 1.6 × LL
Po := Ko × Pv
Po = 256.96
Pv = 321.2
kN
2
m
kN
2
m
CIRIA C671 • Tunnels 2009
46 9
Pv
Po
Diagram of actions on the tunnel lining
In this example the self weight of the lining is neglected for simplicity. This does not take away
generality from the procedure as it can be easily introduced for shallow tunnels.
A line of thrust equilibrating the above loads and compatible with the lining geometry and
material strength can be sought analytically or graphically
For this example it is found analytically.
47 0
Because of symmetry only one quarter of the lining is considered.
The reference frame used is explained in the following sketch
Pv
Po
θ
AA
Reference frame
The radius of the lining centreline is:
r :=
Di + De
2×2
r = 3300 mm
Rex :=
External radius
De
2
Rex = 3600 mm
the thickness of the lining is:
t :=
De − Di
2
CIRIA C671 • Tunnels 2009
t = 600 mm
47 1
the thrust at crown is:
HH := Po × Rex
The horizontal and vertical resultants at any cross section (a section is located by the angle θ) are:
RX( θ ) := HH − Po × ( 1 − cos( θ ) ) × Rex
RY( θ ) := Pv × Rex × sin( θ )
the minimum depth from the extrados at which the line of thrust can be located,
compatible with the masonry strength is:
dcrown :=
0.5HH × m
σ × 1m
dcrown = 101 mm
The bending moment at each section can therefore be expressed as:
⎡⎛ t − dcrown⎞ + ( 1 − cos( θ ) ) × r⎤ × 1m⎤ ...
⎟
⎥
⎥
⎣⎝ 2
⎠
⎦
⎦
Rex ⎞
⎡
2
⎛
⎤
+ ⎢− Pv × ⎜ r −
⎟ × Rex × sin( θ ) × 1m⎥ ...
2 ⎠
⎣
⎝
⎦
Rex + Rex × cos( θ ) ⎞
⎡
⎡
⎛
⎤
⎤
+ ⎢− Po × ⎢⎜
⎟ − r × cos( θ )⎥ × Rex × ( 1 − cos( θ ) ) × 1m⎥
2
⎣
⎣⎝
⎠
⎦
⎦
⎡
⎣
M ( θ ) := ⎢HH × ⎢⎜
The axial force at any cross section has the expression:
AA ( θ ) := ( RX ( θ ) × cos( θ ) + RY( θ ) × sin( θ ) ) × 1m
from which the eccentricity is derived
ecc( θ ) :=
47 2
M ( θ)
AA ( θ )
200
100
ecc( θ )
0
100
200
0
0.5
1
θ
1.5
Diagram of eccentricity of thrust along the arch
⎛ π ⎞ = −140 mm
⎟
⎝2⎠
ecc⎜
ecc( 0) = 199 mm
4000
3000
2000
1000
0
0
1000
2000
3000
4000
Plot of the lining profile and the line of thrust (lengths are expressed in mm)
CIRIA C671 • Tunnels 2009
47 3
9
1.2 .10
9
1.1 .10
AA( θ )
9
1 .10
8
9 .10
0
0.5
θ
1
1.5
Diagram of the thrust (axial force) along the profile of the arch
⎛ π ⎞ = 1156 kN
⎟
⎝2⎠
AA⎜
AA( 0) = 925 kN
From the observation of the line of thrust and the diagram of the axial force we note that the
maximum eccentricity and the maximum axial force are attained at the ring axis.
The minimum depth at which the thrust can be located within the ring from the intrados
(compatible with the masonry strength) is:
d :=
⎛π⎞
⎟
⎝2⎠
0.5 AA ⎜
σ×m
d = 126 mm
The line of thrust found yields a depth of
dem :=
t
2
⎛π⎞
⎟
⎝2⎠
+ ecc⎜
dem = 160 mm
The line of thrust found can be developed within the arch ring and equilibrates the applied loads.
So by enforcing the lower bound theorem of limit analysis, we infer that the tunnel is safe.
47 4