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