Licensed copy:IMPERIAL COLLEGE, 12/02/2009 ...

1 downloads 0 Views 28MB Size Report
Following the defeat of King Harold at Hastings, William marched on London, but was ...... The foundation to the third pier of these arches was stiffened by.
Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

I

B

Y- i

Building response to tunnelling Case studies from construction of the Jubilee Line Extension, London

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

VOLUME 1 PROJECTS AND METHODS

The results of the collaborative research project Subsidence damage to buildings: prediction, protection and repair, carried out by lmperial College with the sponsorship of London Underground Limited, CIRIA, the Geotechnical Consulting Group and other industry organisations, under the DETR-EPSRC LINK Construction Maintenance and Refurbishment Programme.

EDITORS J B BURLAND J R STANDING

F M JARDINE

1' ThomasTelford

Published by Thomas Telford Publishing, Thomas Telford Ltd, 1 Heron Quay, London E 14 4JD. URL: http://www.thomasteIford.com Distributors for Thomas Telford books are:

USA: ASCE Press, 1801 Alexander Bell Drive, Reston, VA 20 191-4400, USA Japan: Maruzen Co. Ltd, Book Department, 3-10 Nihonbashi 2-chome, Chuo-ku, Tokyo 103

Australia: DA Books and Journals, 648 Whitehorse Road, Mitcham 3 132, Victoria

First published 200 1 A catalogue record for this book is available from the British Library

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Building response to tunnelling. Case studies from the Jubilee Line Extension, London Burland, J B; Standing, J R; Jardine, F M (editors)

0CIRIA 2001

ISBN 0 7277 3017 7

CIRIA Special Publication 200

Keywords Building settlements, case studies, tunnelling, damage assessment, protective measures, compensation grouting, precision levelling and taping, ground movements, soil-structure interactions, London Clay, Lambeth Group, Jubilee Line Extension Project, settlement trough

Reader interest

Classification

Tunnel promoters; tunnelling and geotechnical engineers; structural and civil engineers, building surveyors and insurers

AVAILABILITY

Unrestricted

CONTENT

Research results, case studies and technical review

STATUS

Committee-guided

USER

Tunnelling, structural and geotechnical engineers, building professionals, surveyors and insurers

All rights, including translation, reserved. Except as permitted by the Copyright, Designs and Patents Act 1988, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior written permission of the Publishing Director, Thomas Telford Publishing, Thomas Telford Ltd, 1 Heron Quay, London E l 4 4JD.

This book is published on the understanding that the authors are solely responsible for the statements made and opinions expressed in it and that its publication does not necessarily imply that such statements and/or opinions are or reflect the views or opinions of the publishers. While every effort has been made to ensure that the statements made and the opinions expressed in this publication provide a safe and accurate guide, no liability or responsibility can be accepted in this respect by the editors or publishers.

Typeset by CIRIA, 6 Storey’s Gate, Westminster, London SW 1 P 3AU. Printed and bound in Great Britain by MPG Books, Bodmin, Cornwall.

Building response to tunnelling

Summary

The ability to predict the potential for building damage accurately and to have confidence in the chosen protective measures is of increasing importance to the viability of urban tunnelling.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The Jubilee Line Extension Project (JLEP) provided a unique opportunity to capture reliable field measurements about the effects of tunnelling on a wide range of buildings. This book presents the findings from the JLEP-based research collaboration to gather that field data. General information about the relevant parts of the JLEP works and the research methodology is presented in Volume 1, which concludes with the principal findings from the research. Volume 2 contains the series of building and greenfield site case histories that were the subjects of the research. Individual chapters were written by experts who participated in the research, by engineers involved with the JLEP works, and by members of the Imperial College and JLEP-based research team. The book starts with descriptions of the JLEP, of the methods - and the uncertainties of settlement prediction and building damage assessment current during the JLEP works, and of the design behind the research. The following chapters give a desk-study account of the geology and historical development of areas of the case study buildings along the JLE route between Green Park and Canada Water stations. Separate chapters describe the tunnelling methods and the protective measures that were used. The research methodology included formally made best-practice predictions of ground and building movements at two greenfield sites and four buildings. Full transcriptions are included of the reports in which these predictions were made. Other chapters explain the research measurement methods, their precision and data handling. The concluding chapter of Volume 1 by Professor Burland presents the overall findings to date of the research. Volume 2 presents the 27 case studies in their geographical sequence from Green Park in the west to London Bridge all on London Clay, and then eastward on the Lambeth Group to Canada Water. Two of the case studies are of instrumented greenfield sites. The case study buildings include the Big Ben clock tower, public and commercial buildings of central London and Southwark, and residential buildings, including two tower blocks of Bermondsey and Rotherhithe. Important case studies are those where compensation grouting was used. The book is generously illustrated with numerous line drawings, graphs, halftones, and maps. Individual chapter lists of references are backed by a consolidated list. The preparation of this book was in accordance with the review procedures for CIRlA ground engineering reports.

Building response to tunnelling

iii

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Foreword

On behalf of London Underground Limited, I very much welcome publication of this book. During the construction of the Jubilee Line Extension, we gathered an immense amount of data observing and recording ground and building response to tunnels and open excavations. We wanted to learn all we could from this great project and initiated this research to collate, review, present and disseminate the data as meaningful technical information. These research results are important, of course, to the future of underground development in London. They will help to provide more economically the essential upgrades to the underground network, and the cross-London rapid transit systems that are so much needed for commuters and visitors to London, indeed for the wealth of this city. The results will reassure owners and occupiers of properties above and near proposed new tunnels and stations, because they demonstrate the methods and controls that can be applied to the protection of buildings otherwise at risk. We also have a better knowledge of how buildings behave - buildings of all sorts from the great heritage buildings of Westminster to the dwellings of Bermondsey. A particular reason we at LUL have for welcoming this book is that from the outset of the research proposal we wanted as many as possible to be given the opportunity to share in the research. For that reason we were pleased that in addition to the intellectual collaboration between LUL, Imperial College and the Geotechnical Consulting Group, CIRIA was brought in both to mobilise contributions from many other organisations and to manage the processes of disseminating the results more widely. We must also acknowledge the substantial support from the EPSRC-DETR LINK programme for Construction Maintenance and Refurbishment. Publication of this book is one of the results of that collaboration. What has been learnt from the research is already influencing practice; this book will extend that sharing of knowledge much more widely. We believe, too, that not only will the case studies be referred to and used for many years to come, but they will also set benchmarks for future practice. I should like to re-emphasise one point. We did not “keep a score” of the time that people put into the work represented here, but it was much more than the time of the researchers, editors and writers. We can say that the book is an acknowledgement of the splendid work of a great many people and organisations - of LUL and the whole JLE Project team, of the technical contractors and designers, of the main works contractors, of the specialist contractors, and of the many individuals who gave technical advice and support to the research. They all contributed to the results contained in this book, which will be of value in the planning, design and construction of urban tunnelling projects throughout the world Keith Beattie The Chief Engineer London Underground Limited

iv

Building response to tunnelling

Acknowledgements

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

This book is the one of the results of a series of related research and information projects to which many people contributed. A considerable number of organisations gave funding, technical information, and the time of their staff to advise the researchers. The fieldwork and factual reporting on which this book is based was chiefly carried out by researchers at Imperial College, staff provided by the Jubilee Line Extension Project, and members of the Geotechnical Consulting Group, under the research project “Subsidence damage to buildings: prediction, protection and repair”. That research formed part of the DETR-EPSRC Construction Maintenance and Refurbishment LINK Programme (EPSRC reference GWK34306) and CIRIA’s Ground Engineering Programme. An associated research project was sponsored by London Underground Limited and EPSRC entitled “Field measurement of ground behaviour resulting from single and twin tunnel excavations” (EPSRC reference GWK38236). These projects were guided by a board and a management committee, whose members (using their affiliations at that time) are listed below. Board of Management Dr P L Bransby (chairman) Professor J B Burland Mr M Gellatley Dr R J Mair Mr P Pullar-Strecker Mr D J Sharpe

CIRIA Imperial College London Underground Limited Geotechnical Consulting Group DOE-EPSRC Programme Co-ordinator Jubilee Line Extension Project

Programme Management Committee Mr J M Anderson Mr C J Barber Mr R C Beckwith Mr K H Bowers Dr A Haimoni Mr M T Hutchinson Mr F M Jardine (chairman) Dr N Jarrett Mr D Kincaid Mr L F Linney Dr B M New Dr N J O’Riordan Mr A J Powderham Mr H Roscoe Dr B Simpson Dr N Sparnon Dr J R Standing Ms J Williams

Building response to tunnelling

Health and Safety Executive Laing Technology Group Union Railways Limited Transport Research Laboratory AMEC Piling Trafalgar House Technology CIRIA Department of the Environment University College, London Jubilee Line Extension Project Geotechnical Consulting Group Rail Link Engineering Mott MacDonald Group Kvaerner Cementation Foundations Ove Amp and Partners EPSRC Imperial College EPSRC

V

The research as a whole was under the direction of Professor J B Burland, with Dr J R Standing leading the research team. ClRIA’s research manager was F M Jardine. Subsequently, with additional funding from London Underground Limited and industry (see below), it became possible to undertake preparation of this book. For this, the editors had the guidance of an Editorial Advisory Committee, whose members were:

K N Montague (chairman) M G Black R M C Driscoll J E Hellings R J Mair J Moriarty R J M Sutherland A D Withers

CIRIA LUL, Infraco JNP and CrossRail Building Research Establishment Maunsell Group Cambridge University London Underground Limited Consultant Bank of New York, formerly of LUL.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

RESEARCHERS Many of those who made up the team of researchers are authors of chapters of this book, but others took part in the work reported here. Their work is gratehlly acknowledged. Dr J B Newman Mr S K Sharma

Imperial College, reader Imperial College, senior lecturer

Mr A D Withers Mr M G Black Mr D P Page Dr E K de Moor Dr R J Nyren Mr G R Taylor Mr T Geilen Dr G M B Viggiani Mr S C Gupta

JLEP, research team member from Sept 1995 CrossRail, research team member from Oct 1996 JLEP, research team member to Aug 1995 JLEP, research team leader to June 1994 Imperial College, research assistant from Jan 1995 Imperial College, research assistant from Mar 2000 University of Illinois, academic visitor University of Rome, Tor Vergata, academic visitor Imperial College, technical assistant

Mr S Ackerley Mr A Bolsher Mr S Pontin Ms R Tot Thomas

Imperial College, chief technician Imperial College, technician IT consultant to JLEP Imperial College, postgraduate survey assistant

Mr M Farina Dr J Harvey Ms R Selman Ms F Deruelle Mr P Genin Ms S Petrova Mr H Zhou

Undergraduate, Imperial College. Undergraduate, Imperial College Undergraduate, Imperial College Visiting student, Imperial College Visiting student, Imperial College Visiting student, Imperial College Academic visitor from China

A great many others helped the research. It would not be possible to list all and, as to list only some would be invidious, acknowledgement is gratefully given to those many people on the staffs of LUL and JLEP, on site and in the office, of the main and specialist contractors of JLE Contracts 101, 102 and 105, and of the technical contractors to the JLEP, who contributed to the research. Their work is acknowledged more fully in Chapter 2 .

vi

Building response to tunnelling

FUNDERS The editors, Imperial College and CIRIA wish to thank all those organisations without whose funding this work would not have been possible. The principal funders were: London Underground Limited The Engineering and Physical Sciences Research Council Department of the Environment, Transport and the Regions Geotechnical Consulting Group CIRIA Core Programme. Through the several stages of the research the following organisations also contributed to its funding and to the work of preparation of this book. AMEC Civil Engineering AMEC Piling Bachy Soletanche (UK) Ltd

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Balfour Beatty Major Projects Carillon Geotechnical Consulting Group Halcrow Group Haswell Consulting Engineers HSE Institution of Civil Engineers R&D Enabling Fund Kvaerner Cementation Foundations, now Cementation Foundations Skanska Maunsell Group Mott MacDonald Group Ove Arup and Partners Trafalgar House Technology Transport Research Laboratory Union Railways Limited.

Building response to tunnelling

vii

ILLUSTRATIONS It is with great pleasure that the editors and publisher acknowledge those organisations and individuals who provided illustrations and allowed their publication here. The lists below are for those in this volume. Separate acknowledgements are made in Volume 2. Photographs of the JLEP works and finished JLE stations were kindly made available by London Underground Limited, ie Figures 2.2, 2.4, 2.5, 2.6, 2.8, 2.8, 2.1 I , 2.12, 16.7, 16.12 and 16.13.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

QAPhotos of Terlingham Manor Farm, Hawkinge, Kent, well known for its superb tunnelling photographs, opened up its JLEP archive and gave us permission to use the following: Figures2.4, 2.10, 2.12, 10.3, 10.5, 16.8, 17.4 and 17.7. A particularly enjoyable part of the preparation of this book was in finding out more about London and how it has changed. We are all indebted to Alan Godfrey for his publication of old editions of OS maps; excerpts from these are reproduced in part in the desk study chapters. We are gratehl to the Guildhall Library of the Corporation of London for its permission to publish these and extracts from older maps of London: Figures6.1,6.2, 6.7, 7.1, 7.4, 8.1, 8.3,9.1, 9.2,9.3,9.4,9.6and9.7. Another wonderful source of information is the National Monuments Record of English Heritage. Its Blandford Street collection contains a splendid library of books, papers and, especially, photographs of London buildings and other historic monuments. Many of the photographs are more than 100 years old. We are very gratefd to the National Monuments Record for permission to reproduce Figures 6.5, 6.6, 7.1, 7.3 and 8.2. We also gratefully acknowledge:

AMECPilingfortheuseofitsslides 11.10, 11.11, 11.13 and 11.14. Bylander 2000 Limited for allowing us to reproduce, as Figures 6.3 and 6.4, copies of lantern slides from the construction of the Ritz Hotel. Dick Millington for permission to use his cartoon, Figure 4.1.

viii

Building response to tunnelling

THE EDITORS J B Burland DSc(Eng) FREng FRS FlCE FlStructE FCGl Professor of Soil Mechanics Department of Civil and Environmental Engineering Imperial College of Science Technology and Medicine John Burland initiated and supervised all the research reported here. An international authority, known for his many contributions to geotechnical engineering theory and practice and for his teaching, he was a geotechnical advisor on the Jubilee Line Extension.

J R Standing BSc MSc DIC PhD MICE CEng

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

University Lecturer in Geotechnical Engineering Engineering Department Cambridge University Fellow of Downing College Jamie Standing, when a research fellow at Imperial College, was employed by the Jubilee Line Extension Project to take on leadership of the team of researchers and the day-to-day running of the work. The thoroughness, reliability and accuracy of the field data are due in large part to his attention to detail and hard work.

F M Jardine MSc(Eng) Senior Research Manager, Ground Engineering, CIRIA, and Principal Research Fellow, Department of Civil and Environmental Engineering, Imperial College of Science Technology and Medicine Fin Jardine was the Project Manager for the LINK-CMR research project on which this book is based, co-ordinating the collaborative effort of the research team and the industrial partners. Through hrther industry sponsorship, he was seconded to Imperial College to act as co-ordinating editor for the book.

THE RESEARCH TEAM AND CONTRIBUTING AUTHORS

Building response to tunnelling

T I Addenbrooke BEng MSc PhD ACGl DIC Lecturer, Department of Civil and Environmental Engineering Imperial College of Science Technology and Medicine

Chapter 12

M G Black BSc MSc CGeol FGS Geotechnical Design Manager for CrossRail Project, LUL, and seconded to the research team

Chapters 19, 3 I , 44 and 45

C F Field BSc CEng MICE Senior Supervising Engineer Contract 104, London Bridge Station, Jubilee Line Extension Project

Chapter 16

ix

L

T Geilen Former visiting research assistant in the Soil Mechanics Group, Imperial College, from University of Illinois at Urbana, Champaign; now geotechnical engineer with Treadwell and Rollo, California

Chapters33,34 and 35

S Gupta BSc (Civ Eng) Technical Assistant and member of research team, Imperial College

Chapters 23 and 27

D I Harris BSc MSc DIC CEng MICE Director, Geotechnical Consulting Group; member of Geotechnical Engineering Advisory team to Jubilee Line Extension Project

Chapters 1 I , 24 and 28

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Chapters 6, and 36

N Kovacevic MSc PhD DIC Associate Director, Geotechnical Consulting Group

L F Linney BSc MSc CEng MICE Principal Geotechnical Engineer, Montgomery Watson; formerly Senior Geotechnical Engineer, JLEP

Chapter 13

Chapter 5

I I \

X

R J Mair MA PhD FREng FlCE Professor of Geotechnical Engineering and Head of Civil and Environmental Engineering, Cambridge University; Master of Jesus College; Director, Geotechnical Consulting Group

Chapters 10,13, 14 and 15

R Nyren PhD BSc(Eng) Research Assistant, Soil Mechanics Section, Imperial College; instrumentation and monitoring of the two greenfield sites; now geotechnical instrumentation engineer Geocomp-Brown JV

Chapters 18 25 and 37

D P Page BSc MSc CEng CGeol MlMM FGS Formerly Geotechnical Engineer, JLEP, seconded to research team; now senior engineer with High Point Rendel

Chapter 5

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

S Petrova Visiting student in Soil Mechanics Section, Imperial College, from Czech Technical University, Prague

Chapter 32

S Pontin Formerly IT Consultant to JLEP, developing database; now Software Developer, CADAC

Chapter 19

D M Potts BSc PhD DSc CEng FlCE Professor of Theoretical Soil Mechanics Department of Civil and Environmental Engineering Imperial College of Science Technology and

Chapters 12 and 13

D Riley BSc(Hons) Dip Bldg Cons ARCS Construction Building Surveyor, Contract 104, London Bridge, JLEP; now Project Manager Capital Projects, University College, London

Chapter 8

R Selman MEng(Hons) Undergraduate Imperial College, now Graduate Tunnel Design Engineer, Underground Works and Geotechnics Division, Mott MacDonald Ltd

Chapter 29

G R Taylor BEng MSc CEng MICE MHKIE MlHT

Chapters 32, 33,34 and 35

Research Assistant, Soil Mechanics Section, Imperial

Building response to tunnelling

R N Taylor MA MPhil PhD CEng MICE Professor of Geotechnical Engineering, Director Geotechnical Engineering Research Centre, City University; Secretary General, International Society for Soil Mechanics and Geotechnical Engineering

Chapters 14 and 15

Dr G M B Viggiani Associate Professor of Geotechnics, University of Rome, Tor Vergata Academic Visitor, Imperial College

Chapters 20, 26 and 27

A D Withers BSc MSc CEng MlMM CGeol FGS Geotechnical Engineer for JLEP and LUL Infraco JNP, seconded to the research, field monitoring and development of database; now an analyst at the Bank of New York

Chapters 5, 17, 18, 19, 37, 38, 39,40,41,42 and 43

xi

Contents

VOLUME 1 PROJECTS AND METHODS ...

Summary ..................................................................................................... 111 Foreword ..................................................................................................... iv Acknowledgements ...................................................................................... v Glossary ..................................................................................................... xv .. Acronyms ................................................................................................. XVII

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Notation ....................................................................................................

xix

1

Introduction..................................................................................................

1

2

Structures and contracts of the Jubilee Line Extension ...............................

9

3

Assessment methods used in design ..........................................................

23

4

The LINK CMR research project ..............................................................

45

5

Geology and geotechnical properties .........................................................

57

6

St James’s and St James’s Park: a brief history of their development ....... 83

7

Westminster and Waterloo areas ...............................................................

95

8

The London Bridge station area ...............................................................

103

9

Bermondsey and Rotherhithe...................................................................

115

10 Tunnelling methods .................................................................................

127

11 Protective measures .................................................................................

135

12 Finite element analysis of St James’s Park greenfield reference site ...... 177 13 Finite element analyses of ground movements from tunnelling below Southwark Park ........................................................................................

185

14 Elizabeth House: settlement predictions ..................................................

195

15 Settlement predictions for Neptune, Murdoch, and Clegg Houses and adjacent masonry walls ............................................................................ 217 16 JLE construction works at London Bridge station ...................................

229

17 Some aspects of construction on JLEP Contracts 105 and 106 ...............249 I

18 Measuring techniques and their accuracy ................................................

273

19 Data handling and storage........................................................................

301

20 Grout intensities .......................................................................................

311

21 Results of the research .............................................................................

315

Building response to tunnelling

Previous page is blank

xiii

VOLUME 2 CASE STUDIES 22 Introduction to the case studies 23 Faqade observations of the Ritz Hotel 24 Compensation grouting for the Royal Automobile Club building 25 Surface displacements at St James’s Park greenfield reference site above twin tunnels through the London Clay 26 The Treasury 27 Buildings in Great George Street: The Institution of Civil Engineers, Public Health Engineering building and the Royal Institute of Charted Surveyors 28 The Clock Tower and the Palace of Westminster 29 Measurements on existing tunnels near Waterloo 30 Elizabeth House, Waterloo

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

31

1-7 St Thomas Street: results of precise levelling

32 Post Office, Borough High Street 33 Telephone House, London Bridge Street 34 The BT Building, London Bridge Street 35 Fielden House, London Bridge Street 36 Desk study for the Canada Estate 37 Surface displacements at three reference sites above twin tunnels through the Lambeth Group 38 Keeton’s Estate, Bermondsey 39

128 and 130 Jamaica Road, Bermondsey

40

182-2 10 Jamaica Road, Bermondsey

41 Blick House, Rotherhithe 42 Murdoch, Neptune and Clegg Houses in Moodkee Street, Rotherhithe 43 Niagara Court, Canada Estate, Rotherhithe

44 Regina and Columbia Points, Canada Estate, Rotherhithe 45 Tenants’ Hall and Boiler House, Canada Estate, Rotherhithe References

xiv

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Glossary

angular strain

angular strain a at a point B between Points A and C is the combined angular change (rotations) of the lines AI3 and BC (see Figure 3. I(a)). It is positive for upward concavity (sagging) and negative for downward concavity (hogging)

ashlar

a dressed stone used to face a wall; finely jointed masonry using these stones

compensation grouting

one or more combinations of systems or phases of injecting grout into the ground between a subsurface excavation and a building or structure to offset or mitigate the settlement that the structure would undergo as a result of the excavation

concurrent grouting

the injection of grout in a pre-determined pattern during tunnelling in correspondence with the advance of the tunnel face; the injections may be ahead of and behind the tunnel heading, and there may be an exclusion zone in which no injections are permitted either side of the heading the injection of grout in advance of tunnelling at a uniform intensity over a defined area in order to prove a system and to induce a nominal controlled heave (also termed pre-treatment)

conditioning

control site

see reference site

corrective jacking

the injection of grout subsequent to tunnelling to offset actual or expected hture settlements from consolidation or to adjust a settlement profile (also termed observational grouting)

deflection ratio

deflection ratio (sagging ratio or hogging ratio) is denoted by A/L where L is the distance between the two reference points defining relative deflection A

deformation

an alteration in the size or shape of a body

displacement

the difference in position of a point or a body at a given time from an earlier position

epoch

in photogrammetry, a set of baseline survey and photogrammetric records taken at a particular time a floor formed of steel joists and reinforced concrete that act compositely

filler joist floor greenfield site ground stabilisation groutlground response efficiency factor, GEF

place where there has been no previous surface or subsurface construction the use of a geotechnical process, such as grouting, to make the ground stronger or stiffer an index of the effectiveness of compensation grouting that relates the volume of injected grout to the vertical displacement of the ground surface. While it is generally defined in terms of displacement divided by the volume of grout per unit surface area to achieve that displacement, there may be differing ways to estimate the injected volume of grout per unit

grouting

the controlled injection of material, usually in a fluid phase, into soil or rock to improve the physical characteristics of the ground

grout intensity

the volume of grout injected per unit plan area of treatment the vertical upwards displacement of a point a convex upwards deflected shape rock mill used in the trench excavation for diaphragm wall construction

heave hogging Hydrofraise

Building response to tunnelling

xv

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

xvi

jack arches

a system of shallow brick arches spanning between closely spaced iron beams and supporting a floor, all within the overall depth of the floor construction

micrometer stick

a specially designed extensometer that comprises an invar bar with at one end a regular series of machined holes and at the other a machined slot into which is mounted a piston of a Vernier micrometer. The holes and slot closely fit BRE levelling studs, so that with the play and distance range covered by the slot and holes, the distance apart of two studs can be accurately measured and changes in that distance determined

monitoring

a formal system of observation and measurement over time, here usually of a building

observational grouting

the injection of grout subsequent to tunnelling in order to offset actual or expected future settlements from consolidation or to adjust a settlement profile (also termed corrective jacking)

pass (half pass)

the grouting from a defined set of grout ports in an array of tubes U manchette to cover a certain plan area (note that this might mean every second or fourth port); a half (or quarter) pass would cover the same area, but would use only half (or quarter) as many of the ports as a full pass. They need not be the same ports

permeation grouting

strictly, the injection of grout to replace the water in pore spaces of the ground without causing hydrofracture, but generally used here to mean the grouting of coarse-grained soils (Terrace Gravels) with finely ground cement-based grouts or silicates

phase (of grouting)

one of the component parts of compensation grouting or other form of grouting for a site or structure; eg concurrent grouting during the tunnelling below a particular building would be one phase

plunge column

steel columns with its base cast into the concrete of the bored pile that will bear its load

pre-treatment

the injection of grout in advance of tunnelling at a uniform intensity over a defined area to prove a system and induce a nominal controlled heave (also termed conditioning)

reading

a single observation with an instrument or measurement of a gauge

reference site

a place without existing buildings where settlement and other measurements were made at an instrumented section, usually but not necessarily a greenfield site

relative deflection

relative deflection A is the displacement of a point relative to the line connecting two reference points on either side - see Figure 3. I(b)

relative rotation

relative rotation (angular distortion) p is the rotation of the line joining two points, relative to the tilt w - see Figure 3. I(c)

rotation

rotation or slope 0 is the change in gradient of a line joining two reference points (eg AB in Figure 3.1 (a)).

sagging

a convex downwards deflected shape

settlement

the vertical downwards displacement of a point

slope soil conditioning

the gradient of a line joining two reference points

subsidence tilt

the settling process, usually of a building or the land surface)

see conditioning tilt w describes the rigid body rotation of the structure or a welldefined part of it - see Figure 3.l(c)

tube a manchette

tubing with regularly spaced sleeved ports installed in a borehole to give access to a grout injection pipe system

waybeam

a timber or steel beam parallel with and below each rail where a bridge is being repaired

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Acronyms

AGS

Association of Geotechnical and Geo-environmental Specialists

BBA

Balfour Beatty-AMEC Joint Venture

bgl

below ground level

BGS

British Geological Survey

BRE

Building Research Establishment

BS

British Standard

CAD

computer-aided drawing (design)

CAU

anisotropically consolidated (triaxial test)

CIRIA

Construction Industry Research and Information Association

CIU

isotropically consolidated (triaxial test)

CMR

Construction Maintenance and Rehrbishment

CPT

cone penetration test

csv

comma-separated value

CTW

Costain-Taylor Woodrow Joint Venture

DETR

Department of the Environment, Transport and the Regions

EB

eastbound

ELL

East London Line

EPBM

earth pressure balance (tunnelling) machine

EPSRC

Engineering and Physical Science Research Council

FE

finite element

GEF

grouting efficiency factor

GEOSIS

Geotechnical Spatial Information System

GIS

geographical information system

GS

Glauconitic Sand

HSE

Health and Safety Executive

ID

inside (internal) diameter

ICE

Institution of Civil Engineers

ICFEP

Imperial College finite element programme

IMR

interlocking machine room

JLE(P)

Jubilee Line Extension (Project)

JV

Joint Venture

LINK-CMR

EPSRC-DETR (originally SERC-DOE) LINK collaborative research programme on Construction Maintenance and Rehrbishment

LMC

Lower Mottled Clay

LMC

lower machine chamber

Building response to tunnelling

xvii

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

LSC

Lower Shelly Clay

LB

Laminated beds

LUL

London Underground Limited

MGT

mandatory ground treatment

NA

not applicable

NATM

New Austrian Tunnelling Method

NL

Northern Line

NMR

National Monuments Record (a department run by English Heritage)

OD

outside diameter

OD

Ordnance datum

ODBC

open database connectivity

OPC

ordinary Portland cement

os

Ordnance Survey

PB

Pebble Beds

PCC

pre-cast concrete

PCL

Performance Control Levels

PD

project datum

PFA

pulverised fuel ash

RCHME

Royal Commission on the Historical Monuments of England

sc

station concourse

SCL

sprayed concrete lining

SGI

spheroidal graphite iron (as in segmental tunnel linings)

SPJ

step-plate junction

SPT

Standard Penetration Test

SQL

structured query language

TAM

tube a manchette

TBM

(1) tunnel-boring machine, (2) temporary bench-mark

UMC

Upper Mottled Clay

USC

Upper Shelly Clay

uu

unconsolidated undrained (triaxial test)

WB

westbound

Notation

A

B B

cross-sectional area (also curve-fitting constant) breadth constant in effective stress - permeability model (also curvefitting constant) effective stress strength intercept

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

curve-fitting constants offset from centre-line Young’s modulus Young’s modulus for concrete representative soil stiffness undrained Young’s modulus drained Young’s modulus

HB i

‘Y

Building response to tunnelling

shear modulus height (of beam or building); also half-width of beam in expressions for a* and p* height of building horizontal distance from tunnel centre-line to the point of inflection on the settlement trough horizontal distance from tunnel face to the point of inflection on the longitudinal settlement trough horizontal distance from tunnel centre-line to the point of inflection on the transverse settlement trough second moment of area, moment of inertia coefficient of permeability coefficient of permeability at zero effective stress trough width parameter (also bulk modulus) coefficient of earth pressure at rest length, between two points, of a building, faqade, or beam length of building in hogging zone length of building in sagging zone coefficient of volume compressibility Potts and Addenbrooke modification factor in hogging Potts and Addenbrooke modification factor in sagging mean effective stress effective overburden pressure load at a point curve-fitting constant settlement maximum settlement curve-fitting constant undrained shear strength distance of neutral axis from edge of beam in tension curve-fitting constant pore water pressure

xix

21

VL

a

horizontal displacement volume loss = volume of surface settlement trougNexcavated volume of tunnel per unit length volume of surface settlement trough per unit length settlement maximum settlement horizontal distance from the tunnel face horizontal distance from tunnel centre-line depth depth of tunnel axis angular strain, +ve for upward concavity sagging, -ve for hogging (also curve-fitting constant) relative axial stiffness = EA/E,H relative rotation (angular distortion) bulk unit weight (also curve-fitting constant) dry unit weight

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

saturated unit weight curve-fitting constant change of length relative settlement relative deflection deflection ratio principal strains axial strain in triaxial test critical tensile strain (average tensile strain at onset of cracking) maximum extreme fibre strain resultant extreme fibre strain maximum diagonal tensile strain resultant diagonal tensile strain average horizontal strain = 6L/L limiting tensile strain maximum strain volumetric strain effective angle of shearing resistance angle of dilation curve-fitting constant curve-fitting constant microstrain Poisson’s ratio rotation or slope relative bending stiffness principal direct stresses

= E//E,P

vertical effective stress tilt, rigid body rotation

xx

Building response to tunnelling

I

Introduction

F M Jardine, J B Burland and J R Standing

1.I

SUMMARY

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

This opening chapter explains the reasons for publication of this book. It acknowledges the sponsors of the research and the book, explaining their various roles. Both the research on which it was based and the production of this book were collaborations between many people and organisations. The chapter therefore describes the methodology by which the book has been put together, and makes clear the authorship of the different chapters. By spelling out the scope and coverage of the book, noting any particular exclusions, it is intended that the introduction will also provide for readers a “road-map’’ of the book.

1.2

THE REASONS FOR THE BOOK This book is about many unsung successes of the Jubilee Line Extension. The great achievement of constructing a major underground railway line through the heart of London is already taken for granted. At risk of being forgotten are the numerous individual tasks whose completion was made possible by the ingenuity, care and attention to detail of designers, constructors and supervisors. The imposing architecture of Canary Wharf station, the plain strength of the structural elements of Westminster station, the complexity of the layout at London Bridge station, all these are easily recognised as fine engineering. For the tunnels between them and the great excavations that form the platforms and concourses, there is little public recognition of what was involved, even by those walking into a concourse or speeding by train. There is even less appreciation that this was all done while the life and work of the city went on with minimal disruption at the surface. How that was achieved forms the subject of this book. In part a record of activities that kept the buildings above safe, it is mainly a series of case studies of the interactions between the excavations below ground and the buildings at the surface. The purpose of the book, however, is not to extol success, but rather to record facts facts that will underpin future designs, lead to even better engineering and that will help to save money. The facts were obtained by careful field observations and measurements. They are given in the form of case histories for many of the buildings that were affected by the construction of the Jubilee Line Extension. At the parliamentary hearings in select committees for the bills that would permit construction of the Jubilee Line Extension, the absence of good case history information both lengthened the questioning by third parties and weakened the authority of experts. The cost of the hearings was considerable, the cost of the undertakings given to building owners was high, and the cost of the protective works that had to be instituted was very high. The fundamental purpose of the research was to gather the best possible field data; its overall objective was to put the data together in the form of coherent case histories so that the engineering of future works could be undertaken with more confidence and at less cost.

Ch 1 Introduction

1

From a researcher’s viewpoint, construction of an underground railway is an excellent opportunity to study the series of interactions between works at depth, the ground response and the effects on structures in the ground and buildings at the surface. The range of situations - of different geometries (tunnel sizes, shapes, depths, spacings, alignments); of different building types and foundation; of different geologies and ground conditions; and of different construction methods - gives a multitude of options for study. There is one further advantage: timing - knowing when things will happen.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Opportunity is not enough on its own, however; nor is a wish of a researcher. For that scale of research to be fitted into the demands of a very large construction project and its attendant contracts, there are two other pre-requisites: a permitting authority and a central will to learn. London Underground Limited (LUL), which will commission future tunnels, gave both the drive to learn and authority to proceed. It made the research possible. It provided staff, funding, and IT expertise and hardware. Other organisations supported and took part in the work in various ways so that it had the breadth of coverage and quality that were needed. Nothing, however, could have been done without LUL’s encouragement and permission to measure and record while the works were going on. Thus it was that London became a laboratory. The design of the research project included steps to achieve the overall objective of passing on the lessons for engineering practice. This book is one of those steps. A book was chosen as the main medium of dissemination for several reasons. First, it is a very efficient way of presenting the results. The alternative of writing an equivalent number of journal and conference papers would require wasteful repetition of the scene-settings about the JLEP, about the research, about the geology etc, leaving less space for the case study itself. Time slips away in finding different vehicles for publication, diminishing both the impact and the enthusiasm to publish. Worse, though, is that the information is scattered so widely that potential users must struggle to find it all. By bringing together in one place as many case studies as possible, the full scope of the studies becomes clear to readers. The efficiencies of writing and publication are thus also passed on to users of the information. The aim for this book is that it should be a single source reference document, one to be used for many years by practising engineers and researchers, whether structural or geotechnical engineers, tunnelling engineers or building professionals. If it is so used, the book will have done justice to the faith of LUL and the other sponsors and to the care and hard work of the researchers.

1.3

SPONSORS There is more than one research project reported here. They have various stages and levels of sponsorship. The production of this book was made possible by a further stage of sponsorship. The Foreword makes acknowledgement of the organisations and companies who supported any of those stages. Because a list is insufficient to explain the roles of the sponsors, the following account is given. At the stage of parliamentary select committee hearings for the JLE bill, Professor John Burland of Imperial College, who was one of the experts advising LUL, learnt of the then new LINK research programme on construction maintenance and refurbishment (LINK-CMR). The research programme’s sponsoring organisations were the Department of the Environment and the Science and Engineering Research Council, as

2

Building response to tunnelling

they were then called, or as now, the Department of the Environment, Transport and the Regions (DETR) and the Engineering and Physical Sciences Research Council (EPSRC). After preliminary discussions with Dr Brian Mellitt, Director of Engineering of LUL, Professor Burland and Dr Robert Mair of the Geotechnical Consulting Group, with the backing of Mott MacDonald, drafted an outline research proposal for LUL to consider. LUL gave its support in principle at that stage to the notion that the construction works of the JLEP should be the subject of a large-scale series of studies. It insisted on an important proviso - that the research should incorporate wide industrial collaboration, in order that the knowledge gained would be available for take-up by the contractors and consultants who would be competing for future LUL commissions.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

It was to encourage that wide collaboration and to be able to use its established dissemination routes that CIRIA was asked to help prepare and submit a proposal to the LINK-CMR programme. At about the same time, another research proposal was being submitted to SERC for a responsive mode grant. This, again with large LUL support, was for the installation of surface and subsurface instrumentation near the future JLE running tunnel positions at St James’s and Southwark Parks. In the four-year course of the research, and subsequently for the preparation of this book, there have been many forms and contributors of support. Specific funding was provided by LUL, the Research and Development Enabling Fund of the Institution of Civil Engineers, and several private companies for the secondment of Fin Jardine to work on production of this book. In broad terms, the forms of support can be considered as: provision of and funding for the researchers (LUL and EPSRC) provision of instrumentation, equipment and consumables (LUL and EPSRC) provision of services (LUL and EPSRC) contributions of the time of experts in making predictions (GCG and Imperial College) supervision, oversight of the research, and steering (Imperial College, GCG, LUL, and other industrial participants in the LINK-CMR project) provision of IT expertise and resources (LUL) contributions of data, technical information, and advice (JLEP site supervisory staff, JLEP contractors, specialist contractors, and technical contractors) safety training for researchers (JLEP and JLEP contractors) provision of funding support for CIRIA staff input (DETR, CIRIA Core Programme, LUL, industrial participants in the LINK-CMR programme, and sponsors of the preparation of the book. Chapter 4 describes the research project in more detail and identifies the inputs to it of the many participants.

Ch 1 Introduction

3

1.4

THE PREPARATION OF THIS BOOK

1.4.1

Intended readership

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The primary purpose of the research effort reported here is to help advance engineering practice in relation to urban soft-ground tunnelling. That involves the prediction and management of the effects of excavations on buildings as well. Therefore, the book is intended for use by several groups of construction professionals and researchers, including tunnel designers, geotechnical, civil and structural engineers, and building surveyors. They may be professional advisers to tunnel promoters, contractors undertaking the main or associated specialist works, those working on behalf of building owners and insurers, or members of other research teams. Undoubtedly, there is a degree of bias towards the aspirations of geotechnical engineering. This is because the leaders of the research are geotechnical experts and often it is the geotechnical vision that is the driving force for field measurements at large and full scale. The chief aspiration is to achieve a better understanding of the way buildings perform. To achieve this requires understanding the ground’s behaviour, but that is a means to the end not the end itself.

1.4.2

Aut horship There are three types of material in this book: general and scene-setting chapters; predictions made before the events or without knowledge of what actually took place; and case histories. The authors of the case history chapters are, with few exceptions, the members of the research team who carried out the monitoring, recorded the observations and analysed the results. For each of the buildings monitored under the LINK-CMR research project, the researchers prepared a factual report that CIRIA issued to funders of the project (these are termed interim reports). In some cases, the interim report has been edited or updated by another person to the form used in these case study chapters. There are also two building case studies - those of the clock tower of the Palace of Westminster and the RAC building - that were not within the original research project. In the former case, the research team took many of the measurements on parts of the structure. The author here was a member of the GCG geotechnical advisory team employed by JLEP. The chapters that present predictions for the greenfield sites, which were made by finite element methods, were written by those who carried out the analyses and are abbreviated versions of their original prediction reports. There are two chapters presenting predictions of the responses of buildings. These are full transcriptions of the original prediction reports. They have been given in full to show the methodology used. As noted above, one of the efficiencies of a book is that all the scene-setting and background material common to every case study, needs only to be stated once. Here the authors of these general chapters are those members of the whole team of authors most expert in or appropriate to the subject. Additional searches were made by an architectural historian on the case study buildings and, particularly, about the historical development of the areas around them.

4

Building response to tunnelling

For the above reasons, therefore, it is proper that the authorship of each chapter is acknowledged below the title heading. Reference citations to a specific chapter can therefore be attributed to the author, rather than to the editors.

1.4.3

Review processes As this book is intended for practitioners and researchers, the editors should make clear the review processes for the information presented in it. This is doubly important in that it is an output from both an academic institution and from CIRIA.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

First, at the interim report stage, the drafts from the research team members were reviewed by the team leader and the CIRIA project manager and, in some cases, by other reviewers on behalf of CIRIA. After revision, they were then issued to the members of the project management committee. Some were sent to contractors and specialists or others providing information about that case study. Draft chapters for this book were reviewed by the editors (acting as an editorial panel) and after revision were issued for comment to the editorial advisory committee (see the Acknowledgements for the membership of the committee). Copies were sent to LUL for their approval. After a hrther stage of revision, copies were sent to other relevant parties, such as the main and specialist contractors. As with all CIRIA reports, the book was submitted to a review process additional to the ones listed above to obtain approval from the director general of ClRIA for publication.

1.4.4

Coverage The first point is that the studies included in this book are drawn from only part of the Jubilee Line Extension route, ie the section between Green Park and Canada Water stations. The research team made no case studies of buildings farther east. Even in the route between those two stations, there are considerable lengths without case study buildings, eg from Waterloo to London Bridge, from London Bridge to Old Jamaica Road, Bermondsey. Of the buildings between Green Park station and the start of the section beneath St James’s Park, only two are reported here. In fact, several buildings and groups of buildings were monitored, but there has not been time yet to prepare their case histories. They are part of the content of a current research project being undertaken jointly by Imperial College and Cambridge University. Brief summaries are given of the findings from the two thoroughly instrumented crosssections at St James’s and Southwark Parks, the greenfield sites. The results from St James’s Park and the analysis of the measurements at Southwark Park are subjects of the current joint research project. In most cases, the data presented in the case studies were obtained from measurements made by the members of the research team. Where the case study is based upon, or uses, measurements by others this is made clear in the relevant chapter.

1.4.5

Excluded matters A huge amount of investigation, assessment, analysis and design work went into the construction of the JLE. The archive held by LUL is very large. Inevitably, much of what was done at earlier stages was by its nature preliminary or was later superseded by changes to layouts or methods.

Ch 1 Introduction

5

It has been a matter of balance as to the amount of that material to include or refer to in the case studies. In general, the policy adopted is neither to include it nor to reference it. It would be unfair to refer to (and take too much space to explain) the damage risk assessments usually made quite properly with caution but based on preliminary information or alignments and layouts subsequently changed. Where an indication of the assessment is given in a case study, which is usually only in order to explain whether ground treatment was made mandatory under the contract, the attribution is not given.

Chapter4

Explain the research design

/

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

\

I

'W Chapters Sto 9

I

I

Present project-wide construction information

16and17

I Present research methods and data handling

+

Predictions of

-

ground and building response for the case studies

/

-

>

Chapters 17to20

I

Derive and present the research findings and lessons for practice

------------

Case studies of building response

-

' Case studies of protective measures

Figure 1.I

6

7

Chapters 22 to 46 (Volume 2)

\

r

The structure of the two volumes of this book

Building response to tunnelling

Similarly, there is usually no specific acknowledgement of reports, such as the damage and condition surveys, although occasionally these are quoted. They are reports to LUL and not readily accessible. The descriptions of the ground conditions are also deliberately brief. The method used by the editors is to present a summary log with only the following information: a shortened description of each stratum, the depths to changes of strata, the stratigraphic names, and an indication of the relative vertical position of the tunnel(s). Groundwater conditions are described in the text. Not all the monitoring carried out on the case study buildings is necessarily reported. Usually, only the precision levelling by the research team is used, but there are situations where the contractor’s or JLEP results are given instead or as well.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

As is well known, a great many electrolevels were installed to monitor building movements. Almost none are reported here. Too often, the data require considerable analysis and interpretation. A common difficulty was that of the sensitivity to temperature change, even from draughts, the effect of temperature on the instrument masking any building movement from the tunnelling. Although there were some major changes in the work, such as the effect of the Heathrow collapse on the NATM tunnelling, there is no discussion here of any of the contractual questions that arose on JLEP contracts.

1.5

THE STRUCTURE OF THE BOOK The design here is as follows and as summarised in Figure 1.1. The two volumes differ in content. Volume 1 contains information general to the JLEP, to the research project, and to the case studies, so that it does not have to be repeated in them. Rather, it is cross-referenced in the case studies. The background and scene-setting about the JLEP, the current methods of assessing potential building settlements and damage categories, and a description of the research are given in Chapters 1 to 4. The desk study information about the ground conditions and the historical development of the areas around the case study buildings is contained in Chapters 5 to 9. The construction methods and the techniques used to protect the buildings from damage are described in Chapters 10 and 1 1. An important element of the research design was to include a suite of high-quality predictions made using best practice at the time. These predictions are given in Chapters 12 to 15. As five ofthe case study buildings were above London Bridge station and eight were in the Bermondsey part of the route, there are separate chapters (1 6 and 17) to describe the works there. Chapters 18 to 20 deal with the methods of measurement and their precision and cover the compilation of the database to record all the monitoring and construction records. The final chapter (21) of Volume 1 is an overview by Professor Burland of the findings of the research. In contrast to the first volume’s broad coverage and general relevance, Volume 2 is a compendium of the case histories. Most are of buildings, but Chapter 25 deals with the St James’s Park instrumentation, Chapter 29 records the effects on existing tunnels, and Chapter 37 is about surface settlements at three reference sites on the Lambeth Group geology.

Ch 1 Introduction

The case studies follow the geographical sequence from the western end at Green Park to Westminster, Waterloo and London Bridge, where the tunnels are in London Clay, and from there to Bermondsey and Canada Water, where the tunnelling was in the beds of the Lambeth Group. The building case studies tend to follow the same formal pattern of headings, ie: Summary Situation and ground conditions Description of the building Damage assessments and protective measures Construction methods and sequences Monitoring Settlement of the building and damage Concluding remarks

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

References. This pattern has been modified somewhat for the major case studies, which involved a greater variety of measurements or more complex protective measures.

1.5.2

Information in the database The research project database (see Chapter 19), which is owned and held by LUL, is still being added to. It contains all the data presented in this book and very much more. There are construction records, such as shift reports, tunnelling machine instrumentation records, monitoring by the contractors and specialists, and copies of special reports. That database is a unique archive. It remains as a source of information for further research.

1.5.3

References and project bibliography The style used in the book for referencing is to append to each chapter the bibliographic references cited in the chapter. A combined reference list is given at the end of Volume 2. There also is a supplementary bibliography of publications specific to the JLEP works.

8

Building response to tunnelling

2

Structures and contracts of the Jubilee Line Extension

F M Jardine

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

2.1

SUMMARY This chapter describes those parts of the JLE project relevant to the specific tasks of the research presented in this book. It is included at the beginning of this book for three reasons. The first is context - to show the grandeur of the completed JLE works, both in physical scale and engineering and aesthetic quality. This is best done through pictures. The second reason is to acknowledge and list in one place the designers, supervisors and contractors whose works are implicitly referred to throughout the book. The third reason is more mundane - to explain that the information in the book applies only to the western part of the JLE route, between Green Park and Canada Water stations. The chapter deals only briefly with those few aspects of this immense construction project that have a bearing on the main content of the book. A bibliography is given, therefore, of other publications about the JLEP.

2.2

THE COMPLETED JUBILEE LINE EXTENSION The Jubilee Line Extension (JLE) was built between the existing Green Park station, in the West End, to Stratford, in east London, a distance of 15.5 km, of which the western length of some 1 1.5 km is in twin tunnel. The running tunnels are 4.4 m ID and at depths of 20-30 m below ground level. The full length of the JLE is shown in the simple map of London of Figure 2. I . In all, there are 1 1 JLE stations. Four were constructed in open cut, five as enlarged tunnels or a combination of enlarged tunnel and open cut, and there are two surface stations.

0

1

2km

w

City

Rotherhithe

Deptford

Figure 2.1

The full route of the Jubilee Line Extension

Ch 2 Structures and contracts of the Jubilee Line Extension

9

....

=~

-. ",

The research studies that are the subject of this book were made only on the western part of the route, between Green Park and Canada Water. The descriptions and explanations in the following sections apply only to the works completed between those stations. There have been some splendid accounts of the design and construction of the JLE in journals, which often set aside whole issues or special supplements to feature the work. This chapter has drawn heavily on those accounts. They are acknowledged both in the references at the end of this chapter and the JLEP bibliography at the end of Volume 2.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

If by the exclusion made above this book does not describe Canary Wharf station or the major temporary works that were needed to construct it, few will disagree that it is a superb combination of engineering and architecture. Few would not be overawed by its size and vaulting the first time they use the station. Any engineer would be proud to have been associated with it. It is so impressive that, although there is barely another mention of the station in these volumes, it is illustrated here as Figure 2.2.

Figure 2.2

Canary Wharf station

To a comparable extent, although in different ways, these sentiments also apply to the other subsurface stations, the integrated transport at Canada Water, the complex junctions of the great stations of Waterloo and London Bridge, the massive structural strength and depth of Westminster, the simple functionalism of Bermondsey. Add to those the railway engineering and there is ample justification for concurring with the view of a member of this book's editorial advisory committee that the JLE is the real Millennium project of London.

10

Building response to tunnelling

Taken as a tunnelling project, the JLE was a great achievement, not the least for the river crossings and the tunnelling in the layered sands and clays of the Lambeth Group. If overshadowed by the then recently completed Channel Tunnel, the JLE tunnels were not much less in length, were at relatively shallow depth, and were bored below some of the most important places in London, including the Big Ben clock tower, the vital termini of Waterloo and London Bridge. The businesses and homes of Bermondsey and Rotherhithe may not receive the media coverage of the towers and penthouses of Canary Wharf, but they are probably even more important to their occupants and owners. It will not be forgotten, also, that the JLEP tunnelling in London Clay using NATM techniques was put on hold for more than three months after the collapse at Heathrow in 1994. Overall, the JLEP tunnelling was a great success, not the least for its integration with the works of compensation grouting to protect buildings against possible damage.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The ancillary works, undertaken in support of the tunnelling and open excavations, were also on a very large scale. The total length of near-horizontal drilling for grouting amounted to 100 km. The injections of grout from these arrays of drillhole were numbered in the hundreds of thousands. The amount of instrumentation, the numbers of levelling points installed on and around buildings, and the time expended by survey staff taking measurements on them were enormous.

2.3

THE ROUTE The length of the JLE route between Green Park and Canada Water stations is shown in Figure 2.3, on which the positions of the case studies reported in this book are highlighted. Also shown are other buildings that were the subject of research monitoring, but whose case histories have yet to be finalised. London Bridge

Waterloo

Canada Water I

I

I II

Contract 102

Contract 103

lI

Contract 104

lI

Contract 105

I

i

I

0

I

II Figure 2.3

0

1 km

1

Contract 106

Shafts Areas of case studies

U

The length of the JLE studied and the locations of case study buildings

Ch 2 Structures and contracts of the Jubilee Line Extension

11

2.3.1

The stations Green Park

i

r

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The junction here with the existing Jubilee Line involved construction of a step plate junction (Contract 102), linking pedestrian tunnels between the Jubilee and Piccadilly lines, and shafts (Contract 101). Figure 2.4 is a view at the step plate junction.

Figure 2.4

12

Green Park step plate junction

Building response to tunnelling

Westminster

As well as being one of the deepest open excavationsin London and next to arguably the most precious symbol of Britain - the Big Ben clock tower - Westminster station had to be built while keeping safe and operational the existing District and Circle Line stations at a bigher level within the station box. The new platforms for the Jubilee Line are stacked one above the other outside the box, but with eight tunnelled connections into it fbr escape and ventilation.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The wall construction,the tunnelled props at the base of the walls, the massive columns to support the New ParliamentaryBuilding (PortcullisHouse),the wide concourses and escalator systems are each in themselves a major construction task. Together they coxnppise a hugely strong structwe, plain to see to any passenger, without false fiills and finishes,just pure engineering.

Figure 2.5

Westminster station

Ch 2 Structures and contracts of the Jubilee Line Extension

13

Waterloo

Under this important railway terminus,the new Jubilee Line station was a fully

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

tunnelled layout. Each of the two station tunnels was provided with emergency escape and ventilation tunnels at both ends. As well as escalator and mnmurse tunnels, there was also a tunnelled interchangeto the existing Northern and Bakerloo lines with a ‘moving walkway.

14

Building response to tunnelling

. -,.

London Bridge

The work here involved the construction of the new Jubilee Line station complex and a new southbound platform tunnel for the Northem Line. With a new ticket hall, central concourse, many cross-passages, east and west vent, escape and escalator shafts to the JLE station, and escalator shafts to the Northern Line platforms, the NATM-tunnelled excavations were large and very complex.

South escalator shaft to Borough High stnretfidcethall

East escalator shaff

.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Jubilee

Jubilee! Line

k I Existina Northem Lm,

to Joiner

Back shaft

Figure 2.7

London Bridge station, schematic view of the station complex (after Costain-Taylor Woodrow)

Figure 2.8

London Bridge station concourse

Ch 2 Structures and contracts of the Jubilee Line Extension

15

Bermondsey

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Bermondsey station again epitomises simplicity of design, transparency of function, bare bones of structure. It is, perhaps, a plain station for a plain place, but its openness and light make it quite different fiom earlier underground stations. There is little consciousnessof descent to the bowels of the earth. Even the metal linings of the platform tunnels seem to reflect more light than they absorb despite their grey colour (again a reality of both their nature and function).

I

I Figure 2.9

16

Bermondsey station showing trusses

Building response to tunnelling

Canada Water

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Canada Water station is impressive. Joining the Jubilee Line and the East London Line it incorporates a bus station at the surface as well. As with Bermondsey, it gives the feeling of space and openness, but the scale approaches that of Canary Wharf station. There is nothing claustrophobic, nothing ungenerous in its dimensions.

I Figure 2.10

?am”- Wafr- -fatio-

Figure 2.11

Canada Water station with one of the tower blocks of the Canada Estate, Columbia Point, on the left

,

Ch 2 Structures and contracts of the Jubilee Line Extension

17

2.3.2

The tunnels

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The methods used in tunnelling that are relevant to the case studies are described in Chapter 10. Here only a flavour is given by a photograph of a finished tunnel. The four main types of tunnel excavation were: hand mining; open-face shield; NATM, which was later prescribed as the sprayed concrete lining (SCL) method; and earth pressure balance tunnelling machine (EPBM). Linings (Chapter 10) were of sprayed concrete and segments of spheroidal graphite iron or precast concrete. The use of SGI linings with bolted-on facings provides in many platform and escalator tunnels a dramatic modem fin&, but one that suggests strength and security.

Figure 2.12

2.3.3

Emergency escape tunnel at Waterloo station

Ancillary works It is difficult to know where to draw the line between main and ancillary works, even just for the civil works. The temporary works cover a wide range of types of operation. Many were to greater or lesser extent incorporated into the permanent works, even if their purpose was temporary. This book contains case histories that include building protective measures. While the generic method of ground treatment (specifically, compensation grouting) was the most widely applied, many other forms of protection were used. As pointed out in Chapter 11, which describes these methods in more detail, the first protective measure was that of design to avoid the risk, and the second was good workmanship. Other methods included underpinning,foundationjacking, tie rods, soil nailing, a pipe arch canopy over an upper tunnel, and flying shores, among others.

18

Building response to tunnelling

The quantities of work on compensation grouting were extraordinary. Table 2.1 gives some estimates of these quantities on Contracts 101 to 105 made during the works. Associated with the building protection was monitoring to ensure that settlements and building movements were not untoward or were being properly controlled by grouting. Table 2.1 also provides an indication of the survey work and instrumentation data handling involved. Table 2.1

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

2.4

An indication of some of the quantities for grouting works and monitoring on JLEP Contracts 101 to 105

Item

Contract I02

Grout hole drilling length (m)

103 980

Precision levelling points (number)

7810 (and 3000 others)

Levelling survey readings (number)

970 400 readings

ORGANISATIONS AND PEOPLE This section acknowledges the organisations and some of the people whose work went towards the completion of those parts of the Jubilee Line Extension which are referred to in the case studies of this book. Again, the list is not that for the whole JLEP.

2.4.1

LUL and the JLE Project team The first acknowledgement has to be to the support of LUL for the research. Initially from the then director of engineering, Dr Brian Mellitt, and subsequently from Keith Beattie, Mike Gellatley and Jim Moriarty, it was their support that made the research and the writing of this book possible. In parallel, and essential because the research had to take place in the context of the rapidly changing conditions of ongoing design and construction, was the active help of the Jubilee Line Extension Project team. To have the support of Hugh Doherty, JLE Project director, and Mike Smith, project construction manager, when running the largest and most demanding construction project in the UK at that time was a privilege. David Sharpe, the JLEP chief engineer, gave his encouragement and advice throughout the course of the research, both of which were of great value. Most of the burden of procurement, administration, persuasion, communication and management of JLEP aspects of the research was taken up by Lionel Linney, JLEP senior geotechnical engineer. It was his drive that maximised the information and resources available to the research project. At the same time, he understood and took into account the constraints on the site supervisors. Members of the research team came from his staff. Without his efforts, not only would the research results have been much thinner, but the unique database of construction records would not have been compiled for use. Over the several contracts of the JLE, many staff provided information, physical help, training, and encouragement to the research team. The senior supervisory staff of the JLE contracts are named in the next section, but it is right to acknowledge that many others in their site teams were permitted to help the research. As theirs was a capacity general to the JLEP and not specific to contracts, acknowledgement should be made here of the specialist geotechnical advisers to the JLEP, Professor Robert Mair and his colleagues at the Geotechnical Consulting Group (GCG). Their knowledge of the whole project, their technical contributions, eg of the predictions of ground and building response, were an integral and very important part of the research.

Ch 2 Structures and contracts of the Jubilee Line Extension

19

2.4.2

Construction contracts The following lists are to a standard format, giving the organisations that were involved as technical contractors, architects, main contractors, and specialist ground treatment contractors. There are references to the case study chapters that are concerned with structures or work carried out on these contracts.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

2.4.3

2.4.4

20

Contract 101: Green Park station modifications Technical contractor

L G Mouchel and Partners

Architect

Manser Associates

Contractor

Tarmac Construction

Work description

At existing Green Park station, linking pedestrian tunnel between Jubilee and Piccadilly lines, escape shaft

Ground treatment

Bachy-Cementation Joint Venture

Case study chapters

23. The Ritz

Contract 102: Green Park to Waterloo Technical contractor

G Maunsell and Partners

Architect

Michael Hopkins and Partners (Westminster station) JLE Project design team (Waterloo)

Contractor

Balfour Beatty-AMEC Joint Venture

Work description

Westminster and Waterloo stations, running and platform tunnels, escape and ventilation tunnels

Ground treatment

AMEC-Geocisa Joint Venture

Case study chapters

12. Predictions, St James’s Park 14. Predictions, Elizabeth House 24. The RAC building 25. St James’s Park greenfield site 26. The Treasury building 27. Great George Street buildings 28. The Clock Tower of the Palace of Westminster 29. Existing LUL tunnels 30. Elizabeth House

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

2.4.5

2.4.6

Contract 104: London Bridge station Technical contractor

Mott MacDonald

Architect

Weston Williamson

Contractor

Costain-Taylor Woodrow Joint Venture

Work description

London Bridge Jubilee Line and Northern Line southbound stations and associated works

Ground treatment

Keller (under subcontract to Soletanche)

Case study chapters

16. London Bridge station 3 1. 1-7 St Thomas Street 32. London Bridge Post Ofice 33. Telephone House 34. The BT Building 35. Fielden House

Contract 105: London Bridge to Canada Water Technical contractor

Sir William Halcrow and Partners

Architect

Ian Ritchie Associates

Contractor

Aoki-Soletanche Joint Venture

Work description

Bermondsey station and running tunnels from London Bridge to Canada Water

Ground treatment

Keller-Soletanche

Case study chapters

13. Predictions, Southwark Park greenfield site 15. Predictions, Moodkee Street buildings 17. Contract 105 methods 37. Surface displacements 39. Keetons Estate 40. 128-130 Jamaica Road 4 1. 182-2 10 Jamaica Road 42. Blick House 43. Moodkee Street buildings 44. Niagara Court 45. Columbia and Regina Points 46. Tenants’ Hall and Boiler House, Canada Estate

Ch 2 Structures and contracts of the Jubilee Line Extension

21

2.4.7

Contract 106: Canada Water Technical contractor

Benaim - Works Joint Venture

Architect

Heron Associates with JLE Project design team

Contractor

Wimpey Construction

Work description

Canada Water Jubilee and East London Lines station

Case study chapters

45. Columbia and Regina Points

46. Tenants’ Hall and Boiler House, Canada Estate

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

2.5

THE CONSTRUCTION PROGRAMME How the research work fitted with the JLE construction programme for the relevant contracts is shown in Figure 2.13. That shows the durations of the several activities relevant to the areas of the case studies. Inevitably, these were somewhat different from the original programme, but there was benefit to the research as a result. It was not only possible to monitor more structures than had originally been intended, but also to monitor them over a longer period than had at first been intended.

1994 1 1995 I 1996 IllQ2lQ3lQ4lQllQ2lQ3lQ4lQllQ2lQ3lQ4

Green Park station

Westminster station Waterloo station

Contract 104

London Bridge station

Contract 105

Bermondsey station

Figure 2.1 3

2.6

The part of the JLEP programme relevant to the monitored case study buildings

JLE DESIGN AND CONSTRUCTION BIBLIOGRAPHY At the end of Volume 2 there is a bibliography listing publications about the design and construction of the JLE, particularly those concerned with tunnelling, the protection of buildings and geotechnical matters.

22

Building response to tunnelling

3

Assessment methods used in design

J B Burland

3.1

INTRODUCTION

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

In recent years, assessment of the environmental impact of major construction projects has become a normal and required procedure. The construction of tunnels in urban areas, while having many long-term environmental benefits, can also create significant environmental problems. During construction, such problems would include construction traffic, noise, vibration and dust as well as temporary restrictions on access to certain roads and other public areas. Longer-term problems could involve land and building acquisition, traffic and ventilation noise and vibration levels, pollution, ground water changes and effects on ecology. An environmental impact of tunnelling that is the subject of increasing public awareness

and concern is subsidence and its effects on structures and services. Construction of tunnels and deep excavations is inevitably accompanied by ground movements. It is necessary, both for engineering design and for planning and consultation, to develop rational procedures for assessing the risks of damage. Coupled with such assessments is, of course, the requirement for effective protective measures that can be deployed when predicted levels of damage are judged to be unacceptable. This chapter summarises the approach that has been adopted for assessing the risks of subsidence damage for the London Underground Jubilee Line Extension, and the proposed CrossRail project. Both projects involve tunnelling under densely developed areas of central London. The approach draws on the results of a number of studies including the prediction of ground movements, the description and classification of damage and limiting distortions of brickwork cladding and masonry walls. These studies are discussed separately and then combined to give the overall approach.

3.2

DEFINITIONS OF GROUND AND FOUNDATION MOVEMENT A study of the literature reveals a great variety of conhsing symbols and terminology describing foundation movements and structural distortion. Burland and Wroth (1 974) proposed a consistent set of definitions based on the displacements (either measured or calculated) of a number of discrete points on the foundations of a building. Care was taken to use terms that do not prejudice any conclusions about the distortions of the superstructure itself because these depend on many additional factors. The definitions appear to have been widely accepted and are illustrated in Figure 3.1. The following are a few points to note: (a) Rotation or slope 8 is the change in gradient of a line joining two reference points (eg AB in Figure 3.1(a)). (b) The angular strain a is defined in Figure 3.l(a). It is positive for upward concavity (sagging) and negative for downward concavity (hogging). (c) Relative deflection A is the displacement of a point relative to the line connecting two reference points on either side (see Figure 3.1(b)). The sign convention is as for (b).

Ch 3 Assessment methods used in design

23

(d) Deflection ratio (sagging ratio or hogging ratio) is denoted by AIL where L is the distance between the two reference points defining A. The sign convention is as for (b) and (c). (e) Tilt w describes the rigid body rotation of the structure or a well-defined part of it see Figure 3.l(c). (f)

Relative rotation (angular distortion) p is the rotation of the line joining two points, relative to the tilt w- see Figure 3.l(c). It is not always straightforward to identify the tilt and the evaluation of p can sometimes be difficult. It is also very important not to confuse relative rotation p with angular strain a.For these reasons Burland and Wroth preferred the use of NL as a measure of building distortion.

(g) Average horizontal strain Eh is defined as the change of length 6L over the length L. In soil mechanics it is customary to take a reduction of length (compression) as positive.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

A

Figure 3.1

B

C

A

Definitions of ground and foundation movement (a) settlement s, relative settlement 6s, rotation 8, and angular strain a: (b) relative deflection A and deflection ratio AIL (c) tilt o and relative rotation (angular distortion) p

The above definitions only apply to in-plane deformations and no attempt has been made to define three-dimensional behaviour.

24

Building response to tunnelling

3.3

GROUND MOVEMENT DUE TO TUNNELLING AND EXCAVATlON The construction of tunnels or surface excavations will inevitably be accompanied by movement of the ground around them. At the ground surface these movements manifest in what is called a “settlement trough”. Figure 3.2 shows diagrammatically the surface settlement trough above an advancing tunnel. For “greenfield sites”, the shape of this trough transverse to the axis of the tunnel approximates closely to a Gaussian normal distribution curve - an idealisation that has considerable mathematical advantages. Figure 3.3 shows such an idealised transverse settlement trough.

%\

\+x Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

U

/

Extent of surface ,settlement trough

>I S

-X

Figure 3.2

3.3.1

Surface settlement trough above an advancing tunnel

Settlements caused by tunnel excavation Attewell et a1 (1986) and Rankin (1988) have summarised the current widely used empirical approach to the prediction of immediate surface and near-surface ground displacements resulting from tunnel excavation. The settlements is given by:

where smaX is the maximum settlement and i is the value o f y at the point of inflection. It has been found that, for most purposes, i can be related to the depth of the tunnel axis z, by the linear expression:

Ch 3 Assessment methods used in design

25

The trough width parameter K depends on the soil type. It varies from 0.2 to 0.3 for granular soils through 0.4 to 0.5 for stiff clays to as high as 0.7 for soft silty clays. As a general rule, the width of the surface settlement trough is about three times the depth of the tunnel for tunnels in clay strata. It is important to note that, although the value of K for surface settlements is approximately constant for various depths of tunnel in the same ground, Mair et ul (1 993) have shown that its value increases with depth for subsurface settlements. The immediate settlements caused by tunnelling are usually characterised by the “volume loss” VL,which is the volume of the surface settlement trough per unit length V, expressed as a percentage of the notional excavated volume of the tunnel. Integration of equation (3.1) gives:

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

so that: 3.192 i smaX

VL=

(3.4)

D2

Where D is the diameter of the tunnel. Combining equations (3.1), (3.2) and (3.4) gives the surface settlements at any distancey from the centre line:

(3.5)

3.3.2

Horizontal displacements due to tunnelling Building damage can also result from horizontal tensile strain, and therefore predictions of horizontal movement are required. Unlike settlements, there are few case histories where horizontal movements have been measured. The data that do exist show reasonable agreement with the assumption of O’Reilly and New (1982) that the resultant vectors of ground movement are directed towards the tunnel axis. It follows that the horizontal displacement U can be related to the settlements by the expression:

Equation (3.6) is easily differentiated to give the horizontal strain Eh at any location on the ground surface. Figure 3.3 shows the relationship between the settlement trough, the horizontal displacements and the horizontal strains occurring at ground level. In the region i > y > -i, the horizontal strains are compressive. At the points of inflection the horizontal displacements are a maximum and Eh = 0. For i < y < -i, the horizontal strains are tensile.

26

Building response to tunnelling

7 Horizontal displacement U

Figure 3.3

3.3.3

Transverse settlement trough

Assessment of surface displacements due to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The above empirical equations provide a simple means for estimating the near-sur,,ce displacements due to tunnelling, assuming greenfield conditions, ie ignoring the presence of any building or structure. A key parameter in this assessment is the volume loss VL. This results from a variety of effects, which include movement of ground into the face of the tunnel and radial movement towards the tunnel axis due to reductions in supporting pressures. The magnitude of VL is critically dependent on the type of ground, the groundwater conditions, the tunnelling method, the length of time in providing positive support and the quality of supervision and control. The selection of an appropriate value of VL for design requires experience and is greatly aided by well-documented case histories in similar conditions. A number of other assumptions are involved in the prediction of ground displacements due to tunnelling. For example, in ground containing layers of clay and granular soils there is uncertainty about the value of the trough width parameter K . When two or more tunnels are to be constructed in close proximity, the assumption is usually made that the estimated ground movements for each tunnel acting independently can be superimposed. In some circumstances, this assumption may be unconservative and allowance needs to be made for this. It is clear from the above that, even for greenfield conditions, precise prediction of ground movements caused by tunnelling is not realistic. However, it is possible to make reasonable estimates of the likely range of movements provided tunnelling is carried out under the control of suitably qualified and experienced engineers.

3.3.4

Ground movements due to deep excavations Ground movements around deep excavations are critically dependent both on the ground conditions (eg stratigraphy, groundwater conditions, deformation and strength properties) and the method of construction (eg sequence of excavation, sequence of propping, rigidity of retaining wall and supports). In general, open excavations and those supported by cantilever retaining walls give rise to larger ground movements than strutted excavations and those constructed by top-down methods. In urban areas, the latter are clearly to be preferred if building damage is to be minimised.

Ch 3 Assessment methods used in design

27

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The complexity of the problem means that calculation of ground movements is not straightforward, and much experience is needed to make any sensible use of complex analyses. It is therefore essential that optimum use is made of previous experience and case histories in similar conditions. Peck (1969) presented a comprehensive survey of vertical movements around deep excavations, which was updated by Clough and O’Rourke (1990). Burland et a1 (1979) summarised the results of more than ten years of research into the movements of ground around deep excavations in London clay. The Norwegian Geotechnical Institute has published several case histories of excavations in soft clay in the Oslo area (eg Karlsrud and Myrvoll, 1976). For well-supported excavations in stiff clays, Peck’s settlement envelopes are generally very conservative with settlements rarely exceeding 0.15 per cent, but movements can extend to three or four times the excavation depth behind the basement wall. Horizontal movements are generally of similar magnitude and distribution to vertical movements, but may be much larger for open and cantilever excavations in stiff clays. Advanced methods of numerical analysis, based on the finite element method, are widely used for prediction of ground movements around deep excavations. Such analyses can simulate the construction process, modelling the various stages of excavation and support conditions. However, comparison with field observations, shows that successful prediction requires high-quality soil samples with the measurement of anisotropic small-strain stiffness properties using local strain transducers mounted on the sides of the samples (Jardine et al, 1984) and geophysical testing methods. As for tunnelling, it is essential that deep excavation work be carried out under the close supervision of an experienced engineer. Unless positive support is provided rapidly and groundwater is properly controlled, large unexpected ground movements can develop.

3.4

CLASSIFICATION OF DAMAGE

3.4.1

Introduction Assessment of degree of building damage can be a highly subjective, and often emotive, matter. It may be conditioned by factors such as local experience, the fimction of the building, the caution of a professional engineer or surveyor concerned about litigation, market value and saleability of the property. In the absence of objective guidelines based on experience, extreme attitudes and unrealistic expectations towards building performance can develop. It is worth stressing that most buildings experience a certain amount of cracking, often unrelated to foundation movement, which can be dealt with during routine maintenance and decoration. Clearly, if an assessment of risk of damage due to ground movement is to be made, the classification of damage is key. In the UK, the development of an objective system of classifying damage is proving to be very beneficial in creating realistic attitudes towards building damage and also in providing logical and objective criteria for designing for movement in buildings and other structures. This classification system can be described as follows.

3.4.2

Categories of damage Three categories of building damage can be considered that affect: (i) visual appearance or aesthetics, (ii) serviceability or function and (iii) stability. As foundation movements increase, damage to a building will progress successively from (i) through to (iii).

It is only a short step from these three broad categories to the detailed classification given in Table 3.1. This defines six categories of damage, numbered 0 to 5 in increasing severity. Normally categories 0, 1 and 2 relate to “aesthetic” damage, 3 and 4 relate to “serviceability” damage and 5 represents damage affecting “stability”. It was first put forward by Burland ef a1 (1977), who drew on the work of Jennings and Kerrich (1962), the UK National Coal Board (1975) and MacLeod and Littlejohn (1974). Since then it has been adopted with only slight modifications by BRE (1981 and 1990), the Institution of Structural Engineers, London (1978, 1989, 1994 and 2000) and by the Institution of Civil Engineers and BRE again in Freeman et al(1994). Table 3.1 ~~~~

Classification of visible damage to walls with particular reference to ease of repair of plaster and brickwork or masonry

~

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Category Normal degree Description of typical damage (ease of repair is in bold type) of of severity Note: Crack width is only onefactor in assessing category of damage and damage should not be used on its own as a direct measure of it.

0

Negligible

Hairline cracks less than about 0.1 mm wide

1

Very slight

Fine cracks that are easily treated during normal decoration. Damage generally restricted to internal wall finishes. Close inspection may reveal some cracks in external brickwork or masonry. Typical crack widths up to 1 mm

2

Slight

Cracks easily filled. Redecoration probably required. Recurrent cracks can be masked by suitable linings. Cracks may be visible externally and some repointing may be required to ensure weather-tightness. Doors and windows may stick slightly. Typical crack widths up to 5 mm.

3

Moderate

The cracks require some opening up and can be patched by a mason. Repointing of external brickwork and possibly a small amount of brickwork to be replaced. Doors and windows sticking. Service pipes may fracture. Weather-tightness often impaired. Typical crack widths are 5-15 mm or several > 3 mm.

4

Severe

Extensive repair work involving breaking-out and replacing sections of walls, especially over doors and windows. Windows and door frames distorted, floor sloping noticeably’. Walls leaning’ or bulging noticeably, some loss of bearing in beams. Service pipes disrupted. Typical crack widths are 15-25 mm, but also depends on the number of cracks.

5

Very severe

This requires a major repair job involving partial or complete rebuilding. Beams lose bearing, walls lean badly and require shoring. Windows broken with distortion. Danger of instability. Typical crack widths are greater than 25 mm, but depends on the number of cracks.

’ Note: Local deviation of slope, from the horizontal or vertical, of more than 1/100 will normally be clearly visible. Overall deviations in excess of 1/150 are undesirable.

The system of classification in Table 3.1 is based on ease of repair of the visible damage. Thus, to classify visible damage it is necessary, when carrying out the survey, to assess the type of work necessary to repair the damage both externally and internally. The important points below should be noted. (a) The classification relates only to the visible damage at a given time and not to its cause or possible progression, which are separate issues. (b) The strong temptation to classify the damage solely on crack width must be resisted. It is the ease of repair that is the key factor in determining the category of damage.

Ch 3 Assessment methods used in design

29

(c) The classification was developed for brickwork or blockwork and stone masonry. It could be adapted for other forms of cladding. It is not intended to apply to reinforced concrete structural elements. (d) More stringent criteria may be necessary where damage may lead to corrosion, penetration or leakage of harmful liquids and gases or structural failure. Besides defining numerical categories of damage, Table 3.1 also lists the “normal degree of severity” associated with each category. These descriptions of severity relate to standard domestic and office buildings and serve as a guide to building owners and occupiers. In special circumstances, such as for a building with valuable or sensitive finishes, this ranking of severity of damage may not be appropriate.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

3.4.3

The division between categories 2 and 3 damage The dividing line between categories 2 and 3 damage is particularly important. Studies of many case records shows that damage up to category 2 can result from a variety of causes, either from within the structure itself (eg shrinkage or thermal effects) or associated with the ground. Identification of the cause of damage is usually very difficult and frequently it results from a combination of causes. If the damage exceeds category 2 the cause is usually much easier to identify and it is frequently associated with ground movement. The division between categories 2 and 3 damage therefore represents an important threshold, which is referred to later.

3.5

CONCEPT OF LIMITING TENSILE STRAIN

3.5.1

Onset of visible cracking Cracking in masonry walls and finishes usually, but not always, results from tensile strain. Following the work of Polshin and Tokar (1957), Burland and Wroth (1974) investigated the idea that tensile strain might be a fundamental parameter in determining the onset of cracking. A study of the results from numerous large-scale tests on masonry panels and walls carried out at the UK Building Research Establishment showed that, for a given material, the onset of visible cracking is associated with a reasonably welldefined value of average tensile strain that is not sensitive to the mode of deformation. They defined this as the critical tensile strain &,,it , which is measured over a gauge length of a metre or more. Burland and Wroth made the following important observations: (a) The average values of at which visible cracking occurs are very similar for a variety of types of brickwork and blockwork and are in the range 0.05 to 0.1 per cent. (b) For reinforced concrete beams the onset of visible cracking occurs at lower values of tensile strain in the range 0.03 to 0.05 per cent. (c) The above values of to tensile failure.

are much larger than the local tensile strains corresponding

(d) The onset of visible cracking does not necessarily represent a limit of serviceability. Provided the cracking is controlled, it may be acceptable to allow deformations well beyond the initiation of visible cracking. Burland and Wroth (1 974) showed how the concept of critical tensile strain could be used in conjunction with simple elastic beams to develop deflection criteria for the onset of visible damage. This work is discussed in more detail in Section 3.6.

30

Building response to tunnelling

3.5.2

Limiting tensile strain - a serviceability parameter Burland et a1 (1 977) replaced the concept of “critical tensile strain” with that of “limiting tensile strain” qirnThe importance of this development is that &lim can be used as a serviceability parameter that can be varied to take account of differing materials and serviceability limit states. Boscardin and Cording (1989) developed this concept of differing levels of tensile strain. Seventeen case records of damage due to excavation-induced subsidence were analysed. A variety of building types was involved and they showed that the categories of damage given in Table 3.1 could be broadly related to ranges of E ~ ~ , , These ,. ranges are tabulated in Table 3.2. This table is important as it provides the link between estimated building deformations and the possible severity of damage.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Table 3.2

Relationship between category of damage and limiting tensile strain (of,,,)(after Boscardin and Cording, 1989)

Limiting tensile strain (ei,,,)

Category of damage

Normal degree of severity

0

Negligible

0-0.05

1

Very slight

0.05-0.075

2

Slight

0.075-0.15

3

Moderate*

0.15-0.3

4 to 5

Severe to very severe

> 0.3

(”/.I

*Note: Boscardin and Cording (1989) describe the damage corresponding to qimin the range 0.15-0.3 per cent as “moderate to severe”. However, none of the cases quoted by them exhibits severe damage for this range of strains. There is therefore no evidence to suggest that tensile strains up to 0.3 per cent will result in severe damage.

3.6

STRAINS IN SIMPLE RECTANGULAR BEAMS Burland and Wroth (1 974) and Burland et a1 (1977) used the concept of limiting tensile strain to study the onset of cracking in simple weightless elastic beams undergoing sagging and hogging modes of deformation. This simple approach gives considerable insight into the mechanisms controlling cracking. Moreover, it was shown that the criteria for initial cracking of simple beams are in very good agreement with the case records of damaged and undamaged buildings undergoing settlement. Therefore, in many circumstances, it is both reasonable and instructive to represent the faGade of a building by means of a simple rectangular beam.

3.6.1

Sagging and hogging Figure 3.4 illustrates the approach adopted by Burland and Wroth (1974), where the building is represented by a rectangular beam of length L and height H. The problem is to calculate the tensile strains in the beam for a given deflected shape of the building foundations and so obtain the sagging or hogging ratio A/L at which cracking is initiated. Little can be said about the distribution of strains within the beam unless its mode of deformation is known. Two extreme modes are bending only about a neutral axis at the centre (Figure 3 . 4 ~and ) shearing only (Figure 3.4d). In bending only, the maximum tensile strain occurs in the bottom extreme fibre, which is where cracking will initiate. For shear only, the maximum tensile strains are inclined at 45: initiating diagonal cracking. In general, both modes of deformation will occur simultaneously and it is necessary to calculate both bending and diagonal tensile strains to ascertain which type is limiting.

Ch 3 Assessment methods used in design

31

(a) Actual building

(c) Bending deformation with cracking due to direct tensile strain

I

(b) Beam-simple idealisationof building

A

Deflected shape of soffit of beam

Figure 3.4

(d) Shear deformation with cracking due to diagonal tensile strain

Cracking of a simple beam in bending and in shear

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Timoshenko (1 957) gives the expression for the total mid-span deflection A of a centrally loaded beam having both bending and shear stiffness as:

Where E is Young's modulus, G is the shear modulus, I is the second moment of area and P is the point load. Equation (3.7) can be re-written in terms of the deflection ratio A/L and the maximum extreme fibre strain &bm,x as follows:

31

E

where t is the distance of the neutral axis from the edge of the beam in tension. Similarly, for the maximum diagonal strain &&,,a,Equation (3.7) becomes:

"=( $ 1+-.-

L

Edmax

(3.9)

Similar expressions are obtained for the case of a uniformly distributed load with the diagonal strains calculated at the quarter points. Therefore, the maximum tensile strains are much more sensitive to the value of A/L than to the distribution of loading. By setting E, = ~ ~ Equations i ~ , (3.8) and (3.9) define the limiting values of A/L for the deflection of simple beams. For a given value of qim, the limiting value of A/L (whichever is the lowest in Equations (3.8) and (3.9)) depends on L/H, E/G and the position of the neutral axis. For example, Burland and Wroth showed that hogging with the neutral axis at the bottom edge is far more damaging than sagging with the neutral axis in the middle - a result that is borne out in practice. Figure 3.5 shows the limiting relationship between A/L normalised by EI~, and L/Hfor an isotropic beam (E/G = 2.6) undergoing hogging with its neutral axis at the bottom edge. For values of L/H< 1.5 the diagonal strains from Equation (3.9) dominate, whereas for L/H > 1.5 bending strains dominate.

32

Building response to tunnelling

2.01.5-

;1$

1.0-

Bending strain controlling

0.5"

I

I

I

I

I

I

I

0

1

2

3 Ll H

4

5

6

Figure 3.5

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

3.6.2

Relationship between and U H for rectangular isotropic beams with the neutral axis at the bottom edge

The influence of horizontal strain It is shown in Section 3.3 that ground surface movements associated with tunnelling and excavation not only involve sagging and hogging profiles but significant horizontal strains as well. Boscardin and Cording (1989) included horizontal tensile strain Eh in the above analysis using simple superposition, ie it is assumed that the deflected beam is subjected to uniform extension over its full depth. The resultant extreme fibre strain &br is given by: (3.10)

E b r = Ebrnax 4-Eh

In the shearing region, the resultant diagonal tensile strain &dr can be evaluated using the Mohr's circle of strain. The value of cdris then given by:

(3.1 1)

where v is Poisson's ratio. The maximum tensile strain is the greater of &br and &dr. Thus, for a beam of length L and height H, it is a straightforward matter to calculate the maximum value of tensile strain E, for a given value of A/L and Eh, in terms oft, E/G and v. This value of emaxcan then be used in conjunction with Table 3.2 to assess the potential associated damage.

1 -1

0

Figure 3.6

-1

Eh/Elim

1

Influence of horizontal strain on (dL)/~i,,,for (a) bending strain controlling, (b) diagonal strains controlling, and (c) combination of (a) and (b)

33

Ch 3 Assessment methods used in design

~

The physical implications of Equations (3.8) to (3.1 1) can be illustrated by considering the previous case of the isotropic beam undergoing hogging with its neutral axis at the bottom edge and v = 0.3. By combining Equations (3.8) and (3. lO), the influence of E,, on the limiting values of A/L can be examined for bending strains only, by setting Figure 3.6(a) shows the normalised relationship between A/L and Eh. &b,,,ax = For Eh = 0 the limiting values of A/L are the same as given in Figure 3.5 for various values of L/H. It can be seen that, as Eh increases towards the value of the limiting values of A/L for a given L/H reduce linearly, becoming zero when Eh = E~~,,,.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Similarly, Figure 3.6(b) has been derived from Equations (3.9) and (3.1 1) for the diagonal strains only. Once again, for Eh = 0 the limiting values of A/L are recovered from Figure 3.5. As &h increases, the limiting values of A/L decrease non-linearly at an increasing rate towards zero. It is of interest to note that the values of A/L are not very sensitive to the values of L/H between 0 and 1.5. Figures 3.6(a) and (b) can be combined to give the resultant normalised limiting relationships between d L and Eh for various values ofL/H as shown in Figure 3.6(c). It can be seen that, for L/H > 1S , the bending strains always control. Also, for lower values of L/H, as Eh increases, the controlling strain changes from diagonal to bending. It must be emphasised that Figure 3.6 relates to the specific case of hogging with the neutral axis at the lower face and with E/G = 2.6. There are similarities between Figure 3.6(c) and the Boscardin and Cording chart of angular distortion p against Eh. The latter chart has the following limitations. (a) It only relates to L/H = I . (b) Maximum bending strains &b,,,ax are ignored. (c)

p was assumed to be proportional to A/L whereas the relationship is in fact very sensitive to the load distribution.

(d) As mentioned in Section 3.2 the evaluation of p is not always straightforward. , with the various categories of damage given in By adopting the values of ~ ~ i , ,associated Table 3.2, Figure 3.6(c) can be developed into an interaction diagram showing the relationship between A/L and Eh for a particular value of L/H. Figure 3.7 shows such a diagram for L/H = 1, which is directly comparable to the Boscardin and Cording diagram.

0

0.1

0.2

0.3

Horizontal strain (%)

Figure 3.7

34

Relationship of damage category to deflection ratio and horizontal tensile strain for hogging (UH = 1)

Building response to tunnelling

3.6.3

Relevant buiIding dimensions

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

An important consideration is definition of the relevant height and length of the building. A typical case of a building affected by a single tunnel settlement trough is shown in Figure 3.8. The height H is taken as the height from foundation level to the eaves. The roof is usually ignored. It is assumed that a building can be considered separately either side of a point of inflexion, ie points of inflexion of the settlement profile (at foundation level) will be used to partition a building. The length of building is not considered beyond the practical limit of the settlement trough, which for a single tunnel can be taken as 2.5i (where ,s/, = 0.044). In a calculation of building strain, the building span length is required and is defined as the length of building in a hogging or sagging zone (shown as Lh or L, on Figure 3.8) and limited by a point of inflexion or extent of settlement trough.

I-

Figure 3.8

3.7

Building deformation

@

EVALUATION OF RISK OF DAMAGE TO BUILDINGS DUE TO SUBSIDENCE The various concepts discussed above can be combined to develop a rational approach to the assessment of risk of damage to buildings due to tunnelling and excavation. The following broadly describes the approach that was adopted during the planning and enquiry stages of the Jubilee Line Extension underground railway and the proposed CrossRail project.

3.7.1

Level of risk The term “the level of risk”, or simply “the risk”, of damage refers to the possible degree of damage as defined in Table 3.1. Most buildings are considered to be at “low risk” if the predicted degree of damage falls into the first three categories 0 to 2 (ie negligible to slight). At these degrees of damage, structural integrity is not at risk and damage can be repaired readily and economically.

Ch 3 Assessment methods used in design

It will be recalled from Section 3.4.3 that the threshold between categories 2 and 3 damage is a particularly important one. A major objective of design and construction is to maintain the level of risk below this threshold for all buildings. It should be noted that special consideration has to be given to buildings judged to be of particular sensitivity such as those in poor condition, containing sensitive equipment, or of particular historical or architectural significance. Because of the large number of buildings involved, the method of assessing risk is a staged process as follows: preliminary assessment; second-stage assessment; detailed evaluation. These three stages are described briefly as follows.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

3.7.2

Preliminary assessment To avoid a large number of complex and unnecessary calculations, a very simple and conservative approach is adopted for the preliminary assessment. It is based on a consideration of both maximum slope and maximum settlement of the ground surface at the location of each building. According to Rankin (1988), a building experiencing a maximum slope 8 of 1/500 and a settlement of less than 10 mm has negligible risk of any damage. By drawing contours of ground surface settlement along the route of the proposed tunnel and its associated excavations it is possible to eliminate all buildings having negligible risk. This approach is conservative because it uses ground surface, rather than foundation level, displacements. Also, it neglects any interaction between the stiffness of the buildings and the ground. For particularly sensitive buildings it may be necessary to adopt more stringent slope and settlement criteria.

3.7.3

Second-stage assessment The preliminary assessment described above is based on the slope and settlement of the ground surface and provides a conservative initial basis for identifying those buildings along the route requiring further study. The second-stage assessment makes use of the work described in the previous sections of this chapter. In this approach the faqade of a building is represented by a simple beam whose foundations are assumed to follow the displacements of the ground in accordance with the greenfield site assumption mentioned in Section 3.3. The maximum resultant tensile strains are calculated from the pairs of Equations (3.8) + (3.10) and (3.9) + (3.1 I), if necessary partitioning the building as described in Section 3.6.3. The corresponding potential category of damage, or level of risk, is then obtained from Table 3.2. The above approach, though considerably more detailed than the preliminary assessment, is usually still very conservative. Thus, the derived categories of damage refer only to possible degrees of damage. In the majority of cases the likely actual damage will be less than the assessed category. The reason for this is that, in calculating the tensile strains, the building is assumed to have no stiffness so that it conforms to the greenfield site subsidence trough. In practice, however, the inherent stiffness of the building will be such that its foundations will interact with the supporting ground and tend to reduce both the deflection ratio and the horizontal strains. Potts and Addenbrooke (1 996 and 1997) carried out a parametric study of the influence of building stiffness on ground movements induced by tunnelling using finite element methods incorporating a non-linear elastic-plastic soil model.

36

Building response to tunnelling

The building was represented by an equivalent beam having axial and bending stiffness EA and El (where E is the Young’s modulus, A the cross-sectional area and I the moment of inertia of the beam). The relative axial stiffness a* and bending stiffness p* are defined as:

a* = EA/E,H and p* = EI/E,P where H i s the half-width of the beam ( = B/2) and E, is a representative soil stiffness. The eccentricity of the tunnel centre-line is defined as e.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 3.9 shows the influence of relative bending stiffness on the settlement profile for a 20 m deep tunnel excavated beneath a 60 m wide building with zero eccentricity. Figure 3.10 shows modification factors to the deflection ratios N L that would be obtained from the greenfield site settlement profiles for sagging and hogging modes of deformation at different eh3 ratios. It can be seen that, for hogging in particular, the building changes from relatively very flexible to relatively very stiff over quite a small range in relative bending stiffness.

0 0 ‘

2

10

Distance from centreline (m) 20

1

I

30 I

40

I

-1

Figure 3.9

Influence of relative bending stiffness on settlement profile (Potts and Addenbrooke, 1997)

The approach of Potts and Addenbrooke of assessing the influence of global stiffness of a building is a most valuable addition to the existing methodology for assessing the risk of damage. The results can be used to make more realistic assessments of relative deflection and hence average strains within a building. It is only the first step and much has still to be done to include three-dimensional effects of building geometry and in assessing the equivalent flexural and axial stiffness of buildings.

Ch 3 Assessment methods used in design

37

''

0.8

1.2

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 3.10

3.7.4

Modification factors for deflection ratio in sagging and hogging (Potts and Addenbrooke, 1997)

Detailed evaluation Detailed evaluation is carried out on those buildings that, as a result of the second-stage assessment, are classified as being at risk of category 3 damage or greater (see Table 3. I). The approach is a refinement of the second-stage assessment in which the particular features of the building and the tunnelling and/or excavation scheme are considered in detail. Because each case is different and has to be treated on its own merits it is not possible to lay down detailed guidelines and procedures. Factors that are taken more closely into account include the following. Tunnelling and excavation. The sequence and method of tunnel and excavation construction should be given detailed consideration with a view to reducing volume loss and minimising ground movements so far as is practical. Structural continuity. Buildings possessing structural continuity - such as those of steel and concrete frame construction - are less likely to suffer damage than those without structural continuity - such as load-bearing masonry and brick buildings.

Foundations. Buildings on continuous foundations such as strip footings and rafts are less prone to damaging differential movements (both vertical and horizontal) than those on separate individual foundations or where there is a mixture of foundations (eg piles and spread footings). Orientation of the building. Buildings oriented at a significant skew to the axis of a tunnel may be subjected to warping or twisting effects. These may be accentuated if the tunnel axis passes close to the corner of the building. Soil/structure interaction. As pointed out in Section 3.7.3, the predicted greenfield displacements will be modified by the stiffness of the building. The detailed analysis of this problem is exceedingly complex and resort is usually made to simplified procedures such as that of Potts and Addenbrooke (1 996, 1997). The beneficial effects of building stiffness can be considerable, as demonstrated by some recent measurements on the Mansion House, in the City of London, during tunnelling beneath and nearby - see Figure 3.1 1 from a paper by Frischmann et a1 (1994).

38

Building response to tunnelling

Previous movements. The building may have experienced movements due to a variety of causes, such as construction settlement, groundwater lowering, and nearby previous construction activity. It is important that these effects be assessed as they may reduce the tolerance of the building to future movements.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

In that many factors are not amenable to precise calculations, the final assessment of possible degree of damage requires engineering judgement based on informed interpretation of available information and empirical guidelines. Because of the inherently conservative assumptions used in the second-stage assessment, the detailed evaluation will usually result in a reduction in the possible degree of damage. Following the detailed evaluation, consideration is given as to whether protective measures need to be adopted. These will usually only be required for buildings remaining in damage categories 3 or higher (see Table 3.1).

l01

\\ \

3.8

I

‘t’

(x 12

Figure 3.1 1

i II

-o-

Actual

- - - Predicted greenfield

I

Comparison of observed and greenfield site settlements of the Mansion House from driving a 3.05 m-diameter tunnel at 15 m depth (Frischmann et al, 1994)

PROTECTIVE MEASURES The range of possible protective measures is summarised briefly as follows. Tunnelling. Before considering near-surface measures, consideration should be given to measures that can be applied from within the tunnel to reduce the volume loss. There are a variety of such measures such as increasing support at or near the face, reducing the time to provide such support, the use of forepoling or soil nailing in the tunnel face, or the use of a pilot tunnel. These approaches tackle the root cause of the problem and may prove much less costly and disruptive than near-surface measures. If, for a particular building, tunnelling protective measures are considered either not technically effective or not viable in terms of cost, it will be necessary to consider protective measures applied near the surface or to the building itself. However, it must be emphasised that such measures are generally disruptive and can have a significant environmental impact.

Ch 3 Assessment methods used in design

39

The main forms of protective measures available fall into the following six broad groups. 1. Strengthening the ground by means of grout injection (cement or chemical) or by ground freezing. It is normally undertaken in granular water-bearing soils. Its primary purpose is to provide a layer of increased stiffness below foundation level or to prevent loss of ground at the tunnel face during excavation.

2. Strengthening the building so that it may safely sustain the additional stresses or accommodate deformations induced by ground movements. Such measures include the use of tie rods and temporary or permanent propping.

3. Structural jacking to compensate for settlement. 4. Underpinning by introducing an alternative foundation system that eliminates or minimises differential movements caused by tunnelling. 5 . Installation of a physical barrier between the building foundation and the tunnel.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Such a barrier is not structurally connected to the building’s foundation and so does not provide direct load transfer. The intention is to modify the shape of the settlement trough and minimise ground displacements adjacent to and beneath the building. 6. Compensation grouting, which consists of the controlled injection of grout between the tunnel and the building foundations in response to observations of ground and building movements during tunnelling. As its name implies, the purpose is to compensate for ground loss. The technique requires detailed instrumentation to monitor the movements of the ground and the building. The technique was used with success at Waterloo station, London, for the construction of a new 8 m dia tunnel passing a few metres beneath two sensitive masonry structures, the Victory Arch and the Waterloo and City Line tunnels (Harris et a1 (1994)). It was here that the term “compensation grouting” was coined. The later sections of this book describe variations of grouting technique and process that were used in the construction of the Jubilee Line Extension and present case histories of their use to protect specific buildings.

It cannot be emphasised too strongly that all of the above measures are expensive and disruptive and should not be regarded as a substitute for good quality tunnelling practice aimed at minimising settlement.

3.9

GAPS IN KNOWLEDGE This chapter summarises a rational and coherent approach to the assessment of risk of damage to buildings due to tunnelling and excavation. The approach is based on the integration of studies relating to prediction of ground movements, categorisation of damage and the factors controlling cracking of masonry and brickwork. Although the approach has an analytical framework it is evident that much reliance is placed on experience and case records. In developing the approach it became clear that there is a conspicuous shortage of well-documented case histories of measured building response to ground movements. The gaps in knowledge can be described in the following terms.

3.9.1

Ground-st ructure inte raction There are few reliable case records of the influence of building stiffness on the shape and magnitude of ground subsidence or heave. At present, a conservative approach is adopted in which the stiffness of the building is neglected. This leads to over-estimates of building damage, unnecessary concern by building owners and unnecessarily

40

Building response to tunnelling

expensive methods to minimise ground movements. A particularly important gap in knowledge exists on the response of piled foundations to ground movements, both vertical and horizontal.

3.9.2

Damage Present methods of damage assessment rely on the work of Burland (described previously), which was originally directed towards construction settlement and not subsidence or heave. Only very limited progress has been made in addressing the special problems of subsidence in which, unlike settlement, horizontal ground strains play a major role.

3.9.3

Protective measures

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

A number of traditional and novel measures are available for minimising the damaging effects of subsidence. For example, recently the use of compensation grouting has been widely advocated as an effective approach. There is almost no published information on the effectiveness of the various measures and designers have to depend largely on the experience and expertise of contractors.

3.9.4

Remedial measures This is a controversial area where different specialists may advocate a wide range of remedial procedures for any given circumstance. Advice could range from a small amount of patching and repointing to major underpinning. Once again, there are very few well-documented case records that can be referred to - particularly on the medium to long-term effectiveness of remedial measures.

3.9.5

Time There is much uncertainty about the magnitude and distribution of time-dependent movements due to tunnelling and deep excavations. Although this is believed not to be a problem, there is a dearth of reliable long-term (up to ten years, say) measurements of subsidence, particularly above tunnels.

3.9.6

Subsidence trough Gaining as much knowledge as possible of the ground strains that affect a building remains an important research aim. Reasonably reliable methods exist for predicting the vertical surface displacements for single tunnels, but there is considerable uncertainty about the lateral displacements that are of major significance to building response. There are also considerable uncertainties as to whether superposition can be used for multiple tunnels and large openings. Comprehensive studies of the development, shape and amount of subsidence are needed.

3.9.7

The research opportunity In a concerted effort to remedy this lack of case histories to provide data on the above subjects, the opportunity offered by the construction of the Jubilee Line Extension was used to carry out the major co-operative research programme into the behaviour of selected buildings along the route that forms the subject of this book. The research project, which cost more than E1 million and was jointly funded by the UK Government, London Underground Ltd and industry, is described in Chapter 4.

Ch 3 Assessment methods used in design

41

3.10

REFERENCES ATTEWELL, P B, YEATES, J and SELBY, A R (1986). Soil movements induced by tunnelling and their effects on pipelines and structures. Blackie, Glasgow BOSCARDIN, M D and CORDING, E G (1989). Building response to excavationinduced settlement. JGeotech Engg, ASCE, 115 (1); pp 1-21 BUILDING RESEARCH ESTABLISHMENT (198 1, revised 1990). Assessment of damage in low rise buildings with particular reference to progressivefoundation movements. Digest 25 1, BRE, Garston BURLAND, J B and WROTH, C P (1974). Settlement of buildings and associated damage. State of the art review. Conf on Settlement of Structures, Cambridge, Pentech Press, London, pp 61 1-654

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

BURLAND, J.B, BROMS, B B and DE MELLO, V F B (1977). Behaviour of foundations and structures, State of the art report, Session 2, Proc 9th Int ConsS W E , Tokyo, 2, pp 495-546 BURLAND, J B, SIMPSON, B and ST JOHN, H D (1 979). Movements around excavations in London Clay. Invited National Paper. Proc 7th European Conf on S W E , Brighton, 1, pp 13-29 CLOUGH, G W and O’ROURKE, T D (1990). Construction induced movements of insitu walls. Design and Pe~ormanceof Earth Retaining Structures, ASCE Geotechnical Special Publication No 25, pp 439-470 FREEMAN, T J, LITTLEJOHN, G S and DRISCOLL, R M C (1994). Has your house got cracks? The Institution of Civil Engineers, London, and Building Research Establishment, Garston FRISCHMANN, W W, HELLINGS, J E, GITTOES, S and SNOWDEN, C (1994). Protection of the Mansion House against damage caused by ground movements due to the Docklands Light Railway Extension. Proc Instn Civ Engrs GeotechnicalEngineering, 102,2,65-76 HARRIS, D I, MAIR, R J, LOVE, J P, TAYLOR, R N and HENDERSON, T 0 ( 1994). Observations of ground and structure movements for compensation grouting during tunnel construction at Waterloo station. Giotechnique 44 (4), pp 691-7 13. INSTITUTION OF STRUCTURAL ENGINEERS (1978,1989 and 2000). State of the Art Report - Structure-Soil Interaction. Revised and extended in 1989,2nd edn August 2000, The Institution of Structural Engineers, London INSTITUTION OF STRUCTURAL ENGINEERS (1994). Subsidence of low rise buildings. The Institution of Structural Engineers, London JARDINE, R J, SYMES, M J and BURLAND, J B ( 1 984). The measurement of soil stiffness in the triaxial apparatus. Giotechnique, 34, (3), pp 323-340

ENNTNGS, J E and KERRICH, J E (1962). The heaving of buildings and the associated economic consequences, with particular reference to the Orange Free State Goldfields. The Civ Engr in Sth Africa, 5; (5); p 122

42

Building response to tunnelling

KARLSRUD, K and MYRVOLL, F (1 976). Performance of a strutted excavation in quick clay. Proc 6th European Conf SMFE, Vienna, 1; pp 157-1 64. MACLEOD, I A and LITTLEJOHN, G S (1974). Discussion on Session 5. Conf on Settlement of Structures, Cambridge, Pentech Press, London, pp 792-795. MAIR, R J, TAYLOR, R N and BRACEGIRDLE, A. (1993). Subsurface settlement profiles above tunnels in clay. Geotechnique 43; (2); pp 3 15-320

NATIONAL COAL BOARD (1 975). Subsidence Engineers Handbook. National Coal Board Production Dept, UK. O’REILLY, M P and NEW, B M (1982). Settlements above tunnels in the United Kingdom - their magnitude and prediction. Tunnelling ‘82, London, pp 173-1 8 1

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

PECK, R B (1969). Deep excavations and tunnelling in soft ground, State of the art report, Mexico City, State of the Art Volume, Proc 7th Int Conf SMFE, pp 225-290 POLSHIN, D E and TOKAR, R A (1957). Maximum allowable non-uniform settlement of structures. Proc 4th Int Conf SMFE, London, 1; p 402 POTTS, D M and ADDENBROOKE, T I (1996). The influence of an existing surface structure on the ground movements due to tunnelling, GeotechnicalAspects of Underground Construction in Soft Ground (Mair, R J and Taylor, R N eds) Proc Conference at City University, Balkema, Rotterdam, pp 573-578 POTTS, D M and ADDENBROOKE, T I (1997) A structure’s influence on tunnellinginduced ground movements, Proc Instn Civ Engrs GeotechnicalEngineering, 125, pp 109-125

RANKIN, W J (1988). Ground movements resulting from urban tunnelling; predictions and effects. Engineering Geology of UndergroundMovement, Geological Society, Engineering Geology Special Publication No 5, pp 79-92. TIMOSHENKO, S (1957). Strength of materials - Part I , D van Nostrand, London.

Ch 3 Assessment methods used in design

43

4

The LINK CMR research project

F M Jardine 4.1

SUMMARY

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

This chapter is an overview of the research projects on which this book is based. It explains how LINK CMR research project was set up and its relation to the associated research at greenfield sites in St James’s and Southwark Parks. The collaboration between sponsors, industry participants and the research team was an important feature of the project, and the way in which this was effected is described. The research project also involved gathering construction information and monitoring data from the construction contracts and its compilation into the GEOSIS database.

4.2

THE RESEARCH OBJECTIVE

4.2.1

Background A nation’s buildings represent its life and work, its past, present and, in great degree, its future. Almost without exception the principal asset of individuals and companies, their buildings are insured not only against loss but also against damage. One of the major risks for a building, regardless of whether it is a house, a factory, an office block or an ancient monument, is that of subsidence (or heave), ie ground movement below and around the building so as to strain and damage it. As is explained in the previous chapter, the response of buildings to ground subsidence is not well understood. The effect of subsidence is that the building is forced to adjust to the ground movements. This differs from the initial process of settlement, where the structure, its foundation and the ground move to compensate for the loading applied by the building. There are many causes of subsidence and heave, including the shrinkage or swelling of clay soils, changes in groundwater conditions, vibration, collapse settlement, and when ground nearby or below is excavated for deep basements or tunnels. The planning stage of the Jubilee Line Extension Project (JLEP), particularly during the Parliamentary hearings, not only drew attention to the need for greater understanding of the response of buildings to excavation-induced ground movements, but it also showed what an excellent opportunity the JLEP would provide for gaining information. There were several reasons why it was so important to take this opportunity. 1

A wide range of building types would be affected; with different combinations of structure and foundation type.

2

There would be considerable knowledge about each of the buildings, their form and condition.

3

There would be different tunnelling methods and protective measures.

4

There are distinctly different geologies along the tunnelled part of the route.

5

There was one construction project, one promoter, a unified construction team, and there was a central will to obtain and maximise the research effort and the benefits of its results.

Ch 4 The Link CMR research project

Previous page is blank

45

Without this last driving force it would have taken decades to obtain the amount and quality of data that has been gained -and retained - in relatively few months.

4.2.2

The need for the research

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Ground subsidence, its interaction with buildings and the associated repair and maintenance, is a major and growing problem. The value of insurance claims due to shrinking and swelling clays alone amounts to E400 million or more per annum in UK. The amount of excavation for basements in urban areas is increasing. The construction costs of protecting buildings from excavation-induced ground movements are very large and frequently dictate the excavation and support method. Also there is a significant increase in planned tunnelling for transportation in urban areas in south-east England. The prediction of ground movements caused by clay volume changes and excavation has improved in recent years. The associated interactions with buildings, however, and the resulting levels of damage are not well understood. That lack of knowledge has proved costly because professionals tend to adopt a very conservative approach both in the repair of subsidence damage and in the design of protective measures. There is a need to evaluate the effectiveness of different protective measures. The gaps in knowledge about the specific problems of excavation-induced settlement are discussed above in Section 3.9.

4.1.2

Commercial importance The commercial relevance of the research can be considered under the following headings: 0

benefits for insurers, owners, mortgagers, developers of buildings and building professionals reduced costs on capital projects increased competitiveness of UK main works and specialist contractors increased competitiveness of UK consulting structural and tunnelling engineers.

1. Buildings

In respect of aspects of building insurance and the quantification of risk, the results of the research increase knowledge of building response and progression of damage. For building owners and funders (eg mortgage lenders), the case study results will lead to more realistic assessment of the impact of proposed construction works, which could have a significant effect on reducing blight. For building developers, better understanding could reduce the costs of conservative design and expensive preventative measures for property adjacent to deep basement construction.

2. Capital projects On major projects such as a new underground railway, the promoter and owners of potentially affected buildings incur considerable expenses. Greater understanding will lead to more clearly focused efforts. It is understood that a substantial proportion, perhaps up to 20 per cent (some E3 million), of the costs of the Parliamentary process of the Bill for the JLEP was related to the investigation and assessment of the effects of

46

Building response to tunnelling

ground movement on buildings near the route. As well as the promoter’s costs, there are cumulatively very large costs expended by petitioners. For the whole E1 .8 billion project, the construction element, ie the civil engineering works, was initially of the order of E650 million, of which about a sixth would be spent on protective measures such as underpinning and compensation grouting. Improved understanding of the effectiveness of protective measures will lead to their more efficient use and the potential for considerable savings on hture major transit schemes. For the far less expensive but numerous smaller-scale construction projects, whether a deep basement, shallow sewer tunnel or services trench, the costs of untoward damage and over-conservative protective measures can be very wasteful on a national scale. 3. Increased competitiveness

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

UK consulting engineers and main works contractors have to compete internationally and, to an increasing extent, in the home market. Involvement with the research and early access to its results could give them a technical lead. The same could apply to specialist geotechnical contractors, although the benefits are rather different. First, new construction techniques (particularly geotechnical ones) need demonstrations of their effectiveness to gain wide acceptability and overcome engineers’ conservative bias. Rigorous observational research can provide this validation. Second, such research can provide the information that will lead to improvements.

4.2.3

The overall objective In view of the above justification for the research, its overall objective was expressed as follows: By means of field measurement, to improve understanding of (1) building responses to known ground movements ( 2 ) damage that results from such responses (3) measures for protection and repair.

4.2.4

Detailed objectives The overall research objective was focused on the specific research needs described above in Sections 3.9.1 and 3.9.2. These are presented below under the same six headings. Those paragraphs are from the original research proposal, so the aims can be compared with what the research achieved. Several of these refer to gathering data. What started out as an intention to bring the field measurements made by the research team together in a database was developed by JLEP staff into a much more comprehensive system of recording contract monitoring data, construction records, and the operational parameters of grouting and tunnelling. This became the GEOSIS database, which is described in Chapter 19. It contains information relevant to all the detailed objectives below, which is available therefore for later studies.

I . Ground-structure interaction In all, monitoring and other information was obtained for more than 30 structures and is recorded in the GEOSIS database that JLEP staff developed (see Chapter 19). Table 4.1 lists those structures that are included as case studies in this book. Most were monitored by the research team, but some of the case studies use the results of the monitoring by JLEP contractors and specialists. For six of the monitored structures, predictions of their responses were made in terms of vertical displacements, using the results of recent research from Imperial College about the influence of relative building stiffness.

Ch 4 The Link CMR research project

2. Damage There was little reported damage to buildings from the JLEP works, and many of the buildings that would otherwise have been most susceptible were protected, but the research has obtained considerable information on the strains that the buildings experience, including horizontal strains at different heights.

3. Protective measures A considerable amount of data has been gathered in relation to the use of compensation grouting, which was widely used throughout the JLEP contracts. The research has provided the data for several substantial, and likely to be influential, case studies. 4. Remedial measures

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The research could do little on this aspect, other than some continuing observations of cracks in a few structures, mainly because this has not been significant in the JLEP work for the buildings monitored under the research project. Separate arrangements have been made, however, to obtain and record information for additional case studies. 5. Time A substantial set of research data exists over periods of more than six years on some

buildings. The measurements from the greenfield site in St James’s Park cover 60 months after passage of the first tunnel. There are comparable observations on buildings protected by compensation grouting. It is believed that not only are these data unique in comprehensiveness and quality, but also in the way they demonstrate longer-term trends. Separate arrangements have been made to continue to obtain and record longterm measurements in the project database. 6. Subsidence trough Information has been obtained from instrumented sections at greenfield and other sites about the ground displacements of the surface and in the subsurface. Other data relate to the effects of different alignments of the two running tunnels and the timings of their excavation. Comparison has been made of the observed and predicted building responses for twin tunnels and large openings with staged excavation sequences.

4.3

METHODOLOGIES AND INVESTIGATIONS The following sections outline how the research was done and list the specific investigations that are reported in this book. (Note that other structures were monitored under the research project but are to be reported separately.) The aim is to show examples of the type of information obtained and by so doing to demonstrate the methodology and scope of the research.

4.3.1

Selected structures The monitored structures described in the case studies in Volume 2 are listed in Table 4.1, which also summarises their form, foundation and the types of measurements taken on them. Only some of the data obtained over the four years’ currency of the research grant and since can be presented in this book. All the data is collected in the GEOSIS database (see Section 4.3.6 and Chapter 19). Protective measures were applied to rather more buildings than had been anticipated. In some cases, such as the Treasury building, the protective measure of compensation grouting was applied after the first tunnel drive and during excavation of the second

48

Building response to tunnelling

tunnel. Examples of buildings without protective measures given in this book are Elizabeth House near Waterloo station and the three buildings in Moodkee Street, Bermondsey. For these four buildings, predictions of the building responses to tunnelling were made by the Geotechnical Consulting Group.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Table 4.1

Case study structures

Building

Structure type

Foundation type

Protective measures

Monitoring

The Ritz

Steel and masonry frame

Pads

Compensation grouting

FM

RAC Building

Steel and masonry frame

Pads

Compensation grouting

PL

The Treasury

Brick and stone masonry

Raft

Compensation grouting for EB tunnel

PL, TE, FM

Great George Street, south side buildings

Brick and stone masonry

Mixture of shallow strip and piles

Compensation grouting for EB tunnel

FM

Houses of Parliament

Brick and stone masonry

Raft

Compensation grouting

PL, TE, D, El

Elizabeth House

Reinforced concrete

Raft

1-7 St Thomas Street

Reinforced concrete frame, masonry cladding

Raft

Compensation grouting

PL

London Bridge Post Office

Brick and stone masonry

Stepped mortar raft

Compensation grouting

Fielden House

Reinforced concrete

Piled

Compensation grouting

PL, PL, FM

BT Building

Reinforced concrete

Piled

Compensation grouting

PL, FM

Telephone House

Brick masonry

Shallow strip

Compensation grouting

PL, FM

Keeton’s Estate

Brick masonry

Deep strip

Permeation grouting

PL, TE, D

128-130 Jamaica Road

Brick masonry

Shallow strip

182-2 10 Jamaica Road

Brick masonry

Shallow strip

Blick House

Brick masonry with reinforced concrete slabs

3-m deep foundation

PL, TE

Murdoch, Clegg and Neptune Houses

Brick masonry

Shallow strip

PL, TE, FM, D

One free-standing wall, Moodkee Street

Brick masonry

Shallow strip

PL, TE, D

Niagara Court

Brick masonry with reinforced concrete slabs

Short bored piles

Two free-standing walls, Niagara Court

Brick masonry

Shallow strip

Regina and Columbia Points

Reinforced concrete

Reinforced concrete raft

PL, El

Tenants’ hall and boiler house, Canada Estate

Brick masonry

Shallow strip

PL, TE, D

PL, TE, FM, D, RE, BHEL

PL Tie rods and angle brackets

Permeation grouting

PL, TE, D

PL, TE, D

:J&I

BHEl electrolevel inclinometer in borehole; D Demec; El electrolevel; FM faqade monitoring; PL precise levelling; RE rod extensometer; TE tape extensometer

4.3.2

Monitoring and precision The techniques of measurement were straightforward and essentially conventional, except as noted below. The emphasis was on accuracy. The precision typically obtained in levelling by the research team, for example, was of the order o f f 0.3 mm. A key reason for this was proper use of the modified BRE levelling sockets, which give consistent, repeatable positioning of the staff (BRE, 1993)

Ch 4 The Link CMR research project

49

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The measurement methods used by the research team for observations of building response are listed below. See Chapter 18 for a description of the methods and their precision. 1

Precise levelling. The equipment was a LeicaTMNA 3003 precise level with barcode invar staff.

2

Facade monitoring (total station surveying). The equipment was a LeicaTMTC2002 total station with integrated distance measurement. The total station results were analysed by a computer program specifically written for this research by S K Sharma of Imperial College. The initial intention in the research had been to make more use of photogrammetry, and some surveys were commissioned externally. Its use was discontinued, partly because it was considered that total station surveying gave better base station control, and partly because of cost. The in-house research team could respond more quickly to the construction programme and carry out more total station surveys within the allowance for photogrammetry in the research budget.

3

Precise taping. A digital-display Ealey tape extensometer was used to measure between demountable eye-bolts screwed into BRE sockets. Temperatures were recorded to 0.1"C by NAMAS-calibrated digital thermometer.

4

Demec gauge measurements. This standard gauge was used to measure between studs glued to wall faces ether side of cracks

Chapter 18 includes descriptions of the subsurface instrumentation used at the greenfield sites and at Elizabeth House. Note, however, that in the case studies reported in this bookthere are relatively few instances where the results from electrolevels are used. This is mainly because the effects of temperature dominated the electrolevel readings.

4.3.3

Predictions Part of the research was to use best current practice to predict the response of selected buildings to the tunnel excavation. It was intended that the predictions should be made before the event, ie Lambe's Class A (see Section 14.1 for classification of predictions). This was not possible in every case, but all the predictions were made without knowledge of the actual responses. It was important that the predictions were based on correct construction details, eg tunnelling methods, tunnel alignments, excavation sequences - matters that would not have been known if the predictions had been made too early or that had changed from the original scheme. As part of its contribution to the research, the Geotechnical Consulting Group (GCG) made the following predictions. 1

The vertical displacements at ground level of Elizabeth House (near Waterloo Station) at three stages of the JLEP tunnelling beneath the building in London Clay.

2

The vertical displacements of Murdoch, Clegg and Neptune Houses (Moodkee Street, Bermondsey) that were affected by the two running tunnel drives in the Lambeth Group strata.

3

Two masonry walls; one about 1 m high, and the other about 1.8 m high, close to the Moodkee Street buildings.

4

The surface and subsurface ground displacements at the instrumented section in Southwark Park.

Imperial College similarly made predictions of the surface and subsurface ground displacements for the instrumented section at St James's Park.

50

Building response to tunnelling

Chapters 12 to 15 are transcriptions of the reports giving the prediction and explain the methods used for making them.

4.3.4

Protective measures Several methods of protection were used on the JLEP works and these are described in Chapter1 1. Many of the case study buildings were protected from tunnelling-induced ground movements by compensation grouting (see Table 4. I). The specialists carrying out this work use different techniques and the ground situations vary along the JLEP route. A huge amount of data generated from the monitoring in real time of the grouting process and the ground and building responses has been entered into the GEOSIS database. The substantial case study of the Treasury building (Chapter 26) shows not only how effective the technique was, it also gives an indication of what is involved in reducing the data.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

4.3.5

Observations of damage and repairs As noted above, there have been relatively few instances of damage, but that which affected the study buildings was recorded, usually with crack measurements taken at appropriate times. Generally, there seems to have been very little damage requiring immediate repair.

4.3.6

Data collection and assembly The design, construction, testing and setting up of a database was one of the major tasks for the research team. This was made possible by LUL through the JLEP’s IT specialists. The GEOSIS database is massive and still being enlarged. It represents a uniquely comprehensive record of construction and contract monitoring data, research observations and records, and other relevant information. It was used by the research team in preparing the case studies in this book. The database is administered by Infraco JNP on behalf of LUL. An associated IT development of use to industry is the protocol for the transfer of electronic information. Known as GEMINI, the protocol is being tested and refined in conjunction with the Association of Geotechnical and Geoenvironmental Specialists on whose data exchange format it is based as part of DETR-sponsored Partners in Innovation project managed by CIRIA.

4.4

THE RESEARCH SPONSORSHIP

4.4.1

LINK CMR programme The Construction Maintenance and Refurbishment (CMR) research programme was one of the first LINK schemes to be sponsored by the then Department of the Environment. Its address was towards buildings and particularly the existing building stock, as the costs of maintenance and repair have been steadily increasing over recent years.

In view of the need improve understanding of the causes of damage to buildings, and with the anticipated opportunity to begin a database on repair methods, Professor Burland considered that the JLEP-based research he had envisaged would be a suitable candidate project for sponsorship under the LINK-CMR programme. The EPSRC coordinator of the programme, Mr Peter Pullar-Strecker, was especially helpful during the proposal development. His advice later during the research was always welcome.

Ch 4 The Link CMR research project

51

4.4.2

Proposals and grant awards The collaborative research Proposal 174 was submitted to the EPSRC-DETR (formerly DOE) LINK Construction Maintenance and Refurbishment Programme in March 1994. The grant was awarded by letter of 28 July 1994 and work began in the following August. Construction programme changes affected the research, increasing the duration of active observations, and EPSRC allowed the grant-holders an extension of time of one year to 2 1 August 1998.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The research project was widely promoted, even to featuring in a cartoon in New Civil Engineer.

Figure 4.1

Cartoon by Dick Millington in New Civil Engineer, May 1994

The other main research application was made to EPSRC in 1994. This collaborative study was to install and monitor surface and in-ground instrumentation at the two greenfield (control) sites of St James’s and Southwark Parks. Its title was Field measurement of ground behaviour resulting from single and twin tunnel excavations. LUL sponsored the research by funding the installation and provision of instrumentation at the Southwark Park site.

4.5

THE RESEARCH COLLABORATION To be eligible for funding under the conditions of the LINK programme, a project has to be a collaboration between industry and a science-base partner. One partner has to take the role of lead partner. In this research the initial partners were London Underground Limited, the Geotechnical Consulting Group and ClRlA on the industry side and Imperial College for the science. As the intention was to bring in more industrial partners in the project, CIRIA took on the project management as lead partner. The organisations that combined to sponsor the research are listed in Table 4.2. CIRIA managed the research collaboration and programme (Burland et al, 1996) through the following mechanisms: a top-level board of management, representing the initial sponsors and participants; a programme management committee, representing all

52

Building response to tunnelling

the participants and researchers; and a progress group, comprising the research supervisors, the research engineers and assistants and technical advisers from industry. The progress group held more than 50 minuted meetings over the course of the research, controlling technical inputs, quality, analysis and management, including the management of health and safety, expenditure and resources. Members of the board and the management committee are listed in the Acknowledgements.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Table 4.2

4.5.1

Research collaboration

Organisation

Role

Imperial College

Science-base partner

CIRlA

Lead partner

London Underground Limited (JLEP)

Industry partner

Geotechnical Consulting Group

Industry partner

AMEC Piling Mott MacDonald Ove Arup and Partners Trafalgar House Technology TRL Union Railways

Industry partners

Industrial support Much of the research was made possible by the collaboration with personnel from industry. CTRIA’s research manager for ground engineering, Mr F M Jardine, was the project manager for the LINK CMR research project. Mr L F Linney, senior geotechnical engineer of the JLEP organisation, provided the central liaison from before the start of the project to the end of 1996, when he left JLEP, by which time LUL provided the major proportion of the industrial contribution. Professor R J Mair, of the Geotechnical Consulting Group, was LUL’s geotechnical adviser for the Jubilee Line Extension Project. His advice to the research team throughout the research was an important technical contribution, especially with the Class A predictions on the study buildings. He and the Geotechnical Consulting Group have made a substantial and continuing contribution throughout the research. The industry-side contributions took several forms: financial support to the research team, in particular for the purchase of high-quality measurement systems provision of two members of the research team installation of instruments time of senior professionals in technical steering and individual advice to the research team project management and administrative services dissemination of interim and final results liaison services between the research and the construction works provision of supporting construction data.

Ch 4 The Link CMR research project

53

4.5.2

The research project team Professor Burland led the research. Input was also given to the work by Dr J B Newman and Mr S K Sharma of Imperial College, who were co-applicants in the research proposal, in relation to the structural monitoring and to the surveying, respectively. The research team essentially comprised three people, ie the team leader and one researcher, who were employed by LUL, and one RAlB funded by EPSRC over two years. In addition to the Imperial College technicians used part time, other staff were taken on specifically as technical/survey assistants. An engineer from CrossRail worked for three days a week with the research team.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The list of researchers and those who have joined the team for different tasks and at different stages are listed in the Acknowledgements. The research therefore not only contributed to the training of many young people, it also both reached outwards to researchers from other countries and benefited from their work on the project. Much of their work was at relatively little cost to the project or to the project funders. One of the benefits of this kind of large research project is that its interest and momentum attracts, at no extra cost, undergraduates seeking vacation work and visiting academics. This, indeed, is what has happened with this project.

4.5.3

Interim reporting and seminars A series of interim (factual) reports was prepared during the course of the research to make information available to the research sponsors. These were at first restricted to the sponsors, but have since been available on request to CIRIA’s Core Programme members. The interim reports presented information about the buildings being studied, described the monitoring carried out, summarised the construction events relevant to the building, and presented the results of the monitoring. They form the basis of the case study chapters in this book. Additional outputs from the research were procedural notes to record the methods used with Demec gauges, and for precise surveying with level, tape extensometer and total station. These were issued to industrial partners as statements of good practice for measuring building movements. In addition to these formal reports, some results have been reported in conference proceedings (some of which are listed in the references). A considerable effort has been put into technology transfer. The experts employed by the major client, LUL, by some of the leading consulting engineers, by major contractors and specialist contractors were closely involved with the research and saw its outputs. They were in a position to implement the findings directly. Several papers have been published at conferences and in the technical press. Six seminars for funders were organised by CIRIA and held at Imperial College. The first was on 5 December 1994. Some 30 people attended. At each of the four project management committee meetings the research team presented their findings in the form of a seminar to participants in the research. The sixth seminar was held at Imperial College on 1 1 December 1996. Some 50 people from 14 organisations attended. Presentations by the researchers included a demonstration of the GEOSIS database.

54

Building response to tunnelling

4.6

ASSOCIATED RESEARCH Throughout the research project, there have been close links to other research (see Figure 4. I). Thus the Class A predictions used the work by Potts and Addenbrooke (before its publication in 1996 and 1997) on the influence of building stiffness and showed it to be a powerful method of estimating building responses. Their work was also sponsored by LUL. The results from the field observations have informed the continuation of their work. Another beneficial research link is the use of the temperature-independent electrolevel system developed at Imperial College (Barakat, 1996). At the London Clay control site in St James’s Park (Nyren, 1998), very thorough instrumentation extends to depths very close to the tunnels; this has provided extremely detailed control data for this greenfield site. Here, TRL and JLEP instrumented the tunnel lining at a matching section (Davies and Bowers, 1996). Further control data for another greenfield site have been obtained at Southwark Park in the Lambeth Group deposits (Standing et al, 1996).

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

A study was made of the effects on the linings of existing LUL tunnels near Waterloo and which is reported in Chapter 29. The volume losses experienced during the driving of both the east- and west-bound running tunnels in the Westminster and St James’s Park area were much higher than had been expected. The volume losses for both tunnels reduced to about the expected values as they were driven farther west (and north of the lake) in the park. Imperial College undertook an additional research investigation (Standing and Burland, 2000) to compare the soil profiles of the areas of the high and of the expected volume losses. These findings are referred to in Chapters 5 and 25, and are of significance in reaching an understanding of the geology of the London Clay Formation.

4.7

REFERENCES BARAKAT, M A (1 996). Measurements of ground settlement and building deformations due to tunnelling. PhD thesis, Imperial College, University of London BUILDING RESEARCH ESTABLISHMENT (1993). Monitoring building and ground movement byprecise levelling. BRE Digest 386, 8 pp BURLAND, J B, MAIR, R J, LINNEY, L F, JARDINE, F M and STANDING, J R (1 996). A collaborative research programme on subsidence damage to buildings: prediction, protection and repair. In: Geotechnical aspects of underground construction in soft ground (Mair, R J and Taylor, R N eds), Balkema, Rotterdam DAVIES, H R and BOWERS, K H (1996). Design and installation of special tunnel rings to monitor long term ground loadings. Proc. Int. Symp. on Geotech. Aspects of Underground Construction in Soft Ground, City University, London, pp 257-262. MAIR, R J (1996). Settlement effects of bored tunnels, Session report. In: Geotechnical aspects of underground construction in soft ground (Mair, R J and Taylor, R N eds), Balkema, Rotterdam NYREN, R J (1998). Field measurements above twin tunnels in London Clay, PhD thesis, Imperial College, University of London

Ch 4 The Link CMR research project

POTTS, D M and ADDENBROOKE, T I (1996). The influence of an existing surface structure on the ground movements due to tunnelling, GeotechnicalAspects of Underground Construction in Soft Ground (Mair, R J and Taylor, R N eds) Proc. Conference at City University, Balkema, Rotterdam, pp 573-578 POTTS, D M and ADDENBROOKE, T I (1997). A structure’s influence on tunnellinginduced ground movements, Proc Instn Civ Engrs GeotechnicalEngineering, 125, pp 109-125 STANDING, J R, NYREN, R J, LONGWORTH, T I and BURLAND, J B (1996). The measurement of ground movements due to tunnelling at two control sites along the Jubilee Line Extension. In: Geotechnical aspects of underground construction in soft ground (Mair, R J and Taylor, R N eds), Balkema, Rotterdam

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

STANDING, J R and BURLAND, J B (1999). Ground Characterisation to explain JLEP tunnelling volume losses in the Westminster area. Unpublished Imperial College report to London Underground Limited

Building response to tunnelling

5

Geology and geotechnical properties

A D Withers, D P Page and L F Linney 5.1

SUMMARY

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The parts of the Jubilee Line Extension route in which the case s dy buildings are situated cover two contrasting geological settings. From Green Park to east of London Bridge, the tunnels are in the London Clay Formation. Eastwards from there to Canada Water station, the tunnels are in the beds of the Lambeth Group. Made ground, recent alluvium and Terrace Gravels occur over the whole route. This chapter is an overview of the ground and groundwater conditions along the route and summarises the typical geotechnical properties of the deposits.

5.2

INTRODUCTlON

Figure 5.1

5.2.1

Solid geology of London and the Jubilee Line Extension route

The influence of geology on London Underground’s tunnel network Much of London Underground’s tunnel network is north of the River Thames and therefore most of its tunnels are in the London Clay Formation. Indeed, it is partly because of this favourable tunnelling medium that the deep network could be developed. Construction of the Jubilee Line Extension (JLE) meant tunnelling beyond the London Clay area into south-east London, where the solid geology includes the Lambeth Group and Thanet Sand Formation. (The solid geology of much of London and the JLE route is shown in plan on Figure 5.1 .) The Lambeth Group (formerly known as the Woolwich and Reading Beds) comprises a complex sequence of deposits that includes clays, waterbearing sands, gravels and silts. The term Lambeth Group has been in the public domain

Ch 5 Geology and geotechnical properties

57

only since 1994; it was introduced to clarify the stratigraphy shown on British Geological Survey maps, initially in the London area. The Thanet Sand Formation is a relatively permeable but very dense fine sand. In the past, these water-bearing strata presented considerable difficulties for tunnelling, involving hazardous and timeconsuming techniques. Modem tunnelling systems can cope much better with these ground conditions, and their greatly improved performance and reduced costs allowed the tunnels to be constructed within a reasonable programme and budget.

5.2.2

Ground investigation for the Jubilee Line Extension The ground investigation for the JLE was carried out during 1990 and 1991, its primary aims being to define: the geological boundaries between strata at a representative number of locations to enable preparation of geological sections

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

the material and engineering properties of the individual strata throughout the route and provide parameters for use in design 0

the groundwater conditions throughout the route

0

geological hazards.

The ground investigation employed some of the most advanced techniques available, including thin-wall sampling, wire-line core drilling, downhole geophysics, and selfboring pressuremeter tests in the field and the measurement of small-strain stiffness in the laboratory. Approximately 150 boreholes were sunk using cable percussion techniques, about 40 of which were extended using rotary coring techniques. Further details of the investigation works are reported by Linney and Page (1 996). An important element of the investigation as a whole was to obtain from the results of the fieldwork a greater understanding of the stratigraphy of the Lambeth Group. The British Geological Survey was commissioned to carry out detailed logging and analysis. The new scheme for classification of the Lambeth Group that was devised as a result has (with some minor revision) become generally adopted. Overall, the ground investigation was a great success, the ground conditions anticipated from the findings of the investigation being confirmed during construction with only minor variation.

5.3

GEOLOGY OF THE JLE ROUTE The general geological succession along the Jubilee Line Extension is shown in Table 5.1 together with the nature of the boundaries between the different formations. Table 5.2 amplifies the Lambeth Group stratigraphy, giving the names for the beds developed by the British Geological Survey over the last few years and which are used throughout this report. This table also shows the ranges of thickness of these beds that were encountered over the length of the JLE. The JLE route passes through gently folded Palaeocene and Eocene deposits that are overlain by a blanket of Quaternary river terrace gravel and sand and topped with recent alluvium and made ground. Figure 5.2 presents a geological section along the route (see Figure 5.1 for the line of route). Between Green Park and London Bridge the tunnels were constructed within the London Clay Formation. Beyond London Bridge, heading eastwards, they enter the Lambeth Group on the western limb of the Millwall anticline. The tunnels stay within

58

Building response to tunnelling

the soils of the Lambeth Group until Canada Water station (which is the boundary of the area containing the study buildings). East of Canada Water (see Figure 5.2), the tunnels are in the Thanet Sand Formation until rising towards Canary Wharf station they re-enter the Lambeth Group. To the east of Canary Wharf the tunnels go back into the Thanet Sand Formation until the North Greenwich peninsula where they again enter the Lambeth Group and then, via the Greenwich syncline (the western limb of which may be faulted), they re-enter the London Clay Formation. The final section of the tunnelled route is within the London Clay Formation to Canning Town. From Canning Town the route continues northwards at surface level to Stratford via West Ham. Table 5.1

Geological succession along the Jubilee Line Extension

I Group

Time

System

1 Formation

Holocene

No group

Alluvium

Pleistocene

No group

River Thames Terrace Gravel

Eocene

Thames

Quaternary

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

,,-A-

London Clay Formation v

w

.

.*

.

Harwich Formation

,-,-

,-,

Reading Formation

Palaeogene Palaeocene

-

' A/--.

Lambeth

70, 4ZA'

7/2

T--x .' Woolwich

' /

Formation /z?/'05

.wfi

Gk

Upnor Formation

/-

Cretaceous

Green Park

Westminster

I

River

Southwark

WatjrlooI

London Bridge

Bermondsev m PD

I

110 -

- 100

100 -

90 -

- 90

80 -

- 80

70 -

- 70

60 -

- 60

50 -

- 50 Key

Figure 5.2

Made ground and alluvium

London Clay

Thanet Beds

Terrace Gravels

Lambeth Group

Upper chalk

Geological section along route of the Jubilee Line Extension

Ch 5 Geology and geotechnical properties

59

5.4

INFLUENCE OF GROUND CONDITIONS ON CONSTRUCTION TECHNIQUES Along the part of the route containing the study buildings, the tunnels and associated structures were constructed in and through London Clay, the Lambeth Group strata, and the superficial deposits. Table 5.2

Formation (Ellison et af, 1994)

Previous usage (London Basin)

Units (Ellison, 1991: work by King, 1981: and Page and Skipper 2000)

Reading Formation

Reading Beds

Upper Mottled Clay (UMC)

0-7.5 m

Lower Mottled Clay (LMC) (including “ferruginous sand”* Lower Mottled sand)

0-2.5 m

Upper Shelly Clay (USC) (“striped loams”*)

0-2.0 m

Laminated beds

0-9.0 m

Lower Shelly Clay (LSC)

0-2.5 m

“Glauconitic Sand (GS)” “Pebble Bed” (at top)

3.5-6.0 m 0-2.5 m

Woolwich Formation

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Lambeth Group and encountered thicknesses

Upnor Formation

Woolwich Beds

Bottom Bed

Basement Bed (of Reading Beds)

Thickness range encountered along JLE route

* units not defined or otherwise referred to in this book The ground, and particularly the groundwater conditions, dictated the choice of tunnelling method. In general terms it was possible to excavate within the London Clay using open-face tunnelling methods, such as shields with backhoes, road-headers, NATM and hand mining. Excavations within the Lambeth Group were in places more difficult. The main running tunnel drives between London Bridge and Canada Water were driven using earth-pressure-balance machines (EPBMs). Other excavations required compressed air support, grouting and/or dewatering measures to stabilise the ground around the excavations, eg the enlargement of the running tunnels, at Bermondsey, to form platform tunnels using hand-mining and compressed air. In order to gain access to the tunnels from the surface level, station and shaft structures had to be constructed through the superficial materials (made ground, alluvium and Thames Gravel). The various methods used included excavation within propped diaphragm, and bored and sheetpile walls. Chapter 10 gives a fuller, but general, description of the tunnelling and construction methods. Specific features of the construction methods near the case study buildings are given in Chapters 16 (for buildings near London Bridge station) and 17 (for buildings on the Lambeth Group strata).

60

Building r e s p o n s e t o tunnelling

5.5

ENGINERING DESCRIPTIONS AND GEOTECHNICAL PROPERTIES The following sections (5.5.1 to 5.5.8) give typical descriptions of the strata encountered with a summary of their index and engineering properties and the parameters adopted for design. The description of the Lambeth Group (Section 5.5.6), however, has been amplified by information recently published by Page and Skipper (2000) and which is in CIRIA Funders Report CPl83 Engineering in the Lambeth Group (Hight et al, 200 1). Each of the separate building case history chapters has a description of the ground conditions local to the site. Throughout this book, the nomenclature of Tables 5.1 and 5.2 is used when reference is made to strata of relevance to the case studies and that of Table 5.3 when referring to their geotechnical properties. Table 5.3

Geotechnical parameter nomenclature Symbol

Unit

Bulk unit weight

Y

kN/m’

Undrained shear strength

S”

W/m2

Effective cohesion intercept

C’

W/m2

Effective angle of shearing resistance

@’

degrees

Undrained Young’s modulus

E”

m/m2

Drained Young’s modulus

E’

m/m2

Coefficient of permeability

k

m/s

Coefficient of earth pressure at rest

KO

-

I

I

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Parameter

The tables that follow are usually in two parts; the left-hand side lists typical ranges of index properties of a particular unit determined from the site investigation, and the righthand side lists suggested values of design parameters. These are values chosen by the JLEP staff and suggested to the technical contractors as being appropriate, but they were not necessarily the values used in the designs. The made ground materials were so variable and difficult to test that it is not considered worthwhile to give ranges of properties, but the suggested design parameters are listed.

Ch 5 Geology and geotechnical properties

61

5.5.1

Made ground Occurrence Made ground of various ages is present to varying thickness at surface level along almost the entire length of the JLE route, reflecting the long history of urbanisation and industrial development. Some is the result of demolition following bomb damage. The made ground is particularly thick near London Bridge station, which is one of the earliest developed parts of the south bank of the Thames. Typical thicknesses here are 3-5 m, but elsewhere they are of the order of 1-2 m. Description and properties

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The made ground is of so variable a nature that a general description would be inappropriate. The more recent material commonly consists of a mix of concrete, rubble, brick, gravel and sands, often in a clay matrix. Its heterogeneous composition and density range mean that its engineering properties are highly variable. Table 5.4 summarises the design parameters adopted by the project for the materials encountered. The made ground can contain archaeological artefacts and in places may be contaminated from former land use. Both these factors impacted on the JLEP. Table 5.4

Suggested design parameters: made ground

Design parameters (units)

Suggested values

Bulk unit weight, y(kN/m3)

15 to 19

Undrained shear strength, S, (kN/m2)

I5 to 70

Effective cohesion intercept, c' (kN/m2) Angle of shearing resistance, 9' (")

22 to 35

Undrained Young's modulus, E, (MN/m*)

5 to 30

Drained Young's modulus, E' (h4N/m2)

3 to 25

Coefficient of permeability, k ( d s ) Coefficient of earth pressure at rest, KO

62

0

1 x 10-2to1 x 10-6

0.4 to 0.6

Building response to tunnelling

5.5.2

1

Alluvium Occurrence Over much of the section of route there are recent river deposits, ie alluvium, derived from the River Thames and its former tributaries Tyburn (in the Westminster area) and Neckinger (between London Bridge and Bermondsey). These deposits consist of a variety of materials ranging from soft compressible variable clays to silts, sands and gravels. They also commonly contain organic material in the form of peat and the remains of vegetation. The thickness of the alluvium can be up to 5 m. Table 5.5

Index property units)

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Bulk unit weight (kN/m3)

Index properties and suggested design parameters: alluvium

Typical range of clayey material

Design parameters (units)

Suggested values

12.5 to 20

Bulk unit weight, y (kN/m3)

I6 to 20 25 to 50

Natural moisture content (%)

18 to 5 5

Undrained shear strength, S,,(kN/m2)

Liquid limit (%)

30 to 80

Effective cohesion intercept, c’ (kN/m2)

Plastic limit (%)

I5 to 25

Angle of shearing resistance, I$’(”)

22 to 30

Plasticity index (%)

14 to 55

Undrained Young’s modulus, E,, (MN/m2)

5 to 15

Liquidity index

-

NIA

0.9Eu

Drained Young’s modulus, E’ (MN/m2) Coefficient of permeability, k ( d s )

depends on type of alluvium

Coefficient of earth pressure at rest, KO

0.5 to 0.8

Typical descriptions and geotechnical properties The alluvium has been broadly divided into four lithologies as follows: 0

soft light grey-green to dark brown, very silty, sandy, CLAY with rare rootlets and carbonaceous material and occasional shell gravel.

0

soft to firm dark grey to black organic silty CLAY

0

very soft to soft dark brown clayey PEAT medium dense yellow brown fine to medium SAND, locally clayey.

These lithologies are commonly restricted in geographical extent. Each lithology has its own geotechnical properties. For the purposes of this study they have been grouped together, the index properties of the clayey units being summarised in the first two columns of Table 5.5 and the design parameters for the JLE Project suggested prior to design and construction in the second two columns.

Ch 5 Geology and geotechnical properties

63

5.5.3

River Thames Terrace Gravel Occurrence

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The River Thames Terrace Gravel (referred to below as Terrace Gravel and Thames Gravels) is a series of terrace deposits of the historic River Thames and its tributaries at a time of active uplift and erosion in a high-energy fluvial environment during the Pleistocene. In the area studied several terraces have been identified (based on elevation), with the youngest generally at the lowest elevation. In the study area the base of the Terrace Gravel is at its lowest on the route between the Thames at Westminster and the southern part of St James’s Park (Figure 5.2). This appears to have affected the strength of the underlying London Clay Formation (ie it has a lower strength than clay elsewhere at the same depth from the ground surface) and for the tunnel-related ground deformations (ie greater volume losses). These matters are discussed in Chapter 25. The Terrace Gravel is a mixture of quartz sand, comminuted quartz and mainly brown flint and chert gravel. The proportions of sand and gravel vary considerably in short lateral and vertical distances, depending on the local conditions at deposition. There are also frequent zones of finer-grained material, eg of clayey and silty sand and even occasional organic deposits. The Terrace Gravel was encountered over the entire study area and varies in thickness from about 1 m to more than 6 m.

0.001

0.01 Silt

0.1

1 Sand

10

100

Gravel

Typical descriptions and geotechnical properties The Terrace Gravel is typically medium dense to dense orange brown, very sandy (medium to coarse) sub-angular to sub-rounded, fine to coarse, flint GRAVEL with occasional cobbles and rare pockets of grey and red clay. Figure 5.3 shows the grading envelope for these materials. The suggested design parameters for the Terrace Gravel are listed in Table 5.6.

64

Building response to tunnelling

Index properties and suggested design parameters: Terrace Gravel

Table 5.6

~

~

~~~~

Index property units)

Typical range

Design parameters (units)

Suggested values

Bulk unit weight (kN/m3)

Not measured

Bulk unit weight, y (kN/m3)

I9 to 20

Natural moisture content (%)

N/A

Undrained shear strength, S, (kN/mZ)

Liquid limit (%)

N/A

Effective cohesion intercept, c’ (kN/m2)

Plastic limit (“h)

N/A

Angle of shearing resistance, Q‘ (“)

Plasticity index (%)

Non plastic

Liquidity index

N/A

Undrained Young’s modulus, E, (MN/m2) Drained Young’s modulus, E’ (MN/mZ) Coefficient of permeability, k ( d s )

Coefficient of earth pressure at rest, KO

5.5.4

N/A 0 32 to 40

N/A 30 to 160 5 x 10-3t05x 10-6

0.35 to 0.5

London Clay Formation

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Occurrence The London Clay Formation is present along the section of the JLE route between Green Park and Bermondsey Stations. Its thickness is as much as 45 m in the Westminster area. The London Clay was deposited in a marine environment and attained thicknesses greater than 100 m in places. In the studied area, much of the top surface has since been eroded away by uplift and the fluvial and glacial activity associated with deposition of the Terrace Gravel. The geology of the London Clay was defined by King (1 98 l), who identified five units within the London Clay (A to E), with firther subdivision of each unit into two or three parts. These units were based on sedimentary cycles that gave rise to subtle variations in lithology and fossil fauna. Each sedimentary cycle involved an upward coarsening of the material. No attempt was made to identify these units during the pre-construction ground investigation phase of the project because the relatively crude sampling methods of standard ground investigation practice would not detect the subtle variations that define the units. The variations can be defined by detailed logging of continuously sampled boreholes. This method was used for a recent investigation at St James’s Park into the reasons for larger than expected settlements in the Westminster area (Standing and Burland, 2000). King’s lithological units A and B and their subdivisions were identified from a combination of detailed geological and geotechnical description and also from variations in moisture content. Further details are given in Chapter 2 5 . For the purposes of the JLE Project, and based on description and overall engineering properties, the London Clay was considered as being in two units. These were referred to as the “Unweathered London Clay” (or “main body”) and the “London Clay Basal Beds” (or “basal unit”). The basal unit was defined as “a very silty and sandy zone.. .up to 10 m thick.. .towards the base of the London Clay”. Along the line of the JLE, the London Clay is weathered at its interface with the Terrace Gravel to a maximum depth of about 0.5 m. The weathered material is brown rather than dark grey and has a higher moisture content than the unweathered clay

65

Fissuring is a persistent feature within the London Clay and evident at two scales: extremely closely spaced and randomly orientated persistent sub-vertical shear surfaces (termed “backs” or “greasy backs” by tunnellers and in Ward et al, 1959). These fissures relate to the structural depositional and erosional history of the London Clay. Their existence has a significant influence on the engineering behaviour of the clay, eg the backs cause the clay to fail in a blocky manner in unsupported tunnel faces.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Another feature of the London Clay is the occurrence of “claystones”. These layers of indurated nodules within the London Clay occur at specific horizons within the sedimentary cycles. It is usually not possible to trace them laterally, even between closely neighbouring boreholes because, as has been seen in the faces of exposures, they are not continuous. The permeability of the London Clay depends on the direction in which it is measured and the orientation and persistence of silt and sand partings. The silt and sand are often in the form of laminations between clay layers, so that the horizontal permeability is very much greater than the vertical permeability. In addition, the lower parts of the basal beds tend to be more permeable as they contain a greater proportion of sand and silt, dispersed within their fabric. These differences in structure, composition and permeability of the different units of the London Clay affect both their short- and long-term responses to tunnel excavation and associated works Typical descriptions and geotechnical properties The main body of the London Clay (ie excluding the basal unit) is typically a very stiff, thinly laminated, very closely fissured, dark grey and grey-brown CLAY of very low to medium compressibility and high to very high plasticity. Commonly, it also contains partings of silt and sand, traces of gypsum and pyrite, and irregularly dispersed layers of moderately strong to strong claystone and phosphatic and calcareous nodules. The basal zone is of a similar description but becomes dominated by silt and sand partings and layers, commonly being (or becoming with depth) clayey, silty SAND. Typical grain size percentages are 15 per cent clay, 35 per cent silt and 50 per cent sand. Figure 5.4 shows a range of grading curves for the London Clay as a whole, although most of the samples were derived from the “unweathered” London Clay. Table 5.7 summarises the index properties of the London Clay, showing some of the differences between the materials of the main body and the basal unit. This table also includes the suggested design parameters for the London Clay as a whole. Figure 5.5 presents a plasticity chart for the London Clay as a whole. It can be seen that the majority of samples lie in the high to very high plasticity range.

66

Building response to tunnelling

1

0

C .-

v) U)

m a

(I)

0

m C

a,

2 a, a

Silt

Sand

I

Gravel

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Grain size (mm) Figure 5.4

Envelope of particle size distributions for London Clay

Figure 5.5

Plasticity chart for London Clay

Table 5.7

Index properties and suggested design parameters: London Clay and London Clay basal beds [in parentheses)

Index property units) Bulk unit weight (kN/m3)

Typical range

Design parameters (units)

Suggested values

18 to 20.5

Bulk unit weight, y (kN/m3)

18 to 20.5

( 1 8 to 20.5)

Natural moisture content (%)

20 to 30

Undrained shear strength, S, (kN/mZ)

( 1 5 to 30)

50 + 8z * [h20 per cent] See Figure 5.6

Liquid limit (%)

54 to 85 (44 to 76)

Effective cohesion intercept, c' (kN/mz)

0 to 12

Plastic limit (%)

20 to 30 (20 to 30)

Angle of shearing resistance, 4' (")

24 to 28

Plasticity index (Yo) Liquidity index

25 to 50 -0.2 to 0.2 (-0.4 to 0.2)

Undrained Young's modulus, E, (MN/m2) Drained Young's modulus, E' (MN/m2) Coefficient of permeability, k ( m / s ) Coefficient of earth pressure at rest, KO

See Figure 5.7

0.75E, 1 x 10'8to 1 x 10-'O

0.8 to 2

Where z = depth below top of London Clay

Ch 5 Geology and geotechnical properties

67

A key parameter for assessing tunnelling stability and deformations in clay is undrained shear strength, S,. The relation between undrained strength (in kPa) and depth, z, from the top of the London Clay (in metres) given in Table 5.7 of S, = 50 + 8z was adopted as a design line for the Green Park-Bermondsey section. This relation is shown in Figures 5.6 (a) and (b) with SPT and UU triaxial test data from the JLE ground investigation. The line is generally on the conservative side of the mean with respect to field and laboratory data to take account of the mass operational strength of the clay. The implication of this design strength profile is that depth from the Quaternary erosion surface is taken as the determining factor rather than the differences in the lithology, fabric or index properties of the materials. The recent ground investigation at St James's Park (Standing and Burland, 2000) has provided additional data that corroborate this design line. The range of results fiom measuring undrained Young's modulus on triaxial samples of London Clay at different levels of axial strain is presented as an envelope in Figure 5.7, moduli being normalised by undrained shear strength from UU, CIU and CAU triaxial tests. (a)

(b)

SPT N Value

0

20

40

60

80

100

0

50

Undrained shear strenglh , Su (kNlm2) 150 200 250 300

100

350

400

450

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

0

5

--E

10

.-

2.

-

0 6 15 n

.. 6

; a

20 .

B

.

5,

d

*..

I

. \ ' . . .:.

25

30

SPT N Value --Design

line (Su=50+8z=4.5N)

35

__

f

UlOOsamples

1

U38samples

-Design

Figure 5.6

line (Su=50+8~=4,5N)

Strength profiles of the London Clay plotted with depth from top of the Clay, (a) SPT N values, and (b) undrained strengths from triaxial testing

\ \

\

0.01

0.1

Local axial strain (%)

Figure 5.7

68

1.o

10

Envelope of EJS, ratios with axial strain for London Clay

Building response to tunnelling

5.5.5

The Blackheath Beds (Harwich Formation) Occurrence This stratum lies beneath the London Clay and is generally less than 1 m thick (and often is not present). The construction works encountered it only in the area between London Bridge and Bermondsey station, where the tunnels cross the interface between the London Clay and the Lambeth Group; they are not relevant to the case study buildings study. Typical description The Blackheath Beds typically consist of dense to very dense black rounded and subrounded, medium to coarse flint GRAVEL (with occasional cobbles) in a dark brown silty or clayey fine to medium sand matrix.

5.5.6

The Lambeth Group

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Description and occurrence The Lambeth Group comprises a variable sequence of sands, silts, gravels and clays with occasional limestone beds. Much was learnt about the group (formerly known as the Woolwich and Reading Beds) during the JLE site investigation phase in 1990/1991 and during the subsequent deep excavations. A new informal lithostratigraphical system was devised by Ellison (1991) specifically for the JLE Project on the basis of existing information and the ground investigation findings. The informal units were later placed in their stratigraphical context (Ellison et al, 1994). While that remains an informal scheme (in that it does not follow the formal system of using the place name of the type rock) it is that used on the British Geological Survey lexicon and is likely to be used in the London memoir which is to be prepared shortly. For these reasons the names used by Ellison et a1 are followed here and in CIRIA Funders Report CP/83 Engineering in the Lambeth Group (Hight et al, 2001). Note, however, that Page and Skipper (2000) have suggested that it would avoid confusion if the names of the units of the Lambeth Group were termed “beds” because very often they are sands or interlayered sands and clays rather than the “Clay” or “Sand” of current nomenclature. The classification systems are summarised in Tables 5.2 and 5.8, the latter incorporating information from Page and Skipper (2000).

Ch 5 Geology and geotechnical properties

69

Table 5.8

Lithostratigraphy of the Lambeth Group

Formation (Ellison et al, 1994)

Lithostratigraphical unit

Dominant sediment type(s)

Environment of deposition

Reading Formation

Upper Mottled Clay

Pedogenically altered sediments Typically mottled clay, silt and sand

Alluvial plain (overbank deposits plus channel sand)

Lower Mottled Clay Woolwich Formation Upper Shelly Clay

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Upnor Formation

Laminated shelly clay, silt and fine sand Brackish, shallow water, restricted marine, tidal estuary Estuarine sediments plus channel sands

Lower Shelly Clay

Laminated shelly clays

Laminated beds

Laminated silts, clays and fine sands

Pebble Bed

Locally significant gravel component

Shallow marine - high energy

Glauconitic Sand

Heterolithic clays, silts and sands containing the green clay glauconite

Shallow marine - shore face

The deposits were laid down in one or more embayments on the edge of the North Sea basin during an unstable period in which sea levels rose and fell. This gave rise to rapid changes in lithology as the sea level advanced and retreated. When the sea level fell and the sediments emerged, they developed as soils but were cut through by stream channels. Inundation by a rising sea level led to new phases of sedimentation. The resultant depositional environments included terrestrial mud flats (including soil profiles and sand-filled channels), brackish water lagoonal deposits, marine sands and pebble beds. The materials encountered along the JLE are typically in the form of intercalations of the Woolwich and Reading Formations with boundaries between them that cut across time-planes (ie they are diachronous) and underlain by the Upnor Formation. The thickness of the entire sequence, where it was proved, ranges between 12 m and 20 m in the Bermondsey area, where there was the best development of the sequence in that all seven component units were encountered. The individual units also vary rapidly in thickness, but locally some are absent from the sequence. Table 5.2 summarises the variations in thickness of the different sub-divisions as found along the JLE route. To the east of Bermondsey station, with its youngest deposits having been eroded, the Lambeth Group is directly overlain by the Terrace Gravel, as shown in Figure 5.2 The recent work by Page and Skipper (2000) has shown the significance of immediate post-depositional and early burial effects (including bonding, cementation, fissuring and biogenic activity of pedogenesis) and uplift, weathering and periglaciation on the engineering properties of the materials. Typical descriptions and geotechnical properties

The boreholes drilled for the JLE investigation were logged in accordance with the informal lithostratigraphical units devised by Ellison (199 I). Below, typical descriptions of the Lambeth Group, beds (with the uppermost first) are given with table summaries of their index properties and the suggested design parameters. Note the abbreviations for the individual beds, which are used later in this book.

70

Building response to tunnelling

Figure 5.8 is a plasticity chart that includes the Atterberg limit results from all the clayey units. As with the London Clay, the ratios of undrained Young's modulus to undrained shear strength from UU triaxial tests have been plotted against axial strain and shown as an envelope containing all the results in Figure 5.9.

* D

Upper Shelly Clay Upper Mottled Clay Laminated beds Lower Shelly Clay

b 0

Lower Mottled Clay Pebble Bed Glauconitic Sand

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

I

\

Upper plasticity range

I

7 \ \ i \ \i ...... ....

...

....... ................................. I

0.01 0.1 Local axial strain (%) Figure 5.9

1.o

10

Envelope of EJSu ratios with axial strain for Lambeth Group clay soils

In the following pages, there are summaries of the properties of each of the seven units of the Lambeth Group; each unit is given a separate page.

Ch 5 Geology and geotechnical properties

71

Reading Formation (upper) - Upper Mottled Clay (UMC)

This unit is typically stiff to very stiff and hard, extremely closely fissured, red brown and grey mottled CLAY with occasional patches of sand and weak calcareous nodules. The unit also contains water-bearing sand-filled channelshars of silty fine to medium SAND. The grading envelope for these materials is shown in Figure 5.10 and reflects their variable nature. Table 5.9 lists index properties and design parameters. Index properties and suggested design parameters: Upper Mottled Clay, Lambeth Group

Table 5.9

Index property units)

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Bulk unit weight (kN/m’)

Typical range

Design parameters (units)

Suggested values

18 to 22.5

Bulk unit weight, y (kN/m3)

I9 to 21

Natural moisture content (%)

9 to 5 8

Undrained shear strength, S,(kN/m2)

Liquid limit (%)

20 to 84

Effective cohesion intercept, c’ (kN/mZ)

0 to 15

Plastic limit (%)

11 to 53

Angle of shearing resistance, 4‘ (“)

28 to 32

Plasticity index (%)

1 1 to39

Undrained Young’s modulus, Eu(MN/mZ)

see Fig 5.9

1.44 to 2.42

Drained Young’s modulus, E’ (MN/m2)

0.8 E ,

Liquidity index

Coefficient of permeability, k ( m / s ) Coefficient of earth pressure at rest, KO

100 to 400

1 x 1 0 8 to 10-6 1 to 1.5

Grain size (mm)

Figure 5.10

72

Envelope of particle size distributions for Upper Mottled Clay

Building response to tunnelling

Reading Formation (lower) - Lower Mottled Clay (LMC) These beds are typically very stiff and hard, extremely closely fissured, purple, red, brown and grey mottled CLAY with patches of sand. Calcareous concretions are common, especially at the top of this unit, and form thickly bedded limestone in places. Page (1999, and Page and Skipper (2000) identified these concretionary layers as being a type of duricrust. A grading envelope for these materials is shown in Figure 5.1 1, but it is based on only five samples. Table 5.10 lists index properties and design parameters. Table 5.10

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Index property units)

Index properties and suggested design parameters: Lower Mottled Clay, Lambeth Group

Typical range

Design parameters (units)

Suggested values

Bulk unit weight (kN/m3)

about 20

Bulk unit weight, y (kN/m3)

19 to 21

Natural moisture content (%)

17 to 29

Undrained shear strength, S, (kN/m2)

Liquid limit (%)

33 to 75

Effective cohesion intercept, c' (kNlm2)

0 to 15

Plastic limit (%)

I9 to 32

Angle of shearing resistance, $' (")

28 to 32

Plasticity index (%)

22 to 44

Undrained Young's modulus, E, (MNlm*)

Liquidity index

100 to 400

see Fig 5.9

Drained Young's modulus, E' (MNlm')

-0.03 to 0.00

0.8 E,,

Coefficient of permeability, k ( d s )

1 x 10-8to 10-6 1 to 1.5

Coefficient of earth pressure at rest, KO

0.001

0.01

I

Silt

0.1

I

100

10

1 Sand

I

Gravel

I

Grain size (mm)

Figure 5.11

Envelope of particle size distributions for Lower Mottled Clay

Ch 5 Geology and geotechnical properties

73

Woolwich Formation (upper) - Upper Shelly Clay (USC) These materials are typically either very stiff, dark grey and grey-brown very silty CLAY or very clayey SILT with fine sand, shells and flint pebbles and thin impersistent limestone beds. In many localities between Bermondsey and Canada Water this unit is not present, as it has been eroded away. Table 5.1 1 lists index properties and design parameters.

Table 5.11

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Index property units)

Index properties and suggested design parameters: Upper Shelly Clay, Lambeth Group

Typical range

Design parameters (units)

Suggested values

Bulk unit weight (kN/m3)

about 18

Bulk unit weight, y(kN/m3)

19 to 21

Natural moisture content (%)

27 to 30

Undrained shear strength, S, (kN/m2)

Liquid limit (%)

29 to38

Effective cohesion intercept, c' (kN/m2)

0 to 15

Plastic limit (%)

14 to 28

Angle of shearing resistance, @' (")

28 to 32

Plasticity index (%)

9 to 15

Undrained Young's modulus, Eu(MN/m2)

see Fig 5.9

0 to 1.07

Drained Young's modulus, E' (MN/m2)

0.8 E,,,

Liquidity index

Coefficient of permeability, k ( d s ) Coefficient of earth pressure at rest, KO

74

100 to 400

I

x

10-8to 10-6 1 to 1.5

Building response to tunnelling

Woolwich Formation (upper) - Laminated beds (LB) These materials range from dense to very dense light brown-grey fine to medium SAND to very stiff dark grey silty CLAY. They are thinly inter-bedded and inter-laminated and contain occasional shells and plant debris. The grading envelope for these materials is shown in Figure 5.12, although it is based on only 10 of 13 samples. The gradings of individual laminations are likely to be towards either the upper or lower bound of the envelope. Table 5.1 1 lists index properties and design parameters. Index properties and suggested design parameters: Laminated beds, Lambeth Group

Table 5.12

Typical range

Design parameters (units)

Suggested values

I9 to 22

Bulk unit weight, y (kN/m3)

19 to 21

20 to 35

Undrained shear strength, S, (kN/m2)

Liquid limit (%)

27 to 84

Effective cohesion intercept, c’ (kN/m2)

Plastic limit (%)

I4 to 25

Angle of shearing resistance, @’ (”)

Plasticity index (Yo)

I9 to 37

Undrained Young’s modulus, E, (MN/m*)

N/A

-0.13 to 0.53

Drained Young’s modulus, E’ (MN/m2)

80 to 200

Index property units) Bulk unit weight (!AVm3)

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Natural moisture content

(Yo)

Liquidity index

N/A 0 33 to 40

I

Coefficient of permeability, k ( d s )

x

10-3to I O - ~ 1 to 1.5

Coefficient of earth pressure at rest, KO

1

0

C .-

U) U)

m

a a,

.0

m c

a,

2 (I) a

0.001

0.01

I

Silt

I

1

0.1 Sand

I

10

Gravel

I

100

Grain size (mm) Figure 5.12

Envelope of particle size distributions for Laminated beds

Ch 5 Geology and geotechnical properties

75

Woolwich Formation (lower) - Lower Shelly Clay (LSC) These materials are typically described as very stiff and hard, dark grey silty CLAY with laminations of light grey silt and a variable amount of shells which may form thin weak impersistent limestone beds in places. Generally the shell content decreases upwards in this unit. The grading envelope, based on only ten samples of these materials, is shown in Figure 5.13. Table 5.13 lists index properties and design parameters. Table 5.13

Index property units)

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Bulk unit weight (kN/m3)

Index properties and suggested design parameters: Lower Shelly Clay, Lambeth Group

Typical range

Design parameters (units)

Suggested values

20 to 22.3

Bulk unit weight, y (kN/m3)

19 to 21

Natural moisture content (%)

18 to 29

Undrained shear strength, S,, (kN/rn’)

Liquid limit (%)

45 to 71

Effective cohesion intercept, c‘ (kN/m’)

0 to 15

Plastic limit (%)

I9 to 28

Angle of shearing resistance, @’ (”)

28 to 32

Plasticity index (%)

26 to 45

Undrained Young’s modulus, E, (MN/m’)

see Fig 5.9

-0.14 to 0.57

Drained Young’s modulus, E’ (MN/m’)

0.8 E,,,

Liquidity index

100 to 400

1 x 10-*to 10-6

Coefficient of permeability, k ( d s )

1 to 1.5

Coefficient of earth pressure at rest, KO

0.001

0.01

Silt

76

0.1

1 Sand

10

100

Gravel

Building response to tunnelling

Upnor Formation - Pebble Bed (PB) The Pebble Bed consists typically of very dense blue-green and grey, fine to coarse flint GRAVEL with occasional cobbles in a matrix of clay, silty clay or fine to medium sand. The flints are predominantly well rounded. The grading envelope for these materials is shown in Figure 5.14. Table 5.14 lists index properties and design parameters. Table 5.14

Index property units)

Typical range

Design parameters (units)

Suggested values

Bulk unit weight (kN/m3)

Not measured

Bulk unit weight, y (kN/m3)

19 to 21

Natural moisture content (%)

10 to 22

Undrained shear strength, S, (kN/mZ)

Liquid limit (%)

25 to 35

Effective cohesion intercept, c' (kN/m2)

Plastic limit (%)

I3 to 22

Angle of shearing resistance, @' (")

Plasticity index (%o)

12 to 32

Undrained Young's modulus, E,(MN/mz)

NIA

-0.28 to 0.00

Drained Young's modulus, E' (MN/m2)

80 to 200

Liquidity index

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Index properties and suggested design parameters: Pebble Bed, Lambeth Group

Coefficient of permeability, k ( m / s ) Coefficient of earth pressure at rest, KO

N/A

0 33 to 40

1

x

I O - ~to I O - ~ 1 to 1.5

Grain size (mm) Figure 5.14

Envelope ofparticle size distributions for the Pebble Bed

Ch 5 Geology and geotechnical properties

77

Upnor Formation - Glauconitic Sand (GS) These materials are predominantly very dense green-grey clayey, very silty tine to medium SAND, with clay laminae and thin beds of flint gravel. The green coloration is provided by the mineral glauconite that occurs as sand-sized grains. The grading envelope for these materials is shown in Figure 5.15. Table 5.15 lists index properties and design parameters. Table 5.15

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Index property units)

Index properties and suggested design parameters: Glauconitic Sand, Lambeth Group

Typical range

Design parameters (units)

Suggested values

Bulk unit weight (kN/m3)

20 to 22

Bulk unit weight, y (kN/m3)

I9 to 21

Natural moisture content (%)

I5 to 35

Undrained shear strength, S, (kN/m2)

Liquid limit (%)

23 to 41

Effective cohesion intercept, c’ (kN/m2)

Plastic limit (%)

16 to 23

Angle of shearing resistance,

Plasticity index (%)

6 to 23

Undrained Young’s modulus, E, (MN/m2)

N/A

-0.17 to 0.64

Drained Young’s modulus, E’ (MN/m2)

80 to 200

Liquidity index

Coefficient of earth pressure at rest, KO

Figure 5.1 5

5.5.7

Sand

0

+’ (“)

Coefficient of permeability, k ( d s )

Silt

N/A

33 to 40

I

I O - ~to 1 0 . ~

x

1 to 1.5

Gravel

Envelope of particle size distributions for Glauconitic Sand

Thanet Sand Formation This deposit is water-bearing and in hydraulic continuity with the Upnor Formation of the Lambeth Group above. It therefore influenced the deeper construction works of the JLE, such as the box excavations and the tunnels beneath the Canada Estate. The Thanet Sand Formation is a consistently very dense dark grey silty fine SAND. The grain size becomes slightly finer downwards in the sequence as the clay content increases towards the unconformity with the Chalk.

78

Building response to tunnelling

5.5.8

Chalk Chalk sampled in the ground investigation is generally described as white, slightly to moderately weathered CHALK, weak to moderately weak, with nodules and bands of flint withrandomly oriented closely to very closely spaced fractures that are open, closed or in-filled with chalk putty. Chalk was not encountered in the works near the case study buildings.

5.6

HYDROGEOLOGY

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The hydrogeology of the London area is generally considered to be controlled by two aquifers, known as the upper and lower aquifers. In addition, along the JLE route, waterbearing horizons were encountered within the upper granular units of the Lambeth Group, within the Blackheath Beds and also occasionally perched in the made ground. Two different hydrogeological regimes exist along the JLE route. In the west, where the London Clay is present, the upper and lower aquifers are not in hydraulic continuity, and the lower aquifer is more deeply buried, remote from the engineering works. East of Bermondsey, however, where the London Clay is absent, the upper aquifer is close to the surface and may directly overlie the lower aquifer, such that it significantly affects the engineering works.

5.6.1

The upper aquifer The upper aquifer consists of the Alluvium and the Terrace' Gravel. The aquifer is unconfined, with the elevation of the piezometric surface at approximately 100 m PD across the route. The aquifer has a widespread hydraulic connection with the River Thames and is affected to some extent by the river's tidal movements between about + 97 and 103 m PD. At locations where excavation was required through these materials, groundwater was excluded using cofferdams, by ground treatment with grout, or by a combination of the two methods.

5.6.2

The lower aquifer Over the length of the JLE route that contains the case study buildings, the lower aquifer is generally confined. Three strata comprise the aquifer body: the Chalk, the Thanet Sand Formation and the Upnor Formation. At the start of the 20th century, the water pressures in the centre of the London Basin were artesian, linking to water levels in the Chalk in the surrounding hills. As well as the numerous wells for domestic and industrial water supply, some were constructed as drinking fountains in public places. Extensive pumping of water from this aquifer over a period of 150 years, particularly for industrial use, resulted in the piezometric surface being depressed by the order of 50 m in central London. By the mid-l960s, extraction had greatly reduced, and since then the piezometric surface has been rising at a rate of approximately 1 &year. This rise in water level and its potential effects upon civil engineering structures are described in a CIRIA report (Simpson et al, 1989). An informal organisation known as GARDIT, which includes LUL and the Environment Agency, continues to monitor the rise in water level. At the beginning of the JLE construction works, the piezometric surface was between 60 m PD near Green Park and 95 m PD in the Bermondsey area.

Ch 5 Geology and geotechnical properties

79

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

As these water-bearing units would have affected construction of the proposed JLE structures, dewatering operations were employed to reduce piezometric pressures in the Lambeth Group and the underlying Thanet Sand Formation and Chalk along the JLE route between Bermondsey and North Greenwich. The locations of the dewatering sites are shown on Figure 5.16. Of particular relevance to the case-study buildings are dewatering schemes installed at Bermondsey and Canada Water stations and at the shafts in Culling Road, Ben Smith Way and Druid Street. The settlement caused by the dewatering affected the absolute level of local benchmarks and the buildings themselves. Measurement of the tunnelling-induced settlement of the ground and buildings is thought not to have been markedly affected in relative terms. The reasons for this are that the dewatering-induced settlement affected a very large area and the rate of settlement from the dewatering was also very much slower than the tunnellinginduced settlements, or had largely ceased when the tunnelling took place. In general terms, the dewatering depressed the piezometric surface in the lower aquifer by up to 20 m over a large area of east London. The effects of these schemes are reported in greater detail in Chapter 17. In the east of the study area, where the lower aquifer is closer to the surface, the aquifer may be affected by the proximity of the River Thames as a source of recharge. In the west such recharge is not readily possible because of the presence of the London Clay.

I

Figure 5.16

5.6.3

I

Location of sites of dewatering operations

Other water-bearing horizons The granular units within the upper levels of the Lambeth Group, ie in the Upper Mottled Clay and the Laminated beds, are commonly water-bearing. Although they can be present as isolated lenses confined within clay strata, in some zones they are in direct hydraulic connection to the upper aquifer or to the River Thames. The dewatering measures were implemented specifically in order to reduce water pressures in these units at the Bermondsey station complex and at the Druid Street shaft.

80

Building response to tunnelling

5.7

REFERENCES ELLISON, R A (199 1). Lithostratigraphy of the Woolwich and Reading Beds along the proposed Jubilee Line Extension, south-east London. British Geological Survey Technical Report W N 9 l/SC, unpublished ELLISON, R A, KNOX, R N 0, JOLLEY, D W and KPJG, C (1 994). A revision of the lithostratigraphical classification of the early Palaeogene strata of the London Basin and East Anglia. Proceedings of the Geologists 'Association, 105, 187-197 HIGHT, D W, ELLISON, R A and PAGE, D P (2001). Engineering in the Lambeth Group, Funders Report CPl83, CIRIA, London

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

KING, C (1 981). The Stratigraphy of the London Clay and associated deposits. Tertiary Research Special Paper No 6. Dr W Backhuys, Rotterdam LTNNEY, L F and PAGE, D P (1996). Site investigation for the tunnels and stations of the Jubilee Line Extension, London. Geotechnical Aspects of Underground Construction in Soft Ground (Mair, R J and Taylor, R N eds), Balkema, Rotterdam

MOTT MACDONALD GROUP (1991). Jubilee Line Extension, Geotechnical Investigation, Sectional Interpretative Report No 1: Green Park to Canada Water, Unpublished report to Jubilee Line Extension Project, Mott MacDonald, Croydon PAGE, D P (1995). The engineering geolosy of the Lambeth Group (Woolwich and Reading Beds) Unpublished MSc thesis, University of Surrey PAGE D P and SKIPPER, J (2000). The Lithological Characteristics of the Lambeth Group. Ground Engineering, Vol33 No 2, February, pp 38-44 SIMPSON, B, BLOWER, T, CRAIG, R N and WILKINSON, B R (1989). The Engineering Implications of Rising Groundwater levels in the Deep Aquifer Beneath London, Special Publication 69, CIRIA, London STANDING, J R and BURLAND, J B (1999). Ground Characterisation to explain JLEP tunnelling volume losses in the Westminster area. Unpublished Imperial College report to London Underground Limited WARD, W H, SAMUELS, S G and BUTLER, M E (1959). Further Studies of the Properties of London Clay. GCotechnique, Vol IX No 2, June, pp 33-58

Ch 5 Geology and geotechnical properties

81

6

St James’s and St James’s Park: a brief history of their development

G Johnston 6.1

INTRODUCTION

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

This chapter is a background note about the historic development of the area near Green Park station called St James’s and of St James’s Park. It includes short accounts of the site histories and construction of two of the case study buildings, the Ritz Hotel and the RAC building. The wealth of history in this area means that the chapter can only be a broad overview. Much of the information was obtained from the Survey oflondon. References, however, have been checked and are listed for further study. The first of the two following sections is about the development of the St James’s area in general. While its relevance is mainly to the building development, it also refers to the early information about the land that later became the present St James’s Park. As the two case study buildings are in the St James’s area north of the park, they are described in Sections 6.3 and 6.4. Section 6.4 is specific to the history of St James’s Park, which bears on the research results at the greenfield reference site.

6.2

ST JAMES’S The highways and the boundaries in this area were probably in existence in medieval times. The road linking Charing Cross with St James’s Hospital (the site of St James’s Palace) certainly existed by the 12th century, and the road called Piccadilly existed before the 16th century as the highway to Colnbrook and Reading - some theories suggest it might have formed part of the old Roman road, Watling Street. Tiswell’s map of 1585 shows both the Haymarket at the eastern boundary and St James’s Street near the western boundary; St James’s Street may have been formed in the reign of King Henry VIII. There is, however, no positive evidence of building before the 17th century. There is some evidence that the land was divided among several, mostly corporate, owners, including the Abbey of Westminster, the Convent of Abingdon and the Hospital of St James’s. The last-named occupied - probably since the 1 1th century - the area adjacent to where St James’s Palace now stands (see Note 1 at end of this chapter). St James’s Fair, established in 1290 for the benefit of the hospital, was probably held near by. Very little more is known about the medieval history of the area except that previously it was arable. Between about 153 1 and 1536 probably all this area was surrendered to the crown and formed into the Bailiwick of St James’s. St James’s Field (or Fields) (see Figure 6.1) was enclosed by Henry VI11 and turned into meadow land. The exact date of the enclosure of the field is not known but would have been between 1536 and 1547. Later sources give its measurements as 43 acres (Note 2).

Previous page is blank

Ch 6 St James’s and St James’s Park: a brief history of their development

83

The survey of common land in the parish of St Martin in the Fields, made in 1549, states that since 1485 (the date of the previous survey):

There ys A feld Called Sent James feld by estimacon XI Acres wicche was Comen And owght to be Comen And in Clossed by Kyng Henvy viijth .... Wiche was Arrable And now ys Meadowe. [Note 31 During this period the main references to St James’s Field are as a mustering ground for royal troops (1 55 I): ...there was a muster before the Kinges Maiestie in St James Geld) beyonde

Charing Crosse, the Kinges Maiestie sittinge on horse-backe on a hill by St James with his maiesties Privie Counsel1 with him. [Note 41

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

St James’s Palace was built by Henry VI11 between 1532 and 154 1 on the site of a female lepers’ hospital called St James’s in the Fields. It was built initially as a royal manor for Henry Fitzroy, Duke of Richmond, the King’s natural son by Elizabeth Blout. The creation of St James’s Palace did not stimulate building in the area significantly. However, some buildings may have appeared in the 16th century opposite the palace, where the partly timbered houses of Sir Henry Henne and Sir William Poultney survived until after the Restoration and where, in Charles 1’s reign, Berkshire House was built. Some development began near the palace in the 1620s and is shown on the Faithorne and Newcourt map surveyed in the 1640s (see Figure 6. I).

Figure 6.1

84

St James’s Field on Faithorne and Newcourt’s map (surveyed in 1643-7 and published in 1658)

Building response to tunnelling

The most important of these new buildings was the large mansion facing the palace built in 1626-7 by the Earl of Berkshire. Little is known about the history of this house; some of its fabric possibly survived into the 19th century, but it is important to note that the site of the house and garden determined the intricate layout of the area.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

At about the same time, St James’s Field, on the other side of St James’s Street, was appropriated for royal recreation. This saw the establishment of a tennis court, a physic garden to the south and a pall mall alley on the line of the present Mall. (The old game of pall mall involved driving a ball through an iron hoop with a mallet.) Post-Restoration period changes removed the tennis court and physic garden, but the area was generally left unbuilt until Hugh Woodward (appointed keeper of St James’s in 1637) bought the field in 1651 from trustees of the late King’s lands. By April 1661 there were 220 or more houses in the field, mostly brick-built (Note 5), most of which must have been built under Woodward’s tenure. Woodward’s fellow speculators experienced many difficulties when the Restoration invalidated titles to Crown lands acquired under the “usurped authority”. The disputes over title ownership caused much litigation, and when the material of the houses was mentioned it was said to be brick (Note 5). Litigation over title ownership, it may be remarked, has carried on up to the present day. In March 1661, Henry, Earl of St Albans was granted a lease until 1691, and he was responsible for the street-pattern east of St James’s Street. On 1 April 1665, he was granted freehold of half the field, including the site of St James’s Square. The tenure from 1662 was sufficiently long to give incentive to develop the whole area of St James’s Field with new streets as well as to complete the frontages towards the surrounding streets. Altogether, the post-Restoration development of the area took about 20 years. There are references to schools in the 17th and 18th centuries, which would suggest that this was a residential area with domestic dwellings (Note 6). In the 1660s, not long after Charles I1 ascended the throne, access to St James’s Park was extended and it became a fashionable meeting place. (See Section 6.5 for an account of the history of St James’s Park.) Also during this period, Sandpit Field, part of the Poultney Estate, was surrendered to Charles I1 in 1668 to form Green Park. A small strip of land, not needed for the park, now forms the western end of St James’s Place. Proximity to the royal palace influenced the St James’s area from the start: lodgings were sought here for officers of the Court, especially in the properties around St James’s Square. The official removal of the Court from Whitehall to St James’s Palace in 1698 further affected the social character of the area. One result was the increase in the number of coffee houses, particularly in St James’s Street and, to a lesser extent, Pall Mall. They are especially important in a history of the fabric of this area because they led to the development of the 18th-century subscription clubs. Whites Club can be linked with certainty to a 17th-century coffee house. Most of the clubs, however, were not settled and established in the area until the second half of the 18th century. The 25-year period from 1762 to 1787 witnessed the transition from coffee houses to subscription clubs, and many of the new clubs came into existence in these years, among them Almack’s, Brooks’s and Boodle’s. The building now known as St James’s Club was built in 1892. The Royal Automobile Club, which occupies the site of the old War Office buildings, was built in 1907 and occupation took place between 1907 and 1910 (see Section 6.4 and Chapter 24). Extensive rebuilding took place in the 18th century and very little remains of the first domestic buildings. In the 18th and 19th century there were many hotels in the area, particularly in St James’s Street and Jermyn Street, some of which survived long enough to appear on the 1869-74 Ordnance Survey map (see Figure 6.2).

Ch 6 St James’s and St James’s Park: a brief history of their development

85

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 6.2

St James’s from the OS map of 1869-74

What survives of the area is quite diverse in character, largely reflecting the differences between crown property and the private freeholds. The older areas of St James’s tend to be in private freehold areas: on the south side of Jermyn Street, at a few sites in the square, and in St James’s Street - Pickering Place, Berry’s shop, Lock’s, Whites club house and Blue Ball Yard. Crown Passage, a narrow arched passage-way of small shops, has changed little since the 1790s. Since the 1930s, the appearance of the area has been transformed as buildings used as private apartments have been converted to, or replaced by, offices. Damage caused by Second World War bombs also brought about change; sometimes the damage was so extensive it was necessary to redevelop the site (Note 7). Change continues today and in the past ten years alone several new office developments have been constructed.

6.3

THE RlTZ HOTEL The Ritz Hotel (Figure 6.3) stands on the corner of Piccadilly and Arlington Street. The Piccadilly faqade is about 70 m long; its Arlington Road frontage is 35.1 m, but the Green Park side is only 26.5 m. The Ritz is built on the site of one of the most famous coaching inns, The White Horse Cellar, then No 150, Piccadilly. According to a pictorial guide to London printed in 1850, “This house is well known to the public on account of the great number of stage coaches which regularly call there...”. A few years later, the OS map of 1869-74 records that there was a Bath Hotel on the corner of Arlington Street and, also within just part of the current plan area of the Ritz, there were the premises of coach-builders, a coal merchant, a stationer, and Cockbum’s wine merchant. The survival of these small businesses in an area of clubs and aristocratic mansions was probably because of the restricted nature of the site - at the Green Park end it tapered to only 5.5 m.

86

Building response to tunnelling

Redevelopment of the site became a viable proposition with the purchase of a 20 m strip of freehold land backing on to Wimborne House. In 1886, the old buildings were demolished and Lord Walsingham built Walsingham House at the (then) enormous cost of E300 000. The massive eight-storey, red-brick block of service flats was designed by Alfred Burr and completed in 1888.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

In 1902, the Blackpool Building & Vendor CO Ltd purchased the site of Walsingham House for construction of the Ritz Hotel. The appointed architects, Charles Mew& and Arthur Davis, designed an eight-storey steel-frame building that belonged both visually and structurally to a Franco-American tradition (steel-frame buildings having been developed for Chicago). The Ritz was the first steel-frame building of any substance to be put up in London. It created such interest that between 1904 and 1905 The Builder’s Journal published a series of articles and photographs recording its progress month by month. The magazine describes the co-ordination of many new techniques and the solution of new problems. Excavations for the Ritz began in June 1904; the building was completed by 1 October 1905 and the hotel was opened in May 1906. Waring White Building CO Ltd was the main contractor. The total cost of the building was estimated at E345 227, of which more than E15 000 was allocated to English decorators, E49 000 to French decorators, and E102 000 to Waring and Gillow. John P Bishop supervised erection of the structural steelwork, the designer of which was the Swedish-born Sven Bylander, who had worked previously in America. The engineers were M A Potts & CO of Oxford Street. The steel frame was to be founded on pad foundations below the level of the existing 0.9 m thick raft belonging to Walsingham House. According to The Builder (26 September 1904, p 165), the contractors were “. . . troubled to know how to remove it [ 1500 m3 of the raft]. It was eventually overcome by the use of pneumatic drills . .. Cartridges with small charges were used for blasting.” The retaining wall of Walsingham House on the Piccadilly side was retained ... so as to avoid the trouble and danger of disturbing the many telephone cables, electric-light wires, gas pipes, etc., embedded in its concrete foundation.” In all, some 15 000 m3 of soils and debris were removed from the site, deepening it to 8-10 m or so below ground floor level. “

The new foundations were also stanchion bases. Each base was constructed as a concrete pad on which was placed a two-layer grillage of steel beams, the plan area of the pad depending on the column load. For all except the pads on the Wimborne House side there was a cast iron column base set on top of the grillage. On the Wimborne House side, to maximise space without interfering with the adjoining property, the foundation pads were kept back some distance from the boundary. They were constructed as for the other pads, ie with a similar grillage, but generally somewhat deeper and with what was called a cast steel pin support on top rather than a cast iron column base. A plate girder was placed to span from the nearest inner column base over the outer pad and extending as a cantilever towards the boundary. The outermost columns on this side were set to bear on the ends of these cantilevers. The cantilever was therefore fixed by the inner column and with a pin joint at the outer support pad. The structural steel was from Germany. Thus all the steelwork and the foundations were constructed to the metric system. These details of the foundations are given in The Builder’s Journal of 2 November 1904.

Ch 6 St James’s and St James’s Park: a brief history of their development

87

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

,

Figure 6.3

-

- -

Ritz Hotel: cantilever foundation

The exterior of the building is designed in the Beaux-Arts tradition using influences from French Classicism - mansard roof, rusticated masonry thickly cladding the steel W e . Norwegian granite was used on the ground floor and Portland stone on the upper levels. The cladding was thickest on the arcade running along Piccadilly. The arcade enabled the road to be widened without loss of space to the hotel except on the ground floor, as well as providing more space for bedrooms.

,

Ann Davis Thomas, the daughter of the architect Arthur Davis, in a conversation with the author, recalled lunching at the Ritz with her father, who told her that it was a most difficult building to design because there was never enough area to work with. They had thought that the area of Wimborne House would be part of the development, but Lord Wimborne refused to sell. It was for this reason that the arcade was incorporated into the design, which allowed use of the air space to create extra bedrooms. Mirrors were placed in the “enchanting” dining room, a narrow oblong area, to give the illusion of being oval and of greater size. The constructors had to overcome the relatively new problems, now commonplace, of restricted space in city redevelopment. There was little room, for example, to store materials. Mortar was mixed in the basement and the stone was dressed on a platform with a watertight roof over the pavement. They also had to find the way to hoist 12 mlong steel joists on a long narrow site with American derricks, each one weighmg 20 t (Figure 6.4). The Ritz was opened in May 1906 (Figure 6.5) and still epitomises elegant grandeur.

88

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

a

Figure 6.4

A

Ritz Hotel: steelwork erection

1

Figure 6.5

The Ritz Hotel shotfly afier its oper...

Ch 6 St James’s and St James’s Park: a brief history of their development

in 1906

89

6.4

THE ROYAL AUTOMOBILE CLUB The Royal Automobile Club on the south side of Pall Mall, numbers 83-85, has a frontage of 87 m extending back 43 m in the centre. It was built between 1908 and 191 1 on the site of Cumberland House.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Cumberland (originally York) House was completed by 1763 to the designs of Mathew Brettingham as the residence of Edward Augustus, Duke of York and Albany (1 7391767), younger brother of George 111. A town house with a courtyard facing Pall Mall, it was erected on the site of three houses with a total frontage of 28 m that were demolished in 176 1. It was renamed on being inherited by the Duke of Cumberland, another brother. Cumberland intended to enlarge the courtyard fronting Pall Mall by removing the adjoining houses on either side and rebuilding them as slightly projecting wings. In 1773, he had the house on the west side pulled down and rebuilt, and in 1788 he purchased the house on the east side. Robert Adam was engaged to prepare schemes for refurbishing in 1780-2 and 1785-8, but the rebuilding to match the west wing was not carried out until after the Duke’s death in 1790. In 1806, the lease of all three sections of the house was acquired by the Board of Ordnance Survey, which occupied the central part from 1807. The eastern wing was rebuilt in 1809, and two years later the Board occupied both this and the western wing. In 1855, the Ordnance Survey became a responsibility of the Secretary of State for War; and from then until 1906 the War Office used the whole of the house. Between 1907 and 1910, the Offices of Woods, Forests and Land Revenues occupied the western half of the building. In 1908, the Royal Automobile Club signed a building agreement with the Commissioners of Woods and Forests; by that September the site had been cleared, and excavations for the basement of the new building were in progress. The western part of Cumberland House was pulled down in 191 1-12 for the Royal Automobile Club extension. The architects of the new building were the architects of the Ritz Hotel, Charles Mewes and Arthur Davis, in conjunction, here, with E Keynes Purchase. Mewes and Davis designed a steel-frame and concrete structure cased in Portland stone with a slate roof (Figure 6.6). It has three main storeys, basement and dormered mansard. The slightly projecting pedimented central feature of three bays is flanked by wings each six bays wide. The upper floors are faced with giant Ionic order columns that support the entablature with its carved frieze and figure-sculpted pediment. The core of the design is a large central vestibule of oval plan, surrounded by a wide gallery at first-floor level with the principal club and public rooms on the ground and first floors. Recreational rooms in the basement include the elegant swimming pool (which is surrounded by a peristyle of widely spaced pairs of Doric columns), Turkish baths, gymnasium and three squash courts. The four upper floors are given over to members’ bedrooms and staff rooms. The work was completed in the spring of 191 1 at the cost of E250 000.

90

Building response to tunnelling

I

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

F

Figure 6.6

The Royal Automobile Club m 1912

On 20 F & q 1944, a bomb of about 50 kg hit the building, making a hole of about 0.36 m in diametw in &e flat roof. The explosion blew down five inside walls. The

resulting fm,and water fiom fire-hoses, caused

damage.

ST JAMES’S PARK

6.5

St James’s Park is a broad area of trees, lam and gardens with the Mall and Birdcage Walk as its northern and southern boundaries. To the west are Buckingham Palace and the Queen Victoria Memorial; and Horse Guards Road forms its eastern boundary.

As noted in Section 6.2, very little is known about the medieval history of the area except that it was arable land, divided among several owners. St James’s Park was ‘

originally the Royal Park to St James’s Palace, which was built between 1532 and 1541.

In the 1660s, not long after he ascended the throne, Charles I1 undertook to redesign St James’s Park around a monumental canal of French design by Andd and Gabriel Mollet (Figure 6.7). The constructionof the canal had already begun before the French garden designers were appointed. The garden was intended to form part of the proposed development of Whitehall Palace, which never took place, leaving the circle of trees at the east end of the canal relating to little more than a flight of steps down to Horse Guards Parade. The park had decreased in size since 1647, but hunting continud, access to the park however, was extended and it became a fashionable meeting place, where the upper classes could meet their king.

Ch 6 St James’s and St James’s Park: a brief history of their development

91

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

It is known that Capability Brown took over control of the park in 1766, which would perhaps indicate the park having a more natural design, but no plans exist. During the 19th century, the park was extensively redeveloped to complement Buckingham Palace. All restrictions on access were lifted, making it the first royal park to be completely opened to the public. The park was redesigned by John Nash in 1827. The canal was converted into a serpentine lake with islands, and surrounded by a less formal landscape with clumps of trees and bushes. In 1855, the lake was dredged and its base concreted so that it had a uniform depth of 1.2 m.

Figure 6.7

St James’s Park, showing the canal in Restoration times

Various garden features survive, most from the later development of the park, but a terrace in the north-western part of the park may be contemporary with the 1660s’ design. This is up to 1.5 m high and runs roughly south-west to north-east, aligned with the edge of the ornamental canal to its south. Three mounds, presumably all largely artificial, are probably associated with Nash’s redesign of the park. Next to the Mall is a mound approximately 85 m by 50 m and 1.5 m high, with mature plane trees growing on it. A smaller circular tree mound, 50 m in diameter and about 1 m high, stands just north of the Cake House, with a ring of mature trees on top. The largest mound is about 2 m high and located to the south of the canal. It is irregular in shape and has a steep northern side and a shallower southern side. The presence of large tree holes indicates that mature trees formerly grew on it, but most of those growing there now are young. There is little evidence of military activity in the park during the Second World War, but some areas were levelled in 1946, suggesting the possible removal of defence structures. Although the area around the park has been extensively developed since the 17th century; historical maps indicate that the corner of the park near Storey’s Gate is free from previous man-made structures or building loads. It can, therefore, be considered a greenfield site.

92

Building response to tunnelling

6.6

NOTES TO CHAPTER 6 Note 1.23 HENRY VIIIcl (House of Lords Act No.33), 23 HENRY VIIIc 4 (House of Lords Act No.30), 28 HENRY VIIlc (House of Lords Act No.48), 28 HENRY VIIIc (House of Lords Act No.34) Note 2. Sutton deeds, box 1 bundle I , deed of 1607 attached to deed of 13 Feb. 27 Eliz. Note 3. Survey oflondon, vol XXIX, 1960, pp 26-7 Note 4. Ibid., C66/3 181, no 19:Survey of London, vol XXIX, 1960, p 378 Note 5. Calendar of State Papers Domestic, Crown Estate Office, Whitehall. 1668-9, pp 111-12. Note 6. Wren Society, vol xviii, 1941, pp 18-19

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Note 7. Civil Defence Report 1617 (l), (2), (3), (4), Civil Defence Reports 1696, 1710, 1026,O 152, Westminster City Archives. The listed reports relate to bomb damage in the vicinity of the Jubilee Line Extension.

Ch 6 St James’s and St James’s Park: a brief history of their development

93

7

Westminster and Waterloo areas

F M Jardine

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

7.1

GREAT GEORGE STREET Many of the UK’s temporal and spiritual seats of power are in Westminster. Westminster buildings have long been associated with Palace or Abbey or, nowadays, with government itself or with central organisations that benefit from being near to government. This is particularly true of Great George Street. None of its current buildings is a residence. On the north side, the Treasury offices are being refurbished under a Private Finance Initiative for the building to be operated by a private consortium that will lease the offices back to the Treasury. On the south, three adjoining buildings house the Institution of Civil Engineers (ICE), the Public Health Engineering Department (PHED) and the Royal Institution of Chartered Surveyors (RICS). The range of buildings occupies the built-up length of the street between Storey’s Gate and Canning Green on the side of Parliament Square. St James’s Park has been maintained as a limiting bound to the built-up area. Other fixed points are buildings such as Westminster Abbey, the church of St Margaret and the Palace of Westminster. There have been, however, a great many changes in the area, some quite sweeping. In the late 17th century, what we now know as Parliament Square was built up, and King Street (a street roughly parallel to and to the west of the present Parliament Street) continued right up to St Margaret’s. The 1682 map shows many alleys, but no straight road through from New Palace Yard to the Storey’s Gate corner of St James’s Park. Therefore, Great George Street does not follow the route of an earlier street. Great George Street was certainly in existence in the 1760s. The libertarian but libertine John Wilkes and the government ministers who ordered his imprisonment lived in Great George Street at that time. The street would have rung to the cries of “Wilkes and Liberty!” and been daubed with the slogan. James Boswell, Dr Johnson’s biographer, called on Wilkes at Great George Street when, as a young man in 1762 (the year he first met the great doctor) he had lodgings in Downing Street. The latter was not then the place of the official residence of the prime minister. A sweeping change to Westminster was the creation of Victoria Street - an entirely new road. It was opened in 185 1, but its development continued for many years thereafter. As a new route, it cut across old streets and required the demolition of many old houses. By the 1840s, when the new Victoria Street was projected, Parliament Square was an open area, but split by a central road. With one exception, the buildings now in Great George Street date from the late 19th and early 20th centuries. The exception is the western end of the RICS, which belongs to the mid- 1 8th century. Previous buildings go back at least to Queen Anne; perhaps a few are substantially older. Victorian drawings of London as views from a balloon, show the buildings to have had four or five storeys.

Westminster and Waterloo areas

Previous page is blank

95

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 7.1

View (c 1900) along Great George Street in which can be seen the previous buildings on the northern side and the first constructed (eastern) end of the Government offices

Figure 7.1, a photograph taken at the turn of the 20th century, shows plain terraced houses on either side, up to four storeys with a dormer roof, almost certainly with a basement (as remain on the south side). Basements extending out below the pavement or roadway are common in central London and are legacies of previous property boundaries and subsequentroad widening. As CIRIA’s address is in the current Storey’s Gate, it may be noted that on that 1682 map, this street is the Long Ditch. Even as late as the turn of the 20th century it was called Princess or Princes Street, which is how it is variously shown on 1839,1869 and other maps. The major change in Great George Street came about with the construction at the end of the 19th century of the great government buildings at the end of Whitehall and in Parliament Street. In Chapter 26, there is an account of the construction of the Government Offices, Great George Street (GOGGS). Protracted negotiations took place between government and the two institutions- of Chartered Surveyors and of Civil Engineers, both of which had their headquarters buildings on the north side of Great George Street - to purchase the land for the new government offices.

96

Building response to tunnelling

The OS map of 1869 (Figure 7.2) shows the ICE to have been opposite the opening of Little George Street (above the “OR’ of “Great George Street”). Note on this map that Her Majesty’s Stationery Office was on the site of the recently built Queen Elizabeth I1 Conference Centre (Burland and Kalra, 1986). This building was constructed over a tunnel running under the centre of the site as well as a then-operational underground telephone exchange (Halfin, 1988; and Burland and Kalra, 1988).

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The undated photograph of Figure 7.1 not only shows the Institution building set among the older simple terraced houses on the left, but also the first-constructed part of GOGGS (ie the Treasury) in the background.

Figure 7.2

Westminster and Waterloo areas

Great George Street on the 1869 OS map

97

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 7.3

The former building of the Institution of Civil Engineers on the north side of Great George Street in 1896

The 1896photograph Figure 7.3) shows the fine building that housed the Institution of Civil Engineers on the north side of Great George Street. The ICE sold this north-side building in 1908 for E40 OOO; it purchased the new site at the corner with Storey’s Gate for E15 000. It expected that the cost of the new building would be €1 10 000, although the out-turn cost at the time of occupation in 1913 was nearer €150 000.

7.2

PALACE OF WESTMINSTER Too much history is associated with the Palaces of Westminster, the Houses of Parliament, to attempt an outline here. This section, therefore, gives a few facts about the Palace, but mainly the Clock Tower. It is mainly because of J M W Turner’s brilliant watercolour drawings and oil paintings that we recollect that the Houses of Parliament were largely destroyed by fire in 1834. The replacementbuildings were designed in a Gothic style by Charles Barry and Augustus Pugin, who won the competition for its design. Construction started in 1837 and the House of Lords was opened in 1847, although it was not wholly completed until the 1860s, the direction of the work being taken over by Barry’s son.

98

Building response to tunnelling

The Clock Tower, which was completed in 1858, is made from Cornish granite, Caen stone (limestone) and Anstone stone from Yorkshire. There are 334 steps that spiral up the 3 15 ft-high tower. It is the bell that is called Big Ben. Its name comes either from a famous Victorian heavyweight boxer, Benjamin Caunt, or from Sir Benjamin Hall, Commissioner of Works. The bell is 2.7 m in diameter, 2.2 m high and weighs 13.5 t. More details of the structure are given in Chapter 28.

7.3

WATERLOO

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Like London Bridge station, Waterloo is a hub of London’s communications systems. The main line and suburban stations of the surface railway, the Bakerloo, Northern, and Waterloo and City lines of the underground railway link readily by Westminster, Hungerford and Waterloo bridges to the City of Westminster and to the Strand and Fleet Street. There are, however, distinct differences between the historical development of this part of Lambeth and the London Bridge part of Southwark. Until well into the beginning of the 19th century, most of the area now considered as Waterloo was cultivated, other than on the riverside, which, according to a 1799 map, was almost entirely given over to timber yards. Thereafter, building was rapid. Networks of roads and estates of terraced houses were built on the land side of Belvedere Road, the riverside retaining the wharves, and yards for timber and building materials. By 1872 the Lying-in Hospital had been built at the junction of York Road and Addington Street (which although long disused still exists), but few other buildings of this age remain. The 1872 map shows the area near the station (Figure 7.4). Major changes since then include the rebuilding of Waterloo station (completed 1922), the construction of County Hall (l922), the clearance and use of the South Bank site for the Festival of Britain in 1951, and the construction of the Shell Centre and other buildings in the early 1960s.

Westminster and Waterloo areas

99

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 7.4

100

The Waterloo area in 1872

Building response to tunnelling

The following notes and material about the development of the area on which Elizabeth House is built were provided by Mr Chris Holland of London Transport. One of the early reasons for expansion of Waterloo station was the construction by the London Necropolis Company of two tracks and a special station, with one platform for coffins and another for mourners, called London Necropolis station. This was so that they could run funeral trains out to their cemetery at Brookwood, in Surrey. The station had three positions over the years, being moved farther west towards York Road.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The rebuilding of the Waterloo main line station complex lasted from about 1880 through to 1922 and has continued recently with construction of the Waterloo International Terminal just behind Elizabeth House. The great Victory Arch was opened by H M Queen Mary in May 1922. The eastern side of York Road (from about Mepham Street or the taxi road that goes into Waterloo station, down to about opposite Chicheley Street) was occupied by offices built towards the end of the 19th century (and development continued into the 20th century) by the London & South Western Railway. When the L&SWR was merged into the Southern Railway in 1923, the York Road offices became an important part of the new company’s set-up. The area may also have been used as a working site for construction of the Waterloo and City line, which opened in 1898. Until Waterloo International was built in the 1990s, a lift on the west side of the station was used to raise and lower rolling stock to and from the W&C line. Towards the back of the site (roughly where Waterloo International now is) an Underground station was constructed and opened in March 1906. There was an entrance from York Road (about where the passageway from York Road through Elizabeth House is today). The station had lifts leading down to the Baker Street and Waterloo Railway (now called the Bakerloo Line). The lifts were themselves replaced by escalators during the 1920s or 1930s, probably when the section of Northern Line between Charing Cross and Kennington opened in September 1926. At the same time, the interchange between Northern, Bakerloo and the W&C lines with the main line station was also improved. The top-level boarded-up lift shafts could still be seen when one walked through the passageway between York Road and the Underground station under the main line platforms until it was closed and demolished for Waterloo International. During the Second World War the station was badly bombed, as were the York Road offices, whose site was left derelict for a number of years. It may have been used as a queuing area in conjunction with the Festival of Britain in 1951 (the actual Festival site occupied the western edge of York Road through to the river and up to Waterloo Bridge). There were occasions, however, in the1950s - on Saturdays at holiday times - when the area was also used to marshal queues in and around Waterloo station.

Westminster and Waterloo areas

101

Elizabeth House was built in the early 1960s, at which time the Shell Centre Building on the west side of York Road was also constructed and another entrance to the Underground was erected on the western side of York Road as part of the Shell Centre complex. At this time, the current footbridge was also erected, with a walkway at the same level that went straight into Waterloo Station. The tower of the Shell Centre, when built the tallest in London, is supported on a raft approximately 23 m x 50 m and 1.35 m thick at a level of about 15 below the ground surface. The columns are on a grid of 7 m square and beneath them under the raft are 1.8 m-diameter, 24 m-deep piles whose bases are enlarged to 4.5 m (Williams, 1957).

7.4

REFERENCES BURLAND, J B AND KALRA, C J (1986). Queen Elizabeth I1 conference centre: geotechnical aspects. Proc instn Civ Engrs, Part 1, vol 80, pp 1479-1503

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

BURLAND, J B AND KALRA, C J (1988). Closure to discussion on “Queen Elizabeth I1 conference centre”. Proc Znstn Civ Engrs, Part 1, vol 84, Feb, p 121 HALFIN, A J (1 988). Contribution to discussion on “Queen Elizabeth I1 conference centre”. Proc Instn Civ Engrs, Part 1, vol 84, Feb, 98 WILLIAMS, G M J (1957). Design of the foundation of the Shell Building, London. Proc 4th int Conf Soil Mechanics and Foundation Engineering, Buttenvorths, London, vol I, pp 457-46 1

102

Building response to tunnelling

8

The London Bridge station area

D Riley

8.1

INTRODUCTION

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The new LUL Jubilee Line station that opened at London Bridge in 1999 is only the latest transport development in an area that has served as an essential link to the capital city for centuries. Developments in travel -by foot, carriage, boat, train and underground railway - have transformed it, often radically. The process is set to continue with the fiture expansion of the Thameslink railway. The railways have had the greatest impact on the topography and development of London Bridge, but it has always been an important centre of communications.

8.2

EARLY HISTORY Over the past 3000 years the topology of the area around London Bridge has changed, and the level of the River Thames is now some 4.5 m (15 ft) higher relative to the land than in Roman times. This is due in part to the tilt of the south-east landmass. More particularly, it is the result of land reclamation by embankment and containment of the Thames by the construction of river walls - a continuing process. The river was much wider, shallower and slower-flowing than it is today and many areas of land to the south were marshy and waterlogged. Nevertheless, this marshy area was inhabited before the Roman occupation, and archaeological investigations have revealed evidence of settlements in the London Bridge area since Neolithic times. Remains c 1000 BC have been uncovered in Southwark Street, and a Bronze Age ring ditch found at the site of Fenning’s Wharf near London Bridge. Later Iron Age remains (c 600-800 BC) have also been found in St Thomas Street and Southwark Street.

8.3

ROMAN LONDON BRIDGE It was the third Roman invasion by Aulus Plautius in 43 AD that brought the first major settlement to London Bridge, on the north and south sides of the river. The Thames was only tidal as far as Chelsea, and was fordable at Thorney Island (Westminster), so this was the farthest upstream location for cargo arriving by sea. While much traffic would have been by boat, the Romans appear to have constructed the earliest bridge to connect the two settlements. Much of Southwark was marshy, so drainage channels and revetments were constructed to reclaim land, particularly at St Thomas Street, Joiner Street and at Guy’s Hospital. The settlement at London Bridge would have been little more at first than a garrison, but it contained temples and villas, and it is thought that by 65 AD a wealthy merchant class had settled there. The north (Londinium) appears to have been the civic centre, whereas the south side had strategic military importance as the junction of two major highways, Watling Street from Dover and the Channel ports, and Stane Street from Chichester and the Sussex coast.

Ch 8 The London Bridge station area

103

The remains of a major Roman military building and a villa have been excavated near Southwark Cathedral. A substantial riverside building (near the south approach to the current London Bridge on the site of the 15th-century palace of the Bishop of Winchester) contained at least seven rooms and a bath-house. It was in use c 150-330 AD, and a marble slab records the names of several soldiers who had lived there during the early part of the second century. Works for the Jubilee Line Extension uncovered an ornate mosaic floor in vaults beneath London Bridge station, and excavations at Redcross Way exposed timber-lined wells and burials indicative of a late Roman cemetery. Archaeological investigations beneath Borough High Street found evidence of part of the Roman high street. The remains indicated a row of narrow-fronted properties built of timber, wattle and daub, with thatch or tiled roofs incorporating individual shop units to the frontage with a covered arcade on to the street. These dense plot ratios with narrow buildings extended to the rear, continued through to medieval times. Indeed, some remain to this day.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The archaeology also revealed that the high street was burnt to the ground, probably during a raid by Boudicca (Boadicea), who laid waste to several Roman settlements. She would probably have destroyed the bridge as well. The Romans finally left Britain c 410 AD, and with their departure the bridge would not have lasted long. The walled city of London was abandoned for several hundred years with the Saxons preferring to settle further upstream near the present day Strand, although there is evidence of some Saxon activity in Southwark.

8.4

THE DARK AGES During much of this period, the River Thames became a barrier between the Jutes to the north and the Saxons to the south. The Roman city was largely abandoned until 886 AD when Alfred the Great created a defence point here to repel invaders. It is thought that the name Southwark derives from a term used to describe a fortification of about this era - a timber and earth reinforcement known as the “southern bulwark”. This fortification may have been a timber-reinforced earthwork situated near the river crossing, but no trace of it has yet been found. What is now Borough High Street was formerly known as Long Southwark. The first post-Roman bridge is thought to have been built c 944 AD by the Saxons, when Southwark was a market town. This timber bridge was rebuilt at least five times to repair fire, storm and tidal damage, and a report of 1010 tells of its apparent destruction through warfare. The Danes had occupied London and were defending it at the bridge, thus barring the way to Aethelred, who was trying to recapture the City. His ally, the Norwegian King Olav 11, proposed that they tie the long-ships to the bridge supports with rope and pull it down by rowing away. It is not known if he carried out this threat, but Olav’s legacy remains to this day with St Olave’s (House) and Tooley Street, which derive from his name. The first written record of Southwark appears in the 10th century in the Burghal Hiduge as Suthringu geowurche (“the defensive work of the men of Surrey”). By this time, it had become a royal borough with its own market and fair, and had acquired a reputation for gambling, drinking and wild behaviour.

104

Building response to tunnelling

8.5

THE NORMAN INVASION Following the defeat of King Harold at Hastings, William marched on London, but was held back at London Bridge by Aethreling, the last Saxon claimant to the throne, supported by Earls Edwin and Edgar, the Bishop of York and by London’s citizens. The Normans went on to attack London from the west and were eventually successful in capturing the city. Recent archaeological investigations near the Fleet indicate that many were killed in this fierce battle. The Normans took their revenge on Southwark by burning it to the ground. London’s need for good, safe road communication with the south and the coast, and the Normans’ own initial defeat, had demonstrated how vital were London Bridge and the lands adjacent to the defence of London.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

8.6

LONDON BRIDGE In 1176, Henry I1 instructed that a more permanent stone bridge be constructed and he appointed Peter of Colechurch to supervise the works. It was completed in 1209. This bridge was 20 ft wide and comprised 20 arches of varying span - many were given names - with immense piers supported on revetments or “starlings” built up from the river bed. The eighth arch from the south side - Draw Lock - incorporated a drawbridge, which provided protection to the city and allowed tall ships to pass through. The murder of Thomas a Becket in 1 170 had left Henry I1 with a need mitigate his involvement. Recognising the prime role played by London Bridge in the pilgrimages to Canterbury, Henry assuaged his conscience by building a chapel dedicated to St Thomas at the centre of the bridge. It was here that Peter of Colechurch was interred in 1204 before the bridge’s completion. This chapel generated its own income in the form of endowments and the increased traffic from pilgrims. The financial upkeep of the bridge provoked recurring disputes until 1281, when major damage resulted in the destruction of five of the arches. This was resolved by the creation of a trust fund from the income from tolls, rent from buildings on the bridge, and from endowments. In 1282, the Bridge-House Estate was formed and London Bridge became the responsibility of the City. Further to protect the bridge, and thus London, the City also bought the land immediately at the bridge foot - Guildable Manor - at the same time. More land was acquired in 1550 when the adjoining Manors of Kings and Great Liberty were purchased by the City. In 1305 Edward I granted a charter that detailed the tolls to be levied on pedestrians, livestock, carts, and the ships that passed through the drawbridge. This income was insufficient, however, so leases were created to permit shops to be built on each side of the bridge. This resulted in the very distinctive appearance of the bridge for, by 1358, there were 139 shops and houses - many of two storeys - providing rental income, but also reducing the road width to just 12 ft. To minimise this loss, the houses were built jutting out over the water, and overcrowding and congestion must have been commonplace from the outset. Guildable Manor included the Bridge-House Yard, off Tooley Street, from where the bridge masters operated. The Yard held materials stores, timber and masonry workshops, the City granary, a brew house, a lock-up and courtroom, as well as official residences. Their main responsibility was to maintain the bridge, but the bridge masters also hired out plant and equipment, and undertook building contracts for third parties including the construction of whole houses in kit form. The Bridge-House Yard was one of the largest civilian industrial complexes in the country. It moved to Guildhall in 1837 and the site was redeveloped.

Ch 8 The London Bridge station area

105

8.7

MEDIEVAL SOUTHWARK Unlike the City of London, the south shore was not under such pressure to reclaim land. Recent excavations at Bankside indicate only sporadic reclamation, principally in the mid-14th and mid-I 6th centuries. Maps of this period record a rural environment with market gardens, orchards, and fishponds at Pike Gardens. The houses and shops along Long Southwark (Borough High Street) had market gardens to the rear and supplied fresh food. The settlement remained relatively rural in contrast to the City, which, being still contained within the fortified city walls, was overcrowded.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The City’s cramped conditions are presumably why several important people had homes on the south side of London Bridge - close to the commercial centre, but in a more amenable setting. There had been important religious buildings here since the 10th century with the arrival of the Cluniac monks, and in 1086 the Domesday Book recorded 5 1 houses, a monasterium, a dock and a herring fishery. By 1295, Southwark had two members of Parliament - a privilege shared by no other settlement outside the City of London. The Augustinians had built their priory in 1086 on the site of the Saxon nunnery of St Mary Overie. The original church was destroyed in the first Southwark Fire of 1213 and was rebuilt and united with the hospital of St Mary, a new foundation endowed by the Bishop Peter de Rupilous. During the Reformation it was given to Viscount Montague (hence Montague Close) and made into the parish church. The only remaining parts dating from this time are the Retrochoir and part of the walls to the north transept. The south transept and lower part of the tower are 14th century, the upper part of the tower was rebuilt in the 16th century, and the pinnacles erected in 1689. It was this foundation which led to the creation of a separate church with hospital, almshouse, orphanage and school dedicated to St Thomas. This complex of buildings, which became St Thomas’s Hospital (at the site of the present St Thomas Street and London Bridge station), developed as a series of courtyards and incorporated wards, a pharmacy, and rooms where crude operations were performed. It was funded by the priory and by donations from merchant benefactors, and from the pilgrims for Canterbury who started their journey from the nearby coaching inns. One of the most important medieval buildings was the townhouse of the Prior of Lewes, now beneath London Bridge station. During the Reformation it was confiscated, and by 1545 it had become the Walnut Tree Inn. Its vaulted undercrofts were uncovered during demolition for London Bridge station in 1832. Another manor house nearby was owned by the Earls of Warenne and Surrey, which apparently survived largely intact and was in use as the Queen Elizabeth School until it too was demolished for the new railway. On the “Southwarick Surry” map of 1542 with several other important buildings of the time, there are coaching inns (identified by name), several of which survive today even though the properties have been rebuilt.

106

Building response to tunnelling

8.8

TRESWELL’S SURVEYS

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Medieval gardens in the City of London had all but disappeared by the time Ralph Treswell carried out his surveys between I585 and 1616. In contrast, Southwark dwellings had open spaces and gardens. At 46 Blackman Street (site of present 291-299 Borough High Street) the lease of 1585 letting it from Christ’s Hospital to Gilbert Appleton required him to rebuild it of oak, to a height of two and a half storeys plus the roof. All the dwellings contained two principal rooms and a kitchen and privy at the rear. Staircases were located centrally although the location varies, and the frontage of buildings varied from 7 ft 6 in to 12 ft in width, although 11 ft was more common. The numbering of buildings on Borough High Street today still reflects this plot-width pattern, even where the properties have been rebuilt and combined. The buildings demolished for the new Borough High Street ticket hall, for example, nos 3 1 through to 37, had a total street frontage ofjust 39 ft, which can also be seen in Treswell’s plan. Stone walls and brickwork were uncommon, although brick was used for chimneys, wells, ovens and garden walls. Most buildings were timber-framed and jettied (ie with a projecting-out upper storey) with lath and plaster infill. The commonest plan type shows a shop, warehouse or tavern on the ground floor with tenants on the upper storeys; cellars and roof spaces were used for storage. Even the smallest houses had at least one chimney, and several had heating in each room - an unusual feature in other cities. Roofs were usually pitched, but some had a flat roof at the front - as can still be seen at 53 Borough High Street. By the late 16th century there was a comprehensive system of street water pumps with conduits for domestic supplies. Remains of these were found during the JLE archaeological dig on Borough High Street. Wells were also still used, typically sited in common alleyways and shared by groups of dwellings

8.9

OVERCROWD ING AND EXPANS 1 0N The population of London was expanding as people arrived from the country attracted by work and immigrants fleeing religious persecution on the continent. By 1580, the foreign population had doubled to 40 000 within 13 years, and there was a proclamation prohibiting any new buildings within three miles of the City Gates - which included Southwark. Similar legislation was repeated over the following century, to no avail. Fire and disease were the principal hazards of overcrowding. There were two major outbreaks of plague in 1603 and 1625 before the devastation of the 1665 outbreak, which was in turn followed by the Great Fire of 1666. The buildings on the north side of London Bridge had been destroyed by a previous fire in 1632 and were not replaced. This acted as a fire-break, saving both the bridge and Southwark from destruction in 1666. Nevertheless, Southwark did not escape having its own Great Fire just ten years later, on 26 May 1676, which destroyed the Meat Market, the Borough Compter, many of the coaching inns and more than 500 homes.

Ch 8 The London Bridge station area

107

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 8.1

108

Extract from John Rocque map, 1747

Building response to tunnelling

8.10

THE JOHN ROCQUE MAP The map published by John Rocque in 1747 (Figure 8.1) was a badly needed update from the last survey carried out for Morgan’s map, which was originally published in 1692. At this time, the population of the City, Southwark and Westminster formed the largest conurbation in the world. The fieldwork was carried out between 1739 and 1746. New techniques were employed to make the survey as accurate as possible. Surveyors took bearings from landmark buildings using a theodolite and measured angles at street corners. The streets and courtyards were physically measured using chains. This was a mammoth undertaking, and unsurprisingly there were errors, but when it was published after seven years of fieldwork and checking, it was the most detailed map of its period. It is not wholly accurate - some small alleyways are missing, perhaps unintentionally, because they were too difficult to incorporate, or too dangerous to survey. (For a greater level of detail, Ogilvy and Morgan’s map of 1676 is more accurate.)

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

8.1 1

INDUSTRIAL REVOLUTION AND URBAN EXPANSION The Georgian and Victorian industrialists swept away the last remnants of rural life in Southwark by constructing new roads and railways, destroying most of the remaining medieval and Tudor structures. It is difficult today to appreciate just how stark is the contrast between London now and as it was at the turn of the 19th century. In 1802 the River Thames was still largely unembanked and was described by Wordsworth as “All bright and glittering in the smokeless air”. It was still possible for the City’s populace to walk into the countryside, and three windmills were visible from the Strand. Yet by 1841 the population of London had reached over two and a quarter million people, 14 per cent of the total population of England and Wales. By 1850, this explosive growth in population and industry had created an unhealthy combination of mist, fog, coal dust and fumes, which polluted the air. The American novelist Henry James, arriving in 1868, wrote: “the low black houses were as inanimate as so many rows of coal scuttles, save where at frequent corners, from a gin shop, there was a flare of light more brutal still than the darkness”. The first significant change was perhaps the destruction of Old London Bridge in 1832. As always, overcrowding was the problem, and as London’s wealth grew Southwark was transformed into a suburb of the City. Despite the construction of other river crossings (Westminster Bridge was completed in 1752, Blackfriars Bridge in 1769) and the removal of all the shops and houses on London Bridge between 1758 and 1762, traffic congestion remained severe. Tolls had been reintroduced (after being abandoned in the 14th century), but were dropped again in 1782 as they had little effect on reducing the congestion. It was eventually decided to build a new bridge. Construction began in 1825 and was finished in 1832, when the old bridge was dismantled. To clear the approaches to the new bridge, more than E1.5 million-worth of property was purchased and demolished, displacing thousands of mainly poor people. The once modest rural homes in Gravel Lane were becoming surrounded by factories and there was massive overcrowding. The 1850 census reveals that No 9 housed eight families and 16 individuals, 43 people in all.

Ch 8 The London Bridge station area

109

8.12

THE RAILWAYS The development of the railways caused the single most radical and rapid change to London topography. The impact of this new transport was felt at London Bridge with the opening of its first railway terminus in 1836. The route, originally to Deptford, was extended to Greenwich by 1838.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

All the railway developments had the immediate effect of cutting great swaths through existing urban and rural populations. One consequence was the displacement of thousands of mainly poor people (more than 20 000 by the London and Birmingham Railway alone), who could not afford the suburbs being built along the new routes. Many of the poorest who had lived by the riverside became displaced as buildings were cleared wholesale, just as had happened with the rebuilding of London Bridge 30 years previously. Near London Bridge station, even the grand limestone buildings of St Thomas’s Hospital’s new courtyard development, built between 1820 and 1840, could not survive (Figure 8.3). Further expansion of London Bridge station occurred in 1 8 4 3 4 as more railway companies began to use and enlarge the facilities. By 1852, the break-up of the St Thomas’s Hospital site had begun with the sale of the land at Joiner Street to the Brighton railway company. Four companies sought London termini, each requiring a separate Act of Parliament for new railway bridges over the Thames. The expansion from London Bridge into central London required the purchase and demolition of firther portions of the hospital in 1856, and by July 1862 the hospital had moved to a new site at Lambeth. Most of the hospital buildings were demolished, and the church was deconsecrated in 1896. Only the south-west portion, at 9-1 9 St Thomas Street and 19a Borough High Street, remained. Thus south and east London, Kent and parts of Sussex were linked from London Bridge to the new stations at Charing Cross and Cannon Street, which opened in 1866. What is known today as Mary Sheridan House in St Thomas Street was used as railway offices (OS maps of 1872, Figure 8.2, and 1894). Further expansion followed in 1894-1 9 16 with the completion of additional links to the railway terminus at Waterloo. Remaining hospital land adjacent to the surviving south block of the hospital was redeveloped in 19 15 for another form of communication - the south block became the post office and a telephone exchange was built on the site to the north. The railways stimulated rapid development of the suburbs, enabling the better-off to escape both the capital’s crowded, polluted conditions and its poor. Public transport was still limited to those who could afford it, so increasingly the poor were confined to the inner areas of the city in the worst housing. London’s original settlements either side of London Bridge were rapidly depopulated, and the decline seemed complete.

110

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 8.2

Ch 8 The London Bridge station area

The London Bridge area in 1872

111

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

8.1 3

THE UNDERGROUND Charles Pearson, Surveyor to the City of London, proposed a radical solution to the congestion of traffic above ground: transporting people beneath it. Having overcome initial fears that such an idea was against the laws of nature, or that it would help the French to invade, the North Metropolitan Railway opened in 1863 beneath the streets between Paddington and the City. This was built by the “cut-and-cover” method: the ground was dug out, the tunnel lining constructed in brick, the trench backfilled and the road surface replaced. This revolution in public transportation made its debut at London Bridge with its tunnels constructed entirely underground using mining techniques. The underground railway arrived at London Bridge in the form of the City and South London Subway’s cablehauled line from King William Street to Stockwell. Opened on 4 November 1890 by the Prince of Wales (later King Edward VII), it was renamed the City and South London Railway and converted to more reliable electric traction. From 25 February 1900, the CSLR abandoned the sharply curved, steeply graded King William Street in favour of a more northerly terminus at Moorgate, reached by new tunnels from London Bridge. The Moorgate-Stockwell route eventually became part of the Northern Line’s City Branch.

112

Building response to tunnelling

The world’s first tunnelled underground railway was actually a little farther to the east, connecting Tower Hill with the warehouses and docks off Tooley Street. The Tower Subway opened on 2 August 1870 and ran from Lower Thames Street in the City to Vine Street, near Pickle Herring Street (the site of London’s new Mayoral Assembly). Lifts from street level took passengers 5 1 ft (1 5.5 m) down to the platform, where they boarded a narrow carriage (the gauge was just 2 ft 6 in - 0.76 m). This was then pulled by cables via winding engines at each end. Operating difficulties caused it to close after only four months. It enjoyed a second lease of life as a foot-tunnel, accessed by staircases, until the opening of Tower Bridge on 30 June 1894 rendered it redundant. The London Hydraulic Power CO then took it over as a service tunnel for pipework. By 1920, the present basic framework of the London Underground system was in place, most having been built between 1890 and 1907.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

8.14

THE TWENTIETH CENTURY The Second World War transformed the area yet again, as London Bridge and dockland warehouses were prime targets for German bombing. Records of the direct and indirect hits of incendiaries and other bombs show how much of the railway station and surrounding streets were destroyed. Remarkably, the Chapter House, the remaining hospital church of St Thomas’s, survived. The worst single incident, on 17 April 1941, was a direct hit on an arch beneath London Bridge Station Hotel, which was being used as an air raid shelter. The water main in the adjacent St Thomas Street was hit in the same attack and of the 63 people killed, several were drowned. The railway station was extensively damaged and the hotel was destroyed. The extension of the Jubilee Line into south London has brought prosperity to the landowners and businesses of London Bridge. The improved transport and interchange links to the rest of the Underground network and within the station have stimulated new developments. Old hop warehouses have been converted into penthouse apartments; derelict railway arches have become a wine museum. The once declining Borough Market has found a new lease of life as a specialist food outlet hosting weekly food fairs; the warehouse of Neal’s Yard Dairy now sells more cheese than its Covent Garden headquarters. Europe’s largest office block is planned to replace the current block on the site of the London Bridge Station Hotel. London’s new Mayoral Assembly is to make its home at Tooley Street, overlooking the river, on the site of former dockside warehouses and terraced slum housing. At the start of the 2 1st century, as so often before, the Southwark side of London Bridge is in transition. The reason is the same, too: because it is the southern gateway to the City of London and the open door to the south and east. Through the hub of London Bridge, the Jubilee Line Extension links West End and East End, north and south London, and symbolically brings the old financial centre, the City, to the new one of Canary Wharf.

Ch 8 The London Bridge station area

113

8.15

SELECT BIBLIOGRAPHY CHAPMAN, H (comp) (1986). Discoveries, Museum of London CONNOR, J E (2000). Abandoned Stations on London ’s Underground - a photographic record, Connor and Butler Ltd (ISBN 0 947699 30 9) FISHER, J (1979). Notes in: The A to Z ofElizabethan London (A Prockter and R Taylor, comp), London Topographical Society HYDE, R (1982). Note in: The A to Z of Georgian London, London Topographical Society MAYHEW, H (1 861). Mayhew’s London (P Quennell, ed), Bracken Books, 1984, first published 1861

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

PHILLIPS, J F C (1976). Shepherd’s London, Book Club Associates RASMUSSEN, S E (1961). London ;the Unique City, Penguin 1961, first published Jonathan Cape 1937 SANCTUARY, G (1 994). Shakespeare’s Globe Theatre, Shakespeare Globe Trust SCHOFIELD, J (ed) (1987). The London Surveys of Ralph Treswell, London Topographical Society

114

Building response to tunnelling

9

Bermondsey and Rotherhithe

G Johnston 9.1

INTRODUCTlON

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Bermondsey and Rotherhithe are in the present Borough of Southwark on the south bank of the River Thames, and form much of the borough’s river frontage of 4.5 miles. The borough was amalgamated in 1965 from the former Metropolitan Boroughs of Southwark, Bermondsey and Camberwell. These in turn were amalgamations of ancient civil parishes, which included St Mary’s Rotherhithe. However, Rotherhithe did not appear in the Domesday Book in 1085, when it was presumably considered part of Bermondsey. From Tudor times until 1900, Bermondsey was a separate unit of local government as a civil parish run by a body called the Vestry. In 1900, Rotherhithe was again merged with Bermondsey as the Metropolitan Borough of Bermondsey. Throughout their history Bermondsey and Rotherhithe have had distinctive characteristics: Bermondsey has been noted for industry, particularly tanning and food processing, while Rotherhithe has been closely linked with the sea - seafaring, shipbreaking and ship-repairing.

9.2

TOPOGRAPHY The riverine geomorphology of the area is an important factor in the development of this part of Southwark, influencing the pattern of settlement. In prehistoric times the natural landscape was very different. The Thames was then much wider - up to half a mile across - and the area up to a mile from the river’s edge was a large, undrained, fertile flood plain. This large marshy area was broken by a number of small gravel islands. There were three islands roughly along the line of Borough High Street. Another lay to the east of the present-day London Bridge station, once known as Horselydown (founded in 1733, the parish of St John Horselydown appears on the 1872 and 1894 OS maps), and there was the island in Bermondsey. There were other islands at the present Elephant and Castle and at St George’s Circus. These islands were crucial to the development of this area of London. The area bordering the Thames from Rotherhithe downwards differs considerably from the river gravels of Bermondsey, although gravel is usually found beneath it at some depth from the surface. On the alluvial plain, which would often have been under water (the river extended without check during floods and high tides), there were marshy places traversed by creeks and dotted with stagnant pools and brackish lagoons. Here, as the accumulation of mud banks continued, some would become a little dryer than the rest of the plain, trees would take root on them and they would be covered by a marshy vegetation - home to wild animals and birds, but not often of men. These woods and marshlands were then frequented by elk, mammoth, deer, horse, boar, beaver, bison, bear and wolf. In 1875 during the excavations for Canada Dock, many bones of these animals were found in the alluvium.

Ch 9 Bermondsey and Rotherhithe

115

9.3

BERMONDSEY From the early Middle Ages, Bermondsey constituted a separate manor in terms of landownership. Bermondsey’s principal feature, Bermondsey Abbey, founded in 1082, occupied a site on a low gravel island surrounded by marsh - the “ey” in Bermondsey refers to this island. The abbey was an important centre of learning, but it was also responsible for the draining of the marsh and maintenance of a river wall. This and other abbeys were dissolved by Henry VI11 in 1538. Almost nothing remains of its fabric, but it is thought that Bermondsey Square could be the site of the inner courtyard. The church of St Mary Magdalene, however, which originated as the church for the Abbey’s lay workers, contains parts that date back to the 15th century.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Until well into the 19th century, Bermondsey’s development was essentially close to the river, although urbanisation was spreading outwards to the east and south-east from the area of Southwark around London Bridge. London’s first railway was opened in 1836 between Bermondsey and Deptford. It was built on viaduct in a straight line, cutting across “gardens and meadow land”. Quite soon the route was taken in to London Bridge and out to Greenwich. That route, with many more tracks but still on viaduct (part of which can be seen on Figure 9. I), dominates this part of Bermondsey to this day.

Figure 9.1

116

Map of part of Bermondsey in 1875

Building response to tunnelling

By the third quarter of the century, there were large estates of terraced houses with gardens, often surrounding local works, such as timber yards, tanneries, and glue and size factories. Near the river, the buildings and works were still linked to the shipping industry as, for example, warehouses and a ropewalk. Increasingly, however, there was more traffic along Jamaica Road to and from the docks of Rotherhithe. This road remains a very busy thoroughfare and link to south-east London and north Kent.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

While along the route of the Jubilee Line Extension there remain some older buildings, such as those in Old Jamaica Road and the elegant St James’s Church, the redevelopment of the area has been almost total. Partly the result of a deliberate policy of social re-housing, particularly in this part of London, it was also stimulated by the need to repair war damage. A legacy of old features remains in the street names, such as Neckinger Road (after the old river), Spa Road (after a former health spa), and of the streets around the first residential estates, eg Major Road, Keeton’s Road. These can be seen on the 1875 OS map (Figure 9. l), which includes the area of the case studies at 128-130 Old Jamaica Road and Keeton’s Estate, near the new Bermondsey JLE station and the surface reference site in Old Jamaica Road.

9.3

PARISH OF ST MARY’S ROTHERHITHE St Mary’s Street marks the centre of the medieval village of Rotherhithe. Some or all of the settlement stood on ground that was relatively high, but the rest of the district was low-lying - a patchwork of marsh and pasture intersected by drainage ditches that were liable to flood. Rotherhithe is almost a peninsula, bounded on the east by Limehouse Reach and on the north by the Lower Pool. The land side hinges on Jamaica Road, which leads into Bermondsey, and on Lower Road, which goes to Deptford. Rotherhithe Street follows the peninsula’s waterfront for more than two miles and, after the historic heart of the parish, was the first district to be developed. Parts of Rotherhithe Street were once called Rotherhithe Wall or Shipwright’s Street, Trinity Street, Queen Street and Lavender Street. The south-eastem continuation of Rotherhithe Street, Redrith Road, maintains the Saxon name for Rotherhithe and was widely used in the 17th and 18th centuries. Samuel Pepys wrote of going “to Redrith by water and from thence walked over the fields to Deptford with my wife and maid agathering of cowslips”. Rotherhithe Street and Redrith Road provided access to the shipyards and to the pioneering wet dock, known first as the Howland Great Wet Dock. Built in 1700 and enlarged later towards the end of the 18th century when the whaling trade grew, the dock was renamed Greenland Dock. In 1807, the Commercial Dock Company was formed and new docks were built in 1820, increasing the area of docks to 50 acres. The Surrey Canal Company had also been engaged in the construction of new basins, inner docks and timber ponds and in 1864 amalgamated with the Surrey Commercial Dock Company. By the end of the 19th century, the parish of St Mary Rotherhithe extended over 754 acres, of which 360 acres were occupied by the Surrey Commercial Docks and a further 60 acres by Southwark Park. Most of the parish lies between 1.2 m and 2.1 m below Trinity High Water, and so has suffered from flooding for centuries. St Mary’s Church was rebuilt by John James in 1714-15, partly because of the effects of the flood in 1705. Local people still remember the floods of 1928 and 1953.

Ch 9 Bermondsey and Rotherhithe

117

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

From the beginning of the 19th century, the western boundary of Rotherhithe was clearly defined by the mill stream that ran into the Thames at West Lane and flowed along the eastern side of the present Southwark Park Road. The historic route into Rotherhithe from Bermondsey crossed the mill stream at Mill Pond Bridge. It did not run along what is now Jamaica Road as one might have supposed, but turned left into West Lane and along Mill Pond Bridge over the stream into Paradise Street. In the 18th century, the built-up area went no farther south than Paradise Street. The name “Paradise” often meant an enclosure, and it is possible that its use at Rotherhithe derives from the enclosure of King Edward 111’s moated manor house, which was built in the 14th century between Paradise Street and the river. In 1914, the gate from Union Road (formerly Paradise Row) in the north-west corner of Southwark Park still had the name Paradise Gate (see Figure 9.2).

Figure 9.2

Map of parf of Rotherhithe in 1914

The mill stream flowed alongside Southwark Park Road - once called Jamaica Level towards Galleywall Road. To the east of Southwark Park Road, on the Rotherhithe side, there was a labyrinth of streams, ponds and islands. This district of market gardens was known as Seven Islands (Figure 9.3), although there were more than seven islands between the streams. The water from these streams was used to work the flour mill. The miller had the right to cast mud from the pond on to the gardens at any time (without notice) between 25 October and 25 March. He also had the key to every floodgate so that his men could always open them. As a consequence, the gardens were liable to be flooded at any time without notice.

118

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 9.3

The Seven Islands area in 1827, covered today by Southwark Park

Bermondsey and Rotherhithe grew rapidly from the beginning of the 19th century. The former became intensely industrial and urban, and the latter was dominated by docks and wharves that filled the inside of the sweeping bend of the Thames. The inevitable consequences of industrialisation were a rise in population and high-density housing. By the early years of Victoria’s reign many of the grand middle class terraces had succumbed to subdivision, poor maintenance and the addition of new structures in previously spacious front gardens. It became apparent that the traditional two-up twodown terraces were inadequate to meet the demand for new housing. Accordingly, developers started to erect taller buildings: three-storied houses became common, as did tenement buildings. By the middle years of the century, Seven Islands remained the only area of open space, still used for market gardening, with ditches running across it and a series of mill ponds. In 1857, following a deputation from Rotherhithe Vestry, it was concluded that Bermondsey and Rotherhithe were the areas in greatest need of a public park. Ten years later, work began on forming a park from the Seven Islands area and two years later it opened to the public (see Section 9.6).

Ch 9 Bermondsey and Rotherhithe

9.4

THE THAMES TUNNEL AND THE EAST LONDON LINE

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

No book that concerns tunnelling can refer to Rotherhithe without reference to the Thames Tunnel, for this was where the start was made to this historic and daunting achievement of the two Brunels. The tunnel crossing was authorised by Act of Parliament in 1824. Isambard Kingdom Brunel (1 8 0 6 1859), as the resident engineer, completed the work planned and supervised by his father, Sir Marc Brunel (1 769-1 849). Work started in March 1825 when a shaft 42 ft in diameter was sunk at Rotherhithe. Tunnelling began from that end, using a shield. After many difficulties, tunnelling reached the Wapping shaft on 16 November 1841, and then the whole tunnel had to be tiled and paved. For a fascinating examination of the ground conditions and a comprehensive set of references about the work, see Skempton and Chrimes (1994). The tunnel comprises two arched passages 1200 ft long between Rotherhithe and Wapping. Each passage is 14 ft wide and 16 ft 6 in high and is divided from the other by a wall 4 ft thick punctuated with 64 arched openings. The head of each passage has a minimum cover of 16 ft to the bed of the Thames. The tunnel was formally opened on 25 March 1843 at the cost of about E600 000. Planned to be a road tunnel, it was used solely by pedestrians until the East London Railway purchased it. The route of the tunnel and the position of air shafts, still extant, can be seen on Figure 9.2. The East London Railway COwas incorporated in 1865 and was intended to link the Great Eastern Railway with Brighton and the south-east of England. The company’s line from Whitechapel through the Thames Tunnel to New Cross Gate was opened on 7 December 1869. A station was opened at Rotherhithe, in Brunel Road (formerly Adam Street), in 1905. The next station down the line (and the only other station within Rotherhithe) was originally called Deptford Road. It was renamed Surrey Docks station on 17 July 191 1 and the station has been subsequently changed to Surrey Quays. The Thames Tunnel has recently been refurbished - not without controversy because of its engineering and historical significance - and is still in use. The new JLE station of Canada Water incorporates platforms of the East London Line as well.

9.5

THE INDUSTRY OF ROTHERHITHE Water - the natural resource of the area - was put to use by the Hydraulic Power CO, which built a pumping station in Renforth Street next to the Surrey Docks. The mains went westward from the station towards Bermondsey and also across the river through the Rotherhithe Tunnel. The heyday of hydraulic power was between the wars and was used primarily for power cranes and lifts. The company ceased operations completely in 1977, but the pumping station building has survived and is being converted into flats. Its chimney can be seen in Figure 2.1 1. Although nearly all of Rotherhithe’s economy was devoted to activities associated with the river, there were exceptions - the gas works and Brandram’s paint factory. The gas industry began in Rotherhithe when the Surrey Consumers Gas CO set up its works in 1854 (visible on the 1868 OS map). The site was taken over by the South Metropolitan Gas CO in 1879 and eventually extended to 6.5 acres including a frontage to the Lower Pool. After the Second World War, the capacity of the works rose to 9.6 million cubic feet per day. However, all production ceased in Rotherhithe in 1959.

120

Building response to tunnelling

The South Metropolitan Gas COwas one of the most enlightened industrial employers in south London. In 1926, the gas company built seven houses for workers in Brunel Road, Rotherhithe (destroyed by bombs during the war) and three blocks of flats in Moodkee Street, named Murdoch, Clegg and Neptune - providing 30 flats in all. These flats survived the Blitz with no more than superficial damage. Predictions of the responses of these three buildings to the Jubilee Line Extension tunnelling are recorded in Chapter 15, with the results of monitoring in Chapter 43.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Brandram Brothers and CO occupied the site between the bend of Neptune Street and Canada Dock. The factory can be seen on Horwood’s map of 1813 (Figure 9.4), where it is clearly the only building in an area of meadow land. The firm manufactured white lead and paint, and refined saltpetre and flowers of sulphur (brimstone). The parochial valuation of 1843 distinguishes between the old factory and the new premises to its north. Both sites remained in use until 1958.

Figure 9.4

Ch 9 Bermondsey and Rotherhithe

Brandram Brothers‘ factory on Honuood’s 1813 map

121

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The site of the Brandram Brothers’ factory was redeveloped for housing, which became the Canada Estate. In 1962, the two tower blocks Regina and Columbia Points were built on the site (see Chapter 36 for a note about the development of this estate). There were also several three- and four-storey blocks of flats. A relative of an early estate resident states that they were warned not to eat produce grown in the soil on the Canada Estate because of the toxic nature of the materials used and made in the factory. Chapters 44 to 46 present the results of monitoring the tower blocks and some of the low-rise buildings of the Canada Estate as the JLE tunnels were advanced towards Canada Water station. Chapter 37 includes the measurements of displacements of the ground surface and a low wall in the estate. Figure 9.5 shows the relative plan positions of the Canada Estate buildings, the tunnels and the old factory buildings.

Figure 9.5

Canada Estate, the old Brandram’s Works and the tunnel alignments

Another lead factory in Rotherhithe owned by H T Enthoven & Sons, was sited at Upper Ordnance Wharf and manufactured lead, tin and antimony. The factory closed in 1980. The whole area has been redeveloped since the war following the destruction caused by air-raids. Industry, though battered, soon resumed trading again. By 195 1 the Surrey Docks were working to their pre-war levels and reached a peak two years later. The post-war industrial boom was not to last, however. Gradually industry started to desert Southwark and half its manufacturing jobs were lost. Traditional industries disappeared rapidly: by 1986 only one tannery remained in Bermondsey, the Surrey Docks had closed earlier in 1970, the Surrey Canal at the same time. The borough has not filly recovered from this decline and high levels of unemployment persist.

122

Building response to tunnelling

After the Second World War there was also a desperate need for housing; in 1948, for example, 8 per cent of Bermondsey housing was unfit for habitation. In the period to 1955,9600 new homes were provided in the area of the present borough, representing 11 per cent of all new homes in London. These dwellings were provided in roughly equal numbers by the boroughs and London County Council. Many of them were intended to be temporary, but the prefabs became more enduring and popular than their builders could have imagined. However, the post-war period is most associated with the building of flats, both low- and high-rise, on an overwhelming scale.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Over the past ten years and since the development of the old docks, the Docklands Light Railway and the Jubilee Line Extension, the area has become very popular with young executives who work in the city. Wharfside Warehouses have been converted to provide luxury apartments and, as mentioned above, the Hydraulic Power Company’s Pump House is being converted into luxury flats. Many of the developments that have sprung up in the area are too expensive for the local inhabitants. Property coming on to the market is usually “yuppified” - as the locals put it - such that the housing department finds it cannot provide the homes needed. From being an area of slum clearance, it is now very desirable - a return to Southwark’s 13th-century roots when the area provided houses and palaces for London’s wealthy citizens.

9.6

SOUTHWARK PARK Southwark Park is situated west of the area that used to be the Rotherhithe docklands and a few hundred metres south of the River Thames in the densely populated, highly urban district of Bermondsey. It was opened in 1869 in response to a pressing need for a public park in south-east London. Along with Finsbury Park, it is the earliest park to be opened by the Metropolitan Board of Works. Formation of the park began in 1867 on the land known as Seven Islands. This area had previously been used for market gardens and by the middle of the 19th century constituted the only open space in the district. The ornamental portion of the park was planned by the superintending architect of the Metropolitan Board of Works, Mr Vulliamy, and the work was carried out by Murray Anderson. Although the design has always been attributed to Alexander McKenzie (1 829-1 893), his input was ultimately in the fine-tuning of the basic layout. In his report of 1868, McKenzie admitted that he considered the plan produced by Mr Vulliamy to be “on the whole very satisfactory and would suggest only a few minor alterations to detail”. These alterations, he maintained, would “in no way materially alter the design in principle, but will, when carried out in all cases give breadth and harmony in design”. Alexander McKenzie was an influential and pioneering landscape gardener who preferred a non-architectural and more natural style. The earliest plan of the park (most probably by Vulliamy) illustrates the basis of Southwark Park’s definitive layout. Heavily influenced by Birkenhead Park (designed by Joseph Paxton), the design has a wide carriage drive skirting the perimeter of the site, with a band of land left between the carriage drive and the park boundary earmarked for building purposes. The main entrance to the park is from Jamaica Level, with a lodge indicated on the right, just inside the gate. There are five other subsidiary entrances, ensuring easy access from all surrounding areas. The rest of the park is composed of two areas to the north and south, divided by the only surviving carriage drive, which connects Jamaica Level in the west (now Southwark Park Road) to Gomm Road in the east. Both areas of open space unfold around oval spaces bounded by perimeter footpaths; the largest of these lies to the south in the form of a 9 acre cricket pitch. To the north, two smaller ovals (one of which has paths crossing its interior in the shape of a cross), are connected by a footpath. All the ovals are connected to the carriage drive via further footpaths.

Ch 9 Bermondsey and Rotherhithe

123

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

124

Figure 9.6

Southwark Park on the 1870/1872 OS map

Figure 9.7

Southwark Park on the 1894 OS map

Building response to tunnelling

The 1870 OS map (see Figure 9.6) shows that the park layout stayed faithfil to the Vulliamy and McKenzie’s original design. However, the Rotherhithe Vestry and the local community opposed the planned building plots facing inwards to the park. In 1870, Colonel Beresford MP moved to insert a clause in the Metropolitan Buildings and Management Bill to prevent building in the park and in 1872 the Metropolitan Board of Works dropped the development proposal. That removed the need for the carriageway, so in 1874 it was narrowed to a footpath with a double avenue of trees (see Figure 9.7). By the mid-1870s the principal features of the park were all complete. The following 15 years saw the addition of the gymnasium, drinking fountains, several refreshment pavilions, urinals and the bandstand. The lake, of concrete and puddle clay, was constructed in 1885, and in 1908 it was enlarged and adapted for boating. A lido was added to the east of the lake in 1923. Funds to enhance Southwark Park were approved in 1934, and two years later an English rose garden was created

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The park was significantly affected by the Second World War. Bombs damaged the lake, and the bandstand (bought in 1883 in the Great Exhibition) was taken away, along with many of the railings, as part of the war effort. Bomb shelters were constructed in mounds along the northern edge and artillery was stationed on the cricket oval. In 196 1, the cricket oval was modified and the pavilion removed. Around the same time the boathouse was also removed and a large part of the lake was filled in. The most recent change came in 1975 when the widening of Jamaica Road necessitated the removal of houses along the northern boundary and the new boundary was realigned with Paradise Gate. Southwark Park is the only piece of land in the area which has not been built on at any time and is therefore suitable as a greenfield experimental site (see Chapter 37, which summarises some of the findings from the instrumented section in the park).

9.7

BIBLIOGRAPHY BECK, E J (1907). History of Rotherhithe, Cambridge University Press HUMPHREY, S (1997). The Story of Rotherhithe, London Borough of Southwark KINGWELL, P, HARRIS, A, HARRISON, P and WIKELEY, A (1999). Southwark Park: A brief history, London Borough of Southwark MARTIN, J E (1 966). Greater London on Industrial Geography, G Bell & Sons Ltd MILLS, M (1999). The early East London gas industry and its waste products: how were they used?, M Wright, London, 168 pp REILLY, I (1998) Southwark: an illustrated history, London Borough of Southwark SKEMPTON, A W and CHRIMES, M M (1994). Thames Tunnel: geology, site investigation and geotechnical problems, Giotechnique, vol XLIV, no 2, June, pp 191-216

Ch 9 Berrnondsey and Rotherhithe

125

I0

Tunnelling methods

R J Mair and F M Jardine 10.1

SUMMARY This chapter describes the tunnelling methods that were used for construction of the Jubilee Line Extension. The geology, which varies over the route, was the primary factor dictating the tunnelling method that was adopted for each section of the project. The influence of the ground conditions on the choice of tunnelling technique is presented, and details are given of the different methods.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

10.2

INTRODUCTION Eleven tunnelling machines were used on the JLE project to excavate the 11.1 km of

twin tube running tunnels (ie a total of 22.2 lan of running tunnels). The tunnels run from Green Park to Canning Town, a di&mce of approximately 11.5 km, and these are shown in plan and section on Figure 10.1.

Figure 10.1

Ran and kmg s&bn of tfte JLE rrwte

The Nnningtunnels have a n e x t m d tihmterSf-3 m and typicallyrunat depths ofbetween 20 m d 30 m below ground level. There are nine stations below the ground surfatce, as shorn on Figwe 10.1, fom 6fwMch w ~ f e as large open-cut boxes, and five either as entarged tunnels or as a combination of enlarged tunnels and qm-eut boxes. As ctlll be fixnu F i g m 10.1, the L E crosses the River Thames in fQWplaws*

Ch 10 Tunnelling methods

127

10.3

INFLUENCE OF GROUND CONDITIONS ON TUNNELLING METHOD The geology of the route is described in detail in Chapter 5. It consists principally of Tertiary sedimentary soils overlain by a blanket of Pleistocene gravel and sand, which itself is overlain by recent alluvium and made ground. As can be seen from Figure 10.1, for the western half of the route the tunnels are in London Clay, the youngest of the Tertiary sediments. This is a stiff to very stiff, fissured, silty clay, which is heavily overconsolidated. Its undrained shear strength at the depth of the tunnels is typically in the region of 200 kN/m2. Ever since the first bored tunnels were constructed for London’s underground railways in about 1880, London Clay has been found to be an ideal tunnelling material. It is generally sufficiently strong and impermeable that the excavated tunnel remains stable in the short term while the linings are installed.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

On the eastern half of the route most of the tunnels are in the Lambeth Group (formerly known as the Woolwich and Reading Beds) or the Thanet Sands. The Lambeth Group comprises interbedded layers of stiff clays, silts, silty sands and gravels. The sands and gravels of the Lambeth Group, and the Thanet Sands, are water-bearing and highly permeable. These are complex and difficult ground conditions for tunnel construction. For the western half of the route, where all the tunnelling was in London Clay, openface tunnelling methods were used. For most of the running tunnels, which were of 5 m diameter, tunnelling machines were used with segmental linings. In some instances, the segmental linings were expanded rather than bolted. In a number of cases, the largediameter station tunnels in London Clay were constructed using sprayed concrete linings (SCL - often referred to as New Austrian Tunnelling Method, NATM). This technique was also used for some of the shorter drives of the 5 m-diameter running tunnels. For the eastern half of the route, where most of the tunnels are in water-bearing permeable deposits, closed-face tunnelling machines were used. Both slurry shields and earth pressure balance (EPB) machines were adopted. Table 10.1 lists the 1 1 tunnelling machines from five manufacturers that were used to excavate the 22.2 km of nominal 4.4 m ID running tunnels. Additional tunnelling shields were used for platform enlargement tunnels and other station tunnel excavations on the project. Further details are given in the following sections. Table 10.1

Tunnelling machines used to excavate the twin tube running tunnels

Contract and tunnelling machines

Ground conditions

102 Green Park-Waterloo: Two shield-mounted backhoe machines (Wirth Howden)

London Clay

103 Waterloo-London Bridge: One shield-mounted roadheader boom machine (Dosco)

London Clay

105 London Bridge-Canada Water: Four EPB TBMs (Kawasaki/FCB/Decon)

London Clay, Lambeth Group and Thanet Sands

107 Canada Water-Canary Wharf: Two slurry mix-shield TBMs (Herrenknecht)

Lambeth Group and Thanet Sands

1 10 Canary Wharf-North Greenwich-Canning Town: Two EPB TBMs (Lovat)

Lambeth Group, Thanet Sands and London Clay

The organisation and details of the construction contracts are summarised in Chapter 2 . The following sections describe the tunnelling methods employed on construction contracts for which case studies of buildings or greenfield sites appear in this book.

128

Building response to tunnelling

10.4

TUNNELLING IN LONDON CLAY ON CONTRACT 102

10.4.1

Running tunnels A Howden tunnelling machine was used for the majority of the 2.5 km length of the 4.85 m 0.d. running tunnels between Green Park and Waterloo. The machine was fitted with a telescopic back-hoe, as shown in Figure 10.2. The back-hoe had a full 360-degree rotating action and an extended reach of some 1.5 m ahead of the leading edge of the tunnelling shield. The back-hoe cut the face of the London Clay and fed the spoil on to a belt conveyor. The machine was powered by electric motors with a total installed power of 280 kW and could provide a total force of 16.5 MN for the 15 shove rams.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

ForeDolinq boxes

Trailing fingers

I

/

I I 1 /

I

Figure 10.2

Rotary segment erector

Schematic of Howden back-hoe tunnelling machines on Contract 102

The excavation method with a back-hoe essentially involves an open face with some support from breasting plates. The shield is fitted with five telescopic poling plates and five breasting plates, as shown in Figure 10.2. There is also a tail hood of trailing fingers with a series of non-connected thin metal plates covering the spaces in between. These fingers are slightly proud of the shield tail skin and cut grooves in the clay as the shield is jacked forward off the segmental lining. The majority of the tunnelling with the Howden shield was undertaken with expanded precast concrete linings, which are well-suited for London Clay. The expanded linings for the 4.45 m ID tunnels were made up of ten lightly reinforced concrete 1 m-wide, 200 mm-thick segments, and these were expanded against the London Clay by two concrete wedges at knee level. Over some lengths of tunnel, for example beneath the river wall at Westminster and at particular locations where later break-out for connection adits would take place, bolted precast concrete linings were used. During construction of the first (westbound) running tunnel, very high progress rates were achieved. A maximum of 204 m was tunnelled in a production week of two 1 1.5hour shifts per day, five days a week, with an average rate of about 70 m a week. With an open face tunnelling machine such as this operating in clay, the principal factors influencing ground movements are the following (Mair and Taylor, 1997): 0

overbreak size of overcutting edge, known as the bead (in this case 20 mm)

0

length of shield (which influences the length of unsupported ground, depending on the bead size) the tendency of the machine to plough or yaw

Ch 10 Tunnelling methods

129

0

0

fkquency of grouting behind the linings (only applicable in the case of bolted linings) face boarding at stoppages consolidation of the clay as pore pressures change to their long-term equilibrium values.

The ground properties are also an important factor. Where the London Clay contains significant silt and sand seams, there is more potential for overbreak, which leads to larger ground movements.

10.4.2

Platform tunnels

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The two 7.3 m OD platform tunnels at Westminster, each of approximately 150 m length, were constructed with spheroidal graphite-iron (SGI) linings. They were enlargementsof pilot tunnels formed by construction of the running tunnels with the Howden machine as described in Section 10.4.1. The Wirth Howden tunnelling shields designed for this purpose were unique in that they incorporated mechanisms to allow the expanded pilot tunnel lining to be safely dismantled and muck to be transported through the face along the existing pilot (running) tunnel (Bailey et al, 1999). The method of excavation was a back-hoe as for the running tunnel construction.A view of one of the shields is shown in Figure 10.3.

Figure 10.3

130

Tunnerring shield for enrargement or me pianotm tunnels ar wesrmrnster station

Building response to tunnelling

NATM TUNNELLING IN LONDON CLAY

10.5

The use of sprayed concrete linings (SCL) for tunnelling in soft ground is reviewed in the publication by the Institution of Civil Engineers (ICE, 1996). The SCL process is often referred to as the New Austrian Tunnelling Method (NATM), and either or both these terms are used elsewhere in this book. The SCUNATM technique essentially involves applying sprayed concrete (shotcrete) to a recently exposed surface of clay so as to form a temporary lining. The sprayed concrete has special retarding additives to prevent it hmhardening before it is sprayed. Accelerating agents are added as it is sprayed so that it starts to harden within a short time (typically about 15 minutes) after spraying. Its great merit is that it offers considerable flexibility in choice of tunnel shape - there is no requirement to have a circular cross-section as required for most tunnelling machines.The technique is particularly well suited to London Clay because this ground is d c i e n t l y strong and impermeable to remain temporarily stable without support.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

On the L E , NATM has been mainly used for station constructionat Waterloo and

London Bridge. Further information about its use for London Bridge station (where several buildings were monitored for the research project) is given in Chapter 16. For the JLE project, sprayed concrete was used only for the primrtry (temporary)tunnel lining. A secondary (final) lining usually of cast-in-placereinforced concrete was then constructed. In some cases, the primary sprayed concrete lining was in place for periods of more than a year before the secondary lining was constructed. NATM was also used to construct the 5.6 m OD running tunnels immediately to the west of Waterloo station, together with the crossover tunnel beneath Elizabeth House. This is described in more detail in Chapter 30. Generally, when the final tunnel diameter exceeds 6 m,and when NATM is used, the tunnel face is divided into two halves. This was the case for the crossover tunnel beneath Elizabeth House, for which the maximum diameter of the completed tunnel was 12.4 m. Figure 10.4 illustrates the principle and Figure 10.5 shows a photograph of one half of the crossover under construction beneath Elizabeth House. Excavation sequence

Cross-section TemDorarv --,

.

I Primarv

-

Temporary-' backfill

Ch 10 Tunnelling methods

!

!

!

!

!

!

I

!

'

8-13 m

Figure 10.4

.

w 2m

2m

w Imlm

Use of divided face in NATM construction (a) in transverse section and (b) in longitudinal section to show the excavation sequence

131

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 10.5

Side drift construction for cross-over construction beneath Elizabeth House

Alternatively, a pilot tunnel is constructed first, as was the case for the platform and concourse tunnels at Waterloo and London Bridge stations. At Waterloo, the pilot tunnels were approximately concentric with the final enlarged tunnel, whereas at London Bridge the pilot tunnel was constructed in the invert of the enlarged tunnel, as shown in Figure 10.6 Shotcrt

w --,:---:.

.

of wire mesh

Figure 10.6

10.6

.

.

Concrete cradle for SGI lining

Construction of NA TM pilot tunnel in invert of platform tunnel

HAND MINING At step platejunctions, where the new running tunnels join into existing tunnels, some cross-passages and adits, and for some station tunnel enlargements, the excavations were made by hand, eg with clay spades. Where space allowed, independent small backhoes were used as well. Close-boarded timbering was used to support hand-mined faces and the permanent lining was with bolted SGI segments.

132

Building response to tunnelling

In the London Clay, the hand excavation was in free air, but at Bermondsey station in the Lambeth Group, where the station tunnel enlargements were formed by excavating around the EPBM-bored running tunnels, the work was done under low-pressure compressed air. Here again, the faces were boarded where they were not being worked. Inclined shafts were also excavated by hand. Permeation grouting was used to treat permeable or weak ground and the shaft was underpinned with ring segments as it advanced. The step plate junctions comprised 10, 8, 6.5 and 5.75 m ID sections, all constructed around the existing tunnels - these having to be kept in operation throughout the construction period. The Green Park step plate junction is 90 m long (and can be seen in Figure 2.4). The junctions were lined with bolted SGI segments. In some cases, the linings were segments of ellipses rather than circular. In the larger sections, pilot tunnels were excavated first and lined temporarily. The shells of the existing tunnels were carefully propped while the bolted segments were erected as the permanent lining.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

10.7

TUNNELLING IN THE LAMBETH GROUP AND THANET SANDS Contract 105

The two 2.8 km-long running tunnels between London Bridge and Canada Water stations were constructed with four earth pressure balance (EPB) tunnelling machines. These machines, manufactured by Kawasaki, work on the principle illustrated in Figure 10.7. -5 m 4 4

A

*

Grout behind rings

/

i--wfl__

Plenum chamber,

5

6.925m

..--Earth pressure sensors

concrete segments 4.9 m 0.d. 4.4 m i.d.

I

.03m Cutting bits V

'Screw convevor

c c

I

1 0

Shield jacks

Figure 10.7

Schematic section through Kawasaki EPB tunnelling machine used for running tunnel construction on Contract 105

The tunnelling machine shown in Figure 10.7 is 6.925 m long and is 5.03 m in diameter. The essence of its operation is to have a mixture of soil and water in the plenum chamber under pressure, and to extract it via the screw conveyor, which allows the pressure to drop progressively over the length of the screw. The spoil is then discharged from the end of the screw on to a conveyor belt. For ideal operation the soil in the plenum chamber needs to be a paste with an undrained shear strength of around 25 kPa. Soil conditioning agents (additives to the spoil usually at the point of cutting) are often needed to convert the excavated soil to such a paste, and this can be problematic particularly if the soil being excavated is a high-plasticity stiff clay.

Ch 10 Tunnelling methods

133

Pre-cast concrete lining segments are erected inside the tailskin of the machine, as illustrated in Figure 10.7. As the machine is jacked forward off the most recently completed tunnel ring, grout is injected through holes in the segments to fill the annulus between the outside of the segments and the soil. Details of the factors influencing ground movements associated with closed face tunnelling machines are given by Mair and Taylor (1 997). The following are the primary components of ground movements, some of which also apply to open face tunnelling machines (see Section 10.4.1): deformations of the ground towards the face resulting from any stress relief the tendency of the machine to plough or yaw will lead to ground movements, as will the development of shear stresses between the perimeter of the shield and the ground as the shield is jacked forward 0

the tail void, which is the gap between the tailskin of the shield and the lining, will tend to allow the ground to move radially into this gap

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

consolidation of clays as pore pressures change to their long-term equilibrium values. When closed-face tunnelling methods such as an EPB shield are employed, good control of the face pressure and maintaining it close to the original in-situ stresses in the ground can result in the component of ground movement resulting from stress relief at the face being very small. In such cases, the principal cause of ground movements is usually the tail void, and this can be minimised by immediate tail void grouting. Recent experiences around the world with EPB (and slurry shield) machines in sands and gravels have generally shown small volume losses, typically less than 1 per cent and sometimes less than 0.5 per cent. Volume losses may be higher in mixed face conditions, particularly where sands or gravels overlie stiff clays, or where the cover of competent soil above the tunnel crown is low.

10.8

REFERENCES BAILEY, R P, HARRIS, D I and JENKINS, M M (1999). Design and construction of Westminster station on the Jubilee Line Extension. Proc Instn Civ Engrs, Civ Engg, Jubilee Line Extension, 1999, 132, pp 3 6 4 6 DAVIES, H R (1999). Design and construction of the Jubilee Line Extension tunnels. Proc Instn Civ Engrs, Civ Engg, Jubilee Line Extension, 1999, 132, pp 26-35 MAIR, R J and TAYLOR, R N (1997). Bored tunnelling in the urban environment, State-of-the-art Report and Theme Lecture, Proceedings of 14th International Conference on Soil Mechanics and Foundation Engineering, Hamburg, Balkema, Rotterdam, vol4, pp 2353-2385

134

Building response to tunnelling

11

Protective measures

D I Harris

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

11.1

SUMMARY This chapter reviews the many methods used on the JLEP works to protect buildings from damage caused by tunnelling or open excavations. The review includes examples used in different circumstances on several JLEP contracts. The JLEP made the largest use in the UK of the relatively new technique of compensation grouting. This chapter concentrates on its use on the project, using examples particularly from Contract 102 Green Park to Waterloo, where it had the widest range of use. As well as explaining what is involved in compensation grouting and its associated works, the chapter distinguishes between the forms and purposes of grouting that together make up the protective systems. The description of compensation grouting outlines the factors influencing selection and design, discusses its management and control, emphasising the requirement for effective monitoring and reporting, and presents some of the difficulties that were encountered in its use. Overall, as confirmed by case histories of Volume 2, the technique was very successful, exceeding the expectations of designers.

11.2

REQUIREMENT FOR PROTECTIVE MEASURES

11.2.1

Potential damage assessments The Jubilee Line Extension (JLE) is the single largest tunnel construction undertaken in London for 30 years. Over that period, there have been substantial advances in tunnelling techniques (see Chapter IO), but there is also a far greater awareness of the potential effects of tunnelling on existing infrastructure. Greater attention is given now to the building settlement that takes place during and after tunnelling. Indeed, the possibility of damage and disruption from construction works is so little tolerated’that it has become a major hurdle to the acceptance of new tunnelling projects in urban areas. A well-defined staged methodology was adopted for assessment of the potential effect of settlement on overlying structures as described in Chapter 3 . Frequently referred to as “settlement” assessment, it is more accurately called potential damage assessment because its purpose is to identify structures at risk of sustaining an unacceptable degree of damage, rather than to assess or predict the magnitude of settlement. Technical contractors to the JLE Project made the initial assessments. “Slight” damage, as defined in BRE Digest 251 (see Table 3.l), was adopted on the project as the maximum acceptable degree of damage. More onerous criteria may apply to specific structures for serviceability or non-technical reasons. Once it had been established which structures were potentially liable to sustain unacceptable damage, it was deemed necessary to implement measures to protect them and restrict damage to an acceptable level.

This chapter describes the methods available to mitigate the effects of the movements that inevitably occur even during well-controlled excavation. Consolidation settlements can continue for several years following tunnel construction. Because these long-term movements could also cause damage to structures, they should be taken into account in the implementation and ultimate decommissioning of protective measures.

Ch 11 Protective measures

135

11.2.2

Design and specification The primary requirement within the JLE construction contracts with respect to overlying structures was that any damage sustained as a result of the works should be no worse than the “slight” category. The contracts thus were consistent with the assessment methodology in that damage to overlying structures was permitted provided it was maintained within the slight category. Other general requirements were also specified, for example that settlement should be minimised. In addition to these general requirements, specific numerical limits on the magnitude of permissible movement were given for some structures. The most common limit was 25 mm, which was applied to all of Railtrack’s rail-carrying structures. The implications of this were significant because the JLE follows the alignment of the railway viaducts over most of the route between Waterloo and Bermondsey stations, for a distance of 2.5 km. A maximum limit on slope of 1:lOOO was also applied to these structures.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

A 25 mm limit was also specified on several other structures along the route, particularly on Contracts 103, 104 and 105. This limit was applied to eight structures of Contract 104 in addition to Railtrack assets, for example. More onerous limits were also specified for London Underground’s existing assets: 0

tunnels: 10 mm movement

0

escalators: 0 mm movement.

Where a technical contractor deemed the potential damage likely to exceed the slight category (or the settlement was likely to exceed permitted values), protective measures were specified within the construction contracts. The works contractors were required to undertake their own assessment of potential damage and to propose additional or alternative protective measures, if necessary. Services were excluded from the technical contractor assessments and therefore protective measures were not specified for them, although some services were diverted before construction as advanced works. In parallel with the development of assessment techniques described in Chapter 3, there has been development in the methods of protecting the affected infrastructure. Compensation grouting was the most widely specified method on the JLE. This technique, which was first used in the UK in 1992 at Waterloo station (Harris et al, 1994; Mair et al, 1994), comprises the injection of cementitious material between the tunnel and the affected structure as the tunnel advances to compensate for the volume loss movements as they occur. In general, the “design” of compensation grouting included in the construction contracts was limited to indicating the areas for such treatment (known as mandatory ground treatment, MGT) on the contract drawings. The responsibility for design of the injection system and the methodology of implementation lay with the construction team. The implementation of compensation grouting is described in Section 1 1.6, concentrating on Contract 102, which was the largest and most diverse contract. Numerous other methods of protecting structures from the effects of settlement are available; these are briefly described in Section 11.3. The implementation of protective measures is discussed in general terms in Section 1 1.4. Examples of the various types of protective measure specified or adopted, generally drawing on experience on Contract 102, are described in Section 1 1.5. These are broadly representative of the types of measures used on all of the construction contracts.

136

Building response to tunnelling

11.2.3

Alternative and additional measures As noted in Section 1 1.2.2, the works contractors were required to carry out their own assessments of all affected structures, including services, before the start of construction. They were required to propose protective measures beyond those specified, as necessary, based on the results of the assessments. In addition, the submission of alternatives to specified methods was encouraged. Examples of these are presented in Section 1 1.5. Not all of the protective measures that were actually implemented were identified before construction. The need for extra protective measures sometimes became apparent in the construction period as a result of additional knowledge gained or to deal with circumstances encountered during the works. Again, examples are given in Section 1 1.5.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

11.3

TYPES OF PROTECTIVE MEASURE There are various methods by which both surface and subsurface structures can be protected from the effects of ground movements generated by tunnelling and deep excavations. The most straightforward method is to make sure that the structure is outside the zone of significant ground movements. If this might seem a statement of the obvious, it demonstrates that consideration of ground movements within the design process can influence the location and layout of stations and tunnels. Tunnel alignment design is therefore considered to be a form of protective measure. A literal interpretation of this idea would be to relocate the structure so that it is unaffected by the excavation activities. This is commonly applied to services (as mentioned above). Although substantial buildings can be moved bodily, this was not done on the JLEP. Another method of reducing the impact of ground movements on an overlying structure is specifying the tunnel construction sequence. This may be appropriate with multiple tunnels when the most critical movements for the structure will be at an intermediate stage of construction. Once the geometry of the stations and tunnels is fixed and the potential damage assessment has identified the need for protective measures, a range of options can be considered. The available protective measures can be considered in three categories. 1

In-tunnel measures. These include all actions taken from within the tunnel during its construction to reduce the magnitude of ground movements generated at source.

2

Ground treatment measures. These include all methods of reducing or modifying the ground movements generated by tunnelling by improving or changing the engineering response of the ground.

3

Structural measures. These methods reduce the impact of ground movements by increasing the capacity of the structure to resist, modify or accommodate those movements.

These divisions are not distinct because many options can be considered to be acting in different or composite ways. Frequently, two or more methods are used in combination. The most common of the available methods are described under these categories in Sections 1 1.3.2 to 1 1.3.4.

Ch 11 Protective measures

137

11.3.1

Tunnel alignment design The design of the layout of the tunnels is a basic method of protecting overlying structures from the effects of construction-induced settlement, either by ensuring that they are outside the zone of influence or by minimising the potentially damaging effect. In inner city areas there is likely to be little scope for this because of the many other constraints on the alignment. Furthermore, changing the alignment to reduce the magnitude of settlements and distortions on one structure will almost certainly increase them on another. Nevertheless, on Contract 102 consideration of the position of particularly sensitive or important structures with respect to the predicted volume loss settlements had an important effect on the geometric layout of the tunnels. Three instances where changes to the originally conceived alignment were made to reduce the impact of ground movements on particular structures are described in Section 1 1.5.1.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

11.3.2

In-tunnel measures A range of methods can be applied from within the tunnel to reduce the magnitude of movements or distortions attributable to volume loss. Most of the available protective measures aim to increase the degree of support provided to the ground and to apply the additional support as early as possible in the excavation cycle. The increasing capabilities of tunnel boring machines (TBMs) (see Chapter 10) offer the potential to maintain a significant pressure on the excavated ground surface throughout the construction cycle. The use of closed-face machines was historically developed primarily to counter unstable ground conditions. The increasing use of earth pressure balance machines (EPBMs), in a wide range of ground conditions, has demonstrated their potential to control volume loss, although in practice their effectiveness is variable. An alternative method of providing a support pressure throughout the excavation cycle is the use of compressed air. In general, where the ground is capable of supporting itself during excavation, advantage is taken of this property by tunnelling in open-face conditions. There is then the possibility of undertaking hrther measures to reduce movements from within the tunnel, which include: tunnelling method/TBM design 0

face support measures

0

excavation in parts pilot tunnels

0

mechanical pre-cutting

0

barrel vaulting.

The tunnelling method for open-face working could be by hand mining with or without a shield, by NATM (shotcrete support), or by TBM. In each case the key factor is the size of excavation required before a stiff support system can be installed. The time period between excavation and installation of the support is only a secondary factor, because by far the most movement occurs immediately during excavation in normal open-face tunnelling operations (Harris et al, 1994). Traditional hand-mined and NATM tunnels will generally have a much smaller excavation length ahead of the last completed “ring” than shield-driven tunnels, tending to reduce the magnitude of movements. If a TBM is used, design features such as dimensions, bead, face support and excavation method can have a potentially significant

138

Building response to tunnelling

influence on the ground movements generated. Note, however, that the overall rate of tunnel advance is not an important factor in determining volume loss movements for TBM drives in London Clay. The type of lining used is potentially significant, as with bolted linings the support to the ground depends on the grouted annulus; the effectiveness of this grouting is therefore an important factor. The increasing need for the control of settlements associated with tunnelling may well lead to the adoption of more advanced TBM systems in London Clay.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Some degree of face support is encompassed within the general requirements of good practice in all tunnelling methods. In this context, face support measures are taken to refer to additional measures adopted specifically to reduce ground movements and can take many forms. Examples are the installation of spiles (rods installed ahead of the tunnel face) to reinforce the ground (see Figure 11. I), and sequential excavation of sections of the face with a positive pressure applied to each section prior to excavation of the next.

Longitudinal section

Cross section

Figure 11.I

Ground reinforcement by spiles ahead of the face

Larger tunnels are often excavated in parts. This can be done using a pilot tunnel for segmental tunnels, bench and invert or side drifts for NATM tunnels or, for very large tunnels, by a combination of headings or multiple small-diameter tunnels or pipes. The primary reason for adopting these types of construction sequence has been to reduce the size of open excavation for safety reasons. There is limited evidence to show that, in some instances, significant reductions in volume loss can be achieved by these methods. Barrel vaulting and mechanical pre-cutting are examples of systems developed especially to install a substantial part of the final support system with the minimum possible amount of excavation. Figure 1 1.2 illustrates a tunnel excavation sequence combining barrel vaulting and excavation by parts.

Excavation

1

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 11.2

11.3.3

2

3

Barrel vaulting and excavation by parts

Ground treatment measures Compensation grouting Grout is injected in the ground at an elevation between the source of ground movement (the tunnel) and the structure to be protected. The grout is injected concurrently with the construction of the tunnel to compensate for the ground loss and stress relief caused by excavation. The basic principle is illustrated in Figure 1 1.3. The volumes and timing of grout injection are based on detailed observation of performance with the aim of controlling the development of settlement and associated distortion of the structure. In practice, it is virtually impossible to inject sufficient grout during the excavation part of the tunnelling cycle, hence compensation grouting necessarily becomes a series of minor jacking operations. The aim is to minimise the amplitude of these cycles of settlement and jacking. Simply allowing a structure to settle when the tunnel is constructed and subsequently returning it to its original position by corrective grout jacking is not consistent with the principle of compensation grouting. In an ideal situation, compensation grouting is used to generate displacements of the ground at the grouting horizon equal and opposite to, and concurrently with, those generated from tunnel excavation below. Thus the grouting techniques employed can broadly be termed “displacement grouting”, because this is the desired effect on the ground into which the grout is injected. Displacement grouting can be employed to achieve several goals, of which settlement compensation is one. The generation of heave, often referred to as grout jacking or corrective grouting, is differentiated from compensation grouting since the timing of injection is not related to the generation of settlements. The other main application of displacement grouting is to improve the strength and stiffness of the ground by compaction. The term “compaction grouting”, however, has become associated with a particular grout injection technique, namely the controlled injection of a mortar to create an expanding bulb, which was originally devised to displace, and hence compact, loose granular soils.

140

Building response to tunnelling

Settlement controlled by compensation grouting e - .

........

-------------*__________________________.

*------------------

T

....... ................................................ Settlement if no compensation grouting

Compensation grouting

--3

- - *

Tunnel heading

Slight settlement with compensation grouting

Grout injection

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

JI **-

- , ~ # - - - - ”

***-*-

Excavated shaft

0

0

I

0

Severe settlement

Tube a manchette (TAM)

grouting

Compensationgrouting zone Tunnel Figure 11.3

Principle of compensation grouting

The terms used to define the purpose of the grouting have thus become intertwined with those used to describe the grouting techniques employed to achieve them. The grouting techniques need to minimise the extent to which grout can penetrate or permeate into the soil structure since filling voids within the ground will not generate displacements. In fine-grained cohesive soils such penetration does not occur, whereas in granular soils a wide range of grout mixes will penetrate the soil to varying degrees. In essence, the grouting techniques used to generate displacements differ in the shape of the grout inclusion formed within the ground. Compaction grouting forms an approximately spherical bulb of grout at the point of injection. Conversely, “fracture grouting” forms a sheet of grout frequently only 1-2 mm in thickness, the extent of which is limited only by the volume of grout injected. The properties of the grout and of the ground determine the shape of the grout inclusion. The grout properties govern the method by which it can be introduced into the ground. In general terms, a mortar - ie a cement-based grout with sufficient silt sizes to provide plasticity together with sufficient sand sizes to develop internal friction - is required to form bulbs of grout. The grout is generally too viscous to be injected through the narrow ports (holes) in the relatively small-diameter multi-access grouting tubes called tubes Ci manchette (TAMs, see Section 1 1.6.2), necessitating use of larger-diameter grout tubes. These are filled with grout during each injection operation, so for successive injections to be achieved, equipment has to be maintained on site to re-drill grout holes. Conversely, the relatively fluid grouts used for fracture grouting can be repeatedly injected through TAM ports.

Ch 11 Protective measures

141

Other methods to give intermediate grout inclusion shapes between a bulb and a fracture are possible (see also Section 1 1.6.2). To achieve displacement by injection of fluid grout in granular material it is generally necessary to carry out extensive pre-treatment to fill void spaces into which the grout can permeate. The technique of “pressure filtration” grouting was developed to avoid such unproductive grouting (Mair et al, 1994). This comprises the injection through TAMs of a fluid grout with a high fines content, which is designed to bleed rapidly, depositing the solids close to the point of injection and losing the water to the surrounding ground. With repeated injections, limited fracturing may also develop. This process is sometimes referred to as intrusion grouting. Figure 1 1.4 schematically represents the difference between compaction grouting, intrusion grouting and fracture grouting.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

n

Compaction grouting

n

Intrusion grouting

Fracture grouting

U

Figure 11.4

Compaction, intrusion and fracture grouting techniques

Compensation grouting was the primary method adopted on the JLEP to reduce the impact of settlement on structures deemed to be at risk of sustaining unacceptable damage; it is described in detail in Section 11.6. Ground improvement A full range of ground improvement techniques can be applied that increase the strength or stiffness of the tunnelling medium in advance of excavation and thereby reduce the magnitude of displacements associated with excavation. Examples of potentially useful methods are: 0

grouting, particularly permeation grouting

0

compaction

0

drainage replacement.

The& are used more as a way of improving the stability of an excavation, by controlling the flow of water into it, rather than as a protective measure. They can also have an important effect in reducing the magnitude of ground movements. Permeation grouting was used widely on the JLEP in the Terrace Gravels. In this application, permeation grouting comprises the injection of grouts (usually silicates) to fill the voids between the soil particles to create a soil mass of significantly reduced permeability. Untreated Terrace Gravels typically have a permeability of 10-*to 10” m/s, which reduces to about 5 x lO-’ m / s when effectively treated. The strength and

142

Building response to tunnelling

stiffness of the ground are also greatly increased, with the unconfined compressive strength of the treated material being up to 1 MPa and its stiffness about 500 MPa.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Permeation grouting is most commonly carried out by controlled injections of known volumes of grout at specific points along the length of the pre-installed tubes U manchette (TAMs). To treat the gravel fully, the maximum spacing between the drill holes in which the TAMs are installed is typically less than 1.5 m. The TAMs for permeation grouting are made of plastic (for use at relatively low injection pressures) and are about 50 mm OD. They are installed and grouted into a drill hole of about 100-120 mm diameter. Ports from which the injections are made are generally at 0.33 m centres along the TAMs. Typically, the volume of grout to be injected is between 25 per cent and 30 per cent of the total volume of ground being treated. The injection pressures and flow rates ideally should be low enough that no displacement of the soil particles occurs, although in practice some displacement is inevitable to achieve an effective treatment of the Terrace Gravels. Permeation grouting can also be used to provide a stiffened “raft” of soil below a building if there is sufficient thickness of sand or gravel below the foundations. Microfine cement grouts are often used in this application because they produce a stronger and stiffer result than silicate grouts. The aim of such treatment is to reduce differential settlements by modifying the distribution of ground movements. This technique is often used in combination with compensation grouting to give a more uniform response to grout injections. Structural strengthening Structural strengthening of the ground is defined as the introduction of structural elements into the ground that are neither part of the tunnel under construction nor attached to the overlying structure to be protected. They are intended to reduce movements of the affected structure(s) in one of two ways: by globally providing a stiffer response of the ground and thereby generally reducing the magnitude of ground movement; or by providing a restraint (or barrier) to movement between the source of movement and the structure to be protected. An array of mini-tunnels or pipes installed as an “umbrella” above the tunnel but separated from it by ground that will not be excavated is an example of the former type. The structural elements span the advancing excavation at the tunnel face and stiffen the response of the soil allowing the imbalance in load produced by the excavation to be redistributed with smaller movements. The second type could take the form of a cutoff or curtain wall where a deep wall or series of discrete piles is installed between the tunnel and the building to be protected. The theory is that the wall will settle less than the ground and thereby reduce the settlements on the side remote from the source of movement, as shown on Figure 11.5 Control of ground water (drainage/permeation grouting) In granular materials, control of groundwater is a prerequisite for effective excavation. This can be achieved either by lowering the groundwater by dewatering from well points or deep pumps or by grouting (as described above). Whichever method is used, the successful control of groundwater significantly reduces the magnitude of ground movements associated with excavation.

Ch 11 Protective measures

143

I

Slight settlement with cut-off wall

\

\

Extent of trough

\

, Tunnel

Figure 11.5

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

11.3.4

Principle of cutoff or curtain walling

Structural measures Structural measures include a range of techniques to reduce the impact of ground movements that are applied to the structure to be protected itself. Their mode of operation can be to: 0

increase the ability of the foundations to resist the predicted movement stiffen the structure such that it modifies the predicted movement make the structure less sensitive so that it can accommodate the anticipated movement control the movement of the structure by isolating it from its foundation mitigate progressively the effects of movements through the instigation of a planned maintenance regime.

1

Reinforced concrete pads

Firm ground

Figure 11.6

Typical layout of structure jacking

Examples of each of these techniques are: 0

deep underpinning such that the piles extend below the zone of ground movements and thereby reduce the movements of the structure increasing the tensile capacity of the structure where this is small or unreliable. This is achieved by installing tension elements such as straps, tie bars or ring beams

144

Building response to tunnelling

reducing the sensitivity of a structure by increasing the bearing shelf of beams, slackening bolts to allow articulation, strengthening connections between structural elements andor finishes, temporarily removing particularly sensitive finishes, installing temporary support installation of jacks within structural elements to enable the movements of the superstructure (or other sensitive features) to be controlled independently of the foundation (see Figure 1 1.6) planned maintenance (such as the fettling of railway tracks) or contingency measures (such as propping or repair) to be implemented on the basis of observed performance.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Shallow underpinning techniques, such as those illustrated on Figure 11.7, can also be used to protect buildings from the effects of ground movements, particularly where the existing foundations are known to be inadequate. The method is most frequently applied where a weak or heterogeneous soil stratum exists below foundations that might respond non-uniformly, which would increase the potential for differential settlement to occur. Underpinning can also be used to increase the tensile capacity of a structure, especially if the beam-and-pad method is adopted.

Cantilever beam

Figure 11.7

Shallow underpinning techniques

11.4

IMPLEMENTATION

11.4.1

Passive/active

- observational techniques

Protective measures can be either active or passive in their implementation. Passive measures are applied and then are complete; there is no intent to modify them during construction. Active measures, on the other hand, rely on observed performance to control their implementation, and successive modification forms part of the method. Both passive and active systems can be part of an observational approach. The relationship in active systems is self-evident in that the feedback from observation is essential to the continuing implementation of the protective measures. Compensation grouting and structural jacking are the most obvious examples of such methods.

Passive methods can also form part of an observational approach: if unacceptable performance develops, pre-defined contingency measures can be implemented. In this case, the contingency measures could comprise the implementation of the passive measure, the extension of a previously installed system, or the provision of an additional, different method. Prediction, monitoring and control of the overall process are the key facets of the successful application of protective measures. These functions are briefly described in the following sections.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

11.4.2

Prediction The assessment of the potential damage to a structure is the initial stage in identifying the need for protective measures. As noted above, the methodology used on the JLE described in Chapter 3 does not aim to predict total movements, but it does indicate the distribution and magnitude of movements. The need for refined predictions depends on the extent to which the efficiency of the protective measure is sensitive to variation from these values. In many instances, sufficient factor of safety can be included in the design relatively easily. This applies particularly to passive techniques where no modifications are intended during construction. For active methods, there is a need to update predictions during construction based on back-analysis of observed performance. In this situation, where time is critical, the simplest appropriate method should be adopted. In complex situations and where changes in the distribution or magnitude are critical to the outcome, more advanced predictive methods may be necessary. The timescale over which movements are expected to develop is a key factor in determining the optimum method of analysis.

11.4.3

Monitoring Monitoring performance involves much more than simply taking measurements. The planning, design and installation of an appropriate monitoring system are key activities that are often carried out under intense time pressure early within a construction project. Such a system must include methods of processing, presenting, interpreting and disseminating the results. The existence of an operational monitoring system with adequate background readings is a prerequisite for the start of construction activities that will generate ground movements. The following stages of design are identified: identify potential measurements and rank in importance, eg as essential, preferable, desirable or for research purposes identify simplest and most robust system of obtaining measurements at the required frequency and accuracy make sure that there is a back-up system for essential measurements ensure that installation, commissioning and establishment of datums are key activities in the overall construction programme 0

consider the amount of data that will be generated and devise systems to process, present and interpret data within required timescales integrate construction activity records within the monitoring and reporting activity.

146

Building response to tunnelling

Some common potential difficulties are: 0

liaison with building owners and occupiers to identify acceptable locations and types of instrumentation

0

provision of power and telecommunications needed for remote logging instrumentation

0

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

11.4.4

devising suitably robust yet flexible software for the processing of data for which computers are essential.

Control Protective measures can influence production and sequencing of the main excavation works. Thus control of protective measures has to be integrated with the overall construction management and should not be considered as a separate operation. Control of observational techniques (particularly active types where the actions to be taken are dependent on feedback) requires accurate records of construction activities, including the protective measures themselves, to interpret the monitoring results correctly. The collection and co-ordination of information, its interpretation and dissemination is a complex process that requires well-defined roles and responsibilities. Many parties within the management structures of contractor and subcontractors, as well as the client’s representatives and engineers representing the owners of affected structures, have an input in the management process (eg tunnelling, grouting, surveying, monitoring, client and third parties). Generally, the only practical way to make decisions (and to have them approved, as necessary) within the time available is for all the interested parties to meet frequently.

11.5

EXAMPLES OF PROTECTIVE MEASURES This section gives examples of the types of protective measure employed on the JLE. It states the nature of the technique employed, briefly describes the circumstances and lists the types of monitoring used (in addition to the ubiquitous precise levelling) to control implementation. The examples given here are mostly from Contract 102, but they are broadly representative of those employed throughout the JLE.

115 1

Tunnel alignment There are three documented examples on Contract 102 where the final tunnel arrangement was at least partially influenced by consideration of the settlement effect on overlying structures. These are at both Westminster and Waterloo stations and below the RAC building in Pall Mall. The design considerations are briefly described below. Westminster station

The designed layout of Westminster station includes: vertically stacked platform tunnels below Bridge Street to maximise the distance of the tunnels from Big Ben Clock Tower and thereby minimise the settlements predicted from volume loss movements the specified location for the shield chambers for the platform tunnels being beneath Parliament Square to avoid the excavation of large hand-dug tunnels close to the Clock Tower and River Wall.

Ch 11 Protective measures

147

The design was influenced by the need to minimise movements of the Clock Tower (Carter et al, 1996, Bailey et al, 1999). A corollary was that the predicted movements of the District and Circle Line tunnel and services directly over the JLE tunnels were increased. Waterloo station

The JLE Waterloo station is close to both the existing mainline station and the viaduct carrying trains to Charing Cross station. The location of the whole station was changed by about 60 m. A major consideration was the need to reduce the predicted volume loss settlements on rail-carrying structures. The minimisation of the potential impact of settlements on specific (rail-carrying) structures was thus a significant design consideration. The relocation of the new station increased the predicted movements of the concourse and office buildings at Waterloo.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

RAC building, Pall Mall The Royal Automobile Club building lies above the alignment of the eastbound running tunnel at the point where the JLE tunnel crosses the original Jubilee Line tunnels. The RAC building has a reinforced concrete swimming pool in its basement and some deep pad foundations to columns. As the new tunnel had to pass over the existing Jubilee Line tunnels in this area and its depth reduced, there was an increase in the potential effect on the swimming pool and deep foundations. An investigation was undertaken to select the vertical alignment that maximised the clearance between the tunnel and swimming pool and minimised the potential distortions. The horizontal alignment was also modified to avoid, so far as practicable, the deep pad foundations. The potential effect on the existing Jubilee Line tunnels was increased by the lowering of the alignment. Thus the alignment was both lowered and moved laterally based on consideration of settlement issues.

11S.2

In-tunnel measures The only in-tunnel measure specified was the use of pilot tunnels for all tunnels with a internal diameter greater than 5.75 m. This is a commonly adopted methodology for large-diameter tunnels and is specified primarily for safety reasons to limit the maximum size of tunnel face excavation. Pilot tunnels were omitted for the 8.0 m ID escalator tunnels at Waterloo station, as a inclined face improves the stability of the excavation. No particular problems were encountered during construction. Face dowels

There was considerable concern about using the specified protective measure (compensation grouting) in an area where the overlying structure was founded on piles terminating close to the tunnel. To reduce the need for compensation grouting it was decided to adopt in-tunnel methods to reduce volume loss. Face dowels 30 m long at 1 m centres were designed to reduce an estimated 20 per cent volume loss.

148

Building response to tunnelling

The full designed pattern of dowels was not installed and the efficacy of the system cannot be realistically evaluated from the available data. Reduced excavation lengths

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Where the JLE running tunnels approaching Waterloo station passed below the existing Bakerloo and Northern Line tunnels and associated escalators, the predicted settlements exceeded the specified values. The NATM excavation sequence was modified to reduce the maximum unsupported length ahead of the last completed ring to 3 m from 4 m in order to minimise the volume loss movements. The standard and modified excavation sequences are shown on Figure 11.8. Real-time monitoring of settlements with strings of electrolevel beams were combined with nightly measurements of track geometry and clearances. A procedure for processing the data and notifying key personnel in the event of trigger levels being exceeded was instigated such that train services could be suspended if necessary.

Standard Figure 11.8

Modified

Standard and modified NATM excavation sequences

Low volume losses of about 1 per cent were recorded, but the extent to which this was due to the reduced excavation lengths cannot be separated from the effects of other beneficial factors. Sequencing of tunnel excavation

The contract imposed very stringent limits on movement of the escalators of the Bakerloo and Northern Line. The need for protective measures was avoided by reexamining the specified limits on movement. Detailed predictions were made of volume loss movements and associated sag, roll and slew of the escalators. The volume losses were derived from back-analysis of observed movements from similar earlier tunnelling works. The predictions included the development of movements as the two running tunnels approached the escalators because the final condition did not give the critical values of distortion. It was demonstrated that no protective measures were required provided the predicted movements were not exceeded. A detailed monitoring regime was instigated that allowed the developing movements to be compared frequently with the prediction based on actual tunnel progress. A settlement control procedure was established, which allowed for control over the progress of either or both tunnels. Contingencies included immediate suspension of tunnelling and closure of the escalators for inspection if trigger levels were exceeded. Specific monitoring included electrolevels and measurements of deflection of the escalator trusses relative to tensioned piano wires. In the event, no delay to the progress of either tunnel was necessary.

Ch 11 Protective measures

149

Ground treatment Compensation grouting Compensation grouting was the most widely specified protective measure. Its use increased significantly because the contractor proposed to protect services by this method, mainly through extending the areas designated for compensation grouting in the contract. The scope and extent of compensation grouting on JLEP and details of its implementation are given in Section 1 1.6.

Permeation grouting

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

I\

115.3

Virtually all the tunnelling on Contract 102 was within the London Clay. Only three tunnels were constructed partially within the overlying Thames Gravels. Two of these were the escalator tunnels at Waterloo between the ticket hall and the upper concourse; the third was a smoke extract tunnel to the north of the Westminster station box. Permeation grouting was specified in these instances to reduce the permeability of the ground, thereby allowing tunnel construction to proceed in effectively “dry” conditions. An important, but secondary, effect is that hlly treated gravels are very stiff and the volume losses produced by excavation in this material are much lower than in London Clay, for example. This is particularly significant, as tunnels within the Terrace Gravels are invariably shallow, with potentially large associated settlements and distortions. Permeation grouting was also extensively specified in Terrace Gravels wherever tunnel excavation was required in the London Clay with less than 6 m clay cover above the crown to the interface with the overlying Terrace Gravels. Where the tunnels are entirely within London Clay, permeation grouting was used for a slightly different reason. The method provides additional protection against tunnel inundation and potential collapse, which would occur if a hydraulic connection were formed through the (relatively low) clay cover. This is achieved by treating the gravels with silicate grouts to reduce the permeability substantially and therefore increase the flow path from any source of water. The two major concerns are that the level of the clay/gravel interface varies significantly between borehole locations, and that fissuring in the clay allows a block failure to develop in the crown of the tunnel. The adoption of 6 m as the criterion meant that large areas had to be treated, because the upper tunnels at both Westminster and Waterloo stations generally have about 5 m clay cover. The treatment did not cause significant ground movements, and therefore is not considered a protective measure in the context of this report. Permeation grouting was used to provide a stiffened zone within the Terrace Gravels as a building protective measure, particularly on Contracts 103 and 105. A microfine cement was used to permeate the Terrace Gravels, resulting in a significantly stiffer material than was achieved with silicate grouts. This type of treatment was mostly used in conjunction with compensation grouting or where there was an initial intention of compensation grouting.

Structural strengthening of ground Secant piled wall: A 7 m ID moving walkway tunnel at Waterloo station runs approximately parallel to the Railtrack viaduct along Mepham Street. The contract allowed for the provision of a 25 m-deep wall over a length of 1 17 m between the tunnel and the structure comprising 750 mm-diameter alternate hard and soft piles. There would be no connection between the wall and the structure, as the purpose of the wall

150

Building response to tunnelling

was to modify the ground movements by constructing stiff structural elements within it. In the event, this protective measure was not used. 1

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Westminster station Box - low-level struts. Westminster station includes a deep diaphragm-walled box to house the escalators. To reduce the deflections of the wall and the associated ground movements, it was specified that low-level struts be constructed across the box at the base of perimeter diaphragm walls before the section between the walls was excavated (Carter et al, 1996; Crawley and Stones, 1996). This was to be achieved by constructing cross-walls dividing the excavation into four “cells”. The cross-walls were to be concreted at low level only, below the final excavation level. Jacks were to be installed to pre-stress the system and allow the stresses within the struts to be controlled. Thus, to minimise movements at the base of the wall the stiffness of the ground was to be enhanced by the inclusion of structural elements within it. Alternative methods of providing struts at low level prior to excavation were considered at the design stage, ie the diaphragm cross-walls specified and struts constructed by tunnelling. At the design stage, the diaphragm-wall option was preferred as it would give better resistance against heave at the base of the excavation and would provide additional restraint to wall movement over the full depth of the excavation. The contractor proposed the tunnelling option be adopted, mainly because of the limited working room available on site. In particular, the presence of the operational District and Circle Line dissecting the site at high level made the installation of additional diaphragm walls problematic. Tunnelled struts were adopted mainly for constructability reasons. The layout and location of this solution are described in Chapter 27. The deflection of the diaphragm walls was monitored by both electrolevel and manual inclinometers as well as optical surveying methods. The loads in the struts were measured by monitoring pressures in the jacks incorporated in the struts. The struts were successful in reducing recorded movements at the base of the diaphragm walls to a negligible amount. Pipe arch The requirement to provide a blanket of permeation grouting treatment above tunnels with a clay cover of less than 6 m was particularly difficult to achieve above the Westminster station tunnel situated below the very busy Bridge Street. As an alternative, an “umbrella” of five pipe-jacked tunnels of approximately 1 m diameter was proposed to provide equivalent protection against inundatiodcollapse of the tunnel during construction. The geometry is shown on Figure 1 1.9. A total length of 850 m of pipejacked tunnel was required. It was suggested that potential benefit would accrue as a result of the strengthening of the ground in terms of reduced volume loss and a reduction in the potential effect of compensation grouting on the tunnel.

Ch 11 Protective measures

151

I

Westminster station

-

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

.....

Figure 11.9

As-built pipe length

I

Big Ben Clock Tower

Proposed pipe length

Pipe arch geometry

Difficulties were encountered from the start of excavation of the first pipe: the tunnelling machine was unable to cope with the claystones encountered at the elevation of the pipe jacks. A second work-site was started using hand excavation techniques to increase progress, but it was clear that substantial delays to the platform tunnel drive would occur if all of the pipe jacks were constructed as designed. A review of the need for the pipe jacks was undertaken. Several factors were identified. 1

It was known that there was limited variation in the level of the clay/gravel interface because the compensation grouting TAMs had been installed over the entire area about 2 m below the nominal interface level and had encountered no water. In addition, numerous boreholes and excavations had been made from which the elevation of the interface could be collated.

2

The clay cover was approximately 5 m over a 7.4 m-diameter excavation, but a pilot (running) tunnel of 4.8 m diameter significantly reduced the dimensions of the excavated face.

3

The TAMs themselves would act as a form of tensile reinforcement of the ground, reducing the possibility of large-scale block failures from the crown of the excavation.

4

Compensation grouting had been shown to be capable of controlling settlement without adversely affecting the tunnel lining in a similar geometry.

It was concluded that omission of the pipe jacks did not represent an unacceptable risk. Nevertheless, construction of the pipe arch should continue as far as possible before the start of the eastbound platform tunnel drive and be concentrated in areas where there were significant structures above the tunnel, as shown on Figure 11.9. In total, 483 m of the 850 m, or 53 per cent of the proposed pipe jack, were constructed. No particular problems were encountered during the subsequent platform tunnel drive. From analysis of the monitoring results it can be inferred that reduction in volume loss or effect of compensation grouting on the tunnel due to the pipe jacks was minor.

152

Building response to tunnelling

115 4

Structural measures Underpinning The layout of the escape shaft at Storey’s Gate and the provision of an adjacent basement required parts of the existing Storey’s Gate Lodge building to be underpinned. The building is Grade I1 Listed and was in poor condition because of historical settlement. The sections of the walls adjacent to the new build were underpinned with mass concrete footings, with the remainder supported on mini-piles founded in the Terrace Gravels. The underpinning was shallow relative to the tunnels and shaft that would generate the predicted movements. The underpinning was intended to reduce the potential for differential settlements rather than to reduce the magnitude of absolute movements.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

At Waterloo station the twin tunnels for escalators from the Ticket Hall to the Upper Concourse are located in adjacent arches of the existing structure. This required removal of parts of two piers to the arches. The lower part of each of these piers was rebuilt and founded on bored piles. The foundation to the third pier of these arches was stiffened by permeation grouting over the full depth of the Terrace Gravels. Structure jacking Flat jacks were incorporated into the three piers adjacent to the escalators at Waterloo station described above, in conjunction with the underpinning of two of the piers. The intention was to maintain the level of the structure. The adjacent areas were protected by compensation grouting. The underpinning reduced settlement to small values. The jacks had then to be re-set to allow the structure to be lowered as well as raised. This was to enable control of distortions to be maintained while simultaneously reducing the amount of compensation grouting undertaken. This was necessary primarily to accelerate the construction programme, but also partly to ameliorate the build-up of loadings in the temporary works in the upper-level tunnels. The pressures in the jacks, relative movements across the jacks and changes in the width of pre-existing defects were monitored. Structure strengthening Tensile reinforcement, in the form of temporary steel ties, was specified to Storey’s Gate Lodge in conjunction with the underpinning described above. The widths of existing defects were monitored. A range of structural measures was specified to the Railtrack viaduct that carries the Charing Cross to London Bridge railway through the Waterloo area. 1

Ties to piershridge abutments which comprise two distinct structures.

2

Internal bracing to arches.

3

Grouting of cracks and repairs to brickwork.

4

Jacks to main girders.

5

Re-levelling of secondary girders.

6

Renewal of waterproofing.

7

Replacement of bridge bearings.

Ch 11 Protective measures

153

These structural measures were additional to compensation grouting and the need for them was re-evaluated during construction. Only the strengthening measures (numbers 1 to 3) were implemented and the scope of these works was limited.

Flying shores Visible separation of the fagade of a particular building was identified; the specified solution was to provide support to the fagade by propping across the (narrow) street to the building opposite. The alternative of permanent internal ties and straps was preferred by the contractor and was adopted as being equally effective and efficient. It also avoided potential difficulties with installation and subsequent removal.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Lining to Low-level Sewer No 1 The Low-level Sewer No 1 is the main interceptor sewer along the north bank of the Thames; it is a circular 2.6 m-diameter brick-lined tunnel. The sewer traverses Bridge Street and passes above the JLE Westminster platform tunnels; it was therefore in an area of large predicted movements. No protective measures were specified within this contract because services were the contractor’s responsibility. The proposed protective measure was compensation grouting, but a worsening of pre-existing damage to the sewer was noted prior to the grouting arrays being operational. Thames Water’s consultants calculated a factor of safety of unity for the lining; and the damage was indicative of distress due to distortion of the cross-section. As the majority of construction remained to be completed, it was concluded that strengthening of the sewer was the only acceptable protective measure. The sewer tunnel was stiffened by the installation of steel “colliery arches” at 0.5 m centres. This was only acceptable as a temporary solution as it reduced the hydraulic capacity of the sewer. A glass-reinforced plastic liner was installed subsequently as a permanent solution.

Loosening of bolts in Jubilee Line tunnels Where the Extension tunnels commenced, the existing Jubilee Line tunnels were predicted to be subject to significant cross-sectional distortion from construction of both the running tunnels to within 300 mm and the subsequent step-plate junction construction around the operational tunnels. The critical element was identified as excessive stress within the circumferential bolts of the segmental lining. Maximum permissible distortions were set; if movements exceeded these values the bolts were loosened and retightened to a nominal stress.

Re-fettling of LUL railway line tracks In both the subsurface LUL lines affected by movements on Contract 102 (the District and Circle Line at Westminster, and the Waterloo and City Line at Waterloo) track alignment was a matter of considerable concern. It was necessary to establish a robust procedure to avoid unnecessary delays to the works. New design alignments were established based on a survey of the actual track alignment prior to the works. The tracks were then fettled to this alignment. Comprehensive criteria were established for permissible deviations from this alignment based on cant, rate of change of cant and ultimately wheel unload. This required monitoring of track levels at 2 m centres on both rails. Track level surveys were undertaken nightly for a significant proportion of the construction period. When necessary, adjustments to the track were planned and implemented based on observed performance. Criteria and procedures for stopping trains and emergency track work were established.

154

Building response to tunnelling

11.6

COMPENSATION GROUTING ON JLE The principle of compensation grouting is described in Section 1 1.3.3.

11.6.1

Range of application Compensation grouting was by far the most extensively used protective measure throughout JLE Contracts 10 1 to 105. Table 11.1 summarises the areas where compensation grouting was the adopted protective measure. It also details the extent of the grouting area and the grouting undertaken. The need for grouting was identified at all of the station areas and most locations where excavations over and above the two running tunnels were required, eg step-plate junctions and shafts. (See the case histories of Chapters 23,24,26 to 28, and 3 1 to 35 in Volume 2.) Unsurprisingly, these were the areas where the magnitude of predicted movement was greatest. Table 11.1

Extent of compensation grouting on Contracts 101 to 105

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Contracthocation

(m2)

Area

TAMs installed from:

Nollength of TAMs (m)

No of injections

Volumeof No. grout of (m3) shifts

101

4850

2 shafts

541 1700

7620

222

240

102 Step plate

90%

2 shafts

136 14990

2268

161

111

102 RAC building

4140

I shaft

5212250

1758

1 I5

27

102 Great George Street

7052

3 shafts

I38 I4820

3209

340

125

102 Westminster station

12537

5 shafts

363 I 10470

27550

2052

1357

102 Waterloo station

23674

9 shafts & surface

413 I 14360 130 I 1650

41905

3856

2335

103 Southwark station

9563

Surface

3000 I 39000*

nla

nla

nla

nla

nla

nla

nla

nla

103 Wardens Grove 1925

Surface

1031104 Crossover

12263

3 shafts

154 I 7870

104 London Bridge 12100 station

Tunnels

163 I 4700

105 Millstream Road

2470

1 shaft & surface

34 I 1440 229 12980

corrective grout jacking only

105 Bermondsey station

4000

3 shafts & surface

259 15440 nla

stiffened raft and grout jacking

105 ChalfonU Prestwood House area

2280**

2 shafts

198 I 3840**

stiffened raft and grout jacking

105 Kirby Estate

2030

2 shafts

65 / 1640

pretreatment only

I05 Canada Water Estate

1060

1 shaft

25 1830

pretreatment only

Approximate estimate includes for permeation grouting and raft treatment ** Estimated assuming three levels of TAMs for stiffened raft nla not available

Ch 11 Protective measures

155

Locations where compensation grouting was used to protect structures from the effects of running tunnels alone are: contract 102, Great George Street - the tunnels are vertically stacked which increases the total settlement and associated distortions.The coverage of grouting arrays was extended as a result of $reater than expected volume losses recorded on the first running tunnel drive (Chapters 26 and 27) Contract 102, RAC building - a single running tunnel at shallow depth below a basement swinming pool. The a l i m e n t was c o d e d because the JLE tunnel had to pass over an existing tunnel (Chapter 24) Contract 105, Millstream Road - installed after tunnel construction to allow grout jacking to be carried out to revme unexpected settlements

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Contract 105, Chalfont and Prestwood Houses - shallow running tunnels (Limey and Friedman, 1996) . Contract 105, Canada Water Estate - shallow tunnels below piled foundations (Chapter 44).

11.6.2

Grout equipment and mixes Equipment Compensation grouting can be achieved by a range of types of grouting, as described in Section 11.3.3, eg compaction, intrusion or fracture grouting. As compaction grouting requires re-drilling between injections, it was not used on the JLE. Intrusion and fiacture grouting are carried out from tubes U munchette and these methods were used on all the JLE contracts. Figures 11.10 and 11.1 1 show the component parts of a TAM and the insertion of a grout delivery pipe into the end of a TAM in a shaft. Figure 11.12 illustrates the installation and use of tubes U munchette (fiom Rawlings et ul, 2000).

A range of grout mixes and injection equipment was used on the various contracts. The principle was the same in all cases, however: tubes with ports at regular intervals along them are installed and grouted into drillholes (Figure 11.12). The grout is injected by inserting a probe into the tube and isolating the port to be injected by inflating packers at either side of the injection nozzle and then applying sufficientpressure to open the port and initiate flow into the ground. The ports comprise four holes spaced equally mund the circumferenceof the tube and usually covered with a rubber sleeve (the munchette).

Figure 11.I0

156

The components of a tube manchette (TAM) for grouting

Building response to tunnelling

Flgure 11.11

~

Insertion of grout &limy pipe into a TAM installation in a shaft conshytted to give grouting m s s

,

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

London Clay or Lambeth G w p day

Detail m in diagrams blow

TAMreadyfoc grouting

Grout injection

Figure 11.12

Ch 11 Protective measures

Illustration of steps when installing a TAM

157

The tubes have to sustain high pressures during injection potentially to the capacity of the pump if, for example, the packer assembly is slightly misplaced and covers the ports. The tubes are inserted into the drillhole in sections and joined by external couplers. The drillhole diameter has to be sufficient to allow the annulus to be effectively grouted. Details of the TAMs used on the various contracts are given in Table 1 1.2.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Table 11.2

TAM details

Contract

Material

Dia (mm)

Sleeve spacing (m)

Drillhole dia (mm)

101

Steel

nla

0.5

da

102

Steel

70

1 .o

125

103

Steel

60

0.33

110

104

Steel

60

0.5

100

105

Steel

50

0.33

100

Grout mixes The grout mixes that were most commonly used for compensation grouting on the different contracts are given in Table 1 1.3. Desirable characteristics of mixes for use in compensation grouting are: high viscosity - to prevent excessive spread from the point of injection low bleed - to minimise loss of volume prior to setting quick setting time - to maximise the efficiency of multiple injections low strength - to enable numerous repeat injections from an individual port. Table 11.3

Grout mix details

Contract mix code

Water: OPC

PFA: OPC

Bentonite (%water)

Additives

101 BClB

2:1

-

5

None

101 PFASA

11:l

9: 1

2

None

102 fluid

4: 1

-

6

None

102 mortar

1O:l

20: 1

6

None

103 1105.CG1

3: 1

-

4

None

104

1:l

-

5

None

Fracture grouting was used on all the contracts. On Contract 102, the contractor intended to use a very stiff grout mix or “mortar” containing sand, trying to combine the localised effect of compaction grouting with the convenience of injection via TAMs. The injection of such viscous grout through TAMs was a new development. Robust grouting equipment was required, therefore, and the steel tubes had an internal diameter of 70 mm and a wall thickness of 3 mm. Ports were provided at 1 m centres and comprised four equally spaced 20 mm-diameter holes. In the event, a grouting trial showed that repeated injections could not be successfully made from the same port, so this mix was abandoned. A mix utilising 20: 1 PFA:OPC was adopted instead, with the aim of achieving an intermediate result between fracture and compaction grouting, ie a

158

Building response to tunnelling

.

similar effect to the intrusion grouting described in Section 11.3.3; this was also termed a “mortar” grout, but contained no sand. The mix ingredients are given in Table 11.3. Trials confiied that grout thicknesses of up to 10 mm could be achieved compared to the 1 mm or 2 mm normally associated with fracture grouting.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The line losses associated with pumping this grout were high, so it was used only where a concrete pump (Figure 11.13) could be located close to the shaft. A containerised setup for these pumps was developed, which allowed relatively quick transfer of grouting between shafts. A less viscous or “fluid” grout was used exclusively at shafts where o r t a r pump was not possible and intermittentlyat other access for the containerised m shafts depending on resource availability.

11.6.3

Design of grouting facilities Design of thn grouting facilitiescomprises four rrmin elements. 1

Dietermine area to be c o v e d by TAMs.

2

Determinemethod ofinstallingTAMs within these atem.

3

Determineelevation@)for drilling.

4

Determine spacing of TAMs.

1. Determiae area Eo be c o v e d by army8

The mover which TAMs are required is nut neaesmily determined solely by mcpiringprstection. It is 51ece8s81y to anticipate the idesltificatiaaof the uctud s d m t s and hmimntd displslclerreenitsrstb*r than potential damage assewmtprocess. These m d d include other thin volume loss (such grouting facilities). The target conshwtion-rehtedmmes for the &routing in terms of settlement, slope or dist0rtlol;lcontrol needs to be established, particultrrly with respect to &.aceat structures immediately outside the coverage of the gtout map. The mode of implementation has to be For example, if it is considered that mitigation of tumellhg-elated volume loss done is sufficient, it may suffice to install arrays only within the s&tkment trough over &e turinel. This method can be successll 24). Often, lmmmr, it k the provision of the shaft in certain c i r c m m w (m or tunnel from which drilling is undertaken that forms the major part of the cost of establishing grouting facilities, rather than the installation of extra lengths of TAh4s.

Ch 11 Protective measures

159

2. Determine method of installing TAMs within these areas

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The options for the installation of TAMs can be considered to fall into two categories: sub-vertical TAh4s installed from the surface or sub-horizontalTAMs installed from shafts or tunnels. On Contract 102, the shaft option was generally adopted as the only practical solution, although a limited amount of surface drilling was also undertaken. Sites for temporary ground treatment shafts were not identified before the award of the construction contract. Determiningacceptable locations with sufficient working space is problematic in heavily built-up areas, usually necessitating protracted negotiations with the local authority and other third parties. Sufficient shafi sites were eventually secured,

Figure 11.I4

160

Drilling rig being litled into Sh8ft

Building response to tunnelling

..>---

Table 11.1 demonstrates that the vast majority of compensation grouting was undertaken at station locations. On the JLE,permanent works s h a h are required at stations for ventilation, emergency e s q e and &. The presence of these shafts and tunnels at varying levels including escalators impQses a significant constraint, as they will almost certainly intercqt the grouting arrays. The option of utilising the permanent shafts for ground treatment k attractive because it offen, a substautid cost saving. It can create problems, however, as these shafts will have to be completed and fitted out before the end of construction. The timescale for demdssioning the grouting facilities has to be considered at the design stage.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

If access is available a r o u n d and within the struchues to be protected, sub-vertical drilling fiom the surface or from pits is a viable alternative (Figure 11.15). Various factors affect the relative efficiencywith which a given area can be covered by surface or shaft drilling, but in general drilling from a shaft requires fewer individual TAMs and a smaller length of drilling. An example is given by the infornution for Millstream Road in Table 11.I where a very similar area was covered by both surface- and shaft-installed T M . The number ofsusfiu;e-installd TAMs was nine times greater than that required from the shaft and the drilled length was twice as long.

Figun, 11.15

Tpical cross-sections through grouting amys

3. Determine elevation(s) for drilling The two primary constraints on the level at which TAMs are installed are, first, the elevation(s) of the tunnels to be constructed and, secondly,the depths of the foundations of the overlying buildings. In central London, the presence of water-bearing Terrace Gravels over the London Clay provides an additionalconstraint: if nominally horizontal drilling is adopted, it is much easier and cheaper to install TAMs within the London Clay. Ideally, at least one tunnel diameter of cover between the grouting horizon and the crown of the tunnel is desirable to reduce the potential effects of the grouting on the tunnel. In order to allow drilling of u n c d holes, a minimum cover is required to allow for variations in the elevation of the top of the clay. Where tunnels exist at a range of elevations, particularly at stations, array may be required at more than one horizon.

Where compensation grouting is to be combined with permeation grouting of overlying permeable horizons, a single array of TAMs is o h provided to llfil both functions. In this situation, TAMs are installed from an elevation above the compensation grouting horizon to fan out at a range of declinations such that the volume of ground for permeation treatment is covered at the required spacing. The TAMs are extended to the elevation required for compensation grouting.

Ch 11 Protective measures

161

4. Determine spacing of TAMs

Drilling of TAMs from shafts is inherently inefficient because it results in a very high density of ports close to the shaft. The total number of TAMs required to cover a given area is determined by the maximum spacing permitted at the far end of the arrays. The primary criteria used to determine an acceptable spacing are the relative elevations of the grouting horizon the tunnel and the foundations of overlying buildings: the more closely spaced the injection ports, the smaller the grout volumes at each injection point and the less likely it is that undesirable localised effects will occur.

11.6.4

Summary of adopted TAM layout designs

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Contract 101 Green Park On Contract 10 1 two discrete areas of compensation grouting were required. The TAMs were installed from a single shaft in each area. A few vertical TAMs were also installed to reduce settlements during construction of a 9.5 m-diameter shaft before the installation of the horizontal arrays from the shaft itself (Mugaramoorthy et al, 1996). The TAMs from within the shafts were installed within the London Clay with a maximum spacing at the end of the arrays of about 6 m.

Contract 102 Green Park to Waterloo Almost all the TAMs on Contract 102 were installed from shafts. Twenty shafts were utilised, of which 18 were temporary, specifically constructed for the purpose. Of the two permanent works shafts used, both were in the Waterloo station area. Temporary ground treatment shafts were generally constructed of precast concrete segments of 4.57 m diameter. Usually they were extended to 3 m below the designed elevation of the arrays to provide a substantial sump for both the drilling and grouting operations. Shaft sinking was not straightforward in some of the locations, particularly those that had to be sited within buildings. A reduced diameter had to be adopted for the upper section of these shafts and their diameter enlarged within the London Clay. The major problem with shaft sinking in central London is progressing through the waterbearing Terrace Gravels. Both caisson-sinking and permeation grouting solutions were used. The permeation grouting is to form a low-permeability annulus within the Terrace Gravels by injection of chemical (silicate) grouts. This allows excavation in the “dry” with the lining installed by an underpinning method. Permeation grouting was carried out from TAMs. Heave due to grouting (and drilling) could be substantial; for example, where annulus treatments were carried out for three closely spaced shafts at Waterloo station over 30 mm of heave was generated. Settlement caused by shaft sinking was generally small except in locations where heave had previously been generated by the permeation grouting. The shafts were formed of precast concrete segmental linings except at the level or levels of the proposed compensation grouting TAMs where specially fabricated steel segments were used to facilitate the drilling. The TAMs were mostly installed within the London Clay. The elevation was selected to maintain a minimum clearance above the crown of the tunnels of one tunnel diameter and a minimum depth below the Thames GraveMLondon Clay interface of 3 m. Figure 1 I . 15 (a) shows a typical cross-section. In many instances, it was not possible to satisfy these requirements, and the cover above the tunnel to the TAMs was frequently only about half a diameter and the cover above the TAMs was reduced to 2 m or 2.5 m.

162

Building response to tunnelling

In certain circumstances, where the clay cover was particularly thin, eg about one-third of a tunnel diameter, arrays were also drilled at the Thames Gravels/London Clay interface and within the Thames Gravels. Where permeation grouting of the Terrace Gravels had been undertaken horizontal arrays were used, but elsewhere the holes were drilled from above the water table at varying declinations. In two instances where grouting was required in areas with piled foundations, the arrays were targeted at or near the toe of the piles.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Drilling was undertaken from a platform installed in the shaft at an appropriate level. A pedestal rig is ideal for drilling radial holes, and drill lengths of up to 70 m were achieved. The holes were 125 mm diameter (see Table 11.2) and were cased only over the first few metres from the shaft. The vertical alignments of some of the first holes drilled were measured. Sag, ie deviation below the theoretical level, was recorded in all holes, with an average of 300 mm at 30 m length. The amount of sag could increase rapidly with increasing hole length above 30 m; and at 70 m length up to 3 m sag was recorded. Surface and structure movements occurred as a result of TAM installation. For example, up to 10 mm settlement occurred over a wide area at Waterloo. Conversely, heave was generated in some areas, with up to 15 mm below the Waterloo and City Line. Drilling of inclined arrays within the Terrace Gravels was particularly prone to generate settlement, with up to 30 mm recorded. A maximum spacing of 4.5 m at the edge of the “trough” above the tunnel (where the majority of grouting was expected) was adopted initially, increasing to 6 m at the edge of the arrays where these extended farther from the tunnels. The method of concurrent grouting that was developed (see Section 1 1.6.5) meant that a pre-determined pattern of injections relative to the tunnel face was used which advanced in 2 m increments. To accommodate this pattern fully a 2 m maximum spacing is necessary within the trough; additional arrays were installed to fulfil this requirement. The arrays of TAMs installed at Waterloo station are shown on Figure 1 1.16 (a). Contract 103 Waterloo to London Bridge

At Southwark station access was available in and around the railway viaduct, which was the primary structure to be protected, over the area requiring compensation grouting. The arches of the viaduct could be used for drilling and grouting, and consequently TAMs were installed from the surface in vertical and inclined holes. Many holes were cored through the piers and foundations of the viaduct and extended 2-3 m into the London Clay. Most of the TAMs installed in the Southwark station area were intended both to carry out permeation grouting of the Terrace Gravels and to implement compensation grouting within the upper few metres of the London Clay. Over wide areas the spacing was determined by the more onerous requirements for permeation grouting, for which a spacing of 1.3 m was adopted. An indicative cross-section is shown on Figure 1 1.15 (b) and the plan layout of TAMs for part of Southwark station is shown on Figure 1 1.16 (b). TAMs installed from shafts on Contract 103, were drilled at an elevation of 2-3 m below the interface with the Terrace Gravels. The tubes were installed at a declination of 5-1 Oo, giving a minimum cover above the crown of the tunnels of 4 m. The maximum spacing at the end of the TAMs was generally 2.5 m, but increasing to 6 m where the arrays extended beyond the settlement trough over the tunnels.

Ch 11 Protective measures

163

Contact 104 London Bridge

On Contract 104 disused tunnels prodded an alternative method of gaining access to the desired elevation between the tunnels and the €oundations of overlying structures. The CoIlstNCtioIlof an additionaltunnel was also adopted to avoid the use of permanent works shafts or the need to construct temporruy shafts. The elevation of these existing tunnels determined the level of the grouting horizon. The TAMs were installed as deep as 7 rn below the top of the London Clay, although the preferred elevation was 3-5 m higher. The installation of TAMs fhm tunnels allows a parallel m y , and a generally

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

constantspacingof2mw~adoptedasshownonFigure11.16(c).

Bg Ben Clock Tower S W 411

100 rn

w

w

N

T Disused CItyandSOUth LondmLine

.

.

Figure 11.I6

164

Plan layouts of TAM amys (a) at Westminster, (b) at Southwark and (c) at London Bridge

Building response to tunnelling

Contract 105 London Bridge to Canada Water On Contract 105, shallow shafts about 3 m deep were generally adopted, with the arrays drilled at declinations typically between 10’ and 60’ from the horizontal. This configuration was adopted to avoid drilling from elevations below the water table in granular soils. Multiple arrays were provided to accommodate both permeation grouting to form a stiffened raft and subsequent compensation grouting. The stiffened raft was intended to be about 1.5 m thick, and generally three fans of TAMs were installed at varying declinations as illustrated in Figure 1 1.15 (c). The spacing of the TAMs was governed by the permeation grouting and the pipes were extended beyond the zone of treatment to allow compensation grouting injections to be made below the raft.

11.6.5

Implementation strategies

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Three phases of grouting are common in the methodology adopted on each of the construction contracts. The nomenclature varies, as do details of the methodology used to manage the grout injection locations and volumes. The three stages are essentially before, during and between or after tunnelling; and they comprise: Before tunnelling. Generally termed pre-treatment or conditioning, this is a preparatory phase to make the ground “tight”, to ensure that a rapid response is obtained when it is required in the next stage of injections. Pre-treatment is needed to counteract effects from drilling and TAM installation, to compress loose layers and fill voids, and to be sure that the stress conditions will be such that subsequent grouting fractures will be horizontal and produce vertical displacements. Generally, a uniform intensity (or volume per unit area) of grout is injected over the full area of the arrays. The total volume i s injected in a number of passes, usually either two or four using alternate ports andor TAMs. Passes are successively implemented until heave is observed. If necessary, the cycle of passes is repeated. Any ground treatment, generally permeation grouting of granular material, to stiffen

soils between the grouting horizon and the foundation would be carried out at this stage. During tunnelling. This phase is truly compensation grouting, where injections are made contemporaneously with tunnel excavation such that the movements are mitigated as they are created. It is also referred to as concurrent grouting. Management of compensation (concurrent) grouting operations requires rapid production of proposed locations and volumes of injection, which may be modified on a shift-to-shift basis depending on the interpretation of performance. Between or after tunnel drives. Essentially grout jacking, this phase is to reverse settlements that have already occurred, to manage slow ongoing settlement or to pre-lift in anticipation of settlement from a further tunnel drive. This phase is sometimes referred to as observational grouting. The volume and distribution of grout is generally based on the desired distribution and magnitude of heave, making due allowance for the depth of the grouting horizon and the potential effect on the overlying structures. The volumes are selected so that several passes of grouting are needed to achieve the required movement, providing scope for the grouting pattern to be revised based on observed performance.

Ch 11 Protective measures

165

Contract 102 - concurrent grouting The method adopted to manage concurrent grouting was to establish the best estimate of the volume of grout required for a given tunnel advance and to predetermine a pattern of injection locations relative to the tunnel face. The volume of grout required can be assessed fiom the estimated volume loss and a “grout efficiency factor”, which simply represents a factor by which the theoretical volume loss is increased.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Once the total grout volume for a specified tunnel advance is known this can be divided into individual injections. The number of injections depends chiefly on the maximum volume deemed appropriate for a single injection. The location of injections is influencedby the opposing considerations of injecting grout close to the source of movement and the need to avoid generating unacceptable deformationsor stresses within temporary or permanent works supports. The proximity of the grouting horizon to the crown of the tunnel is the most influential parameter. For example, for the eastbound nmning tunnel drive (4.9 m diameter, 3.5 per cent volume loss and a grout efficiency factor of 0.49, the total volume of grout to be injected per 2 m advance was 2870 litres, representing 7.8 per cent of the excavated volume. Forty injections were used,with variable grout volumes, the maximum volume for an individual injection being about 120 litres. The layout of injections was divided into five blocks transverse to the tunnel, with eight injections in each, spaced across a “trough” defined by 45” lines originating from the invert of the tunnel. The proportion of the total grout injected in each block did vary, but generally an equal division was used, at least initially. The blocks were at distances of 16 m, 14 m and 0 m ahead of the ring under construction and 2 m and 4 m behind it. The relative volumes of the eight injections within each block were based on their lateral spacing and offset from the tunnel centre-line in an attempt to produce a distribution of grout consistent with an approximationto a Gawian distribution. A typical layout of injections is shown on Figure 11.17.

I

Location of grout injections relative

Station box diaDhraam walls

to rings under

/

construction

A

16mahead

0

14mahead

Tunnel bring machine I

Diredlyover

2 m behind 4 m behinb

B

3

- - - - - - - - -

I Grwt volumes in l i s

JLE Eastings

Figure 11.17

166

Typical layout of grout injections for concurrent grouting

Building response to tunnelling

The basis of compensation grouting is that it is an observational technique and the volumes of grout injected could be rapidly changed either by omission of some injections or a proportional change in the volume for each injection. Minor variations were controlled on a shift-by-shift basis, but if systematic modification was necessary a new grouting proposal was issued. The number of “blocks” and the number of injections within each block were varied as required (see Chapter 20 in this volume and case study Chapters 24 and 26 in Volume 2). For hand-mined tunnels, a circular pattern of injections was used. This system arose as a result of an “exclusion” zone in plan defined by a fixed distance from the tunnel face. The injections were thus made around the perimeter of the exclusion zone.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Contracts 104 and 105

On Contract 104, at London Bridge, there was a different approach to the grouting. The extent and nature of grouting to be implemented was decided at meetings on a day-today basis. To facilitate implementation a system of uniform injection quantities and spacings were adopted. The grout quantity per injection was held constant (initially at 30 litres) and the number of ports injected varied to give what was termed a full, half or quarter pass. A full pass comprised one injection per 2 m2, a half pass one injection per 4 m2 and so on. The parallel array of TAMs made this system easy to administer. Compensation grouting, in the strictly defined sense of injection during tunnel excavation, was not carried out on Contract 105. Rather there were stages of conditioning and grout jacking. Overall strategies

There is a range of strategies and methods of implementation of compensation grouting. At one extreme, grouting is aimed at controlling an identified mechanism of movement of a structure; at the other, grouting can be carried out extensively in an attempt to maintain the maximum settlement at less than a specific limit. Alternatively, grouting can be managed to minimise the development of distortions. The JLE works clearly demonstrated that compensation grouting is able to control movements to within close tolerances in a range of circumstances and generally with limited clearance between the crown of the tunnel and the grout arrays.

11.6.6

Monito ring requirements Monitoring its performance is an integral part of compensation grouting. Sufficiency, accuracy, interpretation, presentation and dissemination of data are equally important facets of a monitoring system. The JLE contract specifications required real-time monitoring of both ground and structure movements to be undertaken in all areas where there was compensation grouting. Electrolevels were the principal type of instrument used. The contractors, however, being extremely sceptical of the reliability and accuracy of the systems installed, preferred to rely on traditional survey methods. Thus although extensive electronic systems were installed, the works were largely controlled using precise levelling. This resulted in a very high number of readings and very high frequency of readings (every two hours) at critical stages of construction. An indication of the quantity of surveying undertaken is given in Table 1 1.4. On

Contract 102 almost % million readings were taken on over 3000 separate points, and for the six contacts nearly 1 million readings on more than 7500 points. These quantities

Ch 11 Protective measures

167

do not include multiple readings on a single point in one day or any of the extensive monitoring of existing railway tracks. Electronic data were used where access for manual monitoring was restricted, eg in operational LUL tunnels, and other real-time monitoring devices were developed for special circumstances (eg to measure the tilt of the Big Ben Clock Tower, see Harris et al, 1999 and Chapter 28), which proved extremely useful.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Table 11.4

Extent of precise level monitoring

Area

Construction

C 10 1

Green Park passenger tunnels

C 102 North C 102 South

No of readings

No of points 340

37 500

Green Park to Westminster: step-plate junctions; running tunnels, escape shaft; Westminster station

1570

232 200

Westminster to Waterloo, running tunnels crossover and Waterloo station

1720

247 300

Total C102

3290

470 500

C 103

Waterloo to London Bridge

1270

170 100

Cl04

London Bridge

1090

223 300

C 105

London Bridge to Canada Water

1820

69 000

Totals ClOl to C105

7810

970 400

Many other types of monitoring were undertaken that contributed to the control of compensation grouting, including those listed below. In-tunnel

0

crown displacements by levelling cross-sectional distortion by total station prisms stresses in lining by load cells and strain gauges (shotcrete and segments) loads in temporary works by load cells and strain gauges.

Ground 0 0

0

horizontal movements by manual and electrolevel inclinometers vertical movement by electrolevel, rod extensometers and deep pins pore pressure by piezometer.

Structure

0

168

vertical displacements by levelling and electrolevels horizontal displacements by total station prisms and tape extensometer defect monitoring by manual and electronic displacement gauges stresses in structural members derived from strain gauge measurements tiltherticality by total station prism.

Building response to tunnelling

~~

11.7

GROUTING CONTROL AND RECORDS Control and approval

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

All injections of grout to be made were subject to the consent or approval of the engineer’s representative. On Contract 102 grouting proposals had to be in writing and include a statement of purpose of the grouting with details of the injections in terms of shaft, TAM, port and volume. While all injections would terminate when they had reached the specified volume, other termination criteria such as pressure and flow rate limits were also set. Details of the grout mix, any restrictions on the sequence or timing of injections, monitoring requirements, method of calculation of grout volumes were also given as required. Communications procedures were established between the grouters and the tunnellers via the monitoring office. The monitoring office co-ordinated this information with the results of instrumentation and had authority to terminate the grouting or the tunnelling or both if the situation required it (Osborne et al, 1997). At any one time at the peak of tunnel construction activity, up to ten grouting proposals could be implemented simultaneously. Modifications would frequently be required based on observed performance, which had to be evaluated on a shift-by-shift basis. Committees containing representatives of the main contractor (tunnellers, monitoring and design), the grouting sub-contractor, temporary works designers (traditional and NATM) and the supervising engineer met at least daily and more often if necessary to review tunnelling, grouting and settlement progress, give consent to proposals and issue instructions when necessary. Representatives of third parties (eg owners of affected buildings) were often present as well. Similar methodologies were adopted on the other contracts. Grouting records and their presentation The grout injection systems employed automatically record a range of parameters associated with each injection. An example of data from Contract 102 is shown on Figure 1 1.18. As a great many injections were made (Table 1 1. I), presentation of the information for it to be interpreted in conjunction with tunnel progress and monitoring records proved problematic. The data were provided by the contractor in digital format, and spreadsheet-based software was developed to manipulate the data to allow graphical presentation of the distribution and density of grout volumes in plan format. Two types of plot were used. In one, symbols showed the locations of individual ports from which injections were made, the size of the symbol being proportional to either the volume of a single injection or to the sum of injections over a given time period. The other method was to determine contours of grout intensity or thickness. The raw data from which contours are to be produced consists of the co-ordinates of the injection point and the grout volume injected. In order to take into account the varying spacing of injection points, it is necessary to translate the injection volumes into equivalent thicknesses and hence to make an assumption regarding the area over which the grout spreads. This is achieved by dividing the plan area of the grouting into an orthogonal grid and assigning all or part of each injection to a grid node. This can be achieved by: assigning the full volume of grout to the grid square within which the injection lies 0

assuming the grout spreads a fixed distance from the injection point and assigning the injection volumes in proportion to the areas within different grid squares assuming the grout forms a fracture of fixed thickness (such that the distance of grout spread from the point of injection is proportional to the square root of the

Ch 11 Protective measures

169

injection volume) and assigning the injection volumes in proportion to the areas within different grid squares defining other, more complex grout distributions.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Once a grid is defined, contours can be produced using a variety of weighting or smoothing functions. Examples of grout contour plots are given in Chapters 25 and 26, and the derivation of those presented in Chapter 26 is described in detail in Chapter 20.

41

I

41

I

131 151

dCB46%

Figure 11. I 8

11.8

1

551 03/06/951 1525321 15:28:131 551 03/06/951 15:29:481 15:31:53/

16.081 15.691

6.471 6.861

6.471

6.861

201 201

POILirn Vol. 201Lirn Vol.

Example of grout injection records

EXPERIENCES OF COMPENSATION GROUTING ON CONTRACT 102 The high level of compensation grouting employed on JLE Contract 102 successfully limited movements of a wide range of structures in many situations. In particular, volume loss movements were reduced to negligible magnitudes for tunnels of various diameters and lining types (eg Harris et al, 1996). Small clearances between the level of grouting arrays and the crown of the tunnel were accommodated with relative ease. Compensation grouting has in many situations exceeded the expectations of the designers. For example, at Waterloo the omission of numerous other structural protective measures meant that the Railtrack viaduct had to be maintained within the specified limit on a slope of 1:1000 by compensation grouting alone, and this was successfully achieved throughout the construction period. There are many other examples of the fine tolerances on movement that can be achieved using compensation grouting. The most notable example is Big Ben Clock Tower (Harris et al, 1999, and Chapter 28), the tilt of which was maintained for a period within a range of 1 :50 000 almost within the accuracy of the monitoring data. In most situations grouting was directed at controlling distortions with the magnitude of settlement being a secondary consideration. This was achieved by directing maximum effort at the mitigation of volume loss movements with concurrent grouting. Observational grouting to reduce movements after tunnelling or prior to a subsequent tunnel pass also played an important role particularly for structures where there was a plane of weakness and where movements could concentrate. At Waterloo, a limit on absolute settlement was specified and much of the grouting was directed at complying with this requirement prior to its relaxation. It is estimated that

170

Building response to tunnelling

some 20 per cent of the grouting carried out at Waterloo was directed at reducing absolute settlement and can be considered to be superfluous to the primary aim of protecting the structures from damage. Since the grouting at Waterloo comprised 60 per cent of the total grouting carried out on Contract 102, this represents a significant cost. The success of the compensation grouting on Contract 102 is illustrated in the case histories presented in Chapters 25 to 28. The JLE was by far the largest application in the UK of the relatively new technique of compensation grouting and much was learnt during the project. Some of the difficulties that were encountered during construction are considered in the following sections.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Tunnel progress

To be fully effective in mitigating the immediate volume loss movements associated with driving a tunnel, compensation grouting has to be concurrent with the excavation. It is better, therefore, to intercept movements as they develop, rather than to reduce them subsequently by heave-induced grout jacking (or observational grouting). It is not practical to link the two processes (tunnelling and grouting) so that they are truly concurrent, hence some alternation of settlement followed by heave is inevitable. The magnitude of the cycles can be kept to a minimum by making sure that the grout volume required for a specific advance of the tunnel is injected before further advance of the tunnel. This method was adopted on Contract 102 with either a 1 m or 2 m advance of the tunnel used as the control. In general, the rate at which grout could be injected was sufficient to prevent systematic delays to the rate of tunnelling. The exception to this rule was for the shield-driven running tunnels, where very high rates of production (advance) could be achieved in the absence of grouting. The extent of delay was magnified by the high volume losses that occurred; this substantially increased the volume of grout required. Calculations were undertaken to determine the maximum rate of advance that could be accommodated by the grouting in advance of the tunnelling. This was about 8 m per shift, which compares to the maximum production achieved on these drives without grouting of 30 dshift. The primary parameter that determines the time required for grouting is the quantity of grout required to balance the volume loss movements. This in turn is dependent on the actual volume loss created by the tunnel excavation and the “efficiency” of the grout. The calculation then depends on the number of shafts and the number of lines used, the rate of grout flow and the time to position the packer. For a given situation the best achievable rate of tunnelling can be simply calculated. Delays to tunnelling, of course, could occur as a result of other factors associated with grouting. In some instances, the timing of a specific grout injection was related to stages in the cycle of the advancing tunnel. In general, this type of limitation was precautionary and imposed only when a novel situation was encountered. The constraints on the timing of grout injection were progressively relaxed once sufficient observations had been made to confirm that no deleterious effect was likely. Tunnel deformations and loadings

The process of compensation grouting involves the injection of grout into the ground at high pressure. The process is essentially a jacking operation that produces movement of the ground regardless of whether the grout injected is concurrent with tunnel construction or observational afterwards. A reaction force is necessary to generate the required (upwards) movement. There is potential for loadings to be generated in any

Ch 11 Protective measures

tunnel lining or temporary works supports situated below an area of grout injection. The controlling factors can be summarised as: the vertical total stress at the grouting horizon, which has a strong influence on the grout pressure in the ground 0

the vertical spacing between the grouting horizon and the tunnel or excavation, which influences the spread of load the plan location of injections relative to the excavation

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

the properties of the grout particularly with respect to the shape and extent of the grout bulb/fracture formed 0

the quantities of grouting undertaken

0

combination with other factors eg close proximity tunnelling.

In general during concurrent grouting, exclusion zones were imposed around excavation faces, ie no grout was injected within specified plan distances of the tunnel face. The location of injections had a negligible effect on the efficiency of the grouting. Any problems encountered due to the effect of grouting on excavation supports were adequately resolved by either modification of the grouting parameters (mix, volume, timing etc) or by strengthening the excavation supports, or by temporary suspension of grouting operations. Break-ups. At some early stage of construction it is inevitable that an excavation will have only partially completed support, eg during break-ups, support is exclusively provided by timbering with no coherent load paths. This type of support is provided primarily to minimise the potential for local instability and is usually in place only for short periods. In these situations, grouting was suspended until a sufficiently robust structure had been formed, eg after three rings had been constructed at a break-up. The movements associated with such limited excavations do not generally present a problem, but where grouting was necessary the face was boxed up and loads monitored within the temporary supports to enable rapid termination of grouting if necessary. Ring-build; A further example where limitations were required on the timing of grout injections was during the ring-build of the running tunnel expanded linings. The ringbuild area on the running tunnel shield was protected only by flexible “fingers” over the crown section (see Chapter 10). Ground deformations during the build cycle had to be limited to ensure satisfactory completion of the expansion of the lining. NATM temporary lining. The deformation of the Lower Concourse NATM shell at Waterloo station approached the limits set by the designer during its construction. The extent to which these deformations were due to grouting rather than ground loading associated with tunnelling has not been ascertained. However, six large excavations had to be made from this tunnel to form connections to the parallel platform tunnels. The design of the complex three-dimensional structure required by this geometry was largely empirical, hence the prediction of movements of the structure that should be expected was at best approximate. The behaviour of large excavations with complex shapes is inadequately understood, so the additional impact of compensation grouting was avoided because there was little risk of damage to overlying structures. Temporary propping to tunnels. The upper-level tunnels at Waterloo required construction of the large close-proximity tunnels. The size and density of tunnelling in this area can be seen on the isometric reproduced as Figure 1 1.19. High stresses were induced in some parts of the temporary propping in the upper concourse tunnel. The props were necessary to allow the construction of junctions linking adjacent tunnels, but they were installed during tunnel excavation to support the face and to cater for the

172

Building response to tunnelling

expected effects from compensation grouting. The grouting arrays were only 3 m above the crown of the 10 m ID upper concourse tunnel. The early installation of a stiff propping system took load from adjacent excavations as well as from grouting. The main vertical props were able to sustain full overburden pressure, but there was localised loading between the main supports as a result of grouting activities in combination with low resistance to rotation andor shear at joints between segments.

Lift shaft

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Bakerloo

Jubilee Line escalator shafl-

Figure 11. I 9

Isometric of upper level tunnels at Waterloo station

Distortions of the overlying structure were also at the limit of acceptability. The above combination of factors meant that remedial measures had to be implemented, including strengthening of the temporary works, reduced grouting, resetting of structural jacks to allow the adjacent structure to be lowered, and the rebuilding of two rings (see Osbome et al, 1997 for further details). Diaphragm wall movements

In one incident, movements of the diaphragm wall to the Westminster station box were clearly increased by compensation grouting. The locations of the grout injections and the cumulative volume of grout injected from each port between 8 May and 6 June 1996 are shown on Figure 1 1.20. The distribution of grout injected is shown in the alternative form of grout thickness or intensity in Figure 1 1.2 1. The movements recorded on an inclinometer in the wall panel indicated on Figure 1 1.20 are shown on Figure 11.22. Excavation of about 3.5 m of ground directly below the 1.5 m-thick transition slab was carried out within this period (between 10 and 17 May 1996). The recorded movements of the wall were up to 14 mm, which was much greater than expected for this extent of excavation. Figure 1 1.23 is a comparison of the development of maximum wall movement with the total volume of grout injected within a 15 m radius of it. A very close agreement is shown between the two, most of the movement occurring after the excavation had been completed. It is concluded that the grouting substantially increased the wall movements over this period. In general, substantial amounts of grouting were undertaken close to the diaphragm walls with no measurable effect on their movements. In this instance, the combination of large volumes of grout injected and the lack of any props installed below the grouting horizon produced the observed effect.

Ch 11 Protective measures

173

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

.oo

Figure 11.20

Effect on diaphragm wails: location of grout injections

42721

42711

? s

r

Y 7

42701

42691

4267

5

13260

13265

13270

13275

13280

13285

13290

13295

13300

13305

13310

'

15

JLE Eastings

Figure 11.21

174

Effect on diaphragm wails: grout intensity contours

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 11.22

Effect on diaphragm walls: wall displacement profiles

Figure 11.23

Effect on diaphragm walls: correlation of wall movement and grout volume

Ch 11 Protective measures

175

11.9

REFERENCES BAILEY, R P, HARRIS, D I and JENKINS, M J (1999). Design and Construction of Westminster Station on the Jubilee Line Extension. Jubilee Line Extension Supplement, Proc Inst Civil Engineers, Civil Engineering, Vol 132, Special Issue 2, 1999 CARTER, M D, BAILEY, R P and DAWSON, M P (1996). Jubilee Line Extension, Westminster Station design. In: R J Mair and R N Taylor (eds) Geotechnical Aspects of Underground Construction in Soft Ground, Balkema, pp 8 1-86 CRAWLEY, J D and STONES, C S (1996). Westminster Station - Deep foundations and top down construction in central London. In: R J Mair and R N Taylor (eds) Geotechnical Aspects of Underground Construction in Soft Ground, Balkema, pp 93-96

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

HARRIS, D I, MAIR, R J, LOVE, J P, TAYLOR, R N and HENDERSON, T 0 (1994). Observations of ground and structure movements for compensation grouting during tunnel construction at Waterloo Station. Gdotechnique, 44,4, pp 691-714 HARRIS, D I, MAIR, R J, BURLAND J B and STANDING J R (1999). Compensation grouting to control tilt of Big Ben Clock Tower. In: 0 Kusakabe, K Fujita and Y Miyazaki (eds) IS-Tokyo '99. Geotechnical Aspects of Underground Construction .in Soft Ground, Balkema, pp 225-232 HARRIS, D I, MENKITI, C 0, POOLEY, A J and STEPHENSON, J A (1996). Construction of low-level tunnels below Waterloo Station for the Jubilee Line Extension. In: R J Mair and R N Taylor (eds) Geotechnical Aspects of Underground Construction in Soft Ground, Balkema, pp 361-366 LINNEY, L and FRIEDMAN, M (1996). Protection of buildings from tunnelling induced settlement using permeation grouting. In: R J Mair and R N Taylor, eds Geotechnical Aspects of Underground Construction in Soft Ground, Balkema, 399-404 MAIR, R J, HARRIS, D I, LOVE, J P, BLAKEY, D and KETTLE, C (1 994). Compensation grouting to limit settlements during tunnelling at Waterloo Station, London. Tunnelling '94, IMM, London, pp 279-300 MUGURAMOORTHY, C, MURRAY, G S, BALL, P, BRACEGIRDLE, A and LEIPER, Q (1996). Ground movements and vertical compensation grouting during shaft construction. JLE, Green Park Station - Contract 101. In: R J Mair and R N Taylor (eds) GeotechnicalAspects of Underground Construction in Soft Ground, Balkema, pp 4 17-422 OSBORNE, N, MURRAY, K, CHEGINI, A and HARRIS, D I (1997). Construction of Waterloo Station upper level tunnels, Jubilee Line Extension Project. Tunnelling '97, Instn Mining and Metallurgy, London, pp 639-662 RAWLINGS, C G, HELLAWELL, E E and KILKENNY, W M (2000). Grouting for ground engineering, Publication C5 14, CIRIA, London

176

Building response to tunnelling

12

Finite element analysis of St James’s Park greenfield reference site

T I Addenbrooke and D M Potts Note. This chapter is a summary of the report prepared by Dr Addenbrooke about the results of the finite element analysis carried out in 1995 at Imperial College in support of the research project on the instrumented greenfield site in St James’s Park.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

12.1

INTRODUCTION Part of the LINK research programme was to make Class A predictions of ground and building response to tunnelling works on the new Jubilee Line Extension. The intention was to use the best methods available to the profession at the time, including numerical analyses. In the case of the instrumented greenfield site at St James’s Park, it was decided that finite element analyses of the construction of both the westbound and eastbound tunnels would form part of the Class A predictive work. Imperial College Soil Mechanics Section was asked to perform finite element analyses, using the Imperial College Finite Element Program (ICFEP), to predict the ground response to construction of each running tunnel. It employed techniques believed to give the best predictions possible when modelling multiple tunnel construction. At the time of the analyses, October 1995, the westbound tunnel had already passed beneath the control site. The volume loss into the westbound tunnel was therefore known, and had been measured as 3.3 per cent. This was therefore imposed on the first stage of the analysis (the method for achieving a desired volume loss in a plane-strain analysis is discussed in Section 12.2.4). The second stage of the analysis modelled the 8%-month rest period between excavations of the westbound and eastbound tunnels. The third stage modelled excavation of the eastbound tunnel. Predictions were based on two possible volume losses: 2 per cent - the value used for the Parliamentary hearing, and the 3 per cent that the contractor envisaged at that stage. Only the results with 3 per cent volume loss are presented here. The ground response for the westbound tunnel presented in this chapter are Class C 1 predictions, whereas those for the rest period and the eastbound tunnel are Class A predictions based on the definitions of Lambe (1973). (For Lambe’s definitions, see Section 14.1 - Eds.)

12.2

ANALYSIS DETAILS Eight-node plane-strain isoparametric elements with reduced integration were used to represent the soil. Three-node Mindlin beam elements with selected reduced integration were used to model the tunnel lining. An accelerated modified Newton-Raphson scheme with a sub-stepping stress point algorithm was employed to solve the non-linear finite element equations. A coupled consolidation formulation was employed. These features are discussed in Potts and Zdravkovic (1 999).

Ch 12 Finite element analysis of St James’s Park greenfield reference site

177

12.2.1

Geometry and timescales

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The analysis presented in this chapter modelled excavation of the two running tunnels of the Jubilee Line Extension beneath St James’s Park. The soil profile and twin tunnel geometry are presented in Chapter 25 in which some of the results of the field measurements are discussed. The finite element mesh used for this analysis is shown in Figure 12.1 with the soil layering indicated. The axis of the deeper tunnel (westbound) is 3 1 m below ground level and was excavated first. The axis of the shallower tunnel (eastbound) with its vertical axis 2 1 m from that of the deeper tunnel is 20.5 m below ground level. The tunnel diameter modelled was 4.75 m, based on early instrument layout drawings for St. James’s Park, rather than the correct external diameter of 4.85 m. The target volume loss for excavation of the westbound tunnel was therefore 3.4 per cent of the analysed diameter, equal to 3.3 per cent with the real diameter. For the eastbound tunnel the volume loss achieved was 3.1 per cent of the analysed diameter, equal to 3 per cent with the real diameter.

5m 5m L

Figure 12.1

Finite element mesh for St James’s Park site showing tunnels excavated

Table 12.1 shows the construction activities modelled and their respective timescales. The table also draws a comparison with the measurement periods referred to in Chapter 25. These periods are used for the presentation of results in Section 12.3. Table 12.1

12.2.2

Construction activities

Activity

Modelled time period

Equivalent field monitoring period

Initial conditions (steady-state underdrained pore pressure regime)

-

Period 1

Excavation of the westbound tunnel

8 hours

Period 2

Rest period

8% months

Period 3

Excavation of the eastbound tunnel

8 hours

Period 4

Soil parameters for the analyses This section discusses the constitutive models used for permeability and stress-strainstrength behaviour of the soil. Section 12.2.3 contains tables of parameters. The upper strata of made ground and Terrace Gravels were modelled as fully drained non-consolidating materials. The London Clay and Lambeth Group Clay were attributed with a non-linear isotropic permeability based on the model of Vaughan (1989). In this model, the permeability, k, depends on the mean effective stress level, p‘, and is therefore continually being updated during the analysis: k = k, . exp-BP’,where k, is the permeability at zero mean effective stress, and B is a constant incorporating the initial

178

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

void ratio at zero effective stress, and the coefficient of volume compressibility, m,. The premise is that as the voids ratio reduces under increased mean effective stress, so the permeability also reduces. The B value was selected to match the profile of permeability from falling head tests carried out 300 m from this site at the Palace of Westminster, reported by Burland and Hancock (1977). The model gave an excellent prediction of the measured pore water pressure profile (see Section 12.2.4 below). The made ground was isotropic linear elastic. The Terrace Gravels, London Clay and Lambeth Group Clay were elastic perfectly plastic. Pre-yield they were isotropic nonlinear elastic using the model of Jardine et a1 (1984). This model reproduces both the variation in stiffness with mean effective stress and the reduction in soil stiffness with increased strain (the shear modulus reducing with shear strain, and the bulk modulus with volumetric strain). The isotropic non-linear elastic parameters for London Clay were based on curve fitting to stiffness-strain data from high-quality triaxial extension tests. (Mair, 1992, states that extension tests are of most relevance for a tunnel excavation.) Figures 12.2 and 12.3 present respectively the stiffness-strain curve and the stress-strain curve from the model compared with triaxial data. Mohr-Coulomb yield surfaces and plastic potentials defined the plastic behaviour. Addenbrooke (1 998) and Addenbrooke et a1 (1997) present details of the stress-strain-strength models and the parameters utilised here. 600

ICFEP model curve

G -

triaxial extension test

P'

400

G = secant shear modulus p'= mean effective stress

200

-

0 0.001

I

I

I

0.01

0.1

1.o

axial strain (%)

Figure 12.2

Small-strain curves from undrained triaxial tests on London Clay

.._ _...'....._.... _... ,._... 0

0.01

0.02

0.03

0.04

005 A x m Slraln

Figure 12.3

0.06

007

0.08

0.09

I 0.1

(X)

Stress-strain curve and excess pore water pressure-strain curve for undrained triaxial test

Ch 12 Finite element analysis of St James's Park greenfield reference site

179

12.2.3

Model parameters Table 12.2

Linear elastic parameters: made ground

Young's modulus

Poisson's ratio

3000 kN/m2

0.2

Linear elastic parameters: tunnel lining

Table 12.3

Young's modulus (W/m2)

Poisson's ratio

Cross-sectional area (m2/m)

Second moment of area (m4/m)

28 x 106

0.15

0.168

3.95136

x

10'4

Note: unit weight = 24 kN/m3

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Table 12.4

Mohr-Coulomb yield surface parameters, plastic potential parameters, and unit weight Made ground

Gravel

London Clay

Strength parameters

linear elastic

c'= 0 kPa

c'=

Angle of dilation

linear elastic

Bulk unit weight (m/m3)

ydly= 18 'Ysat = 20

Table 12.5

6

6

= 35.0'

o Wa

c'=

0 kPa

6= 27.0'

= 25.0'

v' = 12.5'

v' = 17.5'

Lambeth GrouD clav

v'= 13.5'

Non-linear permeability model

London Clay

k, = 1 x 10.' m/s

B = 0.007

Lambeth Group Clay

k,

10-9m/s

B = 0.007

=2 x

Table 12.6

Small strain stiffness model parameters

Terrace Gravels

1380 1248 5.0 x 10-4 0.974

0.940

8.83346 x 10-4 0.34641

2000

London Clay

1120 1016 1.0 x 10-4 1.335

0.617

8.66025

10-4 0.69282

2333

Lambeth Group 1000 1045 5.0 x 10-4 1.334 Clay

0.591

13.8564 x 10-4 0.38105

2667

Terrace Gravels

275

225

2.0 x 10-3 0.998

1.044

2.1 x 10-3

0.20

5000

London Clay

549

506

1.0 x

2.069

0.420

5.0

0.15

3000

Lambeth Group 530 Clay

460

5.0 x 10'4 1.492

0.678

1.5 x 10-3

0.16

5000

x

x

10"

Note: Where C, a,y, 6, and h are curve fitting constants, q, is the deviatoric strain, E, is the volumetric strain, G is the shear modulus and K the bulk modulus. The constants CI to C6 are given the symbols A, B, C, R,S, and T, respectively in Chapter 13).

180

Building response to tunnelling

\

Top of London Clay

1

ICFEP - non-linear isotropic permeability

-I0

Depth below water table (m)

St James’s Park field data

Westminster field data

-20

.’...

-30

’..

hydrostatic

Bottom of London Clay I - - - - -

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

-40

I

I

0

100

I

1

200

300

Pore pressure (kPa)

Figure 12.4

12.2.4

Underdrained seepage pressure profile

Initial stresses and boundary conditions The initial stresses prescribed an underdrained pore water pressure profile from a water table located 4.5 m below the ground surface (hydrostatic in the Terrace Gravels). Figure 12.4, a plot of pore pressure against depth below water table, shows the agreement between the prescribed underdrained pore water pressure profile and field data from St James’s Park and Westminster. This prescribed profile is consistent with the non-linear permeability defined above. An initial effective stress ratio KOof 0.5 was prescribed in the made ground and gravel, and 1.5 in the London Clay and Lambeth Group clay. In the absence of site measurements of KO,the value adopted in the London Clay is from the “upper-bound” profile given by Hight and Higgins (1995) for London Clay between 10 m and 30 m below the ground surface. Their profile is based on suction measurements in samples of London Clay using the filter paper technique. The displacement boundary conditions permitted no horizontal displacement along the two vertical mesh boundaries, and no displacement along the bottom horizontal boundary. Hydraulic boundary conditions dictated that the pore water pressures in the Terrace Gravels remained hydrostatic. The initial pore water pressure along the bottom boundary to the clay strata was maintained throughout the analysis. The vertical side boundaries were modelled as no-flow boundaries. The tunnels were treated as drains using a special boundary condition that prevents water being drawn across the boundary at nodes where suction exists in the adjacent soil, but permits free flow at nodes where the soil water pressures are compressive. Incremental excavation was used, monitoring the volume of the developing surface settlement profile. The tunnel lining (modelled as a continuous concrete ring for which parameters are listed in Section 12.2.3) was constructed once the desired volume loss had been obtained for each excavation.

Ch 12 Finite element analysis of St James’s Park greenfield reference site

181

12.3

RESULTS Results are presented over a distance corresponding to the field surveying points, not for the full finite element mesh width.

12.3.1

Vertical displacements at the ground surface Figure 12.5 shows the vertical displacement, w,plotted as profiles against the offset from the westbound axis line, for Periods 2,3, and 4. For each profile, the zero datum is the settlement at the end of the previous period.

51

-10

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

i n

Offset from westbound axis (m) 0

10

20

30

40

I

I

I

I

I

. - . 50

*

- Period2

Westbound axis

Figure 12.5

Eastbound axis

I

Vertical displacement profiles from finite element analysis

In Period 2, excavation of the westbound tunnel caused a maximum settlement, w,,,, of 11.1 mm above the tunnel axis. At 50 m from the westbound axis the settlement is less than 1.5 mm. The controlled volume loss achieved for this excavation was 3.6 per cent, equivalent to 3.4 per cent for a 4.85 m-diameter tunnel. During Period 3 settlements continued to develop to a distance of 40 m from the axis line, with the maximum settlement being 6.1 mm. Beyond this distance a very slight heave is predicted. This surface profile is in response to the consolidation and swelling taking place at depth as the excess pore water pressures set up during Period 2 dissipate. In Period 4, the controlled volume loss achieved was 3.1 per cent, equivalent to 3.0 per cent for a 4.85 m-diameter tunnel. Excavation of the shallower eastbound tunnel caused a settlement profile to form with a w,,, of 12.2 mm offset from the tunnel axis line. The offset of w,,, is 5.1 m. In addition to being offset from the tunnel axis line, the profile is not symmetrical about w,,,. This asymmetry can be studied by plotting the natural logarithm of w/w,,,,, against the square of the distance from w,,,, Y’, divided by the square of the depth to the tunnel axis, 22. (If the profile can be matched by an inverted Gaussian curve, this plot will be linear.) This relationship is presented in Figure 12.6 for Periods 2 and 4. The profile for Period 4, due to the eastbound tunnel, has been split into two parts. The dividing point is the position of w,,, and the two parts are the part on the side of the first tunnel (“near first tunnel”) and the part on the side of the second tunnel (“far from first tunnel”). The westbound profile and the eastbound profile far from the first tunnel are very similar in shape. The eastbound profile near the first tunnel, however, is significantly different. The lower values of In(w/w,,,) for any given Y2/Z2 indicate greater settlements on the side of the existing tunnel than on the far side. It is also noteworthy that none of the profiles is linear.

182

Building response to tunnelling

Y2/ZO2

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 12.6

12.3.2

Vertical displacement profiles from finite element analysis

Horizontal displacements at the ground surface Figure 12.7 shows the horizontal displacement plotted as profiles against the offset from the westbound axis line, for Periods 2 , 3 , and 4. Again, for each profile the zero datum is the profile at the end of the previous period. At the end of Period 2, the horizontal displacement profile resulting from westbound tunnel excavation has a maximum value of 6.7 mm occurring 20.2 m from the axis line. On the axis line the horizontal displacement is zero.

101

Period 2

,/-\

-.-.-.

\

Period 3

- - - - _Period 4

\

\ \

\

5

ho

-10

-1 0

J t

Westbound axls Figure 12.7

\ Offset froth\Westbound axis (m)

1o

\ 20\

30

40

50

t Eastboundaxls

Horizontal displacement profiles from finite element analysis

Ch 12 Finite element analysis of St James’s Park greenfield reference site

183

During Period 3 horizontal ground surface displacements continued to develop. The profile shape is very similar to that at the end of Period 2, with a maximum displacement of 4.3 mm. In Period 4, excavation of the eastbound tunnel causes an asymmetric profile to develop. The zero horizontal displacement point is on the tunnel axis line. This is in contrast to the point of w,,, that occurred 5 m nearer the westbound axis line. The horizontal displacement towards the eastbound axis line is a maximum of 10.0 mm on the side of the existing tunnel. In contrast, it is 5.0 mm on the side farther from the existing tunnel.

12.4

REFERENCES ADDENBROOKE, T I (1996). Numerical analysis of tunnelling in stiffclay. PhD thesis, Imperial College, University of London

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

ADDENBROOKE, T I (1998). St. James’s Park:finite element analysis of ground response to tunnelling. LINK CMR Project Report, December 1998, Imperial College, London ADDENBROOKE, T I, POTTS, D M and PUZRIN, A (1997). The influence of prefailure soil stiffness on the numerical analysis of tunnel construction. Gkotechnique, 47, no 3, pp 693-712 BURLAND, J B and HANCOCK, R J R (1977). Underground car park at the House of Commons, London: Geotechnical aspects. The Structural Engineer, vol55, no 2, 87-100 HIGGINS, K G, POTTS, D M, and MAIR, R J (1996). Numerical modelling of the influence of the Westminster station excavation and tunnelling on Big Ben Clock Tower. In: R J Mair and R N Taylor (eds) Geotechnical Aspects of Underground Construction in Soft Ground, London, Balkema, Rotterdam, pp 525-530 HIGHT, D W and HIGGINS, K G (1995). An approach to the prediction of ground movements in engineering practice: background and application. Pre-failure Deformation of Geomaterials, Shibuya, Mitachi and Miura (eds), Balkema, Rotterdam, pp 909-945 JARDINE, R J, POTTS, D M, FOURIE, A B and BURLAND, J B (1986). Studies of the influence of non-linear stress-strain characteristics in soil-structure interaction. Gkotechnique, vol36, no 3, pp 377-396 LAMBE, T W (1973). Predictions in soil engineering. Gkotechnique, vol23, no 1, pp 149-202 MAIR, R J (1992). Unwin Memorial Lecture 1992 - Developments in geotechnical engineering research: applications to tunnels and deep excavations. Proc Znstn Civ Engrs, Civil Engineering, vol93, pp 2 7 4 1 NYREN, R J (1998). Field measurements above twin tunnels in London Clay. PhD thesis, Imperial College, University of London POTTS, D M and ZDRAVKOVIC, L (1999). Finite element analysis in geotechnical engineering: theory, Thomas Telford Publishing, London VAUGHAN, P R (1989). Nonlinearity in seepage problems - theory andfield observations. De Mello Volume, Edgard Blucher, S5o Paulo, pp 501-516

184

Building response to tunnelling

13

Finite element analyses of ground movements from tunnelling below Southwark Park

N Kovacevic, R J Mair and D M Potts

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Note. This section is a transcript, with minor editing and changes necessary for reproduction in this book format, of the report by Geotechnical Consulting Group, dated April 1997 (Geotechnical Consulting Group, 1997), which presents their predictions of the surface and subsurface ground movements to occur at the instrumented greenfield site in Southwark Park, where the tunnelling was through beds of the Lambeth Group. Thus in what follows, the tenses of verbs are the same as those in that report, applying to predictions made at that date. The surface settlements that were measured at this site are presented in Chapter 32. The subsurface movements are the subject of an ongoing research study at Cambridge University.

13.1

INTRODUCTION As part of its contribution to the LINK CMR project, Geotechnical Consulting Group (GCG) has undertaken to carry out finite element (FE) analyses of ground movements due to tunnelling at Southwark Park. As defined by Lambe (1 973), Class A “before event” predictions have been made assuming greenfield site conditions. This report describes the FE analyses and presents the results. (For Lambe’s definitions, see Section 14.1 - Eds.)

13.2

GROUND CONDITIONS Figure 13.1 shows the plan layout of the site together with positions of the analysedinstrumented cross-section and relevant JLEP boreholes. Idealised soil stratigraphy is presented in Figure 13.2. Ground level is generally at about 102.5 m PD. Made Ground and Thames Gravel overlie the various clays and sands of the Lambeth Group, which are underlain by Thanet Beds and Chalk. Initial groundwater conditions are assumed to be similar to those deduced at the nearby Canada Water station (GCG, 1992). Accordingly, two aquifers are present at the site (see Figure 13.3). The upper aquifer is associated with the Thames Gravel. Groundwater level is at an average level of 100 m PD. The lower aquifer is associated with the Pebble Beds, Glauconitic Sands, Thanet Beds and Chalk. In this aquifer the groundwater is at an average level of 94 m PD.

Ch 13 Finite element analyses of ground movements from tunnelling below Southwark Park

185

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 13.1

Plan of the tunnels and instrumented section at Southwark Park

Key A,B,C,D,E,F and G - extensometer locations A, C, D, E and F - electrolevel locations,

Chalk

Y

C

0 ._ c

W

Figure 13.2

Adopted soil profile and instrument locations

Building response to tunnelling

Pore water pressure, U (kPa) 200 300 400 500 105 0, I I I ,,, I ,, I , I I Gro'und level Made Ground .............................................. 100

I

I

I

I

I

,

I

I

I

I

I

,

,

I

I

I

I

3

,

I

I

I

I

I

I

I

I

1

~

I

I

Thames Gravels ..............................................

Upper Mottled Clay, Laminated beds, Lower Shelly Clay, Lower Mottled Clay ............................................

'.

-

Pebble Beds

'.'. Glauconitic Sands '...c..................................... Thanet Beds '.'. '. '.

.......?, ...........................................

85

\

\

a

.Ck

65 1

.-0

;

>

60 L

aJ -

-

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

55 L

Figure 13.3

13.3

Key

- - - - Initial distribution

- After dewatering

Assumed pore water pressure distributions

SOIL PROPERTIES Material properties for each of the soil strata shown in Figure 13.2 are summarised in Table 13.1. They are taken from the report on FE analysis of ground movements around the proposed Canada Water station (GCG, 1992). Only made ground was modelled as linear elastic perfectly plastic material. All other materials were modelled as non-linear elastic perfectly plastic soils. A Mohr-Coulomb yield surface and plastic potential were used to model the plastic behaviour of all strata. To describe the non-linear elastic behaviour, a constitutive model of the form described by Jardine et al(l986) was used. The secant stiffness expressions that describe this behaviour are as follows: 3.G/p'= A + B.cos{a.[logl~(~D/(./3.C))]~}

where: G is the secant shear modulus K is the secant bulk modulus p ' is the mean effective stress given by p ' = (0,l + OZ1+ 03')/3 ED is the deviatoric strain invariant given by &D = [(2/3).(&1-&2)~ + (&2-&3)2 + (&'j-&1)2]1'z and related to the axial strain, E,, in an undrained triaxial test by the expression &D = ./3.&, and are principal strains E,

is the volumetric strain, and

A , B, C, R, S, T, a,y ,6 and p are all constants (A, B, C, R, S, and Tbeing the equivalent of CI to C6 in Chapter 12).

Table 13.1

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Strata (elevation m PD)

Bulk density

Summary of assumed soil propetties

Effective cohesion

Effective angle of shearing resistance

Angle of dilation

Coefficient Young’s of earth modulus pressure at rest

Poisson’s ratio

Y

C’

@’

cv

W/m3

kN/m2

degrees

degrees

Made ground ( 102.5-99.8)

18.0

0.0

25.0

0.0

0.5

5.0

0.2

Thames Gravel (99.8-95.0)

20.0

0.0

38.0

19.0

0.5

non-linear

non-linear

Upper Mottled Clay (UMC) (95.0-93.5)

22.0

15.0

26.0

13.0

1.5

non-linear

non-linear

Laminated Silts and Sands (LSS) (93.5-9 1.2)

22.0

0.0

35.0

17.5

1.5

non-linear

non-linear

Lower Shelly Clay/Lower Mottled Clay (LSC/LMC) (91.2-90.1m)

22.0

15.0

26.0

13.0

1.5

non-linear

non-linear

Pebble Bed (90.1-86.5m)

22.0

0.0

34.0

17.0

1.5

non-linear

non-linear

Glauconitic Sands (86.5-80.6m)

22.0

0.0

34.0

17.0

1.5

non-linear non-linear

Thanet Beds (80.6-70.0m)

22.0

0.0

40.0

20.0

1 .o

non-linear

non-linear

Chalk (70.0-50.0m)

22.0

0.0

40.0

20.0

1.o

non-linear

non-linear

KO

E

U

m/mZ

In the analysis, tangent stiffness expressions are used that can be derived from the above expressions by differentiation. Throughout the analysis, the stiffness at a particular point is continually updated. It depends on the current strain, ED, and the current mean effective stress,^', at that point. Until a specified minimum strain (&,,,in or &,min) is exceeded the stiffness vanes only with the mean effective stress,^'. This also applies once a specified upper strain limit ( E ~ , - or is exceeded. In the analysis, the calculated stiffness is prevented from falling below specified minimum values (Gminor Kmin). The constants and limits of the model for each soil type are given in Table 13.2. They generate the secant shear stiffness curves as shown in Figure 13.4. Only the upper Lambeth Group Layers (Upper Mottled Clay, UMC; Laminated Silts and Sands, LSS; and Lower Shelley and Mottled Clays, LSC/LMC) were modelled as being undrained in the FE analysis. All other soil layers were assumed to be permeable and were therefore modelled as being h l l y drained.

188

Building response to tunnelling

Table 13.2

Coefficients and limits for non-linear elastic bulk modulus expression

Values for strata Identity

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

S

T (%)

Thames Gravel and Laminated Silts and Sands

Upper Mottled Clays, Lower Shelly Clays, Lower Mottled Clays, Pebble Beds and Glauconitic Sands

1104

1300

930

1035

1380

1120

5.0 x 104

1.0 x I O - ~

2.0 x I O - ~

0.974

1.22

1.1

0.940

0.649

0.7

8.8335 x 104

1.9053 X 10-4

3.6373 X 10-4

0.3464

0.1300

0.1645

2000

1000

2000

275

275

190

225

235

110

2.0

x

10'~

Thanet Beds and Chalk

5.0 x 1 0 - ~

1.0 x I O - ~

6

0.998

1.658

0.975

P

1.044

0.535

1.01

&v,nlax

Kmin ( k W

1.1

5.1 x 1 0 . ~

2.1 x 1 0 - ~

&v,min

x

103

0.2

0.3

0.2

5000

3000

5000

3,000

"..+2,500

'Thames Gravcl & LSS

--*--. Thanet Beds & Chalk2

2,000

P

x

L. 0

1,500

f

3 d .

E. 0

:: 1,000

v?

\i:...

500

0 0.0001

~

I 0.001

:

0.01

0.1

1

Axial strain, ~3 ( O h )

Figure 13.4

Assumed non-linear elastic small strain secant shear stiffness

Ch 13 Finite element analyses of ground movements from tunnelling below Southwark Park

189

13.4

FINITE ELEMENT ANALYSES The finite element code ICFEP (Imperial College Finite Element Program), developed by Professor D M Potts of Imperial College, was used to perform the analyses reported here.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

To represent the geometry in the analyses a finite element mesh was developed based on the cross-section shown in Figure 13.2. The finite element mesh is shown in Figure 13.5.

Figure 13.5

Finite element mesh

An extensive dewatering scheme has been undertaken at the Canada Water station site. Given the proximity of the analysed section at Southwark Park to that site, the initially assumed pore water pressure distribution at the Southwark Park site was modified by dewatering the Pebble Beds, Glauconitic Sands and Thanet Beds in a similar way to that for the study of Canada Water station (GCG, 1992). This is depicted in Figure 13.3. The eastbound (EB) and westbound (WB) running tunnels are 28 m apart, and have the same outside diameter (OD=4.9 m) and axis level (8 1.8 m PD). The WB running tunnel was to be excavated first. To model the ageing effects after dewatering, a low-strain stiff response in the soils has been re-invoked before the WB running tunnel excavation. This has also been done before excavation of the EB running tunnel, because of the changes in stress path direction over a substantial part of the mesh during the second (EB) running tunnel excavation. Tunnel excavation was modelled by the incremental removal of the solid elements within the tunnel boundary. The stresses that the soil within the tunnel applied to the tunnel boundary were evaluated and then applied in the reverse direction over ten increments. To approximate the three-dimensional (3D) effects of an advancing tunnel heading, only a proportion of the initial ground stresses was removed prior to installing the tunnel lining. (The tunnel linings are made of bolted concrete segments approximately 250 mm thick. An emphasis of the analyses was on ground movements that are not influenced by the stiff tunnel lining. Thus the tunnel linings were simply modelled using beam elements and as a linear elastic material with a Young’s modulus of 28.0 x 103 MPa and a Poisson’s ratio of 0.15.) This is equivalent to the “hfactor” approach (Panet and Guenot, 1982). As the tunnel is excavated, the ground undergoes stress relief, resulting in ground (volume) loss towards the face with associated ground movements. The ground movements caused by tunnelling are often characterised by this volume loss, VI, expressed as a percentage of the notional excavated volume of the tunnel. Given the complex soil profile through which the tunnels are driven, the drained nature of materials involved, and the use of the earth pressure balance machine for tunnel excavation, three different volume losses have been adopted during excavation of both WB and EB running tunnels. The volume losses are: VI= 0.5 per cent (Run l),

190

Building response to tunnelling

1.2 per cent (Run 2) and 2.0 per cent (Run 3). (Bearing in mind that nearly all the materials involved are modelled to behave in a drained manner, it is worth noting that the area of the surface settlement trough has been used in calculating the volume loss, not the area “over-excavated’’ during tunnel construction. During the tunnel excavation in a drained material these two areas are not necessarily the same.) The amount of ground stresses removed before installing the tunnel lining to achieve the above volume losses were 50, 70 and 80 per cent, respectively.

13.5

RESULTS OF THE ANALYSES

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

An extensive array of instruments for monitoring ground movements has been installed at the site. The location of boreholes containing magnetic extensometers and in-place electrolevel inclinometers is shown in Figure 13.2. Apart from these instruments, some surface settlement points have also been placed. Surface settlement profiles, vertical movements at extensometer locations and horizontal movements at electrolevel locations are predicted for three different volume losses: VI= 0.5 per cent (Run I), 1.2 per cent (Run 2) and 2.0 per cent (Run 3). Movements due to excavation of WB and EB running tunnels are separated in all figures under sub-titles (a) and (b). Accumulative movements due to excavation of both WB and EB running tunnels are presented under (c). Surface settlement profiles are shown in Figure 13.6. Maximum surface settlements during excavation of both tunnels are above the tunnel centrelines. I n the case of the WB running tunnel they are 3.6 mm, 9.3 mm and 16.5 mm for assumed volume losses V, = 0.5 per cent (Run I), 1.2 per cent (Run 2) and 2.0 per cent (Run 3) respectively. During excavation of the EB running tunnel these values are 3.6 mm, 9.4 mm and 16.8 mm, suggesting that the EB tunnel excavation is not influenced by the presence of the WB running tunnel. The shape of the settlement profiles for both tunnels is also the same, but their magnitude is dependent on the assumed amount of volume loss. However, the width of the settlement troughs appears not to be influenced by the assumed value of volume loss. Vertical and horizontal movements along different extensometer locations (locations A, B, C, D, E, F and G in Figure 13.2) were derived from the FE analyses and presented graphically. (For reasons ofspace only one example is given here, that for Location C in Figures 13.7 and 13.8 - Eds.) The vertical movements depend on the amount of volume loss assumed and they are nearly the same during excavation of both tunnels. An example of the vertical movements, those at electrolevel location C, is given as Figure 13.7. Maximum vertical movements are concentrated just above the tunnel crowns (for the WB tunnel: 16.4 mm, 46.0 mm and 103.8 mm, and for the EB tunnel: 17.1 mm, 45.2 mm and 98.8 mm in Runs I , 2 and 3 respectively). At one tunnel diameter above the tunnel crown they are significantly reduced and nearly equal to the maximum surface settlements. Also, the vertical movements are becoming smaller further away from the tunnel centreline. They are very small at a distance 25 m from the tunnel.

Ch 13 Finite element analyses of ground movements from tunnelling below Southwark Park

191

'R

Distance (m) -50

-25

25

50

............

""\

..

run. 1

......

run.3

(a) due to excavation of WB tunnel only

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Distance (m)

Legend: run. 1 run.2 run.3

.............. ...

(b) due to excavation of EB tunnel only Distance (m) -50

-25 1 1 1 1 1 .........________

\.' ---...

50

25 1

...... ----

.......

Legenu: i

\ \

\..

i

.........

run. 1 run.2 run.3

(c) due to excavation of both WB and EB tunnels Figure 13.6

192

Surface settlement profiles

Building response to tunnelling

Vertical movement (mm)

Vertical movement (mm) 10 20

Vertical movement (mm)

-5 0 I

- I

I I

I

go

I

85

I I I

80

.‘

.





. 0

-- Key Runl . - Run2

75

~

90

I

Run3

70 j

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 13.7

(a)due to excavation of westbound tunnel only

- - Key Runl - Run2 .-

-75

Run3

70

(b)due to excavation of eastbound tunnel only



Key Runl

Run2 .Run3

(c)due to excavation of both westbound and eastbound tunnels

Vertical movements at extensometer location C

s

a

E

C

.-.m

W Q)

, --

Key Run 1 Run 2 Run 3

Key

- Run Run Run

7d (a) due to excavation of westbound tunnel only

Figure 13.8

(b) due to excavation of eastbound tunnel only

(c) due to excavation of both westbound and eastbound tunnels

Horizontalmovements at extensometer location C

The horizontal movements are also dependent on the amount of volume loss assumed. An example of the horizontal movements, those at electrolevel location C, is given as Figure 13.8. During excavation of the first (WB) running tunnel, the displacement vector pattern is symmetrical and there are no horizontal movements along the tunnel centreline. However, its excavation has brought about stress relief, and thus slight horizontal movements at the EB tunnel centreline have been predicted during its excavation. In the vicinity of the WB running tunnel, the horizontal movements are concentrated at the tunnel axis (see Figure 13.8a). They are becoming smaller and more uniform at greater distances from the tunnel. It is of interest to note the “kinks” in the distribution of horizontal displacements at the position of the upper Woolwich and Reading Beds layers (UMC, LSS and LSC/LMC), which were modelled in the analyses as undrained materials (see Figure 13.8). Electrolevel locations D, E, C and F are far from the EB running tunnel, and consequently a relatively uniform horizontal displacement pattern has been predicted during its excavation at these locations (see Figure 13.8b).

Ch 13 Finite element analyses of ground movements from tunnelling below Southwark Park

193

13.6

CONCLUSIONS The WB and EB running tunnels at Southwark Park were to be driven through complex soil geology using an earth pressure balance machine. Because of these factors, and in view of the expected drained response of most of the materials involved, three different volume losses have been adopted in the FE analyses: Vi = 0.5, 1.2 and 2.0 per cent. The magnitude (and, to a smaller extent, the pattern) of the ground movements during excavation of both tunnels depends significantly on the amount of volume loss assumed in the analyses. On the basis of previous experience of tunnelling through the London Clay, the predictions made assuming the value of the volume loss of 1.2 per cent is considered to be the most probable. If good control of the face pressure is achieved, it is considered that the lower assumed value of 0.5 per cent could be obtained. The higher assumed value of 2.0 per cent is judged a reasonable upper bound.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

13.7

REFERENCES GEOTECHNICAL CONSULTING GROUP ( 1992). Finite element analysis ofground movements around the proposed Canada Water Station, Report to London Underground Ltd’s Jubilee Line Extension Project Team, April 1992 GEOTECHNICAL CONSULTING GROUP (1 997). Southwark Park - Finite element analyses of ground movements due to tunnelling, Unpublished Report to Imperial College research team for the LINK CMR Research Project, April 1997 JARDINE, R J, POTTS, D M, FOURIE, A B and BURLAND, J B (1 986). Studies of the influence of non-linear stress-strain characteristics in soil-structure interaction, Gdotechnique, vol36, no 3,377-396 LAMBE, T W (1973). Predictions in geotechnical engineering, Gkotechnique, vol 23, no 2, pp 194-202 PANET, M and GUENOT, A (1982). Analysis of convergence behind the face of a tunnel, TunneNing’82, Institution of Mining and Metallurgy, pp 197-204

194

Building response to tunnelling

14

Elizabeth House: settlement predictions

R J Mair and R N Taylor Note. This section is a transcript, with minor editing and changes necessary for reproduction in this book format, of the report by Geotechnical Consulting Group, dated May 1996 (Geotechnical Consulting Group, 1996), which presents its predictions of the settlements to be experienced by this building in response to the complex series of tunnelling operations below it. Thus in what follows, the tenses of verbs are the same as those in that report, applying to predictions made at that date.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The actual settlement of the building is presented in Chapter 30, where it is also compared with these predictions.

14.1

INTRODUCTION As part of its contribution to the LINK CMR project, Geotechnical Consulting Group (GCG) has undertaken to carry out Class A predictions of the performance of a number of buildings affected by construction of the Jubilee Line Extension (JLE). This report summarises the settlement predictions made for Elizabeth House. Both Class A and Class C predictions have been made. Lambe (1973) classified predictions in geotechnical engineering as follows:

Prediction type

When prediction made

Results at time prediction made

Class A

Before event

-

Class B

During event

Not known

Class B 1

During event

Know

Class C

After event

Not known

Class C1

After event

Known

The sequence of tunnel construction beneath Elizabeth House is complex; full details are given in Section 14.3. Three main stages of tunnelling are involved, two of which were completed very early on in the JLE Project. At the time of writing this report the third stage is about to begin. The results to date of the measurements of the performance of Elizabeth House, and of the ground beneath it, have not been made available to the authors of this report, Dr R J Mair and Dr R N Taylor. In terms of the classification of Lambe (1 973), therefore, Class C predictions are presented in this report for the first two stages of tunnelling (already completed), and Class A predictions are presented for the third stage of tunnelling, which is about to begin.

Ch 14 Elizabeth House: settlement predictions

195

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

expansionon ji,t

W

Figure 14.1

14.2

/

‘Cross-over

passage

N

Plan of Elizabeth House and JLE tunnels

THE BUILDING Elizabeth House is located on York Road adjacent to Waterloo station; a photograph of the building is given as Figure 30.1 in the case study chapter. It is a multi-storey reinforced concrete-framed building constructed in the 1960s with two levels of basement and founded on a 1.4 m-thick reinforced concrete raft. The length of building affected by the JLE tunnels has ten storeys above ground level; the number of storeys reduces to seven some way along the building. A plan of the building is shown on Figure 14.1. Also represented on the plan are the JLE

tunnels, details of which are given in Section 14.4. The ten-storey building is approximately 80 m long, and for most of its length is 18 m wide. The underside of the raft foundation beneath the lower basement is at a level of 95.2 m JLE Project Datum (ie at -4.8 m OD), which is approximately 8 m below ground level. There is generally about a 1 m thickness of Thames Gravels beneath the raft foundation,

196

Building response to tunnelling

which is underlain by some 30 m of London Clay. The tunnels are in the London Clay, and their axes are at a depth of about 23 m below foundation level. There is an expansion joint at Row 1 1, which appears not to extend below ground level. The movement joint at Row 21 (where the building changes from ten storeys to seven storeys above ground) is continuous through to the underside of the raft foundation.

14.3

GROUND CONDITIONS JLE site investigation data indicated generally similar soil stratigraphy to that obtained in the site investigation undertaken for the adjacent Waterloo International Terminal (Hight et al, 1993). The general soil profile and observed pore water pressures are shown in Figure 14.2. The top 6 m of the London Clay are weathered. The pore water pressure profile indicates under-drainage due to historic pumping from the deep aquifer. Pore water pressure, U (kPa) 100 200 300 400

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

0 100 R

I

I

I 90 -_

I

500 I

I

Thames Gravels Weathered London Clay

_

'.,

h

n

80-

Unweathered London Clay

v

c

.-0 m

2 70w

60

50

e

I-t

e

Lambeth Group Clay

'*,

Key Distribution assumed

____ Figure 14.2

Hydrostatic distribution Observed pore pressures (Hight et al. 1993)

Elizabeth House: assumed pore water pressure distribution

Ch 14 Elizabeth House: settlement predictions

197

14.4

THE TUNNELS A plan of the tunnels in relation to Elizabeth House is shown on Figure 14.3. Three stages of tunnelling are shown, as follows: Stage A The access adit tunnels (up to a maximum horizontal diameter of 7.5 m) and the start of the eastbound (EB) and westbound (WB) running tunnels (5.6 m diameter) were constructed using sprayed concrete linings, often termed the New Austrian Tunnelling Method (NATM). The tunnels were constructed between July and October 1994.

/

N

0

/

10m

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

U

StageA

I

Figure 14.3

Construction of tunnels beneath Elizabeth House and stages used in analysis

Stage B Following a delay arising from the collapse of the tunnels at Heathrow in October 1994, the EB and WB 5.6 m-diameter running tunnels were constructed beneath Elizabeth House in January to February 1995. NATM techniques were used, except for a 16 m length of the WB tunnel, which was started using segmental linings. Stage C A 5.6 m-diameter crossover tunnel is to be constructed, which will connect the EB and WB running tunnels. At the locations where the crossover tunnel merges into the running tunnels, the latter will be broken out and larger-diameter turnout tunnels formed. A cross-section through the part of one of the turnout tunnels where the horizontal diameter is largest (12.4 m) is shown in Figure 14.4. All of the Stage C tunnelling will be constructed using NATM techniques.

198

Building response to tunnelling

10 m

0

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 14.4

Cross-section through turnout tunnel at largest section (horizontal diameter = 72.4 m)

The axes of all the tunnels are at a depth of approximately 3 1 m below ground level (23 m below the raft foundation of Elizabeth House), which corresponds to a level of about 72 m Project Datum. As can be seen from Figure 14.2, the tunnels are in the lower strata of the unweathered London Clay.

14.5

GREENFIELD SITE SETTLEMENT PREDICTIONS

14.5.1

General The approach to predicting the response of the building to the tunnel construction is firstly predict the greenfield site settlements at building foundation level and then assess how the inherent stiffness of the building is likely to modify these settlements. The greenfield site prediction ignores the presence of the building. The prediction is based on the empirical approach outlined by O’Reilly and New (1982) and Attewell and Woodman (1982).

14.5.2

Ass umpt ions A schematic plan of the Stage A and Stage B tunnels, together with the volume losses assumed for the predictions, is shown in Figure 14.5. Different volume losses have been assumed for different types of tunnel construction. For the 5.6 m-diameter tunnels, a volume loss of 1.3 per cent has been assumed, based on experience of tunnels constructed in London Clay elsewhere using NATM techniques (eg New and Bowers, 1994; Kimmance et al, 1996). In the case of the shorter-length, larger-diameter tunnels (tunnel lengths numbered 4 to 6 in Figure 14.5), a slightly larger value of 1.5 per cent has been assumed. Where the tunnels break out from existing tunnels (tunnel length numbers 7 to 10 in Figure 14.5), a larger volume loss of 2.0 per cent has been assumed to allow for interaction effects.

Ch 14 Elizabeth House: settlement predictions

199

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Tunnel

V, (%)

1

1.3

Equivalent diameter (rn) 5.58

5 6 7 8 9 10

1.5 1.5 2.0 2.0 2.0 2.0

7.45 7.45 5.66 6.02 7.21 7.14

12 13 14 15

1.3 1.3 1.3 1.3

Figure 14.5

Assumed values of volume loss for different parts of the tunnels beneath Elizabeth House; Stages A and B

A plan of the Stage C tunnels, together with the volume losses assumed for the predictions, is shown in Figure 14.6. Using the principle of superposition, the settlement due to the EB and WB running tunnels constructed in Stage B near the turnout tunnels (shown dotted) is removed, and the settlement due to the turnout tunnel construction is added for the entire volume of the turnout, assuming a volume loss of 1.5 per cent. This figure makes some allowance for the “dowel” effect arising from the existing running tunnels (Mair, 1993), at the same time recognising that there will be interaction effects due to the proximity of adjacent tunnels. A volume loss of 2.0 per cent was assumed for the relatively short length of crossover tunnel between the two turnout tunnels to allow for interaction effects.

The volume loss figures assumed for the predictions are relevant for the immediate settlement caused by the tunnel construction. Additional time-dependent settlements associated with pore water pressure changes are allowed for separately, as described in Section 14.5.4 below. In prediction of the immediate settlements caused by tunnel construction, the trough width parameter K applied at raft foundation level was assumed to be 0.45. This is based on other measurements of surface (or near-surface) settlement profiles above tunnels in London Clay (eg New and Bowers, 1994). Subsurface settlements have also been predicted, in order to compare with the extensometer measurements. For these predictions, a variation of K with depth of the form described by Mair et al(1993) has been assumed.

200

Building response to tunnelling

Stage C: Tunnels 16 25

Tunnel

16 17 18 19 20 21 22 23 24 25

VL

("/.I

1.5 1.5 1.5 1.5 2.0 1.5 1.5 1.5 1.5 1.5

Equivalent diameter (m) 7.80 8.40 9.14 10.62 5.58 10.62 9.00 8.40 7.80 7.28

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

0

Scale: rn

20

/

v

Elizabeth House

Figure 14.6

14.5.3

Assumed values of volume loss for different parts of the tunnels beneath Elizabeth House; Stage C

Settlement contours The predicted cumulative settlement contours at raft foundation level (ie 8 m below ground level) are shown for Stages A, B and C in Figures 14.7, 14.8 and 14.9 respectively. These contours are the predicted immediate settlements (ie assuming undrained conditions), ignoring any time-dependent consolidation settlements. Figure 14.10 shows the predicted additional settlement contours resulting from the construction of the Stage C tunnels alone, ie the difference between the cumulative contours for Stage C and Stage B.

Ch 14 Elizabeth House: settlement predictions

201

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

j

202

stage^

j

Figure 14.7

Greenfield site settlement contours: Elizabeth House Stage A (note: 25 m grid lines shown dotted)

Figure 14.8

Greenfield site settlement contours: Elizabeth House Stage A + Stage B (note: 25 m grid lines shown dotted)

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 14.9

Greenfield site settlement contours: Elizabeth House Stage A+ Stage B + Stage C [note: 25 m grid lines shown dotted)

Figure 14.10

Greenfield site incremental settlement Elizabeth House Stage C - Stage B [note: 25 m grid lines shown dotted)

Cross-sections through the building corresponding to the greenfield site settlement contours at the different stages are shown in Figures 14.1 1 to 14.13. The positions of the cross-sections are shown on Figures 14.7 to 14.9. Also shown on Figures 14.12 and 14.13 are the estimated additional greenfield site settlements due to consolidation. These are discussed in the following section.

Ch 14 Elizabeth House: settlement predictions

203

Section a-a

Section b-b

Section d-d

Section c-c

h

E E

v

* c W

-E

5

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

cn

Distance (m)

Distance (m)

Elizabeth House Stage A; cross-sections showing greenfield site settlements

Figure 14.11

Section b-b

Section a-a

0 h

E E

20

v

*

c

W

E

W % (0

40

I I I I I I I ( I I ( 6o0

20

40

60

80

100

Section c-c

Section d-d

0 h

E

E *

20

c W

E

5

40

cn

I

4

' a

I 12 Distance (m)

' 16 ' 20

---

Distance (m) Stage B1 Stage 82

Figure 14.12

204

Elizabeth House Stage 6;cross-sections showing greenfield site settlements plus 10 mm consolidation settlement (note: dotted line corresponds to Stage 61; solid line corresponds to Stage 62, ie Stage B l + I0 mm consolidation settlement)

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Section a-a

Section b-b

Section d-d

Section c-c

Distance (rn)

---

Distance (rn) Stage B1 Stage B2

-

Figure 14.13

14.5.4

Elizabeth House Stage C;cross-sections showing greenfield site settlements plus 10 mm consolidation settlement

Consolidation settlements Estimates have been made of the additional settlements due to time-dependent changes of pore-water pressure following completion of a particular stage of tunnel construction. The basis for these estimates has been field measurements of the increase of ground settlements with time above tunnels in London Clay made at Heathrow (Bowers et al, 1996) and at St James’s Park for the LINK-CMR research project (Nyren, 1996). The three stages of tunnel construction were undertaken during the following time periods: Stage A

Constructed July to October 1994

Stage B Stage C

Constructed January to February 1995 Commencing May 1996.

There was therefore an interval between completion of Stages A and B of approximately six months, and a further interval of about 15 months after completion of Stage B before construction of Stage C. At Heathrow, for the single Type 2 tunnel of 8.5 m diameter at a depth of 21 m below the ground surface, the consolidation settlement after 2 1 months was approximately 7 mm over the tunnel centre-line, reducing gradually to 5 mm at 15 m from the centreline, and to 2 mm at 30 m. In the first six months, approximately 3 mm of consolidation settlement had occurred over the centre-line. At St James’s Park, the EB and WB tunnels are separated horizontally by 21 m, which is similar to their average separation beneath Elizabeth House. Their depths are comparable to those of the tunnels beneath Elizabeth House. To date, the measurements show consolidation settlements above the WB tunnel to be 5 mm over the tunnel centre-line after 250 days, and 5 mm even at a distance of 25 m from the centre-line (the tunnel is at a depth of about 30 m). An almost uniform settlement increase across the whole

Ch 14 Elizabeth House: settlement predictions

205

construction settlement trough is indicated. Above the centre-line of the EB tunnel, the measured consolidation settlement is about 9 mm after 100 days. Extrapolating the St James’s Park data, and taking into account the Heathrow data, it is estimated that a uniform consolidation settlement of about 10 mm will have occurred in the 15-month period after completion of Stage B, before construction of Stage C. This is shown on the cross-sections on Figure 14.13 as Stage B2. Stage B 1 is the predicted settlement immediately after completion of Stage B construction. During the six-month period between completion of Stage A and completion of Stage B, it is estimated that 5 mm of consolidation settlement will have taken place at the centre of the “sink-hole’’ of the immediate settlement contours (see Figure 14.7), reducing to 2 mm at the edge of Elizabeth House. These consolidation settlements will therefore only have had a very small local effect on Elizabeth House, and hence they have not been included on the cross-sections for Stage A on Figure 14.1 I .

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

14.6

PREDICTED BUILDING PERFORMANCE The influence of the building stiffness in modifying the predicted greenfield site settlement profiles has been assessed. Use has been made of the parametric study by Potts and Addenbrooke (1996), based on finite element analyses of a wide range of beam stiffnesses, the building being idealised as a simple beam. Potts and Addenbrooke define the relative bending stiffness p* and relative axial stiffness a* of the building, which are dimensionless quantities expressing the stiffness of the building in relation to a representative stiffness of the underlying soil. Details of the assessment of these relative stiffnesses in the case of Elizabeth House are given in Section 14.8. In assessing the bending stiffness of the building, it has been assumed that the 1.4 m-thick raft foundation and the floor slabs each contribute to the overall stiffness but do not act compositely. This means that the overall bending stiffness is dominated by the stiffness of the 1.4 m-thick raft foundation. Based on this approach, it is shown in Section 14.8 that in the transverse direction (ie over its 18 m width) the building behaves almost rigidly, with negligible deflection ratio (AlL) in response to the tunnel construction. Conversely, in the longitudinal direction (ie over its full 80 m length) the assessment shows the building to behave almost perfectly flexibly in response to the tunnel construction. Figures 14.14 to 14.16 show the predicted response of the building for Stages A, B and C of the tunnel construction, based on the above conclusions relating to the influence of building stiffness. In the transverse direction (cross-sections aa and dd), the predicted greenfield site settlement profiles have been modified to allow for rigid behaviour of the building. In the longitudinal direction (cross-sections bb and cc), no modifications have been made, because it has been assessed that the building will behave almost perfectly flexibly and hence conform to the predicted greenfield site settlement profiles. The likely damage to the building associated with the predicted settlements is best assessed in terms of the induced tensile strains (Burland and Wroth, 1974;‘Burland et al, 1977; Boscardin and Cording, 1989). Treating the building as a simple beam, the maximum tensile strain can be expressed in terms of the deflection ratio, A/L, and the average horizontal strain, Eh (Burland, 1995; Mair et a/, 1996).

206

Building response to tunnelling

Section b-b

Section a-a 0

20

40

uu

0

4

8

12

16

20

60

0

Section d-d

20

40

60

80

100

Section c-c

0

0

20

20

40

40

h

E

E. e

C

a

-E 0

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

6o0

4

8 12 Distance (m)

16

20

Elizabeth House Stage A; cross-sections showing predicted building settlements

Figure 14.14

Section a-a

Section b-b

h

E E

v

V"

0

4

8

12

16

20

Section d-d

Section c-c

h

E E

v

u"O

Figure 14.15

4

8 12 Distance (rn)

16

20

- - - Stage B1 -Stage 8 2

Distance (rn)

Elizabeth House Stage B;cross-sections showing predicted building settlements (note: dotted line corresponds to Stage 81; solid line corresponds to Stage 82, ie Stage B1 + 10 mm consolidation settlement)

Ch 14 Elizabeth House: settlement predictions

207

Section b-b

Section a-a

0

20

40

60’ 0



I

4

I

I

8

I

I

12

I



16

I



20 Section c - c

Section d-d

0 h

E

-

E.

20

c

a,

E

-

%

40

v)

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

60 0

4

8

12 Distance (m)

16

20

---

uv

~~

0

20

40 60 Distance (rn)

80

100

Stage B1

-Stage 82 Figure 14.16

Elizabeth House Stage C; cross-sections showing predicted building settlements

The maximum deflection ratio corresponding to the predicted building settlement profiles shown on Figures 14.14 to 14.16 is about 0.3 x 10”. This is obtained from sections b-b and c-c for Stage C of the tunnel construction, shown on Figure 14.16, and is characterised by a sagging mode. It is considered that the building will experience negligible horizontal strain by virtue of its continuous reinforced concrete raft foundation. The level of damage, therefore, can be directly assessed from the maximum predicted deflection ratio of 0.3 x 10”. Making allowance for the framed construction of the building, the maximum tensile strain associated with this deflection ratio is very low (about 0.03 per cent) and hence negligible damage is predicted. This is consistent with the observations of sagging deflection ratios for framed buildings settling under their own weight, presented by Burland et a1 (1977). Some opening of the movement joint at Row 21 is predicted as the ten-storey building rotates away from the seven-storey building. The movement at the level of the seventh storey is unlikely to be more than about 10 mm on completion of all the tunnel construction. The magnitude of the opening of this movement joint is likely to be significantly influenced by temperature effects, and some seasonal fluctuation can therefore be expected.

14.7

PREDICTION OF SUBSURFACE GROUND MOVEMENTS The locations of five borehole extensometers installed beneath Elizabeth House are shown on Figure 14.1 (BH2-BH6). Subsurface vertical ground movements at these locations have been predicted using the approach given by Mair et a1 (1993). The predicted variation of settlement with depth below foundation slab level is shown on Figure 14.17 for the various stages of tunnel construction. Negligible movements are predicted at these locations for the Stage A tunnels. The predictions are only shown for Extensometers 3, 5 and 6. The predicted settlements at the locations of the other extensometers due to all stages of tunnel construction are negligibly small.

208

Building response to tunnelling

--

Extensometer 4

Extensometer 5

70

60 h

E

50

-E

40

$

30

v

c

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

a, t:

20 10

70 60

E

-

50

v

c

z

40

Q)

$

30 20

_____-

------.---._

- - -- - - - ----_

10

Figure 14.17

Elizabeth House: predicted subsurface settlements at different stages of tunnel construction at the locations of Extensometers 3, 5 and 6

The predicted movements for Stage B1 correspond to immediately after completion of Stage B. Stage B2 corresponds to a further period of 15 months, representing the period up to the commencement of the Stage C tunnel construction. The data from Heathrow (Bowers et al, 1996) and St James’s Park (Nyren, 1996) indicate consolidation settlements to be almost constant with depth for extensometers relatively close to the tunnels. Based on these observations, an estimated 10 mm of consolidation settlement has been assumed between Stages B 1 and B2, as shown in Figure 14.17. The predictions for Stage C shown on Figure 14.17 include this consolidation settlement component.

Ch 14 Elizabeth House: settlement predictions

209

14.8

ASSESSMENT OF INFLUENCE OF STIFFNESS OF ELIZABETH HOUSE Note: this section formed a hand-written appendix to the prediction report. These calculations by GCG are dated April 1996 and initialled RJM. They have been transcribed by the editors. Refer to Potts and Addenbrooke (1996) -paper submitted to ICE Geotechnical Engineering. Cer

ne

i.-.I

Width B

/I\

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Depth z

,

Offset e

Relative bending stiffness p*=-

EI

where H=B/2

EsH4 E, = secant stiffness of soil at depth 0.52 for 0.01% strain (as defined by Potts and Addenbrooke)

Relative axial stiffness

take EcOncrele = 23 x I O6 k

I

Consider transverse section through building, ie Section A-A

Estimate total height of building, HB

=

1 2 ~ 3 . 5+ 1.4 = 43.4 m

LBjloor slab, 1 = 1.43/12m4/m = 0.23 m4/m Floor slab I

= 0.33/12 = 2.3 x I O - ~m4/m

Assume (EI)structure= E

Islab

ie treat as 13 individual slabs and ignore any AH2 terms. This is reasonable for ground jloor upwards, but may underestimate stiffness of substructure (B i-LB). (EI)structure = 2 3 ~ 1 (01~3 ~ 2 . 3 ~ 1 0i-0.23) -” = 23X106 ~ 0 . 2 6=6x106kN-m. H=B/2 =9 m

210

Building response to tunnelling

Tunnel axis at depth 23 m below LBfloor slab, approx. 30 m below ground level. .: at 0.52 below LBfloorslab, 11.5 m. ieapprox. 18.5 m belowgroundlevel (U = 110 kPa), KO = 1 at this depth at Waterloo

.:

p;, = o i =20~18.5-110=260 kPa

A

1

I

t

I

I

7

i

I

1 1 3.5 m (assumed)

6

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

5

4

1

v]

Assum e a c h floor sl$ 0.3 rn

-

/A\

B

Ground level

LB

/

V

1.4 rn-thick slab

At 0.01!% axial strain, 3G 700 P;, P;, :.Es =7O0x26O=182x1O3 kPa -=-=

Taking into account only the LB, B and groundfloor slabs.

a* = 23x 106(1.4 + 2 x 0.3)

=28.1

182X1O3X9

From Figures1 4. I 8 and 14.19 (Figures I O(a) and 10(b) in Potts and Addenbrooke, I996), p g = 0.2, Mhog = 0. Figures 14.I 8 and 14. I 9 apply to zero eccentricity (e = 0). Note that largest eccentricity ratio (e/B) explored by Potts and Addenbrooke is 0.64 Figure 14.20 (their Figure 12) and this was for B= 44 m.

Ch 14 Elizabeth House: settlement predictions

21 1

04

m

0.4-

.c

9

0.8

5

I

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

1.2-

Figure 14.18

m

1.2

-

0.8

-

x

10-6

Variation of modification factors for deflection ratio with relative beam stiffness: variation with a*(Potts and Addenbrooke, 7 996)

a ' =

I

3

0.40 10-8 O A

m

10-7

10-6

10-5

10"

10-3

10-2

10-1

100 p* (m-1) I

0.4-

.c

3

0.8 -

1.2-

Figure 14.19

212

Variation of modification factors for deflection ratio with relative beam stiffness: variation with p* (Potts and Addenbrooke, 1996)

Building response to tunnelling

1.2

0.8

‘ m

I

0.4 0

m .c

0.4

1

1’

-

D

,

,*”

0.8-

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

1.2

-

Figure 14.20

A

elB = 0 Upper bound curve elB = 0.18; B = 44 m

I

elB = 0.36; B = 44 m

+

elB = 0.64; B = 44 m

0

e16 = 0.26; variable Band e

Variation o f modification factors for deflection ratio with relative beam stiffnessfor eccentric beams with practical values o f relative axial stiffness(Potts and Addenbrooke, 1996)

At Section A-A, e/B = I , but e/B reduces rapidly in a north-easterly direction along the building. Concentration of contours (max greenfield site settlement of 45 mm for Stage B) makes problem very localised. .: From the “Design curves’’ (Figure 16 ofPotts anddddenbrooke, 1996, andgiven in this book as Figure 3.10) P

g

= 0.2, P h o g = 0.2

For practical purposes, this meum that in a transverse direction the building behaves almost rigidly, with negligible d L . Consider longitudinal section through building, ie Sections B-B, C-C

Length of building = 80 rn to main movement joint at Row 21 (extending right down through both basement levels). Length of building = 50 m to expansion joint at Row 1I (extends on& down to ground level). Assume B = 50 m forfirst calculations.

.: H=B/2 = 25 m (using Potts and Addenbrooke terminology and same assumptions as for transverse section) p* =

6x106 182~1~ 0~ 2

= 8 . 4 ~ 1 0 -(or ~ 1 . 3 ~ 1 0 -for ~ B/2=40m) 5 ~

a* = 23x106(2.0) =10.1 (or 6.3 forB/2=40m) 1 8 2 ~ 1 x25 0~

Ch 14 Elizabeth House: settlement predictions

213

Stage A: Sections B-B, C-C Equivalent e = 20 m, e/B = 0.4 From Figure 14.21, h.pSag=I.Oforp* = I-8x10” p g = l . O f o r p * = I-8x10” Stage B: Section B-B - similar to above.

Stage B: Section C-C Equivalent e = 25 m, e/B = 0.5 From Figure 14.21, p g = 0 . 6 t o 0.9 forp* = 1-8~10-’ ~ @ ~ ‘ ~ = l . O f o r p *I-8x10” = Conclude that generally building in longitudinal direction behaves almost perfectly flexibly. Figure 4(b) of Potts and Addenbrooke (given as Figure 3.9 in this book) confirms this - settlement profile of building similar to “greenfield site”profi1e.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

14.9

SUMMARY Class A and Class C predictions have been made for the settlement and performance of Elizabeth House, as well as for subsurface ground movements. The Class A predictions are for the Stage C construction (the crossover and turnout tunnels), which has just begun at the time of preparation of this report. The Class C predictions are for the Stages A and B tunnels, which have already been constructed. The results of the measurements for Stages A and B have not been made available to the authors of this report. Using an empirical approach, the maximum predicted settlement of the building is 55 mm on completion of the Stage C tunnel construction. Parametric finite element studies reported by Potts and Addenbrooke (1996) have been used as a basis for assessing the influence of the building stiffness on the greenfield site settlement predictions. The building is predicted to behave almost rigidly in the transverse direction (ie over its 18 m width), and almost perfectly flexibly in the longitudinal direction (ie over its full 80 m length). Negligible building damage is predicted. It is likely that there will be a slight opening of the movement joint at Row 2 I , at the seventh-storey level at the junction between the ten-storey and seven-storey sections of the building.

14.10

REFERENCES ATTEWELL, P B and WOODMAN, J P (1 982). Predicting the dynamics of ground settlement and its derivatives caused by tunnelling in soil. Ground Engineering, vol 15, no 8, pp 13-22,36 BOSCARDIN, M D and CORDING, E G (1 989). Building response to excavationinduced settlement. J Geotech Engg ASCE, 115; 1; pp 1-2 1 BOWERS, K H, HILLER, D M and NEW, B M (1996). Ground movement over three years at the Heathrow Express Trial Tunnel. Geotechnical Aspects of Underground Construction in Soft Ground (R J Mair and R N Taylor eds), Balkema, Rotterdam, pp 647-65 2

214

Building response to tunnelling

BURLAND, J B (1995). Assessment of risk of damage to buildings due to tunnelling and excavations. Invited Special Lecture to IS-Tokyo ’95: 1st Int. Con$ on Earthquake Geotechnical Engineering. BURLAND, J B, BROMS, B and DE MELLO, V F B (1977). Behaviour of foundations and structures - SOA Report, Session 2, Proc 9th Int Conf SMFE, Tokyo, 2; pp 495-546 BURLAND, J B and WROTH, C P (1974). Settlement of buildings and associated damage. SOA Review, Conf Settlement of Structures, Cambridge. Pentech Press. London, pp 61 1-654 HIGHT, D W, PICKLES, A R, DE MOOR, E K, HIGGINS, K G, JARDINE, R J, POTTS, D M and NYIRENDA, Z M ( 1 993). Predicted and measured tunnel distortions associated with construction of Waterloo International Terminal. Proceedings of the Wroth Memorial Symposium, Oxford, Thomas Telford, pp 3 17-338

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

KIMMANCE, J P and ALLEN, R (1996). The NATM and compensation grouting trial at Redcross Way. Geotechnical Aspects of Underground Construction In Soft Ground. (R J Mair and R N Taylor eds) Balkema, Rotterdam, pp 385-390 LAMBE, T W (1973). Predictions in geotechnical engineering. Gkotechnique 23,2, 149-202 MAIR, R J (1993). Developments in geotechnical engineering research: application to tunnels and deep excavations. Unwin Memorial Lecture, Paper 10070, Proceedings of the Institution of Civil Engineers, February, pp 27-4 1 MAIR, R J, TAYLOR, R N and BRACEGIRDLE, A (1993). Subsurface settlement profiles above tunnels in clay. Gkotechnique 43,2; pp 3 15-320 MAIR, R J, TAYLOR, R N and BURLAND, J B (1996). Prediction of ground movements and assessment of risk of building damage due to bored tunnelling. Geotechnical Aspects of Underground Construction in Soft Ground (R J Mair and R N Taylor eds) Balkema, Rotterdam, pp 7 13-7 18 NEW, B M and BOWERS, K H (1994). Ground movement model validation at the Heathrow Express trial tunnel. Tunnelling ’94, Chapman and Hall, London, pp 301-329 NYREN, R J (1 996). Personal communication O’REILLY, M P and NEW, B M (1982). Settlements above tunnels in the United Kingdom - their magnitude and prediction. Tunnelling’82, London, IMM, pp 173-1 8 1 POTTS, D M and ADDENBROOKE, T I (1996). A structure’s influence on tunnelling induced ground movements. Accepted for publication in Proc Instn Civ Engrs, Geotechnical Engineering. The foregoing is the reference given in Geotechnical Consulting Group (1996), but the paper was subsequently published as: POTTS, D M and ADDENBROOKE, T I (1997). A structure’s influence on tunnelling induced ground movements. Proc lnst Civ Engrs Geotechnical Engineering, vol 125, Issue 2, April, pp 109-125

Ch 14 Elizabeth House: settlement predictions

15

Settlement predictions for Neptune, Murdoch, and Clegg Houses and adjacent masonry walls

R J Mair and R N Taylor

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Note. This section is a transcript, with minor editing and changes necessary for reproduction in this book format, of the report by Geotechnical Consulting Group, dated July 1996 (Geotechnical Consulting Group, 1996), which presents its predictions of the settlements to be experienced by these three buildings and nearby walls. Thus in what follows, the tenses of verbs are the same as those in that report, applying to predictions made at that date. The actual settlements of the three buildings are presented in Chapter 43, where they are also compared with these predictions.

15.1

INTRODUCTlON As part of its contribution to the LINK CMR project, Geotechnical Consulting Group

(GCG) has undertaken to carry out Class A predictions of the performance of a number of buildings affected by construction of the Jubilee Line Extension (JLE). This report summarises the settlement predictions made for Neptune House, Murdoch House, Clegg House and two masonry walls in the vicinity of these buildings. Both Class A and Class C predictions have been made. Lambe (1 973) classified predictions in geotechnical engineering as in Table 15.1. Table 15.1

Types ofprediction (Lambe, 1973)

Prediction type

When prediction made

Results at time prediction made

Class A Class B Class B 1 Class C Class C 1

Before event During event During event After event After event

Not known Not known Known Not known Known

The structures will be affected by construction of JLE running tunnels. The westbound tunnel was constructed beneath the structures in February 1996; the results to date of the measurements of the performance of the structures have not been revealed to GCG. The eastbound tunnel is shortly to be constructed beneath the structures. In terms of the classification of Lambe, therefore, Class C predictions are presented in this report for effects of the westbound tunnel and Class A predictions for effects of the eastbound tunnel shortly to be constructed.

Previous page

is blank

Ch 15 Settlement predictions for Neptune, Murdoch, and Clegg Houses and adjacent masonry walls

217

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

I Figure 15.1

5m

0

25 m

Locations of the buildings and running tunnels

15.2

THE STRUCTURES

15.2.1

Neptune House, Murdoch House and Clegg House A plan of the buildings in relation to the tunnels is shown on Figure 15.1. They are located about 200 m to the west of Canada Water station, which is under construction as part of the JLE. Photographs of Neptune House, Murdoch House and Clegg House are given in the case study chapter as Figures 43.2 to 43.4 respectively. All three buildings are of very similar construction, being three-storey load-bearing brick structures.

Neptune House and Murdoch House are very similar. The approximate plan dimensions of each building are 40 m x 8 m, and each building is divided into four flats on each floor. Several load-bearing walls cross between the two main external long walls and divide the building into flats. Details of the foundations are not known, but it is probable that the walls are on concrete strip footings that are founded at a depth of about 1.5 m below ground level (and possibly deeper locally if soft alluvium was encountered at founding level). Clegg House is of similar construction, except that it is half the length of Neptune House and Murdoch House. For the locations of the sockets for precise levelling and taping see Figures 43.5 and 43.6.

218

Building response to tunnelling

15.2.2

Masonry walls

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The two masonry walls are shown on the plan in Figure 15.2. Wall 1 is approximately 22 m long and 1.7 m high; it is immediately adjacent to two blocks of flats at Niagara Court. Wall 2 is approximately 32 m long and 0.8 m high.

-....

5m 0

Figure 15.2

15.3

.

.

.

.

X Surface settlement points

Locations of the two masonry walls

G ROUND CONDITIONS A longitudinal section through the tunnels together with relevant JLE boreholes is shown on Figure 15.3. The approximate locations of the structures are shown on the section. The ground profile is generally about 2 m of made ground and alluvium overlying about 4 m of Thames Gravel. The latter is underlain by the Lambeth Group (Woolwich and Reading Beds) extending to a depth of approximately 22 m below ground level.

The various units of the Lambeth Group are shown on Figure 15.3. The tunnels rise in level from west to east and beneath the structures the tunnels are in mixed ground conditions, with the crown in the Pebble Beds and the majority of the tunnel in the Glauconitic Sand horizon. Prior to construction of the JLE, the observed piezometric levels in the Glauconitic Sand were as shown in Figure 15.3, indicating water pressures of about 100 H a at the level of the tunnel. However, deep well dewatering has been undertaken during construction of Canada Water station, which has resulted in the water pressures in the more permeable strata of the Lambeth Group in the general area being reduced to about zero (Nyren, 1996).

Ch 15 Settlement predictions for Neptune, Murdoch, and Clegg Houses and adjacent masonry walls

219

West

-

-

-

Neptune

Murdoch

wall

*

East 110 100

50

MG AL TG USC UMC LSC PB GS TB BH UC

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 15.3

15.4

Made Ground Alluvium Thames Gravels Upper Shelly Clay Upper Mottled Clay Lower Shelly Clay PebbleBed Glauconitic Sand Thanel Beds Bullhead Beds Upper Chalk

“L

Water strike in bore-hole

Longitudinal section of tunnels

THE TUNNELS Beneath the structures, the axis level of the tunnels varies in depth below foundation level from approximately 17 m beneath Neptune House, Murdoch House, Clegg House and Wall 2 to approximately 15 m beneath Wall I . Both tunnels are at almost the same level and their axes are separated horizontally by 23 m at Neptune House, reducing to 17matWall 1. Both the eastbound and westbound tunnels are being constructed with earth pressure balance (EPB) shields manufactured by Kawasaki of Japan. The shields are 7 m long and 5.03 m OD, with earth pressure being maintained by extruding the spoil through a 6.5 m-long x 600 mm-diameter screw conveyor. The tunnels are lined with 1.2 m-long precast concrete bolted rings, which are erected within the tailskin of the shield.

15.5

GREENFIELD SITE SETTLEMENT PREDICTIONS

15.5.1

General The approach to predicting the response of the structures to the tunnel construction is first to predict the greenfield site settlements at foundation level and then to assess how the inherent stiffness of each structure is likely to modify these settlements. The greenfield site prediction is the empirical approach outlined by O’Reilly and New ( 1 982) and Attewell and Woodman (1 982).

15.5.2

Ass umpt ions The Southwark Park reference site is approximately 300 m to the west of the location of Neptune House, Murdoch House and Clegg House. The ground conditions are generally similar at this site to those beneath the structures, although the level of the tunnel is slightly lower at this site and is entirely within the Glauconitic Sand horizon of the Lambeth Group. The westbound and eastbound running tunnels were driven through the reference site in mid-January 1996 and at the end of June 1996 respectively. For both tunnels the recorded volume loss immediately after the tunnels had passed through the site was about 0.5 per cent, with a trough width parameter K of 0.45 (Nyren, 1996). Negligible interaction was observed (the tunnel axes are separated horizontally by 28 m).

220

Building response to tunnelling

In view of the more mixed ground conditions likely to be encountered by the tunnels beneath the structures, as described in Section 15.3, the volume loss assumed for the settlement predictions has been increased from the observed 0.5 per cent at the Southwark Park reference site to 0.75 per cent. The trough width parameter applied at foundation level has been assumed to be 0.45.

15.5.3

Settlement contours The predicted greenfield site settlement contours for the westbound tunnel construction only are shown in Figure 15.4. The predicted contours corresponding to construction of both the westbound and the eastbound tunnel are shown in Figure 15.5. The principle of superposition has been adopted. It has been assumed that time-dependent ground movements in the six-month period before construction of the eastbound tunnel will have been very small; these have therefore been neglected, particularly in view of the relatively small thickness of clay strata above the tunnels.

/3 Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

i

1

Wall 2

-4

-2

-

Scale: m

--

Figure 15.4

Predicted greenfield site settlement contours after construction of the westbound running tunnel

Figure 15.5

Predicted greenfield site settlement contours after construction of both running tunnels

Ch 15 Settlement predictions for Neptune, Murdoch, and Clegg Houses and adjacent masonry walls

22 1

15.6

PREDICTED BUILDING PERFORMANCE Sections at the structures showing the settlements corresponding to the greenfield site contours after construction of the westbound tunnel are shown in Figures 15.6 to 15.8 and in Figures 15.10 to 15.12 after construction of both tunnels. The sections correspond to the walls of the structures being monitored.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The influence of the structure stiffness in modifying the predicted greenfield site settlement profiles has been assessed. Use has been made of the parametric study by Potts and Addenbrooke (1 996), based on finite element analyses of a wide range of beam stiffnesses, the structure being idealised as a simple beam. Potts and Addenbrooke define the relative bending stiffness p* and relative axial stiffness a* of the structure, which are dimensionless quantities expressing the stiffness of the structure in relation to a representative stiffness of the underlying soil. Details of the assessment of these relative stiffnesses for each of the structures are given in Section 15.8. It is concluded that Neptune House, Murdoch House and Clegg House will all behave almost rigidly, with negligible deflection ratio developing in response to the tunnel construction. These buildings are therefore predicted to show almost uniform tilt, as indicated on the cross-sections in Figures 15.6 to 15.8 and 15.10 to 15.12, where adjustments have been made to the greenfield site settlement profiles. In the case of the masonry walls, more flexible behaviour is predicted. For Wall 1, which is 1.7 m high, the modification factor to the deflection ratio proposed by Potts and Addenbrooke is about 0.7, as shown in Section 15.7.2. This indicates relatively flexible behaviour (a factor of 1 .O corresponding to fully flexible behaviour). In view of the small predicted settlements, for practical purposes it has been assumed that this wall will behave in a hlly flexible manner. No adjustment has therefore been made to the predicted greenfield site settlement profiles in Figures 15.9(b) and 15.13(b). Wall 2 is only 0.8 m high and the assessment of its relative bending stiffness in Section 3 1.7.2 shows it to behave in a fully flexible manner. As for Wall 1, no adjustment has therefore been made to the predicted greenfield site settlement profiles in Figures 15.9(a) and 15.13(a).

222

Building response to tunnelling

Neptune House: West wall,

-N

~

Neptune House: East wall, S-N

0

0

- 2 E E Y 4

2 4

c

a,

$

z

"

6

6

8

8

"0

5

10 15 20 25 30 35 40 45 Distance (m)

Figure 15.6

10

0

Murdoch House: North wall, W-E

Ec

10 15 20 25 30 35 40 45 Distance (rn)

Predicted greenfield site and building settlements after westbound running tunnel excavation: Neptune House (a) West wall, (b) East wall

Murdoch House: South wall, W-E

0

- 2 E E

5

0 2

4

4

6

6

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

(I)

-$

z "

8 10 0

8

5

10 15 20 25 30 35 40 45 Distance (rn)

Figure 15.7

0

10

z "

Clegg House: NW wall, W-E

-

8 I

I

I

I I 10 15 Distance (rn)

I 5

Figure 15.8

10 15 20 25 30 35 40 45 Distance (m)

Clegg House: SE wall, W-E

6 t 10' 0

5

Predicted greenfield site and building settlements after westbound running tunnel excavation: Murdoch House (a) North wall, (b) South wall

- 2 E E = : 4 c

-aE

0

I

20

I

25

10 0

5

10 15 Distance (rn)

20

25

Predicted greenfield site and building settlements after westbound running tunnel excavation: Clegg House (a) NW wall, (b) SE wall

Wall 2: N-S

Wall 1: NE-SW

0

0

- 2 E E

2

Y

- 4 c

4

E

6

a,

z

6

8

" 8 10 0

5

Figure 15.9

10

15 20 25 Distance (rn)

30

35

10

0

5

10 15 Distance (m)

20

25

Predicted greenfield site and wall settlements after westbound running tunnel excavation: (a) Wall 2, (b) Wall 1

Ch 15 Settlement predictions for Neptune, Murdoch, and Clegg Houses and adjacent masonry walls

223

0

Neptune House: West wall, S-N

Neptune House: East wall, S-N

0

- 2 E E =:4

2 4

c

0

-$

6

6 a ( " 8 5

"0

8

5

Figure 15.10

0

0

10 15 20 25 30 35 40 45 Distance (m)

5

10 15 20 25 30 35 40 45 Distance (m)

Predicted greenfield site and building settlements after construction of both running tunnels: Neptune House (a) West wall, (b) East wall

Murdoch House: North wall, W-E

0

- 2 E E zC 4

2

-$

6

6

4

8

8

Murdoch House: South wall, W-E

4

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

W

r:

loo

5

10 15 20 25 30 35 40 45 Distance (m)

Figure 15.11

0

10' 0

"

5

"

"

"

'

10 15 20 25 30 35 40 45 Distance (m)

Predicted greenfield site and building settlements after construction of both running tunnels: Murdoch House (a) North wall, (b) South wall

Clegg House: NW wall, W-E

Clegg House: SE wall, W-E

- 2 E E z 4 C (U

$

6

z a

8 101 0

I

I

Figure 15.12

I

10 15 Distance (rn)

5

I

20

J 25

Wall 2: N-S

C

I

I

10 15 Distance (rn)

I

20

I 25

Wall 1: NE-SW

0

- 2 E E -

5

Predicted greenfield site and building settlements after construction of both running tunnels: Clegg House (a) NW wall, (b) SE wall

0

Y

I

0

2 4

4

W

6

$ 6 z a, ( " 8 "0

8 5

Figure 15.13

224

10

15 20 25 Distance (m)

30

35

10 0

5

10 15 Distance (m)

20

25

Predicted greenfield site and wall settlements after construction of both running tunnels: (a) Wall 2, (b) Wall 1

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The likely damage to the structures associated with the predicted settlements is best assessed in terms of the induced tensile strains (Burland and Wroth, 1974; Burland et al, 1977; Boscardin and Cording, 1989). Treating each structure as a simple beam, the maximum tensile strain can be expressed in terms of the deflection ratio, AlL, and the average horizontal strain, &h (Burland, 1995; Mair et a/, 1996). In the case of Neptune House, Murdoch House and Clegg House, negligible deflection ratios are predicted because of the building behaviour being essentially rigid in response to the tunnelling. In view of the walls probably being founded on continuous strip footings, it is considered that these buildings will experience only very small horizontal strains. As a consequence of the negligible deflection ratios predicted and the very small horizontal strains, the level of maximum tensile strain induced is likely to also be very small and therefore negligible damage is predicted. In the case of the two masonry walls, it is likely that the horizontal ground strain (albeit small) will be transmitted to the walls. Even taking this into account together with the predicted maximum deflection ratios, the maximum tensile strain induced in both walls is very low. In Wall 2, for example, the average horizontal tensile strain on completion of both tunnels is predicted to be 0.017 per cent (in the hogging zone between the two tunnels), giving a maximum tensile strain of 0.023 per cent. Assuming that there were no pre-existing cracks in the walls, the degree of damage associated with these levels of strain is negligible.

15.7

ASSESSMENT OF INFLUENCE OF STIFFNESS OF STRUCTURES Note: the three parts of this section are the calculations given by GCG, dated July 1996 and initialled RJM. These were hand-written and have been transcribed by the editors.

15.7.1

The buildings Assumefoundations are strip footings founded at approx 1.5 m below ground level. Depth of tunnel axis (z) below ground level = I7 m. below founding level, 10 m below ground level, estimate Es at 0.01 !% strain. At 0 . 5 ~ Borehole SM90-417 shows stratum level of 10 m to be Woolwich and Reading Beds Upper Mottled Clay ("stiffbecoming very stiff"). Tunnel axis at 18.5 m below ground level is in Woolwich and Reading Beds Glauconitic sand. For Woolwich and Reading Beds,

_ -- -= 3G P;

1000 for 0.01% strain

P;,

0: =(IOx20)-(5x10)=

pb ="(I

15OkPa

150

0)

+ 2 K o ) = - ~ 4 = 200 kPa 3 3 :. E s = 1OOOx 200 kPa = 200 Mpa Estimate groundwater level at depth of 5 m and KO= 1.5. Assume E for masonry lies between 5 x I O6 kPa and I0 x I O6 kPa. Take E = 7.5x106kPa

Ch 15 Settlement predictions for Neptune, Murdoch. and Clegg Houses and

~

Neptune House Tunnels approximately perpendicular to long dimension of building. Three-storey building, 39 m x 8 m in plan. Estimate height of load-bearing brickwork walls to be 9 m. El - 7 . 5 ~ 1 0~ ~1 x 9 ~ p * E,H = 7 - 2 0 0 ~ 1 ~0 ~1 9 . x5i 2 ~ = 1.6 x 10-* (H = B/2

= 39/2

=19.5 m)

From Figure 3. I0 (Figure I6 of Potts and Addenbrooke, 1996), pg or Mhog are less than 0.2for this value of p* and building will behave rigidly for any value of e/B. Murdoch House

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Building more skew to direction of tunnels. Similar dimensions to Neptune House. Hence likely to behave as rigidly, andpossibly more so, because of influence of shorter dimension (ie 8 m cf39 m). Clegg House 22 m x 8 m in plan, otherwise similar to Neptune and Murdoch Houses. Likely to behave in a substantially rigid manner as for Neptune and Murdoch Houses.

15.7.2

The walls Wall 1

1.7mhigh,21 mlong; .:H=21/2=10..5m p* =

7.5~10 ~ ~1 ~ 1 . 7 ~ =1.3~10-~ 200~1~ 0 1~0 . 5 4 XI2

Eastbound running tunnel, e/B = I .O Westbound running tunnel, e/B = 0.25 main effect isfrom westbound running tunnel: max. settlement = I0 mm, sagging mode. dL~,,,,,~,,d, = 0.1 x I 0-3 Eastbound running tunnel increases settlement to = 1I mm, but reduces d L . Saggingmode, p g = 0 . 7 f o r p * = l . 3 xlO”ande/B ~ 0 . 2 5 .

.:

Wall I likely to behave reasonablyjlexibly.

Wall 2 0.8 m high, 32 m long; .:H

p* =

= 32/2 =

16 m

7.5 x 106x 1 x 0.g3 = 2.4 x 10-5 200~1~ 0 ~ 1 x12 6 ~

Eastbound running tunnel, e/B = 0, sagging mode, pg = 1.0 Westbound running tunnel, e/B = 0.5, hogging mode, n/phos= 1.0

.: expect Wall 2 to behave fully flexibly. 226

Building response to tunnelling

15.7.3

Check axial stiffness Axial stiffness, a* (Potts and Addenbrooke, 1996)

H = Bl2 take E = 7.5 x 1O6 kPa for masonry Neptune, Murdoch and Clegg Houses 9 m high walls, .: A = 9 m2/m B = 30 m (in direction transverse to tunnels) .-.H = 15 m 7 . 5 ~ 1 x0 9~ = 22.5 ff*= 2 0 0 ~ 1 x15 0~

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Wall 1 1.7 m high wall, .: A = 1.7 m2/m B = 15 m (in direction transverse to tunnels) .: H = 7.5 m 7 . 5 ~ 1 0 6x1.7 ff*= = 8.5 2 0 0 1~o3 x 7.5 Wall 2

0.8 m high wall, .:A = 0.8 m2/m B = 32 m (in direction transverse to tunnels) .: H = 1 6 m 7.5 x 106x 0.8 ff*= = 1.9 2 0 0 x 1 0 ~x i 6 Figure 3.10 (Figure 16 of Potts and Addenbroo.,z, 1991 is appropriate for a > Above calculations show that a* > 0.5for UN the structures.

15.8

SUMMARY Class A and Class C predictions have been made for the settlement and performance of Neptune House, Murdoch House, Clegg House and two masonry walls. The Class A predictions are for the eastbound running tunnel, which is being constructed beneath the structures at the time of preparation of this report. The Class C predictions are for the westbound running tunnel, which was constructed in February 1996; the results of the measurements for the westbound tunnel have not been made available to GCG. Using an empirical approach, the maximum predicted settlement of all the structures is less than 10 mm. Parametric finite element studies reported by Potts and Addenbrooke (1 996) have been used as a basis for assessing the influence of the stiffness of each structure on the greenfield site predictions. Neptune House, Murdoch House and Clegg House are predicted to behave almost rigidly in response to the tunnelling. In contrast, the two masonry walls are predicted to behave in a much more flexible manner. For all the structures considered, the level of damage predicted is negligible, meaning that no worse than a few hairline cracks can be expected, assuming that there were no pre-existing cracks in the masonry.

Ch 15 Settlement predictions for Neptune, Murdoch, and Clegg Houses and adjacent masonry walls

227

REFERENCES

15.9

ATTEWELL, P B and WOODMAN, J P (1982). Predicting the dynamics of ground settlement and it derivatives caused by tunnelling in soil. Ground Engineering, vol 15, no 8, pp 13-22,36 BOSCARDIN, M D and CORDING, E G (1989). Building response to excavation induced settlement. J Geotec. Engrg ASCE, 115; (l), pp 1-21 BURLAND, J B (1995). Assessment of risk damage to buildings due to tunnelling and excavations. Invited Special Lecture to IS-Tokyo ’95: 1st Int Conf on Earthquake Geotechnical Engineering

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

BURLAND, J B, BROMS, B and DE MELLO, V F B (1977). Behaviour of foundations and structures - State of the Art Report, Session 2, Proc, Proc 9th Int Conf Soil Mech and Found Engg, Tokyo, 2, pp 495-546 BURLAND, J B and WROTH, C P (1974). Settlement of buildings and associated damage. State of the Art Review, Conf on Settlement of Structures, Cambridge. Pentech Press, London, pp 61 1-654 GEOTECHNICAL CONSULTING GROUP (July 1996). Neptune House, Murdoch House, Clegg House and masonry walls, Settlement Predictions. Geotechnical Consulting Group, London LAMBE, T W (1973). Predictions in geotechnical engineering. Gkotechnique 23, 2, pp 149-202 MAIR, R J, TAYLOR, R N and BURLAND, J B (1996). Prediction of ground movements and assessment of risk of building damage due to bored tunnelling. Geotechnical Aspects of Underground Construction in Soft Ground (R J Mair and R N Taylor eds), Balkema, Rotterdam, pp 7 13-7 18 NYREN, R J (1996). Personal communication O’REILLY, M P and NEW, B M (1982). Settlements above tunnels in the United Kingdom - their magnitude and prediction. Tunnelling ’82, London, Instn Mining and Metallurgy, pp 173-1 81 I

228

POTTS, D M and ADDENBROOKE, T 1 (1996). A structure’s influence on tunnelling induced ground movements. Accepted for publication in Proc Instn Civ Engrs Geotechnical Engineering. (Note: the foregoing is the reference given in Geotechnical Consulting Group (1996), but the paper was subsequently published as: Potts, D M and Addenbrooke, T I (1997) A structure’s influence on tunnelling induced ground movements. Proc Inst Civ Engrs, Geotechnical Engineering, vol 125, Issue 2, April, pp 109-25)

Building response to tunnelling

16

JLE construction works at London Bridge station

C F Field

16.1

SUMMARY

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

This chapter gives a summary of the main subsurface excavations and associated works for the new London Bridge station of the Jubilee Line Extension and the Northern Line. The complex series of NATM excavations for running and platform tunnels, crosspassages and escalator and safety shafts affected a range of buildings at the surface. Compensation grouting and other protective measures were applied to these buildings. The chapter presents an overview of the construction methods and programme and of the protective measures and their use.

16.2

INTRODUCTlON Construction of the Jubilee Line station at London Bridge was part of Contract 104 of the JLEP. In addition to Jubilee Line works, this contract included major improvements and upgrading of both the Northern Line passenger facilities and the interchange with the National Railways station. Figure 2.7 gives an indication of the complexity of these works. They took place in a congested urban area just south of the River Thames around the approach to London Bridge (see Figure 16.1). Changes to the overall scheme layout were made during the works. These included the adoption of the contractor’s alternative based on use of the New Austrian Tunnelling Method (NATM) and layout alterations following a design review after the Heathrow collapse. The Northern Line is sited below Borough High Street running in a north-south direction. The alignment of the Jubilee Line lies a little farther south and runs in a westeast direction at a level below the Northern Line. The eastern end of the Jubilee Line station is located under the London Bridge main line terminus. Many of the buildings in the vicinity are of historic significance, sensitive fabric and structure, or important function. Their proximity to the new construction of the several tunnels and shafts and their associated works meant that they could be adversely affected by ground movements induced by excavation. It was found possible to use an existing pedestrian subway and disused tunnels of the former City and South London Line for grouting access. The JLE works comprised: running tunnels approximately 200 m long eastwards towards the station complex from the surface access compound and work-site; platform tunnels; a central concourse linking the two ticket halls via escalator shafts; connecting cross adits; interchange escape and ventilation tunnels; and associated shafts. A stepplate junction was constructed as part of a crossover at the western end of the contract. The Northern Line works were divided into two stages. The first stage comprised all the elements of works required to switch the operational railway from the existing southbound platform into a new southbound platform tunnel. It envisaged building two step-plate junctions north and south of the station encircling the existing Northern Line

Ch 16 JLE construction works at London Bridge station

229

to allow the switchover; these would connect to the new platform tunnel via running tunnels. Passenger circulation would be provided by cross passages between new and existing tunnels and by means of an overbridge across the existing northbound platform tunnel to connect to a new escalator shaft to the main ticket hall.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The second stage works linked the Northern Line low-level passenger areas to the new ticket hall constructed under Borough High Street.

Figure 16.1

16.3

Aerial view ofwesfem approach to London Bridge station

GROUND CONDITIONS Geology The geology through which the works were to be constructed and the geotechnical assessment of the ground conditions were established by a series of desk studies, field and laboratory investigations and subsequent interpretative reports (see Chapter 5 for an overview). The following’summarisesthe London Bridge station area.

230

Building response to tunnelling

A typical sequence of strata in terms of depths from ground level at London Bridge is: Made Ground

0 to 5.0 m

Alluvium

5.0 to 6.3 m

Terrace Gravels

6.3 to 12.5 m

London Clay

12.5 to 35.0 m

Blackheath Beds

35.0 to 35.5 m

Lambeth Group

> 35.5 m.

Made Ground. This is present on site as a layer of gravel, sand, clay and demolition debris, with inclusions of organic material and remnants of past construction. The layer varies in composition and density across the site. SPT blow counts in the made ground were in the range of 0 to 12.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Alluvium. Recent alluvial deposits up to 1.5 m thick comprise layers of soft to firm silty clay and of sand with gravel. The alluvium contains occasional roots and organic debris. Terrace Gravels. These fluvial deposits, several metres thick, are of medium dense to dense fine to coarse gravel and coarse sand. Typical SPT values are from 20 to 40. London Clay. The stratum is 15-22 m thick in the London Bridge area. The clay is generally homogeneous, being stiff to very stiff grey clay, but with some thin silty sand layers, intermittent clay-stone bands, frequent fissures and occasional slickensides. The top 5-1 0 m is weathered, where the moisture content is about 4 per cent higher and fissures are more frequent than in the unweathered clay. Undrained shear strengths are between 50 and 150 kN/m2. SPT values increase from 20 at about 13 m depth to 50 at the base of the stratum. Blackheath Beds. Where present, this thin stratum (typically less than 0.5 m) comprises dense black pebbles in a dark brown clayey or sandy matrix. Lambeth Group. Upper Mottled Beds of the Lambeth Group were encountered below about 35 m. These are red-brown and blue-grey mottled sandy very stiff clays and dense silts. Undrained shear strengths were in the range 150-550 kN/m2. Groundwater conditions Two major aquifers lie within the area, the upper one being the superficial terrace gravel deposits overlying the London Clay, and the lower aquifer comprising the basal units of the Lambeth Group, the Thanet Beds and Upper Chalk. The lower aquifer is confined by the low-permeability clay layers of the Lambeth Group and the London Clay. The upper aquifer is linked to the River Thames. Standing water level is approximately 100 m PD, but tidal fluctuations may modify this by 5 1 m depending on the attenuation in response with distance from the river. Groundwater present in the permeable basal units of the London Clay caused seepage at claystone horizons and sand pockets. The piezometric surface of the lower or Chalk aquifer is approximately 80 m PD, which is within the lower granular units of the Lambeth Group.

Ch 16 JLE construction works at London Bridge station

231

Construction aspects Lower-level tunnels were sited so that excavation, as far as possible, was carried out within the London Clay. At the eastern end of the site, however, shaft bases and tunnel inverts were expected to encounter the upper layers of the Lambeth Group. To facilitate excavation of the inclined escalator and vertical ventilation and escape shafts through the water-bearing Terrace Gravels, extensive permeation grouting was carried out to strengthen them and reduce their permeability. The actual geological conditions met during the progress of the works were little different from those assessed from the site investigations. The only exception of note was that the Terrace GraveldLondon Clay interface was some 2 m lower than had been expected in the area between St Thomas Street and London Bridge Street.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Forward probing and face logging by the contractor’s geotechnical engineers were undertaken during the construction of all tunnels to gather detailed knowledge of the ground conditions. The exposures and samples were described in accordance with BS 5930. An example of a geological face log is shown as Figure 16.2.

Strata descriptions

Tunnekhaft

j Date

Chainage (m)

, Rib number

-

Crown exposure

Time Logger I^

I Sketches of details

I

Structural details

r----1 Photographs, samples

Comments

Figure 16.2

232

Example extracts from of a face log from the eastbound station tunnel at London Bridge station

Building response to tunnelling

16.4

EXISTING STRUCTURES There is a mix of building types in the area of influence of the works. These range from a modern 25-storey tower block founded on deep piles to the Grade 1 listed Southwark Cathedral possibly founded upon shallow strip footings. Many properties have been repaired, altered or added to during their history. Some have been constructed such that they interact with adjacent properties, for example by sharing remnant foundations. Redevelopment of others has resulted in differing foundation types for different parts of the building. The complications arising from these factors are explored further in Sections 16.10 and 16.1 1.

16.5

PROGRAMME

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

From the outset, the programme for London Bridge station was dominated by several external constraints. These may be grouped as follows: 1.

Remoteness of the main surface work-site (in Redcross Way) from the station tunnels.

2.

Restricted works areas for shafts.

3.

Proximity of and need to maintain continuity of public access, for example, to the Northern Line station, main line station and Borough High Street.

4.

Density and fragility of existing buildings and structures in the area.

5.

Release of areas of the site to trackwork and E&M contractors as soon as possible.

These constraints meant that the critical path for the JLE works was finely balanced between the tunnel and the shafts (including escalators). Overall, however, the critical path for the contract was completion of the Borough High Street ticket hall and the final stage of the Northern Line works. Although substantial delays were experienced on this contract, considerable effort was put into mitigating these delays by optimising the programme. Every time a programme change was proposed, the access and logistics had to be thoroughly reviewed, to the extent that often these became critical factors.

16.5.1

Suspension of NATM work On 2 1 October 1994, the catastrophic collapse occurred in the Heathrow Express tunnels, and the JLEP suspended all NATM work, pending an extensive review of the tunnel design and construction at both Waterloo and London Bridge. Notice of “nonobjection” was then received in stages from the Health and Safety Executive (HSE), which allowed tunnel construction to be restarted, albeit in various modified forms. This had a major impact on the programme. The works had to be reprogrammed to recover the time lost as a direct consequence of the suspension, which now jeopardised a critical date, by re-sequencing or accelerating activities to accommodate a change in methods of working. By July 1995, sufficient redesign and planning had been carried out to produce an acceptable revised programme. The complexity of the situation is reflected in the number of iterations required to reach this position.

Ch 16 JLE construction works at London Bridge station

233

Over the next nine months, however, the design and scope of the tunnel works underwent further major changes. Of these, a revised layout of the tunnels to the east of the JLE platforms and the redesign of the central concourse were the most critical. Consequently, by November 1996, a further overall programme had been developed, which took into account these changes and other delays and which was targeted on the critical dates for start of tracklaying and station fit-out.

16.5.2

Northern Line works

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The incorporation of the Northern Line works into the JLE contract affected the programme significantly, as they had to be phased to provide continuity of operation of the existing station. Initially, there were to have been two stages of construction. The fit-out and commissioning of the new southbound tunnel and the new escalators from the main ticket hall were to be completed before the second stage - Borough High Street escalator and intermediate concourse - could be progressed. In practice, a more complicated sequence evolved. This allowed the new southbound tunnel to be opened to the public well in advance of the much-delayed main ticket hall escalators, and progress to be made on the Borough High Street escalators shaft.

16.6

DESIGN AND CONSTRUCTION

16.6.1

NATM design history The engineer’s design of the works envisaged segmentally lined tunnels constructed by traditional shield andor hand mining methods, although tenderers were encouraged to submit alternatives. The successful bidder’s alternative was based upon use of the New Austrian Tunnelling Method (NATM) for the majority of the low-level tunnelling works within the London Clay. A consequence of this was that the contractor would assume responsibility for design of the primary lining. A large-scale trial in an early phase of the contract proved the success of NATM in tandem with compensation grouting. Works on the Jubilee Line running tunnels, and the access adit, running tunnel and station tunnel of the Northern Line, were in progress using this method when the Heathrow collapse happened. As a result of this and the subsequent HSE investigation, the Northern Line works reverted to the engineer’s design because of programme criticality. The benefits from the use of NATM in constructing the Jubilee Line station tunnel complex with its many tunnel junctions (see Figure 16.3) were seen to outweigh any temporary programme advantage that would be gained by re-adopting the engineer’s design. Both the NATM works and the overall design underwent detailed review to ensure that the HSE would not object to the contractor continuing to use this method. Following the review, the station layout was changed (see Figure 16.4) and construction methods were revised. An indication of the relative scale of the station excavation is given in Figure 16.5.

234

Building response to tunnelling

F Running tunnel

F Eastbound station tunnel

Permanent lining 200

rF

Access tunnel

Separation membrane

estbound

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Interlining4OOJ

16.6.2

Figure 16.3

Detail of a tunnel junction at London Bridge station

Figure 16.4

Revisions of station layout after review of NATM

F Running

Lseparation membrane

NATM design philosophy The overall design philosophy adopted for the shotcrete linings departed from the original NATM observational method, which allows the design to be modified in the light of measurements of the ground response thereby achieving economy while maintaining safety. The JLE approach differed in that observational data could only be used to amend support requirements to increase the stability of the tunnel. A primary reason for adopting this approach was pre-eminent need to maintain tunnel stability in an urban area - especially given the multiple tunnel layout at London Bridge.

Ch 16 JLE construction works at London Bridge station

235

The standard in-tunnel instrumentation consisted of convergence measurement pins (tape extensometer) and radial and tangential pressure cells. The tape extensometer readings were a simple and accurate way to measure lining deformation. Measurement arrays were installed at every 25 m of tunnel advance as well as before and after each breakout for a junction. In the centre of the junction, the roof settlements were monitored. As the installation of the instrumentation was an integral part of the construction cycle, each monitoring station was established as soon as possible behind the tunnel face. In addition, visual inspection of tunnel faces and exposed ground surfaces was an important complement to the instrument readings. Buildings

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

M

Cross-section of station concourse Figure 16.5

Sketch of relative proportions of the tunnelling and the overlying buildings

The original contract required the large-diameter tunnels to be excavated initially as a side drift with a temporary wall, then enlarged to full cross-sectional area after a minimum advance of 20 m. It also envisaged the need to excavate heading, bench and invert as individual components within the sequence. During the reappraisal of NATM tunnelling, a review was also carried out on the methodology of construction of large-diameter tunnels using NATM techniques. Initially an assessment was made of the effective overburden acting on a NATM excavation in London Clay when compensation grouting is applied. This, in conjunction with adoption of the load factors taken from BS8 1 10, gave rise to a concern that an economic and practically viable temporary sidewall, which is inherent in the original excavation sequence, could not be designed to these criteria. Also of concern was the integrity of the joint between the temporary sidewall, tunnel invert and crown, and the possible inadequate thickness and quality of the invert shotcrete.

236

Building response to tunnelling

tunnel T

0

0 Ln

rY

0

0 0 m

!n I-

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

?

Figure 16.6

NA TM construction for large-diameter tunnels

The review proposed the following general sequence of work for large-diameter tunnel construction (Figure 16.6). 1. Construct a NATM pilot tunnel (sized as running tunnel), the pilot tunnel to be located such that its invert was below that of the future enlargement.

2. Within the invert of the pilot tunnel, cast a reinforced concrete section, which will form the invert section of the enlarged tunnel lining. 3.

Enlarge the pilot tunnel to full face using NATM techniques in three stages, ie heading, bench and invert, breaking out the pilot tunnel as the drive progresses.

4.

Erect SGI segmental lining within the NATM shell approximately 30-40 m behind the tunnel face to provide the permanent lining to the tunnel at an early stage.

This general sequence was varied to cater for individual circumstances as work progressed. The use of the pilot tunnels was especially favoured because as well as facilitating formation of the invert of the enlargement, it would also reduce the area of face exposed, shorten cycle times and reduce settlement.

Ch 16 JLE construction works at London Bridge station

237

16.6.3

Methods of NATM construction

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Excavation of tunnels was carried out generally by back-acter type machines (Figure 16.7),with trimming of the advance by hand to achieve the required profile. The number and sequence of benches cut varied according to tunnel diameter and the specific machine being used. In most cases, the length of advance for all sizes of tunnels was 1 m.

Figure 16.7

Eastbound platform tunnel under construction (pilot tunnel and top heading visible)

A sealing coat up to 50 mm thick of sprayed concrete was applied to the excavated profile and face to retard drying of the clay and prolong the “stand-up” time of the excavation.

Sprayed concrete with a specified characteristic compressive strength of 25 MPa at 28 days and 10 MPa at 1 day was applied around the tunnel profile to a pre-determined thickness - which also varied according to tunnel diameter (between 150 and 400 mm).

238

Building response to tunnelling

76.6.4

Conventional methods were used, with excavation carried out by air-driven hand tools, the face supported by timber face-boards with walings strutted off the assembled tunnel lining. Exact details of the face support system changed according to tunnel size, rate of progress and ground conditions. Tunnel linings were erected with the aid of air winches, suitable crown bars and roller bolts. In all tunnels, for every ring, the annulus between the ground and lining was grouted.

16.7

MONIT0RING Measures were introduced to protect third party and LUL interests, to monitor and control the works, and to attempt to prevent, or at least limit, damage to existing structures. These are outlined below.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

16.7.1

Defect surveys Defect surveys were made on all buildings that might potentially be affected by the works, taken in practice as buildings within the predicted 1 mm settlement contour. The intention was that the surveys could be used after completion of the works to ascertain damage. An evaluation was also undertaken to identify those buildings at most risk, ie in categories 3 and 4 of Rankin’s classification (Rankin, 1988). Structural surveys of these properties were also considered necessary.

16.7.2

Settlement monitoring Surface monitoring systems of precise levelling sockets (see Chapter 6) set in the structures to be monitored were installed by the JLEP team and, separately, by the Contract 104 contractor. Approximately 2000 sockets in total were installed over the duration of the contract. All were monitored by the contractor at regular and varying intervals and when required on a daily basis in areas likely to be affected by the works. The results of these surveys were discussed daily with the JLEP site supervisory staff and written reports were also submitted to JLEP on a weekly basis. Independent checks were made regularly by JLEP staff.

16.7.3

Electronic monitoring As well as the general monitoring throughout the area, there was selective monitoring of sensitive structures, particularly for control of the compensation grouting where ground treatment was a mandatory protective measure. Electrolevels theoretically capable of providing thousands of readings daily were attached to these structures at basement level to provide computer-controlled real-time electronic monitoring that would record movements as the tunnelling works approached and passed the structure. In addition, there was further electronic monitoring through numerous vertically installed extensometers and inclinometers, in boreholes at spacings of 20 m along the route of the tunnels to depths of approximately 12 m, ie close to the clay/gravel interface. The whole monitoring system was set up with the intention that its measurements could be integrated with other knowledge to inform engineering decisions. Initially, it was thought that the electrolevels would provide information on movements to assist control in the short term, ie during compensation grouting, and that the results of precise levelling would be more appropriate to long-term control of movements. In the event, the outcome was disappointing, as only about a quarter of the data collected from electrolevels was readily usable. Furthermore, the amount of data generated created an

240

Building response to tunnelling

information overload. Later, it was recognised that the dominant effect of temperature on the electrolevel outputs often made interpretation and back-analysis of the data impossible. This led to greater emphasis on the direct survey methods.

16.7.4

Other monitoring Other monitoring techniques employed included strain gauges and tensioned wire extensometers to measure deformations, and tilt meters, tell tales and pins to measure defect movements. Some of these were subject to similar drawbacks to the electrolevels, and as the contract progressed it became evident that more reliance had to be placed upon traditional direct monitoring methods. This was largely a result of the need to assimilate and disseminate the information rapidly.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

16.8

COMPENSATION GROUTING The contract required that the settlement of 14 specified structures - generally over the Jubilee Line station complex - be limited to a maximum of 25 mm. By inference, this was to be achieved by compensation grouting and, as the implied consequence, building damage would be controlled and limited. Furthermore, the contractor was to assess the need to apply this control to a further 21 structures in order to limit damage to Category 2 “slight damage” (BRE Digest 25 I). The compensation grouting process chosen by the contractor was hydrofracture of the London Clay at a level of approximately 4 m below the top of the stratum. Tubes d manchette (60 rnm-diameter steel pipes) were installed in sub-horizontal 100 mmdiameter drill holes with grouting ports and sleeves at 0.5 m spacings along them. Initial injections were made before tunnelling to “condition” or “tighten up” the ground to the point when movement was detected. The type of grout used varied, but was generally a cement mix with 5 per cent bentonite. The rate of injection was limited to 5-10 litres/min at a pressure up to 40 bar. The injection volume at each sleeve was monitored and the flow stopped when the predetermined volume had been injected.

To install the TAMs in the Jubilee Line station area, the contractor selected three locations from which drilling would be carried out from shafts. These were: 39/41 Borough High Street (temporary) shaft West vent (permanent) shaft 0

East vent (permanent) shaft.

Additionally, TAMs would also be installed from the existing but disused City and South London Railway tunnels to keep drill holes to reasonable lengths. This arrangement is shown in Figure 16.9. The proposed shaft site in Borough High Street was within the footprint of the ticket hall that was to be constructed in this area. Owing to difficulties caused by the existing condition of No 43 Borough High Street and the resulting impact upon construction of the ticket hall, programme constraints required that alternative locations for this grout shaft be considered. At the same time, progress on both the west and east vent shafts was such that programme demands would not allow their use as compensation grouting sites.

Ch 16 JLE construction works at London Bridge station

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 16.9

Original compensation grouting layout for the JLE station complex

The contractor had already seen the benefit of using existing tunnels from which to drill and, in view of the delays, this approach was extended to include use of a disused pedestrian subway tunnel. This together with the City & South London Railway tunnels allowed most of the TAMs to be installed. To reach other areas,a small-diameter tunnel was excavated below St Thomas Street. The layout is shown in Figure 16.10 Disused

City and South LondonLine .

v Figure 16.10

242

Revised compensation grouting layout for the JLE station complex

Building response to tunnelling

The Heathrow collapse prompted the HSE to question the stability of tunnel faces and temporary tunnel linings (ie NATM tunnels) when subjected to the pressures induced in the ground by grouting in close proximity. To assuage these concerns, exclusion zones were introduced to distance the grouting operation from open tunnel faces or newly installed (assumed to be low-strength) tunnel linings. The effect of these exclusion zones partly negated the benefits of compensation grquting, as ground replacement.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Initially compensation grouting was successful in limiting settlements, but as tunnelling progressed, gaining lift on a structure became more difficult. It was therefore decided to determine the need for compensation grouting on a wider set of criteria than only those of settlement and distortionThis review process took account of the settlements recorded, defects observed and the details of compensation grouting carried out, attempting to establish correlations. Further settlement from works still to be constructed was then predicted and allowance made for the effects of any exclusion zones in operation or tubes U manchette no longer accessible as a result of the tunnelling works. The decision was then taken on the need or otherwise to grout. While this was subjective, the introduction of limited compensation grouting reduced damage to buildings. On the other hand, it led to settlements exceeding pre-set formulaic limits, although these were much less than the “greenfield site” predictions.

16.9

OTHER PROTECTIVE MEASURES Measures other than compensation grouting were used in specific locations to protect properties from the effects of tunnelling. These included the following. Underpinning of buildings and other structures was necessary where the works truncated the existing foundations. While not regarded as a protective measure because the underpinning does not transfer the loads away from the zone of disturbance, it proved to be of benefit in limiting damage, as recorded elsewhere. A jacking system was incorporated into a new lift tower to maintain the relative position of the lift core to the adjacent multi-storey building. This proved to be both simple and effective. Structural jacking to compensate for settlement, however, was not considered viable due to the need to underpin the structures in order to decouple them from their existing foundations. Strengthening measures to specific buildings were carried out at a few locations, although not considered as part of the design process. These works included bracing window openings, internal propping, and installation of tie bars. The efficacy of these measures is unknown.

16.10

EFFECTS ON BUILDINGS Most of the structures were monitored by the means listed in Section 16.7 during both the construction and immediate post-construction phases. Perhaps the most valuable data collected, however, was that gained from simple visual inspection, ie observing the response of the building during the works and from measuring joint and crack widths frequently. These methods have the drawbacks noted below, but there is an important benefit in demonstrating to building owners that attention is being given to the care of their property. Table 16.1 lists some of the potentially affected buildings showing their predicted and actual settlements and assessed damage categories. The locations of these buildings relative to the works are shown in Figure 16.11; see also Figure 16.12

Ch 16 JLE construction works at London Bridge station

The predicted data in Table 16.1 are based upon the original scheme layout. The results show that both settlements and damage (in general) were less than originally assessed. In reading the table, note the following additional points. Predicted and actual settlement and assessed damage category for a selection of buildings

Table 16.1 Building'

Approximate dimensions

Name (street number) and description

Height

Foundation

Im)

Predicted

Depth to crown of tunnel

Settlement Slope (max)

(m)

(mm)

Risk category (Rankin)

Actual settlement (max)

Assessed damage category

(mm)

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

St Thomas Street 1

The Bunch of Grapes (2) Masonry Four storeys plus a basement

12

Strip footings

24

80

1:lOO

3

21

See text

2

New City Court Masonry to front portion with RC frame at rear

12

Strip footings to front with piled foundation at rear

22

60

1:325

3

30

See text

3

1-7 St Thomas Street Frame with masonry cladding Four storeys plus single basement

14

Raft foundation level to about 3.2 m bgl

12

130

1:150

3

12

0

4

Mary Sheridan House (1 1-19) Masonry, load-bearing, part with steel frame supports. Four storeys plus single basement at front

12

Strip and mass concrete footings to 3.6 m below ground level

21

115

1:200

4

60

2

16

Mass concrete footings to about 3.0 m deep

22

88

1:300

4

58

2

Borough High Street 5

Post Office (1 9A) Masonry: Five storeys plus single basement

London Bridge Street 6

Telephone House (10-18) Masonry Five storeys plus two-storey basement

20

Strip footings

16

108

1:200

4

65

112

7

British Telecom House (20-26) RC frame with masonry cladding Seven storeys

19

Piled

9

103

1:225

4

65

0

8

Fielden House (28-30) RC frame Seven storeys at rear

20

Piled

21

68

1:300

3

54

1/2

1. See Figure 16.11 for building locations

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 16.11

Location plan in relation to the London Bridge station works of the buildings listed in Table 16.1

Figure 16.12

St Thomas Street, showing many of the buildings listed in Table 16.1

Ch 16 JLE construction works at London Bridge station

245

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

For Buildings 1 and 2, the actual damage is not given a category, as this should not be based purely upon joint widths exceeding the appropriate limit. Building 1was connected to an adjoining property that rotated because of compensation grouting causing the party wall to part from the remainder of the structure. For Building 2, the critical aspect was the rotation of the movement joint between the fagade, which is founded on strip footings, and a modem piled extension at the rear. As can be seen in Figure 16.13, the joint opened up to more than 25 mm (which is one indicator, though not the determiningone, of Category 5 , very severe). This movement is reflected in the recorded settlements, which also illustrate the effect of the piled foundations on the development of the settlement trough - see Figure 16.14.

c

Figure 16.13

Joint measurement

0 ,-I0

E E -20

v

-9 -30 2 -40

.-ca -50

s -60

0

i

6 -70

i

i

i

i

i

-80

0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80 Distance from tunnel centre-line (m)

Figure 16.14

246

Distance from tunnel centre-line (m)

Settlement profiles of New City Court

Building response to tunnelling

The other buildings listed in Table 16.1 all have substantial foundations. While not all of these are original, they have probably had a significant effect in limiting damage. Where damage has been recorded this has again occurred at those parts of the building that adjoin adjacent structures, eg party walls or the junctions of differing foundation types. For most of the buildings monitored by inspection and survey of the non-structural finishes, subsequent investigations in which these were removed and the structure exposed showed that the provisional assessments of damage made as a result of the earlier inspections were pessimistic. Buildings thought to be in a condition worse than actual damage category 3 were proved only to have suffered aesthetic damage. While the severity of this is dependent upon the nature of the decorative finish, this type of damage is relatively easy and economic to rectify - unlike any measures that could be considered to prevent this.

16.11

FURTHER DEVELOPMENT OF PREVENTATIVE MEASURES

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The problems of tunnelling-induced building settlement and potential damage were dealt with in the contract by adopting the generally accepted staged evaluation process, ie: to assess maximum settlements (normally at surface level) assuming there to be “greenfield site” conditions and that superposition can be applied where there are multiple tunnels 0

to calculate maximum slopes

0

to assign the risk category based on Rankin (1988) to determine the need for protective measures.

More detailed assessments were made equivalent to Stage 2. The option for protective measures was limited to compensation grouting and for all structures subject to compensation grouting a single maximum settlement criterion was specified. In the approach to assessment and protection to cater for particular situations the following steps help in tailoring specific measures to specific buildings. 1.

Desk study, supplemented as necessary by fieldwork to determine details of buildings and foundations, with surveys and trial pits to fill gaps in data.

2. Assessment of possible interactions between buildings that could result, for example, from contiguity, a party wall, or parts of buildings acting differently because of discontinuous foundations.

3. In the light of the comments in Section 16.10 about defect surveys tending to record the condition of the decorative fabric, it would be worthwhile to review the scope, role and method of these surveys. Greater emphasis could be placed, for example, upon logging the condition of the structure, especially those parts that tend to be neglected because of access difficulties. 4.

Attention should also be given to pre-construction monitoring of the building, not only for settlement, but more importantly for verticality of major components, in order to establish the existing condition and reaction to movements caused by background effects such as temperature, season, tide etc.

5.

Such a pre-construction study will also assist in determining the type of monitoring to install during the works. There is a case for considering settlement monitoring on all buildings likely to be affected. Further specific evaluation should be considered for each building about the value of installing instrumentation such as electrolevels, tiltmeters, extensometers etc.

Ch 16 JLE construction works at London Bridge station

247

6. Make forecasts of the building’s response to the staged nature of the construction, not just the final predicted maximum movements. 7.

The range of potentially suitable protective measures should be assessed on the basis of information of the kind listed above. Thus, options other than compensation grouting might be more viable or economic overall.

8. When considering compensation grouting as an option, take account of the relatively high establishment cost. Also, bear in mind that the technique may induce additional (and to a large extent unknown) stresses in both building foundations and tunnel linings, that the nature of the ground in the area of injection will alter and subsequently be consolidated, and that the grouting has to be carehlly targeted to avoid deleterious effects on adjacent structures. 9.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

16.12

Underpinning may be the appropriate option, especially if the building rests upon shallow and insubstantial foundations. Such a choice will be based upon the building’s existing condition, age and degree of importance, in addition to commercial factors.

CONCLUDING REMARKS Experience at the JLE works at London Bridge station has shown again that total settlement is only one aspect of the wider issue of building response to ground movements. Control of settlement will not itself nullify concern about the effects on a building. Adequate consideration should be given to evaluating each building individually to determine the movements that it can accommodate rather than attempting to severely restrict or prevent movements. While this study can first be made before work begins on site, the evaluation should be continually reviewed in the light of both visual inspection and monitoring. It is essential to be able to draw upon sufficient staff resources of knowledgeable people able to assimilate and react to this information. A most important aspect of managing excavation work that affects buildings is to keep the building owners and occupiers informed and to retain their co-operation. Regardless of the protective measures chosen, the first and overriding priority is to ensure that the tunnels are built robustly.

16.13

REFERENCES BS 5930: 1981 Code ofpractice for Site Investigation, British Standards Institution, London BS 81 10: Structural use of concrete Part 1 1997, Code ofpractice for design and construction Part 2: 1985, Code ofpractice for special British Standards Institution, London BUILDING RESEARCH ESTABLISHMENT (1 990) Assessment of damage in low-rise buildings with particular reverence to progressive foundation movements, Digest 25 1, Building Research Establishment, Garston

RANKIN, W. J. (1 988) Ground movements resulting from urban tunnelling: predictions and effects, Engineering Geology of Underground Movements, Spec Publn No 5, Geological Society, London, pp 79-92.

248

Building response to tunnelling

Some aspects of construction on JLEP Contracts 105 and 106

17

A D Withers

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

17.1

SUMMARY This chapter outlines some of the methods and events in the construction works of JLEP Contracts 105 and 106 between Old Jamaica Road,Bermondsey, and Canada Water station, Rotherhithe. The account is restricted to aspects relevant to the case studies of the buildings and the instnunented reference sites along this part of the JLE route. From the point of view of the research, one of the distinguishingfeatures is that the tunnelling was through the beds of the Lambeth Group and the Thanet Sands. Thus this chapter sets the scene for Chapters 37 to 46 in Volume 2. It gives brief descriptions of the tunnelling and excavations for the Bermondsey and Canada Water station boxes, the associated shaft constructions and protective measures, and groundwater control. In the part of the route west of the Bermondsey station box, some irregular surface depressions formed unexpectedly above the tunnels. These are described. Southwark

London

BygeBermondsey

I

Westminster

I

t

I

Canada Water

m [own

,

I

1 Contract

I

Contract 105

I

8

Contract 104 Green Watertoo London Park Bridge Westminster Southwark

I*

n

.

-_

106 Canada Water

-

1 km

0 0.5 115/108

NorUl Canafy Wharf Greenwich

Bermondseey

Canning ToWn

A

-

&

%

Contract 102 Contrad 103 contract 104

Contract 105

Contract 107 Contract Contract 108 115/108

Contract 110

Contract 111

Key London Clay

Lambeth Group

Figure 17.1

Thanet Sands

Chalk

Location of reseamh area where tunnelling was in the Lambeth Group and Thanet Sands

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

249

INTRODUCTION

17.2

Several of the case studies in Volume 2 (Chapters 37 to 46) deal with ground and structural displacements caused by deep excavations (ie tunnels, shafts and box excavations) in the beds of the Lambeth Group and Thanet Sands. This chapter therefore describes some the aspects of the construction works relevant to those case studies. The length of the JLE route to which the chapter applies was constructed under JLE Contracts 105 (Old Jamaica Road to Canada Water) and 106 (Canada Water station), whose locations are shown on the JLEP route plan in Figure 17.1. The sketch plans Figures 17.2 (a) and (b) show the part of the route from the Old Jamaica Road temporary access shafts (where the tunnelling began on Contract 105) to Canada Water station. They also indicate the locations of the instrumented Southwark Park greenfield site, two other surface settlement reference sites, and the buildings studied by the research team. Note that the construction works to the west of the Old Jamaica Road shafts (part of Contract l05), and their effect on the overlying ground and structures, was not part of this research.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

N 1000 m

n

Ben Smith Way shaft V

N 0 (b)

I

1000 m w 182-210 Jamaica Road

Blick House

128-130 Jamaica Road

Neptune#Clegg and Murdoch Houses, Moodkee Street

Old Jamaica Road reference site station

Figure 17.2

250

Southwark Park reference site

Columbia and Regina Points

Locations of (a) shafts and stations of JLE Contracts 105 and 106 and (b) the study buildings and reference sites between Old Jamaica Road and Canada Water

Building response to tunnelling

17.3

JLE CONTRACTS Contract 105 of the Jubilee Line Extension Project covered constructionof Bermondsey station, the 2.8 km of twinrunning tunnels between London Bridge and Canada Water, and i n t e d t e ventilation and escape shafts. The contract was awarded to the AokiSoletanchejoint venture. Aoki &ed out the tunnelling and shaft excavations while Soletanche was responsible for diaphragm wall@, protective measures and monitoring. The umstruction of Bermondsey station was subcontractedto O'Rourke and Son. Canada Water station was constructed under JLE Contract 106 by Wimpey ConstructiOn, which was taken over by Tarmac Construction during the works.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

17.4

GROUND CONDITIONS The longitudinal section of Figure 17.3 Summarises the geology along the tunnel route of Contract 105 and at Canada Water. At the positions of the buildings and reference sites, the tunnels were &ly excavated through the layered beds of sands, silts, clays and gravels of the Lambeth Group. The overlying materials that were excavated for the shafts and stations were made ground, recent alluvium and Terrace Gravels. Chapter 5 describes the geology and geotechnicalcharacteristicsof all the materials encountered in the works. Each of the chapters dealing with specific buildings and reference sites includes information about the local ground conditions.

LONDON BRIDGE

BERMONMM

.

Figure 17.3

CANADA WAT€R

I

Geobgyof Contracts 705 and 7 0 6

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

251

17.5

RUNNING TUNNELS

17.5.1

General

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The contractor completed the tunnelling using Kawasaki earth-pressure-balance tunnelling machines (EPBMs), one of which is shown in Figure 17.4 being lowered into the Old Jamaica Road shaft. Three types of spoil conditioning agent were used, starting with bentonite muds, then drilling polymers, but changing to foams, which proved better for cutting, extruding and conveying the cut materials, particularly the clays, of the Lambeth Group. Figure 17.5 shows a simplified long-section through a machine, which is 6.925 m long with a cut diameter of 5.03 m. The running tunnels in the main were lined with 4.4 m ID (4.9 m OD) pre-cast segmental concrete rings, the segment width (or length along the tunnel) being 1.2 m. Spheroidal graphite iron (SGI) segments were used in areas where there were connectionsbetween a tunnel and a shaft and for the station platform enlargements.The SGI rings are 0.6 m wide. The segments of each ring were bolted together within the tail-skin of the machine such that the leading edge of each ring, when erected, was about 5.1 m behind the face of the machine. The annulus between ring and soil was usually grouted after two further advances of the machine, ie for concrete rings, Ring 1 was grouted after excavating for Ring 3, as it emerged from the tail-skin, as illustrated in Figure 17.5. The grout was injected through grout holes in the segments at pressures of about 1 bar above surroundingground water pressures. Secondary grouting of the annulus was carried out later, generally some 50 m behind the tunnel face, as the tunnel rings emerged from the rear of the EPBM back-up train.

-

Figure 17.4

252

Kawasaki EPBM being lowered into the Old Jamaica Road shail

Building response to tunnelling

-5 m

*

4

Grout behind rings

/

6.925 m

A .“.I.””.

,

I

concrete -segments 4.9 m 0.d. 4.4 m i.d.

5

\

Cutting4 bits

V

c c

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Figure 17.5

Screw convevor

Schematic section through the Kawasaki EPBM

During tunnelling, a range of machine performance parameters were automatically recorded by the EPBM instrumentation, including earth balance-pressure, thrust and torque, for each tunnel advance. Earth balance-pressures were recorded by gauges on the left- and right-hand sides of the rear of the plenum chamber. The ranges of these pressures applied during the tunnelling below the buildings are reported in the building case studies.

17.5.2

Tunnelling between Old Jamaica Road and Bermondsey station Tunnelling on Contract 105 started from two purpose-built 14 m ID, 30 m-deep temporary access shafts adjacent to Old Jamaica Road (see Figures 17.2 (a) and (b)). The first tunnel drives began in December 1994, heading eastwards on the 500 m section towards Bermondsey station. It was always intended that these short sections of tunnel should be completed first as they would provide information about the suitability of the earth-pressure-balance tunnelling technique in the Lambeth Group and show whether modifications were needed for the subsequent drives (Anon, 1996). In the first 100 m of the drive, clogging of the plenum and screw-conveyor with spoil occurred and some difficulty was experienced in maintaining the required alignment tolerance. As the drives neared the Bermondsey station box, progress increased to about 50 mlweek. The EPBMs reached the station box at the end of June 1995. For this entire section of tunnelling, the overall average advance rate was between two rings per day (eastbound) and three rings per day (westbound), ie between 2.4 and 3.6 mlday. (Note that periods of complete stoppage are included in these rates.) A 1.2 m-thick ‘‘soft” wall panel had been installed immediately outside the diaphragm wall of the Bermondsey station box for the EPBMs to dock into. This was to facilitate later connection with the tunnel eye that would be broken out of the station wall. The soft wall panels were designed to have strength equivalent to a firm consistency in soil mechanics terms (Dawson et al, 1996). When the EPBMs reached the station box, surface depressions formed above them, suggesting that more of the soft wall material had been excavated than intended. The loss of ground that occurred above the westbound drive on 28 June 1996 was described as a ‘‘large’’ depression. This sank further when the machine was advanced fully into the station box. Other surface depressions appeared above the tunnel centre-lines after passage of the EPBMs (see Figures 17.6 and 17.7), which are described in Section 17.1 1.

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

253

Eastbound running tunnel

0. 0-

c c

T

-

c

-

I

-

Old Jamaica Road shafts

100 m

Figure 17.6

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

c

I

Bermondsey station box

Westbound running tunnel

Locations of surface depressions between the Old Jamaica Road shafts and Bermondsey station (NB: the dots do not represent size)

When the EPBM of the westbound tunnel drive entered the free air of Bermondsey station box, the left-hand gauge in its plenum chamber was still recording about 0.4 bar face pressure. Similarly, on the eastbound machine reaching free air, its right-hand gauge recorded about 0.2 bar. The pressure gauge readings should have been zero. It is not known when these two gauges went out of calibration, but each recorded consistently higher pressures than its counterpart on the other side of the chamber. In the last 5 m of the drives in front of the walls of the Bermondsey station box, for example, the recorded gauge pressures were from 0.4 to 0.5 bar. If the gauges were over-reading by 0.2 or 0.4 bar, the actual plenum pressures would have been that much lower, giving little or no margin with respect to the overburden pressure.

17.5.3

Tunnelling between Bermondsey and Canada Water stations After maintenance and modification by the provision of additional cutting teeth, face pressure sensors and ports for injecting conditioning agent, the EPBMs were used to drive the tunnels eastwards from Bermondsey station to Canada Water station. The westbound tunnel was constructed first between November 1995 and March 1996 and the eastbound tunnel between May 1996 and August 1996. Tunnelling on these drives achieved overall advance rates of between six rings a day (westbound tunnel) and eight rings a day (eastbound tunnel), ie 7.2-9.6 &day. On reaching the pile wall of the Canada Water station box, the EPBMs were dismantled from inside the tunnel after grouting thoroughly around them. This included burning out sections that could not otherwise be dismantled. The outer skins of the machines were left in place and SGI rings of 4.6 m ID were constructed within them. The final stage of the process involved burning out small sections of the cutting head and casting an in-situ concrete lining joined to the station wall piles. There was no need for special wall or face support because the tunnel was within the Glauconitic Sand unit of the Lambeth Group, which had been dewatered to below invert level (see Section 17.10 for details). Along these drives there is no evidence of the face pressure sensors being out of calibration.

17.5.4

Running tunnels between Old Jamaica Road and London Bridge station The tunnel drives between Old Jamaica Road and London Bridge were driven between August 1995 and June 1996 (eastbound) and September 1995 and June 1996 (westbound). None of the structures in this part of the route were studied by the research team and the construction works are not reported in this book.

254

Building response to tunnelling

.._.

17.6

BERMONDSEY STATION The Bermondsey station complex comprises not only the box structure containing the platform concourse and access to the surface, but also, extending beyond the box, the platform tunnels (7.3 m OD, 7 m ID) and an escape shaft (in Ben Smith Way). The construction of these is described below.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

17.6.1

Bermondsey station box The general sectional arrangementof Bermondsey station box is shown in Figure 17.7. The outer walls consist of 1.2 m-thick, 32 mdeep diaphragm wall panels excavated using conventional grab equipment. Bearing piles were installed inside the box to support the internal columns. The box was constructed topdown. The first stage of excavation was down to upper landing level (99.6 m PD, ie to a depth of about 3 m). At this level, a floor slab and internal strutting of the box were cast. The subsequent stage of excavation was down to lower plant room level (91.7 m PD, ie about 11 m below ground level) for casting the lower internal strutting and another floor slab. The third stage of excavation was to base slab level (83.1 m PD, ie about 19.5 m below ground level). This was carried out in small bays to provide some propping to the walls prior to the entire slab being completed. The base slab was 3 m thick to prevent flotation.

Figure 17.7

Enlargement of the bored running tunnel to station tunnel at Bennondsey under compressed air

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

255

The excavation was carried out in more or less dry conditions. The granular beds within the Lambeth Group strata within these walls were dewatered by sump pumping and by drainage resulting from the vertical wells installed within the shaft, which extended into the Thanet Beds. The Thanet Beds were depressurised by pumping from wells situated outside the station walls, which both aided dewatering and prevented base failure of the excavation. More information about the measures used to reduce groundwater pressure is given in Section 17.10.

17.6.2

Bermondsey station platform tunnels (and cross-passage)

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The platform tunnels were formed by enlargement of the running tunnels. The enlargement was made by hand excavation under compressed air applied to exclude groundwater and support the excavation. Figure 17.7 is a view of the enlargement work in one tunnel. Compressed air working was needed because the tunnels are within the water-bearing Upper Mottled Clay and Laminated beds of the Lambeth Group. The eastbound station tunnel was constructed first using compressed air pressures ranging from 0.8 bar to 1.3 bar and averaging 1.02 bar. The work involved breaking up the 4.9 m OD running tunnel rings, starting on 6 October 1995 with three rings at around chainage 1900, ie about 15 m east of Ben Smith Way (see Figure 17.9). Initially, the station tunnel enlargement was driven eastwards towards the station box and then advanced in both directions. After completion of the tunnel to the station box, the crosspassage was constructed to within a few metres of the westbound running tunnel. structures In Major Road and John Roll Way 1200 mm slurry wall for tunnel eye 1200 mm diaphragm walls

p 99.6 m PD, upper landing level 6 m 0.d. running tunnel

-

-

-!\

96.1 upper plantroom level 91.7 lower plantroom level

--

- - - - - - 7.3 m 0.d. platformtunnel

;

88.1 p1atfo.m level

BERMONDSEY STATION BOX 1-72.7

Figure 17.8

Cro s-section through Bermondsey s stion box a its junction with the westbound tunnel (looking northwards)

After moving the compressed air set-up, construction started on the westbound station tunnel and the final section of the cross-passage using the same techniques as for the eastbound tunnel. Compressed air pressures ranged from 0.85 to 1.35 bar and averaged about 1.05 bar. The construction involved breaking up the 4.9 m OD running tunnel rings starting from around chainage 1897, under the junction of Ben Smith Way and John Roll Way (see Figure 17.9). The station tunnel was initially driven eastwards towards the station box and then, as with the eastbound tunnel, advanced in both directions. The compressed air was turned off after completion of the station tunnel.

256

Building response to tunnelling

break-up for eastbound station platform tunnel 27/9/95 to 5/10/95

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

break-up for westbound station platform tunnel 21/4/96 to 26/4/96

Figure 17.9

17.7

A

--

25 m

0

Plan showing construction sequence of station platform tunnels beneath Keetons Estate

CANADA WATER STATION The different parts of the Canada Water station box, the excavation levels within it, and other construction features, are shown in plan in Figure 17.10.

Props a t 100 m

Slurry cut-off wall

Temporary Contiguous Pile Wall

Permanont Contiguous Pile Wall

Figure 17.10

Plan of Canada Water station

The plan of the station box is like a skewed cross. The JLE forms the long, east-west, axis and this cuts across, and about 10 m below, the pre-existing East London Line (ELL), which forms the arms of the cross. The perimeter of the station box consists mainly of contiguous pile walls. To the west of the ELL, these walls are part of the permanent support walls of the box. To the east, and beneath the ELL, the pile walls formed temporary supports and were propped against

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

257

each other, the main structure was then built in a “bottom-up” fashion within these walls (described below). Instead of piles in the perimeter wall north-east of the ELL axis, a slurry wall was installed around the area to reduce groundwater flow into the excavation for the base slab on this northern part of the ELL axis. Here, the excavated slope was ramped down from ground level to give easy access for plant into the excavation.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Excavation along the ELL axis was to a level of about 92 m PD. Within the plan area of the station box, the existing ELL tunnel was exposed and demolished for construction of new platforms and passenger access routes. The ELL excavation was propped at a level of about 101 m PD in the north-west corner of the box and on the entire south side of the JLE axis. In addition, ground anchors were installed around much of this section of the box to reduce wall and ground movements (Jackson and Twine, 1997). The anchors were installed in the broad areas indicated on Figure 17.10. Most of the structure on the JLE axis was built from the base slab upwards, ie from an excavated formation level of about 83 m PD. As the excavation deepened, props were installed between the pile walls at levels of about 96 m and 88 m PD. The base slab was cast against the pile walls forming a third level of internal bracing. The final box structure was built by casting, in situ, the internal slabs and walls (including the intersection between the JLE and ELL axes) working up from the base level, removing propping as required. An exception was part of the west end of the box, as shown in Figure 17.10, where a top-down technique was used. The roof slab in this area was cast first at about elevation 100 m PD; thereafter the excavation was made below this level with propping at levels of 94 m, 91 m and 87 m PD to minimise movements.

I7.8

SHAFT CONSTRUCTION There are two combined ventilatiodescape shaft complexes between Old Jamaica Road and Canada Water; these are shown on Figures 17.2(a) and 17.3. The Ben Smith Way shaft is part of the Bermondsey station complex; the Culling Road shaft is situated just to the east of Southwark Park. In addition, two temporary access shafts (one for each running tunnel) were sunk at the Old Jamaica Road work site to start the tunnelling works and allow material movements into and out of the tunnels. It was from these shafts that the running tunnels were driven first towards Bermondsey station box and then towards London Bridge. The diaphragm walls of the shafts at Old Jamaica Road (45 m deep) and Culling Road (14 m ID and 25 m deep) were constructed in slurry trenches excavated using the reverse-circulation Hydrofraise method. The shaft at Ben Smith Way (total depth of 18 m) was constructed within a sheet-piled cofferdam. The sheet piles, which had a water-jetting facility to assist penetration, were driven through pre-augered material by a silent piling system. All the shafts required some form of groundwater control measure. In most cases, this involved a cutoff through the higher water-bearing layers and pumping from the lower aquifer. These measures are described in Section 17.10. The case study buildings were sufficiently distant not to have been affected by construction of the shafts, although they are likely to have been in the zones of influence of the groundwater control measures.

258

Building response to tunnelling

17.9

PROTECTIVE MEASURES Prior to the construction work, settlement analyses and potential damage assessments were made on all the buildings within the theoretical zone of influence of the tunnel and station excavations (as explained in Chapter 3). Those assessments identified a number of buildings potentially at risk of being in a “moderate” or worse category of damage as defined by BRE Digest 25 1 (BRE, 1990). These buildings were designated, under the contract, as requiring mandatory protective grouting to make sure that any damage during the construction works would be no worse than the “slight” category. Figure 17.1 1 shows the buildings requiring mandatory ground treatment, which were assigned allowable limits of absolute settlement, angular distortion or tilt. N 1000 m w

0

I

Kirby Estate 138 Jamaica Road and 2 Maior Road Keetons I

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

\ \

I Canada station

1

Broomfielld Court

BermondseY station

Chalfont House 1-14 Perryn Road The Surgery and Prestwood House

Niagara court

Buildings requiring mandatory ground treatment along the route of Contract 705

Figure 17.11

The contractor installed several shallow grout shafts near the designated buildings from which to drill and place tubes-ci-manchette (TAMs) beneath the buildings. Two of these buildings were monitored as part of the research: Keetons Estate (Chapter 39) and Niagara Court (Chapter 44).

Grouted zone below building footprint

Figure 17.12

0

.....

London Clay

or Lambeth Group

Station ..... tunnel

Example of the scheme of grouting below a building

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

259

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The contractor’s grouting scheme generally had two distinct phases. The first, using the contractor’s term, was “ground stabilisation”. This involved two stages of injections into the Terrace Gravel. The initial injections were of a bentonite-cement grout placed to form a containing boundary to the treatment zone. The second stage was permeation of the treatment zone using a microfine cement grout, which is sufficiently fine-grained to penetrate pore spaces in gravels and coarser soils, ie permeation grouting. The aim was to create a “semi-rigid grout slab” beneath the plan area of the buildings at a level just above the contact between the Terrace Gravels and the underlying London Clay or Lambeth Group formations. The grouted zone of 1.5-2 m thickness is shown diagrammatically on Figure 17.12. The grouting was primarily intended to even out settlements, but also to stabilise the Terrace Gravels prior to compressed air excavation in the Lambeth Group. If, after settlement of a building had taken place (whether due to the tunnelling or any other reason), it was necessary to bring part or all of the building back within allowable limits of absolute settlement, angular distortion or tilt, the second phase of grouting was employed. Although called “compensation grouting” by the contractor, it is more accurately termed corrective jacking (see Chapter 1 1). A grout paste or mortar was injected below the semi-rigid slab (ie generally into clay by fracture grouting) to cause controlled heave of the building. Following passage of the running tunnels and the appearance of surface depressions at Keetons Estate, there was a substantial period of corrective jacking to reduce the angular distortion. This is discussed further in Chapter 39. For Niagara Court, only the first phase of grouting was carried out, ie ground stabilisation; this was shortly before the passage of the first running tunnel. No further or corrective grouting was needed (see Chapter 44). At Chalfont House, one of the other buildings for which mandatory treatment was required (see Figure 17.10), the contractor followed a grouting scheme similar to the initial phase at Keetons Estate. The work at this building was reported by Linney and Friedman (1 996).

17.10

GROUNDWATER CONTROL FOR SHAFT AND STATION CONSTRUCTION

17.10.1

Genera I In the area of Contracts 105 and 106, and as described in Chapter 5, there are two main aquifers. 1. The lower aquifer comprises the Upper Chalk and Thanet Beds together here with the lowest units of the Lambeth Group (ie Pebble Bed and Glauconitic Sands). 2.

The upper aquifer comprises the Terrace Gravels and Alluvium.

In addition, there are permeable water-bearing horizons contained between clay layers in the upper part of the Lambeth Group (although these units vary in thickness and are not always present). The shaft and station excavations in the study area all extend down through the upper aquifer and either reached the lower aquifer or encountered the permeable strata of the Lambeth Group. Each excavation, therefore, needed some form of groundwater control. This was achieved by a variety of methods. Where possible, the higher water bodies

260

Building response to tunnelling

were cut off by the installation of diaphragm, slurry or pile walls. Control of the lower aquifer was required to prevent ingress of groundwater (where the excavation bottomed out in the aquifer) and base heave (where the base of the excavation was below the lower aquifer) before sealing the base of the excavation by casting the base slab. There were dewatering schemes in operation at all work-sites between Druid Street (ventilation and escape shafts) and North Greenwich station. These sites are shown in plan on Figure 17.13. Durands Wharf

Pioneer

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Road shafts

l Figure 17.13

17.10.2

1 km

Locations where there was dewatering of the lower aquifer

Lower aquifer Reduction of pore pressures in the lower aquifer was achieved by pumping from the Thanet Beds andor the Chalk. The main purpose of the dewateringldepressurisation measures was safety of the excavations. Typical installations are shown in Figure 17.14. Over time, because of the continuity and permeability of the materials of the lower aquifer, the dewatering led to very widespread drawdown. The resulting increase in effective stresses caused settlement over a similar area. These effects are described in more detail in Section 17.10.4. Figure 17.15 represents the groundwater levels in the lower aquifer over the length of the JLE route between the Druid Street shafts and Canada Water station at three key times during the tunnelling works. These times were: June 1995, when the running tunnels were approaching the Bermondsey station box from the west (Curve A) 0

January 1996, towards the completion of the westbound tunnel drive from Bermondsey to Canada Water (Curve B) July 1996, towards the completion of the eastbound tunnel drive from Bermondsey to Canada Water (Curve C).

On Figure 17.15 there are four lines of water levels (including the initial conditions, Curve 0) plotted against chainage to a hugely distorted scale. The vertical alignment of the tunnels is shown by their crown and invert levels. The top surface of the lower aquifer is also shown (Curve D). Although the drawdown was extensive and varied from about 8 m to 25 m over the 2 km of the route, there was little change in level at any particular chainage over the period of the tunnelling works. Curves B and C were almost

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

261

the same and little more than 1 m below the level of Curve A. From a point about halfway between Bermondsey station and the Culling Road shaft, the tunnels driven eastwards towards Canada Water were wholly within lower aquifer in the dewatered lowest units of the Lambeth Group.

17.10.3

Intermediate water-bearing strata

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

During excavation of the Ben Smith Way escape shaft, in addition to dewatering/ depressurisation of the lower aquifer, water was drained from confined water-bearing beds of the Lambeth Group. There were three possible causes of this drainage. 1

The filter zone of the wells installed within the shaft extended up from the lower aquifer to these higher water-bearing beds, so the wells also drained them.

2

There was seepage through the clutches of the sheet pile wall into the excavation (which was then pumped away from sumps).

3

There may have been downwards drainage through the pre-augered (and subsequently loosened) material around the pile walls from the upper to the lower aquifer (which was at a lower piezometric pressure). (b)

(a)

>cKuK?I

$1

,

MadeGround

boo(xx1

Alluvium and Terrace Gravels

pzozozozol

.............. .. . ... . .... .... .... . ... . .. .. .. .. .. .. .. ..

(I

I>:

Filter pack c

. ... . .. ....... .. . . .... .... . .. . .... . .. . . .

0.5 to 1 mrr

Figure 17.14

262

.;; .;, .;, .;, .!#

.:; .;; .', .', .'. .. . ... . .. . .. . ... ... . .. . .. . ... . . . ... ... ...................... . .. .. . . . . . . .

... ... ... . . . . . . .

Typical well installations (a) to dewater the Thanet Beds directly, and (b) to underdrain Thanet Beds by pumping from the Chalk

Building response to tunnelling

Curves A, B and C

65

T.”

Vertical alignment of

I

4

T

Druid Street shafts

Millstream Road

55 J 4900

I

I

IBermondsey Station

Old Jamaica Road shafts

5900

5400

Culling Road shaft

-II Canadal Water Station

6400

6900

Project chainage (m)

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

-

- - + - -Jun-95

-original

_ _ _ _ _ -top of lower aquifer

- -A- - Jul-96

- - Jan-96

-tunnel invert

tunnel crown

Figure 17.15

Groundwater levels in the lower aquifer at various times from Druid Street to Canada Water Intermediate water-bearing strata

/ curve 0

4

I

55 4900

5400

6400

5900

6900

Project chainage (m)

- +( - original - June 1995 - - - top of intermediate water-beraring strata

---* -B

tunnel invert

Figure 17.16

4

-

A-May 1995

4-C-Jan

1996

- - - base of intermediatewater-bearing strata tunnel crown

Groundwater levels in the intermediate water-bearing strata at various times from Druid Street to Canada Water

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

263

These combined effects reduced the piezometric pressures in these confined permeable beds near the Ben Smith Way shaft and the Bermondsey station complex generally. Figure 17.16 shows the groundwater levels in these strata over the same length of route as Figure 17.15. Again, the initial levels are shown as Curve 0; the tunnel’s crown and invert levels indicate its vertical alignment. There are water levels shown for three dates: May and June 1995 and January 1996, Curves A, B and C, respectively. On this plot, the depth range (ie top and bottom) of the permeable intermediate beds in the Lambeth Group are shown as columns where known. As the internal concrete of the shaft was completed in May 1995, the drainage from these beds effectively ceased (so Curve A shows the lowest pressures). The groundwater pressures were recovering as the tunnels were driven the final 100 m of the route from Old Jamaica Road to Bermondsey station. The effect of this is further discussed in Section 17.10.5.

I

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

17.10.4

Effect of dewatering/depressurisation of the lower aquifer The depressurisation of the lower aquifer at the work sites shown in Figure 17.13 created individual cones of drawdown of more than 2 km in diameter. As most of the schemes were operating at the same time, these drawdown cones coalesced to reduce the groundwater level within the lower aquifer over a large area of east London. This drawdown caused settlement at the surface over a similar area. Most of the settlement would have occurred in the lower units of the Lambeth Group and the Thanet Beds where stress changes were greatest. Figure 17.15 shows the difference between the original and the reduced groundwater levels recorded by piezometers during the works along part of the route. Between Old Jamaica Road and Canada Water station, the drawdown was between 8 m and 15 m. There would have been a similar amount of drawdown laterally away from the lines of the tunnels, which would have affected the various local benchmarks used for the precise levelling - probably to almost the same extent as the settlement of the buildings. For this reason and because changes were gradual, it is assumed that the dewatering while affecting absolute levels - had little net effect on the results of the precise levelling surveys at least over the short to medium term, ie during the tunnelling operations and probably for several months to years thereafter. The situation may not be so straightforward in the long term (more than four or five years, say) after the broad recovery of water levels in the deep aquifer and the establishment of local drainage towards the tunnels, shafts and stations.

17.10.5

Effect of dewatering/depressurisation of intermediate waterbearing strata below Keetons Estate Local dewatering during excavation of the Ben Smith Way shaft near Keetons Estate (see Figure 17.2) brought about a reduction in pore pressure of as much as 35 H a . This reduction was measured in piezometer TWPZO1, which was installed to measure pore pressure within one of the confined water-bearing beds of the Lambeth Group. It was caused by drainage towards the shaft. After May 1995, when the shaft dewatering was not needed, the pore pressures started to recover, but their return to pre-existing values was interrupted and reversed by the dewatering measures associated with breaking into the station box. The pore pressures were also affected by the application of compressed air during station tunnel construction. Figure 17.17 represents the changes in piezometric pressure in the intermediate permeable beds that took place over three and a half years from the end of September 1994. On the same plot is the change in level of monitoring point 2001 at Keetons Estate. This monitoring point (shown on the location plan Figure 17.9) is one of the

264

Building response to tunnelling

farthest on these buildings from the tunnels and least affected by tunnel construction or grouting. The time plots of settlement and pore pressure change follow a broadly similar pattern over the period. The recovery of about 35 kPa of piezometric pressure between May 1995 and April 1998 seems to have been accompanied by about 5 mm heave of the monitoring point (although there would also have been movements associated with the tunnelling and subsequent grouting). As the settlement monitoring only started after the water pressure reduction of about 35 kPa had occurred, this movement probably represents the monitoring point regaining its pre-dewatering level. As the reduction of pore pressure in the confined permeable strata was relatively small, it is assumed that there would probably not have been any significant effect on the local benchmark, some 75 m south of Bermondsey station. Change in piezornetric pressure around Keetons Estate Running tunnel construction

Eastbound station tunnel

Westbound station tunnel 6

20

5

10

4

m

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

$

0

e

3

v)

z E

--

e!

.-5 -10

0

*

2:

I

.-

m

.-I-20

W

n

1 2

E .-

m

W

5

-30

0

S

-40

-1

-50

-2 198

231 Date possibly 1 mm settlement due to running tunnel construction +TWPZO~( E ~ . ~ ~ - L S +monitoring C) point 2001

1

Figure 17.17

1

Changes over time in piezometric pressures in the confined permeable strata near Keetons Estate and changes in the elevation of monitoring point 2001 {Keetons Estate) over the same period

17.11

LOCALISED SUBSIDENCES, THEIR LOCATIONS AND POSSIBLE CAUSES

17.1I.I

Locations The appearance of localised surface depressions (subsidences) above the centre-line of the running tunnels between Old Jamaica Road and Bermondsey station is mentioned in Section 17.4. The locations of the 13 depressions that were recorded are shown in plan on Figure 17.5. Other than no 13, their numbering is based on their geographic position rather than their sequence of occurrence. Thus, 1 to 5 were along the route of the eastbound running tunnel, 1 being the most westerly, and 6 to 12 along the route of the westbound tunnel, with 6 being the most westerly of that group. No 13 was over the eastbound tunnel line, but three years later.

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

265

While these subsidence features had little effect on the case study structures, their possible causes and proximity to the structures, and the remedial measures adopted, are relevant to the patterns of deformation recorded in this area. The following sections describe these surface depressions and include discussion about their possible causes.

17.11.2

Descriptions of some of the subsidence features The subsidence features which are described are those that appeared close to the Keetons Estate buildings (see case study Chapter 39), the descriptions of the features being derived from JLE records.

Feature 5

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

This feature was in the rear garden of 2 Major Road and appeared on 14 February 1996 as a depression about 2 m deep and 4 m3 in volume. At the time, the eastbound station tunnel was being excavated under compressed air. Although the east face of this tunnel widening was fairly close in plan below the depression, there does not appear to have been any run of ground at the face or drop in air pressure in the tunnel. The depression was backfilled later the same day with sandy gravel and covered with topsoil.

Feature 9 This was a shallow collapse of a roadway around a telecommunications manhole at the junction of Ben Smith Way and John Roll Way on 20 July 1995 after a heavy goods vehicle crossed over it. The investigation revealed that there was a thin layer (only 75 mm) of reinforced concrete around the manhole forming the road pavement. This was bridging over a void about 150 mm deep. It is not clear if the void formation can be attributed to the tunnelling or, as seems at least as likely, pre-existed as a consequence of the manhole construction.

Feature 10 This was a depression of the pavement in John Roll Way next to the buildings of Keetons Estate, which occurred during the week ending Friday 30 June 1995. It left an elongated shallow trough about 40 mm deep, with its long axis above the centre-line of the tunnel. The paving stones became uneven, but there was no discernible effect on the John Roll Way faqade of the building. There were no voids when the paving slabs were lifted. The area was levelled and repaved by early August 1995.

Features 11/12 The first of these ground collapses was recorded on 28 June 1995 at 1 1:30 am. It was noticed because a subsurface settlement pin (about 0.5 m-long set originally with its top just below the pavement) could not be found. There was also differential settlement of the paving around a telecommunications manhole chamber, with some parts of the manhole proud of the pavement while others had subsided. The zone of ground below the subsided surface was immediately grouted using about 8-9 m3 of grout. It is not clear from the records how this was done. Later that day the footpath collapsed, the hole straight away being backfilled with about 2 m3 of pea gravel. There was some further settlement here on 3 July, after which the pavement was repaired. It was reinstated by 6 July. On the evening of 2 1 July the area subsided again. Figure 17. I8 is a photograph to show the large voids exposed when the paving stones were lifted. More than 8 m3 of grout and

266

Building response to tunnelling

an unknown quantity of ballast were used to fill the hole on 2 1 and 22 July, but more subsidence was reported on 25 July. Remedial measures following these events included constructing a decking walkway over the subsided area.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The appearance of this depression coincided with a loss of ground into the station box on 2 1 July at 17.30 pm as the EPBM entered the station box and the seal on the machine blew. No estimate of the volume of the in-flowing material was made. At a later stage, grouting of the area was carried out from surface level. This involved the injection of about 71 m3 of grout. Both this second phase of subsidence and its associated remedial measures had an effect on the south-east corner of Keetons Estate as described in Chapter 39.

Figure 17.18

Subsidence features 11/12 when pavement slabs lifted

Feature 13 This void appeared over the weekend of 22 and 23 May 1999, ie more than four years after construction of the running tunnel. At ground level, the opening was about 300 mm across, but the hole was much wider below. After breaking out the surface, water was hosed into the hole to wash down loose material. The hole was backfilled with about 6 m3 of sand, which was compacted before replacing the surfacing.

17.11.3

Possible mechanisms There are several possible reasons for the formation of these subsidence features. These are discussed below as mechanisms that could have caused or contributed to their formation. It appears that it was mainly in this length of the JLE route that such features appeared. One of the contributory factors may be the nature of the ground, therefore. Others considered below are EPBM face pressures, the soft wall at the station box, and ground loss into the station.

Ground conditions The ground conditions in this section of the JLE route may have led to a greater likelihood of occurrence of subsidence features for the following reasons.

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

267

The near-surface materials in this area, particularly the made ground, may have been locally very loose, eg debris from the demolition of older buildings (much of the housing being new or relatively recent) or disturbed as a result of Second World War bomb damage. The granular soils - both the Terrace Gravels and the sands and gravels of the Lambeth Group - are more likely than clays to run and ravel. This would encourage the migration of voids, even if not a direct flow into the face of the tunnel. Clay layers within the Lambeth Group would tend to form bridges across voids, but the thinner ones would probably be of relatively short span and unlikely to be long-lasting.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

EPBM face pressure The different readings of the pressure gauges in the plenum chambers of the EPBMs on the running tunnel drives between Old Jamaica Road and Bermondsey station are mentioned in Section 17.5.1. On the basis that the gauges consistently over-recorded by the amount of their reading in free air afterwards (0.4 and 0.2 bar, rather than zero), the face pressures recorded for each running tunnel excavation have been corrected by these amounts. In Figure 17.19, a rolling average of these corrected pressures for each drive has been plotted against distance from the Ben Smith Way faqade of Keetons Estate. The rolling average is over five rings, ie the ring when the face was immediately below the specific (plotted) position and the previous four rings. This averaging was done to smooth out the extremes of the pressure fluctuations. It can be seen that in the final 5 m of both tunnel drives - where subsidence features 5, 1 1 and 12 occurred near the station box wall - the face pressures were below 0.6 bar, ie less than 60 kPa. In June 1995 (Curve B in Figure 17.16), the pressure head of groundwater was typically 9 m and locally 1 1 m at the tunnel invert, ie 90-1 10 H a . This is rather greater than most of the averaged face pressures over the lengths where subsidences 8 , 9 and 10 occurred.

1.;

1

- . m

P

0.8

v)

8

2 0.6 S

5 0 W

$ 4

0.4

0.2

0

-20

-10

Figure 17.19

268

0

10 20 30 40 Distance from Ben Smith Way facade (m)

50

60

70

Rolling averages of corrected EPBM face pressures as the tunnelling machines advanced towards the Bermondsey station box

Building response to tunnelling

Soft wall panels

As described in Section 17.5.1, soft wall panels had been constructed just outside the station box’s diaphragm wall for docking the EPBM and to help make the connection of the tunnel to the station box. The materials actually encountered as both tunnels met these soft panels were slurries (rather than a firm clay-like material) into which it proved difficult to advance without over-excavation. As an example, on 28 June 1995 in the westbound tunnel the machine advanced only 200 mm in three hours of excavation. The spoil, largely silt and water, was of a consistency that could not be drawn up the conveyor. This, and similar experiences on the eastbound tunnel, may have led to the EPBMs drawing in large volumes of material, creating the potential for ground loss at the surface.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Ground loss into station box There were several occasions of ground loss into the station box. The first occurred on 27 June 1995 when probing from the station box towards the westbound tunnelling machine. The probe holes not being fitted with valve heads, water and silt flowed back along them into the station box. One estimate was that about 1 m3 came in. On 28 June, after probing towards the eastbound tunnel boring machine, there was an inflow of about 3 m3 of water and silt. Probing again towards the westbound tunnel on 29 June caused a further ground loss of about 2 m3. Others estimated the inflow volumes to be greater, at about 6 m3 from the eastbound tunnel probing and 5 m3 from the westbound. During the entry of the westbound EPBM into the station box, there were substantial ground losses around the tunnel eye on both 2 1 July (estimated as 15 m3)and 25 July 1995 (some 2 m3). It is thought that the machine was slightly too high for the tunnel eye and the seal had to be loosened to allow the passage of the EPBM.

17.12

SURFACE REFERENCE SITES One of the aims of the research was to examine the patterns of settlement associated with tunnelling using an EPBM in the Lambeth Group. To do this for greenfield situations or at least where there were no buildings would help to put the structural deformations measured on buildings into context. These sites could be considered as controls. One section, in Southwark Park, was fully instrumented on the surface and in the subsurface. This can be taken to be a greenfield site. Two other sections were monitored only by precise levelling of surface points. One at Old Jamaica Road was across paved and concreted areas, and does not, therefore, represent greenfield conditions. The other monitored surface section was in the Canada Estate, Rotherhithe. While this was a grassed area, there had been previous development of the site and it is unclear what the made ground below the site contained. These instmmented sections are termed surface reference sites. Their locations are shown in plan on Figure 17.2(b). The site at Old Jamaica Road to the west of Bermondsey station box is above the first section of tunnel driven on the contract, ie in the operational shakedown stage. The contractor’s staff did the precise levelling of the line of surface monitoring points here. The thoroughly instrumented Southwark Park site, which was monitored by the research team, is just to the west of the Culling Road shaft. The third site was in the public area between the three Moodkee Street buildings (Chapter 43) and Niagara Court (Chapter 44). A cracked masonry wall and a line of settlement pins in the grassed area were also monitored by the research team.

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

269

These reference sites and some of the data derived from them are h r t h e r described in Chapter 37.

17.13

BUILDINGS STUDIED The buildings along this section of the JLE that were the subjects of the research are shown in Figure 17.2(b) and their names, structural form and the sources of monitoring

data are listed in Table 17.1. Chapters 39 to 46 are case histories with descriptions of the buildings. They present the results of the monitoring in terms of building settlements and strains and relate these to the JLEP works that affected them.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Table 17.1

List of the buildings monitored in the Contract 105 part of the JLE route

Name of building(s)

Structural form

Keetons Estate

Load-bearing masonry walls

128-130 Jamaica Road

Foundations

Measurement methods and source

Precise levelling

Precise taping

Demec gauge

Electrolevel

Cand R

R

R

-

Strip foundations about The outer walls are loadbearing brickwork. In 0.6-0.92 m below addition, there is a loadbasement level bearing spine wall consisting of brickwork and timber stud walling at different levels

R

R

R

-

1 82-2 10 Jamaica Road

Two-storey concrete frame construction and brick cladding

R

R

R

-

Blick House

The building has five storeys, Not known, probably load-bearing brick with piled concrete floors

R

R

-

-

Murdoch, Neptune and Clegg Houses in Moodkee Street

Brick construction

Corbelled brickwork above mass concrete strips

R

R

R

-

Niagara Court

Four-storey load-bearing brickwork

Short piles bearing on Terrace Gravels

R

R

R

-

Regina and Columbia Points, Canada Estate

Reinforced concrete frame construction with internal partitions and external walls of brickhlock construction

Concrete raft

R

-

R

R

Canada Estate (single-storey buildings)

Load-bearing brickwork

Strip footings about 0.8 m deep

R

-

R

-

Mass concrete strip foundations approx 2.7 m deep

Assumed to be shallow piling, end bearing on to the surface of the Terrace Gravels

Source of reported measurement: R = Research; C = Contractor

270

Building response to tunnelling

17.14

REFERENCES ANON (1996) Tunnelling Machines, Contract 105, in: Jubilee Line Extension, underground construction, Supplement to World Tunnelling February 1996, pp 19-23 DAWSON, M P, DOUGLAS, A R, LINNEY, L F, FRIEDMAN, M and ABRAHAM, R ( 1996), Jubilee Line Extension, Bermondsey Station box: design modifications, instrumentation and monitoring. In: Underground Excavations in Soft Ground (R J Mair and R N Taylor, ed), Balkema, Rotterdam, pp 99-1 04

BOSCARDIN, M D and CORDING, E J (1 989J Building response to excavation induced settlement. Proc ASCE, Journal of Geotechnical Engineering, Vol 115, No 1, pp 1-21

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

BUILDING RESEARCH ESTABLISHMENT (1 98 I , revised 1990). Assessment of damage in low-rise buildings with particular reference to progressive foundation movement. Digest 25 1, BRE, Garston, Watford JACKSON, R W and TWINE, D ( 1 997) Temporary ground anchors at Canada Water Station. In: Ground Anchorages and Anchored Structures (G S Littlejohn, ed), Thomas Telford, London, pp 272-280 LINNEY, L F and FRIEDMAN, M (1996), Protection of buildings from tunnelling induced settlement using permeation grouting. in: Underground Excavations in Soft Ground (R J Mair and R N Taylor, ed), Balkema, Rotterdam, pp 399-404.

Ch 17 Some aspects of construction on JLEP Contracts 105 and 106

271

18

Measuring techniques and their accuracy

J R Standing, A D Withers and R J Nyren

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

18.1

INTRODUCTION The information presented in this book is based on field measurements carried out mainly by the research team and in some cases by JLEP surveyors or contractors. The measurements, which were made both at greenfield sites and on structures, were primarily to determine displacements and to monitor the changes in position of points in the ground and on the buildings. The research team used four main techniques: precise levelling, precise taping, faqade monitoring and Demec gauge measurements. These are described below. Additional notes describe electrolevel systems and the measurement of horizontal ground displacements with a micrometer stick, as used at the greenfield reference sites. Subsurface vertical displacement measurements with rod extensometers were made at the St James’s Park reference site and below Elizabeth House (see Chapter 30 for the specific details relating to that building). A summary of the monitoring of each research study building is presented in Table 4.1. In almost all cases the primary quantity being determined was vertical displacement (or change in level), which was measured by the precise levelling of a series of points at each of the locations. At some sites additional techniques were used to supplement the levelling data and provide further information about other modes of deformation. The comprehensive instrumentation that was installed at the two greenfield reference sites of St James’s Park and Southwark Park allowed deformations in three dimensions at surface and subsurface levels and changes in total and effective stresses to be monitored. The results from some of the near-surface greenfield site monitoring are presented in this book in Chapters 25 and 37. Further details on the monitoring techniques used and the results obtained are given by Nyren (1 998). The following sections contain a brief discussion on accuracy and some comments relating to monitoring in general. The main measuring techniques used by the research team to record building movements and deformations are then described along with their general accuracy and methods of processing and analysing the field data.

18.2

DEFINITIONS RELATING TO MEASUREMENT It is worth defining some of the terms used when reporting measurements, assessing instrument performance, and estimating the degrees of uncertainty in the methods. The following definitions are quoted from Dunnicliff (1 988). Resolution is the smallest division readable or measurable on the instrument. Interpolation by eye between divisions generally does not improve the resolution, as the estimation is subjective and operator-dependent.

Accuracy is the degree of correctness of a measurement, or the nearness of a measurement to the true quantity. The accuracy of a measurement depends on

Ch 18 Measuring techniques and their accuracy

Previous page is blank

273

the accuracy of each component of the monitoring system. It is generally evaluated during instrument calibration, where a known value and the measured value are compared. Accuracy is expressed as f x units, meaning that the measurement is within x units of the true value. Precision is the reproducibility and repeatability of measurement, or the nearness of each of a number of measurements to the arithmetic mean. Precision is also expressed as f x units, meaning that the measurement is within x units of the mean measured value and is often assessed statistically with a degree of confidence associated with the statistical distribution; a larger number of significant digits reflects a higher precision. Noise describes the random measurement variation caused by external factors. Excessive noise results in lack of precision and accuracy, and may conceal small real changes in the measured parameter.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Combinations of the above measurement uncertainties manifest themselves as measurement error, which is the deviation between the measured quantity and the true value (and is mathematically equal to accuracy). Errors are classified as fo1I0ws . Systematic errors result from improper calibration, changes in calibration, hysteresis and non-linearity, but are generally errors whose magnitude and sign can be determined or estimated. Where appropriate, corrections may be applied to measured quantities to improve the accuracy.

Gross errors are caused by carelessness and inexperience. They include - but are not limited to - misreading, booking errors, misnumbering of monitoring points, use of non-standard monitoring techniques, and incorrect use of instrumentation. All measurements are suspect until gross mistakes are identified and eliminated. Measurements should be checked immediately (usually by repeating the measurement) to spot and correct errors during the survey. This is not always possible, however, because of the transient nature of some types of monitoring, eg as tunnelling progresses. Conformance errors are the result of poor installation procedures and poor selection of instrumentation and/or survey layout, but are minimised by careful supervision of instrument installation. Environmental errors arise from the influence of factors such as heat, moisture, vibration, atmospheric conditions and lighting. Errors arising from temperature changes, for example, can sometimes be quantified and corrected, but many environmental errors are of unknown magnitude and only acknowledged in a qualitative manner. Observational errors result from different Observers using different monitoring techniques. Well-designed automatic monitoring systems have the potential to minimise observational error, as does the use of standard recording forms and marking flags for instrument locations. The monitoring data presented in this book were carefully obtained and, at the outset of any analysis with them, they have been further assessed to eliminate so far as possible the errors described above. This has had varying degrees of success, depending on the monitoring system, site conditions and instrument operators. Even after identifying and correcting errors, variation in the readings will remain as a result of random error, comprising the influence of the errors and uncertainties given above; this variation is representative of the precision of measurement.

274

Building response to tunnelling

It is difficult to ascribe specific accuracies to a monitoring system as these vary from site to site. Accuracy is therefore assessed individually for each case study building and discussed in context with the building behaviour. The accuracies for the different measuring systems given in Table 18.1 give an indication of the best accuracy that was achieved by the research team.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Table 18.1

18.3

Examples of the best accuracies achieved for the various monitoring techniques

Instrument type (monitoring method)

Building example

Resolution

Precision

Accuracy

Precise level (NA 3003)

Treasury Palace of Westminster

0.01 mm

0.1 mm

f0.2 mm

Total station (TC 2002)

Ritz: (vertical displacement)

0.1 mm

0.5 mm

Zt0.5 mm

Ritz: (horizontal displacement)

0.1 mm

1 mm

*1 mm

(angular displacement)

0.1 arcsec

2 arcsec

* 5 arcsec

Photogrammetry

Elizabeth House

1 mm

1 mm

*2 mm

Tape extensometer

Elizabeth House

0.01 mm

0.03 mm

*0.2 mm

Demec gauge

Palace of Westminster

0.001 mm

0.01 mm

ZtO.01 mm

Rod extensometer

Elizabeth House

0.001 mm

0.01 mm

*0.2 mm

Electrolevel

Elizabeth House

2 arcsec

10 arcsec

f10 arcsec

GENERAL COMMENTS ON MONITORING Many factors should be taken into account when planning a monitoring survey, regardless of the measuring techniques employed. Some of those commonly encountered are listed below. Specific details concerning the individual instruments are given in the procedural notes produced by the research team (listed in the references at the end of this chapter).

18.3.1

Layout of monitoring points Careful selection of the layout of the points is fundamental to achieving good quality data and influences the cost and ease of performing a survey. A preliminary site visit to appraise the structure and its features is an essential prerequisite to this activity. When planning the layout of monitoring points for geodetic surveys, one or two monitoring points should be assigned as reference points, installed well outside the area of influence of the construction works where movements are not expected. If the whole structure is expected to move, they may have to be placed on a nearby, uninfluenced structure, preferably of similar construction. These points act as a reference against which the relative movement of the studied structure can be put into context. They might also help with the identification of thermally or seasonally induced movements. The distribution of monitoring points should be sufficiently close to give information to the required detail. When installing points for levelling, apart from practical constraints of performing the survey itself, consideration should be given to the feasibility of taping between the points, where possible avoiding obstructions such as drainpipes. Attention should also be paid to features such as expansion or construction joints, as these often have a strong influence on the way a building behaves.

Ch 18 Measuring techniques and their accuracy

275

It is always a good investment to establish at an early stage the level and position of all monitoring points, in relation to the building itself and to a suitable co-ordinate system. This helps to ensure that coverage is adequate and enables monitoring data to be plotted accurately before and during construction activities.

18.3.2

Background monitoring

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The importance of a good set of base readings cannot be over-emphasised. In all monitoring projects it is advisable to allocate as long a period as possible to establish background thermal and seasonal changes. This facilitates the interpretation of data and allows these movements to be isolated from other effects such as construction activity. Their magnitude and distribution depend on the way the structure was constructed and the nature of the foundations and the building materials. Often, of course, there simply will not be time to obtain as comprehensive a set of data as would be wished (say over a year). To some extent, however, seasonal effects can be established after the construction events. This is not ideal, however, if the works are taking place over a long period, where consolidation might influence the results or when monitoring is required for the implementation of protective measures. Likely external influences are seasonal or tidal changes in groundwater level, pressure, temperature and the daily changes in the direction of maximum solar radiation. In many cases, diurnal influences relating to tides, temperature and solar effects should be assessed by continuous monitoring over one or more 24-hour periods to establish the possible extent of daily changes. Subsequent measurements can then be put into context.

18.3.3

Weather and environmental conditions Weather and environmental conditions strongly influence the quality of data. They affect both the instruments themselves and their operation, the tripods that support them and sometimes those taking the measurements! It is good practice to record the weather conditions and if possible the temperature. Weather conditions that cause particular problems are hot sunny days (even those with intermittent sunshine), strong winds, heavy rain and snow. In some instances, there is no choice but to survey in such conditions. Some commonly encountered problems and possible solutions are listed below. the feet of the tripod sinking into a bituminous surface in hot weather - use pads to spread the load 0

the tripod being shaken by strong wind - greater stability is achieved by reducing the height of the legs and increasing the distance between the feet reduced visibility and hence accuracy from rain, snow, mist or heat haze - reduce sight distances if possible

0

instrument malfunctioning because of strong sunlight entering through the rear of the lens - use a shade tape flutter in windy conditions - use a damping system at mid length non-uniform tape temperature because only parts of the tape are exposed to sunlight - if it is not possible to defer work until conditions are favourable, arrange

equipment and assistance so that each measurement is made as quickly as possible. Make sure that the surface on which tripod feet are placed is firm and stable, avoiding loose paving slabs. Try to keep the position of an instrument clear of where pedestrians might walk into its legs or otherwise interfere with it.

276

Building response to tunnelling

There are particular problems from vibrations in built-up areas caused by vehicular and pedestrian movement at surface and subsurface levels, eg heavy vehicles moving along a narrow road and underground railways. These can cause considerable variations in the accuracy of readings from optical observations with sensitive instruments. When possible, it is advisable to take measurements at quiet periods, such as early morning or late evenings. This may also reduce survey time, as sights are less likely to be obscured by people or traffic. It also helps avoid poor visibility resulting from traffic fumes.

18.4

PRECISE LEVELLING

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Precise levelling can be carried out in most types of structures and involves the measurement of the height of each point to sub-millimetre accuracy. Precise levelling at ground level is the most common and primary form of monitoring and for this reason, considerable detail about the technique is included in this chapter. Vertical movements at ground level are usually good indicators of the extent to which other deformations of the building may occur. Precise levelling can be performed anywhere within a building and can help define any relative differences in deformation at various levels.

18.4.1

Planning the layout of monitoring points for precise levelling As the same points installed for precise levelling can also be used to measure horizontal displacements, their positions are chosen, where possible, to give uninterrupted spans for taping. Ideally, a series of points along one side of a building, wall or structure should be established to give a continuous series of spans. Such a layout allows both horizontal and vertical displacements to be measured relative to the same positions. In cases where the potential damage from building subsidence is being predicted, this is particularly relevant, as building damage is assessed in terms of settlements, slopes and horizontal strains. The layout of monitoring points around the studied structure should be at spacings that give adequate coverage of the building, consistent with its structural form, and reflect the expected form of deformations, eg monitoring points being at closer spacings in areas where large deformations are anticipated.

18.4.2

Precise levelling equipment and its operation There are five components required for precise levelling. 1

The precise levelling instrument is a key component. The research team used a digital level manufactured by Leica, model type NA 3003. Using an internal lightsensitive device, the instrument can measure the height of the plane of collimation on a suitable bar-coded staff to a resolution of 0.01 mm and can take multiple measurements of an individual point. Level measurements are only stored on the module to 0.1 mm, perhaps indicating a more realistic resolution based on the manufacturer’s experience.

2

The tripod has to provide a rigid and secure platform on which to mount the level. Tripods are still generally made from wood, as they are robust and temperatureinsensitive. During surveying it is important to try to follow the same procedure so far as is practicable, and this involves setting up the tripod at the same (preferably stable) position.

Ch 18 Measuring techniques and their accuracy

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The staffused in most cases by the research team is 2 m long, with a bar-coded invar strip. A shorter 0.8 m-long staff was used where there was reduced headroom (in the sub-basement of the Treasury, for example). The use of a bar-coded staff, in conjunction with the light-sensing device within the NA 3003, eliminates parallax errors introduced by incorrect adjustment of the eyepiece. A stable benchmark. It is important that the levelling run starts from a stable datum outside the zone of influence of any activity likely to cause movement. Ideally, this should be a deep datum, eg a rod extensometer with its anchor installed to a depth well below any zone of influence. If this is not possible (for instance, when the survey is within the basement of a building) the starting point, or the part of the structure in which it is installed, should be checked independently. The research team established a series of benchmarks adjacent to the structures of interest but outside the zone of influence of the construction works. Generally, they were located close to or at ground level on structures or pavements that showed no signs of previous deformation and that were not adjacent to trees. Several of the research surveys were within the basements of buildings. In such cases, a benchmark was established in the basement as far away as possible from the construction works, and additional points were established at ground level to extend the survey to unaffected ground. The absolute levels of several structures studied by the research team were affected by the deep dewatering in the vicinity. These activities caused settlement of the ground surface and structures over very wide areas. In these areas, the settlement affected the benchmark to the same extent as the structures of interest and thus absolute settlements were not measured.

Reproducible rneasuringpoints. Selection of the type of monitoring point, its fixity and shape, have an important influence on the accuracy of readings. The most repeatable results are obtained when the staff can be placed on the same point in the same way each time. The Building Research Establishment (BRE) socket and levelling plug have been designed so that the plug is removable and is screwed into the socket for each survey (see Figure 18.l(a)). Details of the BRE socket and levelling plug are given in BRE Digest 386 (BRE, 1993). This arrangement allows the plug to be repositioned each time to within 0.1 mm in both the vertical and horizontal senses. A smaller version of the BRE socket was produced which makes installation easier and less obtrusive (see Burland and Standing, 1997 and Figure 18.1(b)). Care must be taken to keep the machined surfaces clean so that the plug can be repositioned in exactly the same manner. The sockets were grouted in place using a sand-cement mortar, keeping them as horizontal as possible. Some alternative, inferior designs of socket and plug are available that rely on screwing the plug as tightly as possible into the socket. The sockets used by the JLEP contractors often differed from the BRE type adopted by the research team, being of brass with a diameter and length of about 10 mm and 40 mm respectively. These sockets required the use of a 75 mrn extension piece screwed into the levelling plug. A much smaller drilled hole was required for these sockets, which were installed using an epoxy-resin with sand filler. Measurements using this type of socket were less accurate than those obtained with the BRE sockets, the scatter of data being typically at least twice that obtained by the research team.

278

Building response to tunnelling

(a) Original full-size socket

,

I Min. hole dia. 38 mm

to depth of

- 80 mm

Loose thread

(b) Reduced-size version

/

mortar (2.5:l)

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Loose thread

Min. hole dia. 25 mm to depth of 80 mm

-

Figure 18.1

Details of the BRE sockets and plug used in precise levelling (a) original size, (b) miniature version

If an alternative socket design (or surveying technique) is adopted, it is recommended that the reproducibility with which the plug can be repositioned should be established before the survey programme begins to ensure that the accuracy required can still be achieved. The test could be done by taking numerous readings from one station to one or more points, removing and repositioning the plug each time. The variability between different people screwing in the plug and holding the staff should also be assessed. Locations that should be avoided as precise levelling points include marked or scribed lines on ledges, walls or level (approximately) surfaces or n a i l s h d s installed in yielding, soft or friable surfaces. Installation of points in mortar between bricks or joints between paving should also be avoided.

18.4.3

Precise levelling procedures The research team’s method of precise levelling used practices described in BRE Digest 386. Details of the methods used by the research team and suggestions for improved accuracy of measurement are given by Withers and Standing (1999). The research team followed, as far as was practicable, a set survey sequence in which the aim was to set up the tripod at approximately the same location for each survey and use the same monitoring points for change points. Distances to both intermediate and (particularly) change points were generally less than 20 m to maintain accuracy. When the instrument is moved during the levelling run, back- and foresights to change points should be made at similar distances as far as is practicable to minimise any collimation errors. Once the farthest measuring point has been sighted, the survey should be closed, either by returning to the initial datum to check on the closure of the run or to a separate benchmark. Under favourable conditions, the closing error should be within 0.3 mm. A larger error might have to be accepted under adverse surveying conditions.

Ch 18 Measuring techniques and their accuracy

279

18.4.4

Measurement accuracy in precise levelling The monitor of the precise level instrument used by the research team shows individual measurements to 0.01 mm. However, the value taken for the level and continuation of levelling is only stored to 0.1 mm, implying perhaps a more practical degree of accuracy. During levelling, an average of three individual measurements was taken, which would generally be within 0.02 mm of each other. Even if a consistent monitoring procedure is followed and the same equipment is used, however, the absolute field accuracy is also influenced by other factors.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Monitoring point and stafferrors. When using BRE sockets, the levelling plug can be relocated each time to within 0.1 mm. When taking intermediate sights, the accuracy of the measurement does not affect the overall survey. However, for change points, the accuracy of measurement of the fore- and backsight will affect the overall accuracy of the survey. It is good practice to keep the staff on the measuring point throughout the period of both sights. Thus, if either the plug or socket is loose it is more likely that the level would be the same for both sights, although this could result in discrepancies with other surveys. Weather and environmental conditions affect the accuracy of precise levelling, as described in Section 18.3.3. Their effect can be quantified to some extent by the overall closing error and the quality of data. Digital levels are particularly susceptible to strong sun and it is recommended that the instrument is shaded in these conditions. In addition, hot weather often creates heat hazes and associated shimmering, which interfere with the passage of the infi-a-red rays between the instrument and the staff. Traffic vibrations are a common source of error and are best avoided by surveying during quiet periods if possible. Strong or gusting winds may also cause error. Locations particularly susceptible to wind include street corners, the areas around tower blocks and near bodies of water. In adverse conditions it may be necessary to repeat measurements and relax the permissible differences in readings to a single point in order to complete a survey. External influences and instability of the benchmark. Certain external influences may affect the absolute level of monitoring points. While perhaps not significant when assessing differential settlements, they should be taken into account when assessing absolute measurements. The following list represents examples experienced by the research team: 0

seasonal changes in moisture content of the near-surface soils damage of surface monitoring points by traffic

0

changes in water level in bodies of water, eg tidaVseasonal effects

0

construction operations, including groundwater control measures temperature changes, eg freezing of the ground, softening of bituminous surfaces

0

sinking ground during surveying, eg due to rapid tunnelling operations.

These factors should be considered when planning the monitoring surveys and assessing the results of precise levelling. Collimation errors. Precise levelling relies on the instrument having a line of sight in an exactly horizontal plane. Modem digital levels have self-levelling devices. This, however, does not guarantee that the axis of the instrument telescope is in a totally level viewing plane. If the optical axis is not level, errors are introduced which increase with

280

Building response to tunnelling

larger sighting distances from the instrument to the staff. During the course of the research monitoring, checks for collimation error were made periodically using the twopeg test. Additionally, distances to foresights and backsights were kept roughly equal to minimise potential collimation error. Estimating the overall survey accuracy. The simplest and probably the most representative measure of the overall survey accuracy is the closing error. This gives an indication of the sum effects of the various potential sources of error over the length of the survey. However, it does not necessarily represent the accuracy of individual points. Achieving a high level of accuracy is also related to the frequency of measurements and having a good set of base readings to understand background movements. Closing errors in the surveys were generally small, ie within f 0.25 mm. Depending on the particular conditions of the survey and environmental conditions, accuracy was sometimes much better, eg f 0.1 mm, although on occasions it was worse at f 0.5 mm.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

18.4.5

Precise levelling data reduction Analysis of precise levelling data is straightforward. The most common method of reducing the data is to calculate the change in level of a monitoring point in relation to an initial reading or an average from a set of base readings. It is usual to present this as a change in level in millimetres (mm) with positive values representing heave and negative ones settlement. The necessity for and method of correction of data has to be assessed for each case study and time period individually. For structural monitoring, quantities that are important for assessing potential ground movements and structural damage are absolute settlement, differential settlement, slope and deflection ratio (see Chapter 3 for definitions of these terms). These quantities are all obtained using the results from precise levelling.

18.5

PRECISE TAPING MEASUREMENTS Precise taping with a tape extensometer was carried out on several of the structures to obtain changes in spans between precise levelling points. Ideally, a series of points along one side of a building, wall or structure, with distances between them typically of 5-10 m, should be established to give a continuous series of spans. (See also the comments on planning the layout of monitoring points given in Section 18.4.1.) This allows horizontal strains to be determined, which can then be correlated to changes in level. By carefully selecting the location of monitoring points it is possible to obtain a continuous profile of strain over the length of a building. As with precise levelling, a sub-millimetre accuracy of measurements was achieved. The research team used a particular type of extensometer - the Ealey extensometer - but most of the comments below apply to other tape extensometers.

18.5.1

Equipment required for precise taping The two main components of the tape extensometer are the tape itself and the measuring instrument, which is supported on a frame to which the tape is also attached and spooled. These and the other main components are shown in Figure 18.2.

Ch 18 Measuring techniques and their accuracy

28 1

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

..

Figure 18.2

.

Essential components required for precise taping measurements Tensioned to 0.135 kN (30Ibf)

Digital read-out

Fine winder

Figure 18.3

Coarse winder

Sketch of tape extensometer set up ready for measurement of a span

The tapes are made of steel and may be of different lengths but are typically 20 m long. Steel is preferred to other materials (eg invar) because of its known characteristics, quick response to temperature changes, cheapness and availability. Holes are precisely punched in the tape at 50 mm intervals and the free end of the tape has a hook permanently attached. During measurements, the hook at the free end of the tape is connected to a demountable eye-bolt that can be screwed into a BRE socket. A second demountable eye-bolt is located at the other end of the span (see Figure 18.3). These connectors have two axes of rotation that allow them to act as a universal joint (ie the eye-bolt can be positioned at any orientation). The extensometer instrument itself has a yoke at one end with a spigot, which passes through one of the punched holes in the tape at 50 mm spacings. The yoke is attached to sliding shafts, which are connected to a strain-gauged device housed within the instrument. An electrical trigger mechanism illuminates a green display light on the instrument when a tension of 133 N (30 lbf) is applied to it. The tension is applied by means of a winding handle, which retracts (or extends) the yoke. This device enables a constant tension to be applied to the tape for each measurement. A trigger is connected to a red light, which comes on if the tape is over-tensioned. A digital display on the instrument indicates the amount the tape has been wound in or let out.

282

Building response to tunnelling

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

The tape,being of steel, is very sensitive to tempesature, so a good-quality thermometer is required. The research team used a platinum resistance digital thermometer, capable of resolving to 0.1"C.

Figure 18.4

18.5.2

Making a measurement with the tape extensometer

Procedures for taping measurements As there is little available in the literature about guidance for taping measurements, the following paragraphs are provided to explain the rudiments of the methods used by the research team. Causes of inaccuracy and other site potential site problems are also addressed. More detailed information is given by Standing (2000).

The demountable eye-bolts are screwed into the BRE sockets at each end of the span to be measured. The tape is hooked to one end of the span and spooled out to reach the other end where the hook on the instrument is attached (Figure 18.4). The appropriate punched hole is then located on to the spigot and the tape wound in to take up the slack (it is important that the same punched hole is used each time), The thermometer is usually placed on the ground about half-way along the span being measured.

Ch 18 Measuring techniques and their accuracy

283

At the start of a series of measurements, the tape should be left set up on the first span for about five minutes before taking readings to allow it to equilibrate with the ambient temperature. The tape is tensioned to the correct value as indicated by the green light. For each span, a set of three readings should be taken within a range of about 0.03 mm. Readings will vary if the surrounding temperature is changing (or if the instrument is still equilibrating) generally increasing if the temperature is falling and vice versa. During the course of the research monitoring with the tape extensometer, it was found that there is a degree of operator influence on the readings (related to the manner in which the tape is tensioned). A set procedure was adopted to minimise this (for details, see Standing, 2000).

I

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

It helps if two people take measurements: one to operate the instrument and the other to log the readings and observe the temperature gauge. The second person is needed to prevent people walking or driving into the tape during measurements. A tensioned tape is easily broken by someone walking into it. The layout of BRE sockets should provide a continuous series of spans along the building side or faqade being monitored. This is not always possible because of irregularities in the walls (recesses, projections etc) or because of drainpipes. Spans should be between 5 m and 10 m long. Shorter spans can lead to less reliable measurements because the instrument length accounts for a greater proportion of the taped length and temperature corrections are not so readily applied. Longer spans are more prone to tape flutter from breezes. Repeatable measurements are more difficult to obtain with longer spans and so are less reliable. When planning the layout of monitoring points, provision should be made to include one or two control spans, ie that are independent of the agency causing the other spans to strain. If the whole structure is expected to strain, these may have to be located on another nearby, uninfluenced structure, preferably of similar construction. These spans act as reference distances on which the extensometer can be checked each time and also provide a measure of accuracy.

18.5.3

Factors affecting the accuracy of taping measurements Establishing the accuracy of a series of taping measurements is not as straightforward as with precise levelling. Factors concerning the use of BRE sockets are common to both types of measurement. The resolution of the tape instrument is 0.01 mm; if the three readings taken for each span fall within a range of 0.03 mm, accuracy should be enhanced. However, the overall accuracy is highly dependent on environmental conditions, particularly temperature. The weather conditions can have serious adverse effects on the accuracy and reliability of readings. Direct sunlight on any part of the tape should be avoided, radiation heateffects often not being accurately measured by a temperature gauge. The structure being monitored also responds to sunlight, further complicating interpretation of the data. If there is a possibility that spans will be exposed to sunlight, readings should be taken early in the morning, ie after the overnight period of stabilisation of building temperature and before daytime conditions start to affect readings significantly. In exposed areas, even small breezes will cause tape flutter. This effect of tugging at the tape causes the green tension indicator light to come on prematurely. Provided the tape span is not too long (eg less than 10 m), this can be largely avoided if the second person

284

Building response to tunnelling

holds a piece of folded paper to enclose (rather than to grip tightly) a short length of the tape at about mid-span to damp-out the flutter without applying extra tension to the tape. As noted above, it was found necessary to follow a rigorous procedure for these

measurements to avoid the largely operator-dependent gross and observational errors to which the technique would otherwise be prone. Once implemented, the procedure helped improve accuracy.

18.5.4

Correction and analysis of taping results

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

It is changes in span length that are of primary interest. Unless movements are very large, requiring a new punched hole position to be used, these are given by the readings shown on the digital display. The overall span is required for determining strains and this can be taken as the sum of the reading given on the digital display, the length of the tape to the punched hole and the instrument length. The primary correction that has to be applied to raw readings from site relates to thermally induced changes in the tape length. Temperature corrections can be applied using ( I ) a theoretical calculation method, or (2) an empirical approach where the tape has been calibrated by measuring a series of spans at various temperatures and producing a chart with correction curves for different span lengths. The former method has been used for the research measurements. The fabric of the structure is also affected by temperature changes. During the initial processing of taping data, an attempt was made to account for temperature changes to both the steel tape and the building fabric, but with poor results - the two corrections were essentially cancelling each other out. The coefficient of thermal expansion of concrete or masonry is very similar to that of steel (see BRE Digest 228, 1979). The significant factor is the difference in conductivity of the two materials, the tape equilibrating in a few minutes while the structure might take several weeks. Because of the uncertainty about the time-lag and the temperature distribution within a structure, corrections for any thermally induced building movements were not applied. It is also useful to appreciate the magnitude of thermally induced movements of the building, which, after all, are real. The expression used for correcting the raw data for each span at time, t, is given below.

sAl*

is the measured span corrected for the temperature of the tape.

sAt is the measured length of span A at time t (nb sAtis the tape length to the punched hole being used plus the reading taken on digital display, but without taking instrument length into account).

Clapeis the coefficient of thermal expansion of tape (taken to be 0.0000124 E/OC for the steel tape used). is the change in temperature in relation to the average value of that from the base readings.

ATA!

TA, is the measured temperature at time t (over span A ) and TA0 is the average value taken during the base readings at time to. Positive values relate to an increase in temperature. The change in span, with changes in tape length due to temperature eliminated, can be given by bAt*, the difference between the current corrected reading and the original span, SAO.

Positive values indicate that the span length is increasing.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

It is often useful to convert displacements to strains, especially when assessing potential damage to the fabric of a building, damage classifications often being related to strain ranges. The strain at the midpoint of each span at time, t , can be expressed as:

where Linstis the length of the instrument (= 0.544 m for the Ealey extensometer). The expression is set up to give the sign convention used in structural engineering, ie positive values indicate expansion (tension) and negative contraction (compression). The measurements from a series of spans can also be expressed as cumulative displacements relative to one end of the line by summing individual span changes.

18.6

HORIZONTAL DISPLACEMENTS MEASURED USING A MICROMETER STICK Horizontal displacements or strains can also be measured using a micrometer stick. This method was originally considered for the measurement of building strains. It was not adopted because of the practical difficulties of measuring different lengths and sometimes long spans within the buildings, and also because of the cost and difficulty of manufacture of the devices. Tape extensometers were used instead. Although they are less accurate, they provide an efficient means of measuring horizontal displacements to an adequate accuracy between points whose vertical position was also being monitored. Micrometer sticks were used at the two greenfield reference sites at St James’s Park and Southwark Park to measure near-surface horizontal strains between shallow surface settlement points (Figure 18.5). These points were essentially extended BRE sockets installed within a concrete column about 1.5 m deep and 2.5 m apart (see Chapter 25 for more details). In these conditions, the micrometer stick had an added advantage that it did not impose undue force on the ground points (as there could have been from tensioning a tape).

18.6.1

Procedures for micrometer stick measurements At the greenfield reference sites, relative horizontal displacements between adjacent settlement points transverse to the tunnel axes were measured using a 2.5 m-long micrometer stick similar to that described by Burland and Moore ( 1 974). These points extended over at least 60 m. The micrometer stick comprises a long aluminium bar of hollow square section with a permanently mounted micrometer at one end, a machined slot, and a series of five machined holes located near the opposite end (as shown in

286

Building response to tunnelling

Figure 18.6). A thermometer is seed at the centre of the beam beneath a protective clear acrylic window. Two extended posts with enlarged ball seathgs at the top (-270 mm in length) were screwed into adjacent shallow surface settlement points. The threading arrangement for the posts is identical to the BRE monitoring plug described earlier. A micrometer mounted at the slotted end is screwed in until contact is made between the metal micrometer tip and the ball seating. To kilitate repeatable readings between instrument operators, an electrical circuit is set up through the beam so that when contact is made, the circuit is closed and a light is illuminated. Sets of three readings on the micrometer are taken for each span by raising and relocating the slotted end of the beam between each measurement. The temperature is also recorded at each span.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

I

Figure 18.5

Taking a measurement with a micrometer stick

Machined slot 4

Precisely machined holes

Vernier micrometer (50 mm range)

A -

'...U Figure 18.6

Posts with enlarged ball seatings screwed into adjacent monitoring points

Schematic drawing of the micrometer stick

Ch 18 Measuring techniques and their accuracy

287

18.6.2

Uncertainty and sources of error with micrometer stick measurements The micrometer is graduated with 0.005 mm divisions. The micrometer manufacturer quotes an accuracy of measurement of 0.005 mm. Repeated readings at each span are usually within f 0.05 mm, which is representative of the precision for each set. Trials to determine the potential variation for removing and replacing the extended posts into the shallow surface settlement points showed a repeatability between measurement sets of about f 0.2 mm, however. The accuracy of measurement is generally dependent on the care with which the micrometer plunger is brought into contact with the ball. Using an electrical circuit to indicate contact minimises this potential error.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

A potential gross error committed by using the wrong numbered posts is possible and would show differing measurements for all points across the site. Observational errors arising from misreading the micrometer andor temperature, and poor tightening of the posts into the shallow surface points, may also contribute to random errors. The bar-mounted thermometer is graduated to 1°C. Measuring the correct bar temperature at each span is difficult when the temperature changes rapidly (ie in direct sunlight); in such instances the effects on the bar and the thermometer are unlikely to be similar. Generally, the micrometer stick was removed from its protective box and exposed to the ambient conditions for at least half an hour before taking measurements to allow it to equilibrate. Not allowing sufficient time for the aluminium bar to acclimatise could also lead to significant errors. Temperature-controlled laboratory tests were performed to determine the bar’s temperature sensitivity (see Nyren, 1998).

18.6.3

Processing micrometer stick measurements Changes in R, the measurement observed on the micrometer, represent changes in length between the adjacent shallow settlement points. The value has to be corrected to account for potential thermally induced changes in the length of the micrometer stick. An expression for the corrected reading, R,, is given below. Therefore, although R, is not a true measure of the actual span length, it is the change in distance between points, and therefore the change in R,, that is of interest. A temperature correction (relative to 20°C) is applied to each measurement using the measured temperature on the bar, T, its length, L, and the coefficient of thermal expansion, C, to give the corrected reading R,:

R, = R + C.L( T - 20) The average of the corrected readings before tunnelling was established for each span as the base readings, Ro. Change in micrometer reading from the base reading reflects the combination of real change in the span distance between the two shallow surface points (AS), and errors in repositioning the post and ball-seatings described above.

For small changes in span, the measured change divided by the span length, S, gives the average horizontal strain.

288

Building response to tunnelling

As with the taping results, a structural engineering sign convention is used, with positive values indicating extension, and negative indicating compression. The transverse horizontal displacements for each point can be assessed by assuming that one end of the line of monitoring points is stationary and summing the measured span changes (AS). The displacement, vi, at the i'h point in the line from the "fixed" end-point is calculated using: vi =

( A S , + A s 2 + ... + dSi )

Alternatively, a profile of horizontal displacement may be estimated from the average horizontal strains across each pair of shallow surface points by curve-fitting the data and differentiating the expression.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

18.7

FACADE MONITORING Faqade monitoring involves surveying points on the faqade with a high-precision total station (which measures horizontal and vertical angles, and distances to retro-reflective prisms) to establish changes in their positions. Plane faqade movements in a threedimensional Cartesian co-ordinate system can then be determined from these measurements.

18.7.1

Equipment used for faGade monitoring For most of the surveying work, the research team used a total stution instrument was a Leica TC 2002, which is fully electronic and all data are stored on a record module. The instrument can measure angles and distances to a resolution of 0.1 seconds of arc and 0.1 mm respectively. Corrections are entered into the software at the start of the survey to take atmospheric conditions into account (temperature and barometric pressure). The instrument is mounted on a tripod during measurement. A less accurate Leica TC 1610 was used for a short period for some of the monitoring. The targets installed on buildings were retro-reflective prisms mounted on a plastic square laminate with cross-lines marked on them. This type of target allows distances to be measured with the total station. Sizes generally used are 60 mm x 60 mm on upper floors or far-off points and 40 mm x 40 mm on nearer positions. Targets are placed to give good coverage of the building, with a greater density in areas of the building most likely to be affected by construction, and where possible remote from potential vandalism. On certain buildings, an alternative, non-reflective target was used, which was more easily visible with greater contrasting markings on it. These targets were installed on buildings with the intention that they might be surveyed using photogrammetry techniques. The position of such targets could still be fixed with the total station, but measurements were necessary from more than one station.

18.7.2

Procedures for faGade monitoring Measurements to the targets are made from survey stations at ground level opposite the faqade. Reference points affixed to structures outside the zone of influence of JLE construction activities were also established to provide a frame of reference for the angle and distance measurements. These were installed at the same time as the targets on the building itself and were assumed to be stationary during all site monitoring. Survey station positions are carefully selected to maximise the number of targets that can be seen from each. It is important that the targets are seen from at least two stations

Ch 18 Measuring techniques and their accuracy

289

to supply redundant observations, as this considerably increases confidence in the measurements. Also, when the angle of the instrument to the target becomes too oblique it is often not possible to measure distance. It is then essential that angle measurements are made from two stations to such targets. The survey station positions are marked on the ground surface (eg by a nail) and their locations co-ordinated at the start of each survey using the “stationary” reference targets. The stations are also co-ordinated in relation to each other using a tripod-mounted reflector centred over the nail at each location. This reflector is a “cube-corner prism” with cross-hairs for measuring horizontal and vertical angles while the prism returns transmitted waves from the electro-distomat (EDM) back precisely in the same direction as they are received for distance measurement. Co-ordinating between stations also improves the accuracy of measurements. Base lines between stations and reference points are then established after a number of initial base readings.

Licensed copy:IMPERIAL COLLEGE, 12/02/2009, Uncontrolled Copy, © CIRIA

Angle measurements are made on both faces of the instrument many times to eliminate face errors, although in the total station used (ie TC 2002) they can be eliminated automatically. Taking measurements on both faces does give additional redundant readings, thereby increasing accuracy. Distances can only be determined to points marked by retro-reflective targets. The accuracy that can be obtained depends on the distance between the instrument and the point being measured and their relative positions. Measurements to the non-reflective photogrammetry targets have to be made from more than one station to fix their position. After surveying all the visible targets, the measurements on the reference targets were repeated. The survey was then repeated from the other survey stations. Total station measurements were also made to the shallow surface settlement points at the St James’s Park greenfield reference site to obtain the three-dimensional movements of the line. Measurements to the points were made using a demountable retro-reflective prismatic target with the thread system of the BRE monitoring plug incorporated in the design. The top of the target rotates freely around the vertical axis via a precision bearing enabling the prism orientation to be adjusted to face the total station squarely. Details of the procedure used at the greenfield reference sites are given by Nyren (1 998).

18.7.3

Processing and analysis of total station measurements Prior to processing the measurements it is necessary to establish a local co-ordinate system and determine the position of all reference points relative to this local system using triangulation and trilateration. The positions were calculated for several surveys performed before construction activity to establish the initial co-ordinates for each reference point. These were recalculated for surveys performed later in the construction sequence to confirm that the reference targets were stable relative to each other over longer time periods. The results are processed with the assistance of a computer program that adjusts all observed values of horizontal and vertical angles and distances by the method of least squares. The principle uses redundant measurements taken during each survey to adjust the measured quantities to satisfy geometric criteria (eg triangulation and trilateration) while minimising the adjustments made to the actual measurements. The program used for analysis was written by Mr S K Sharma (formerly of the Department of Civil Engineering, Imperial College). Relative weightings are assumed

290

Building response to tunnelling

for measured angles ( 1 second) and distances (1/50 000) for calculating and minimising the standard error of analysis (ie the overall value of the adjustments). These weightings reflect the precision and accuracy for each measurement. However, changing the relative weightings to 2 seconds and 1/25 000 was found to result in only small differences in the calculated co-ordinates (