C743

Reservoir owners and safety engineers have long known or suspected that conduit problems
have led directly to a significant number of dam failures and incidents. The mechanisms and
modes of failure are well understood, but what is lacking is a single guidance document
providing comprehensive information on inspection, monitoring, investigation, maintenance
and repair of conduits in their various forms specific to UK reservoirs and dams.

Dams and reservoir conduits. Inspection, monitoring, investigation, maintenance and repair
A conduit, in the context of this guide, is considered to be a means to convey water from a
reservoir through, around or under an embankment dam in a controlled manner, or to
provide access.
This guide covers a wide range of aspects to conduits such as forms and materials used
in their construction, failure modes, methods for inspection and assessment (ie surveys,
integrity testing, and leakage investigations), and the instrumentation that can be used
for such testing.
There is also guidance on identifying the optimum solution with respect to the renovation,
Dams and reservoir conduits
repair, replacement or abandonment of conduits.
Steps to take in the event of a significant failure (ie emergency actions) are given and the Inspection, monitoring, investigation,
importance of accurate record keeping for current and future management of conduits
are included.
maintenance and repair
This stand-alone guide is aimed at reservoir owners and engineers concerned with
reservoir safety. In addition, reference to other documents from the UK and those
highlighting practice overseas has been provided where appropriate.

CIRIA

C743 9 780860 177487

 Who we are
Established in 1960, CIRIA is a highly regarded, industry-responsive, not for profit research and information
association, which encompasses the construction and built environment industries.
CIRIA operates across a range of market sectors and disciplines, providing a platform for collaborative projects
and dissemination by enhancing industry performance, and sharing knowledge and innovation across the built
environment.
As an authoritative provider of good practice guidance, solutions and information, CIRIA operates as a knowledge-
base for disseminating and delivering a comprehensive range of business improvement services and research
products for public and private sector organisations, as well as academia.

How to get involved

CIRIA manage or actively participate in several topic-specific learning and business networks and clubs:

zzCore membership zzCEEQUAL

Allows your employees to assist with the development of CIRIA co-manages the sustainability assessment, rating
and access to good practice guidance, formal networks, and awards scheme for civil engineering, infrastructure,
facilitation, conferences, workshops and training. landscaping and works in public spaces.

zzAssociate membership zzLACL (Local Authority Contaminated Land Network)

Allows your employees to access CIRIA’s services. LACL helps local authorities address responsibilities
Members are able to access exclusive content via the under Part IIA of the Environmental Protection Act 1990.
CIRIA website.
zzEMSAGG (European Marine Sand and Gravel Group)
zzThe CIRIA Network CIRIA provides secretariat support to EMSAGG, including
A member-based community where clients and professionals management of the Group’s conferences, workshops and
meet, develop and share knowledge about specific topics website and producing its newsletter.
relevant to construction and the built environment.
zzLANDFoRM (Local Authority Network on Drainage
zzCIRIA Books Club and Flood Risk Management)
Members can buy most CIRIA publications at half price and A platform for sharing knowledge and expertise in flood
can attend a range of CIRIA conferences at reduced rates. risk management and sustainable drainage.

zzProject funding zzBRMF (Brownfield Risk Management Forum)

Project funders influence the direction of the research Promoting sustainable and good practice in brownfield
and gain early access to the results. projects in the UK.

Where we are
Discover how your organisation can benefit from CIRIA’s authoritative and practical guidance – contact us by:

Post Griffin Court, 15 Long Lane, London, EC1A 9PN, UK

Telephone +44 (0)20 7549 3300
Fax +44 (0)20 7549 3349
Email enquiries@ciria.org
Website www.ciria.org
(for details of membership, networks, events, collaborative projects and to access CIRIA publications through
the bookshop)
CIRIA C743 London, 2015

Dam and reservoir conduits

Inspection, monitoring, investigation,
maintenance and repair
Dr A K Hughes, D A Bruggemann, K D Gardiner, Atkins Ltd
T Blower, M Deane, Mott MacDonald Ltd
Dr D F T Nash, University of Bristol

Griffin Court, 15 Long Lane, London, EC1A 9PN

Tel: 020 7549 3300 Fax: 020 7549 3349
Email: enquiries@ciria.org Website: www.ciria.org
Dam and reservoir conduits – inspection, monitoring, investigation, maintenance and repair

Hughes, A K, Bruggemann, D A, Gardiner, K D, Blower, T, Deane, M, Nash, D F T

CIRIA

C743 © CIRIA 2015 RP982 ISBN: 978-0-86017-748-7

British Library Cataloguing in Publication Data

A catalogue record is available for this book from the British Library.

Keywords
Asset management, conduits, embankment, concrete and masonry dams, health and safety

Reader interest Classification

forms and materials, failure modes, Availability Unrestricted
inspection and assessment, surveys and Content Enabling document
investigation, instrumentation and testing,
Status Commissioned, committee-guided
emergency actions, renovation and repair,
User Reservoir safety engineers, asset owners and
replacement, abandonment, record keeping.
managers

Published by CIRIA, Griffin Court, 15 Long Lane, London, EC1A 9PN, UK

This publication is designed to provide accurate and authoritative information on the subject matter covered. It is sold and/or
distributed with the understanding that neither the authors nor the publisher is thereby engaged in rendering a specific legal or any
other professional service. While every effort has been made to ensure the accuracy and completeness of the publication, no warranty
or fitness is provided or implied, and the authors and publisher shall have neither liability nor responsibility to any person or entity
with respect to any loss or damage arising from its use.
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, including
photocopying and recording, without the written permission of the copyright holder, application for which should be addressed to
the publisher. Such written permission must also be obtained before any part of this publication is stored in a retrieval system of any
nature.
If you would like to reproduce any of the figures, text or technical information from this or any other CIRIA publication for use
in other documents or publications, please contact the Publishing Department for more details on copyright terms and charges at:
publishing@ciria.org, Tel: 020 7549 3300.

ii CIRIA, C743
Executive summary

Reservoir owners and safety engineers have long known or suspected that conduit problems have led
directly to a significant number of dam failures and incidents. The mechanisms and modes of failure
are well understood, but what is lacking is a single guidance document providing comprehensive UK
reservoir and dam specific guidance on inspection, monitoring, investigation, maintenance and repair of
conduits in their various forms. Efforts have been made to produce a stand-alone guide, but reference to
other documents has been provided where appropriate.

A conduit in this context is considered to be a means to convey water from a reservoir through, around
or under an embankment dam in a controlled manner or to provide access. Arrangements where the
conduit is incorporated in a tunnel through one of the abutments are not considered in this guide.
Drainage and inspection tunnels (galleries), and concrete and masonry dams, are also excluded.

This guide covers the following aspects of conduits:

‰‰ forms and materials

‰‰ failure modes
‰‰ inspection and assessment
‰‰ surveys, integrity testing and leakage investigations
‰‰ instrumentation and testing
‰‰ emergency actions
‰‰ renovation and repair
‰‰ replacement
‰‰ abandonment
‰‰ identifying the optimum solution
‰‰ record keeping.

The guide is aimed at reservoir owners and engineers concerned with reservoir safety for the inspection,
monitoring, investigation, maintenance and repair of reservoir conduits. Reservoir safety engineers
could be those appointed or acting under the Reservoirs Act 1975 or the owner’s reservoir safety
managers. Design guidance for conduits in new dams is not included, but reference can be made to
FEMA (2005) for further information. It is expected that users of this guide are from an engineering
background and are familiar with civil engineering construction.

Dams and reservoir conduits iii

Acknowledgements

This guide has been produced by CIRIA as a result of Research Project 982. The project was carried
out under contract to CIRIA by Atkins Ltd working in partnership with Mott MacDonald Ltd and the
University of Bristol.

Authors
Dr Andrew Hughes BSc (Hons) PhD DMS CEng FICE FCIWEM MIM
Andy is director of the Dams and Reservoirs Group at Atkins Ltd. He has over 30 years’ experience
in the study, design and project management of water supply and flood alleviation schemes, asset
management and risk studies and reservoir safety works. He is an All Reservoirs Panel Engineer, and
is a geotechnical engineer and hydraulic engineer, and hydrologist, and has written a large number of
technical papers and research documents. Remedial works at more than 200 dams have been designed
and supervised under Andy’s direction. He has worked overseas in a large number of countries including
Nigeria, Cyprus, Hong Kong, Laos and Canada.

Tony Bruggemann BSc (Eng) MSc DIC CEng MICE

Tony is technical manager in the Dams and Reservoirs Group at Atkins Ltd. Tony has over 35 years’
experience in the planning, detailed design and construction of embankment dams, geotechnical
engineering and project management in the UK and overseas. He has an extensive experience as a
supervising engineer under the Reservoirs Act 1975, and in the execution remedial works to UK and
overseas dams. Recently he has been involved in dam safety inspections and design reviews in Nigeria
and Pakistan.

Keith Gardiner BSc CEng FICE

Keith is an All Reservoirs Panel Engineer with Atkins Ltd. Keith has 40 years’ experience as a civil
engineer, most of it reservoir related. He was lead design engineer on reservoirs with North West Water
Engineering and Bechtel Water Technology overseeing civil engineering works on 130 dams for 10
years. He then spent 12 years as regional reservoirs strategy manager for United Utilities (UU). At UU,
Keith was responsible for the safety of over 200 reservoirs and developed the portfolio risk assessment
methodology used to justify and prioritise their reservoir capital programme.

Tim Blower BSc MSc CEng FICE CGeol FGS

Tim is a technical director of Mott MacDonald Ltd, with over 30 years’ experience of ground
engineering in the UK and overseas. He has worked on the geological and geotechnical aspects of a
large number of dams, barrages and reservoirs, including various dam types from simple earthfill
embankment dams and service reservoirs, to substantial concrete-faced and clay-core rockfill dams in
Europe, and major gravity dams in Africa and Asia. His most recent project involvements are with dams
in Vietnam, Indonesia, Zambia and Albania.

Martin Deane MEng (Hons) CEng MICE

Martin is a hydraulic structures engineer for Mott MacDonald Ltd, and is currently working on the
Devoll hydropower project in Albania. Martin has been appointed to the Reservoir Supervising Engineer
Panel as established by the Reservoirs Act 1975. He has worked on a variety of dam engineering projects,
including yield and freeboard assessments, feasibility studies, design, and site supervision, primarily in
the UK.

iv CIRIA, C743
Dr David Nash MA MSc DIC PhD CEng MICE
David is a senior lecturer in geotechnical engineering in the Department of Civil Engineering at
the University of Bristol. After gaining wide-ranging geotechnical experience at the Norwegian
Geotechnical Institute and with Arup Geotechnics, David moved to Bristol in 1978 where he has
developed and taught courses in soil mechanics, foundation design, slopes and dams, site investigation,
soil-structure interaction, instrumentation and case studies. His research interests include monitoring
the field performance of geotechnical structures, laboratory measurement of soil stiffness and numerical
modelling of creep settlement in soft soils. He regularly acts as a geotechnical consultant and is currently
working with Arup on the design of a new reservoir at Cheddar in Somerset.

Project steering group

Following CIRIA’s usual practice, the research project was guided by a project steering group (PSG),
which comprised:

John Ackers Black & Veatch

Norman Blacklock Northumbrian Water
Newman Booth Yorkshire Water
David Brown (chair) Canal & River Trust
Tony Deakin Environment Agency
John Gosden Jacobs
David Hart Environment Agency
Dr Des Hartford British Columbia Hydro
David Henderson DARD Rivers Agency
Henry Hewlett Independent Consultant
Bren Hodkinson Environment Agency
Ian Hope Severn Trent Water
Jonathan Howard Dŵr Cymru (Welsh Water)
Gamini Karunaratne Scottish Water
Jasvir Malia Scottish Water
Robert Mann AECOM
Jack McCarey South West Water
Dominic Molyneux Black & Veatch
James Penman CH2M Hill
Andrew Ross Scottish Canals
Ian Scholefield United Utilities
Alan Warren Mott MacDonald Ltd (formerly CH2M Hill)
Denis Wilson Northern Ireland Water
Jarek Zolnacz Southern Water (formerly Environment Agency)

Project funders
Dŵr Cymru (Welsh Water)
Environment Agency
Northern Ireland Water
Northumbrian Water
DARD Rivers Agency
Scottish Water
Severn Trent Water
South West Water
United Utilities

Dams and reservoir conduits v

CIRIA managers
Chris Chiverrell Project director
Philip Charles Project manager

Other acknowledgements
CIRIA would like to acknowledge the following organisations with respect to the content of case studies
within the guide: Canal & River Trust, Dŵr Cymru (Welsh Water), Mott MacDonald Ltd, Scottish
Canals, United Utilities.

CIRIA would like to acknowledge the following organisations for providing photographs and diagrams
within the guide: Advanced NDT Ltd, AECOM, Atkins Ltd, Avongard, P Beckwith, D A Bruggemann,
Canal & River Trust, Cemex, CERAM, Dŵr Cymru (Welsh Water), Electro Scan Inc., K Gardiner,
Geokon, Geo-Observations, Getec UK, Halcrow, IBAK, Impact-Echo, R T Jover, Leica Geosystems,
MacKenzie Construction, MG JV, Moiré Tell Tales, Mott MacDonald Ltd, D F T Nash, National
Trust, OnSite Ltd, Oz Coasts, Pure Technologies, Rock Solid Group, A Ross, Scottish Canals, Severn
Partnership, Severn Trent Water, SewerVUE Technology Corp, Sisgeo, Spiralboy, TesTex, United
Utilities, Wikipedia, Willowstick Technologies LLC.

vi CIRIA, C743
Contents

Abbreviations and acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv

Executive summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Conduit type and uses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 Issues associated with conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4.1 Types of failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4.2 Legislative aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4.3 Downstream discharges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1 Conduit forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2 Conduit materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.1 Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.1.1 Reinforced cast in situ concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.1.2 Precast concrete pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.2 Metal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2.2.1 Grey CI and DI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2.2.2 Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.2.3 Corrugated metal pipes (CMP) and culverts . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.3 Plastic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2.3.1 Thermoplastic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2.3.2 Twin wall PE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2.3.3 Thermoset plastic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2.3.4 Glass reinforced plastic (GRP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
2.2.4 Masonry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2.4.1 Brick . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.4.2 Stone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2.4.3 Mortar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.5 Timber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2.6 Vitrified clay pipes (ceramic) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3 Conduit general arrangements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3.1 Outlet works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3.1.1 Upstream control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3.1.2 Middle control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3.1.3 Downstream control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3.2 Spillways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3.3 Design details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
3 Failure modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1 Main modes of failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1.1 Erosion of soil into a non-pressurised conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.2 Erosion of soil by water escaping from a pressurised conduit . . . . . . . . . . . . . . . . . . . . . 20
3.1.3 Erosion of soil by water flowing outside a conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.4 Erosion of soil near to a conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1.5 Collapse of the conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Dams and reservoir conduits vii

 3.2 Causes of defects in conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2.1 Abrasion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2.2 Corrosion of metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2.3 Concrete deterioration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.3.1 Freeze-thaw . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
3.2.3.2 Sulphate attack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.3.3 Carbonation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.3.4 Chloride attack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.3.5 Alkali-aggregate reaction (AAR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
3.2.4 Defects in precast concrete conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.5 Defects in brick and masonry conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.6 Defects in CI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.7 Poor conduit design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.8 Cavitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Causes of hydraulic fracture and cracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.3.1 Uneven foundation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.3.2 Inadequate or uneven compaction and backfilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.3.3 Inadequate structural capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.3.4 Cut-off collars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4 Conduit examination and cleaning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.1 Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.1.1 The purpose of the examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.1.2 The requirement for/feasibility of human entry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.1.3 Equipment and precautions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.1.4 The preparation of an examination plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.1.5 The data available to inform the examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.1.6 The personnel required on the examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.2 Frequency of examination of conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.2.1 Initial examination on first filling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.2.2 Periodic examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.2.3 Routine examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.2.4 Special examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.2.5 Supplementary or emergency examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.3 External examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.3.1 Ground movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.3.2 Seepage flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.3.3 Normal flow from the conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.3.4 Additional concerns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4 Internal examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4.1 Masonry culverts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4.2 Concrete conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.4.3 CI and steel conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.4.4 Corrugated metal pipes (CMP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.4.5 Pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.4.6 De-watering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.5 Specialised inspections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.5.1 Underwater . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.5.2 ROV examinations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.5.3 CCTV examinations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.6 Cleaning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.6.1 Reasons for cleaning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.6.1.1 Sediments and debris . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.6.1.2 Hard deposits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.6.1.3 Bacterial growths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.6.1.4 Zebra mussels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

viii CIRIA, C743

 4.6.2 Cleaning methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5 Conduit assessments – surveys, integrity testing and leakage investigations . . . . . . . . . . . . . . . . . . . . . 40
5.1 Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.1.1 Conventional surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.1.2 Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.1.3 Laser scanning (LiDAR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5.1.4 Sonar scans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5.2 Conduit integrity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2.2 Non-destructive testing (NDT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2.2.2 Surface hardness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2.2.3 Ultrasonic test methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
5.2.2.4 Electromagnetic test methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.2.3 Other test methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.2.4 Invasive coring and testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.3 Leakage investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.3.1 Initial investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.3.1.1 Flow measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.3.1.2 Water analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.3.1.3 Pore water pressure measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.3.2 Temperature measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.3.2.1 Profile of temperatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.3.2.2 Temperature measurements at specific locations . . . . . . . . . . . . . . . . . . . . . . 51
5.3.3 Electrical resistance measurement (electro-scanning) . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.3.4 Electro-magnetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.3.5 Other geophysical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.4 Void detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.4.1 Ground penetrating radar (GPR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.4.2 Other test methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
6 Instrumentation and monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
6.1 Purposes of monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
6.2 Identification of key parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
6.2.1 Prediction of values and planned response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
6.3 Specification of instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
6.4 Types of instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
6.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
6.4.2 Water measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
6.4.2.1 Water ingress into conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
6.4.2.2 Seepage flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
6.4.2.3 Pore water pressures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
6.4.2.4 Water temperature measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
6.4.2.5 Reservoir level and toe drain water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
6.4.3 Total stress measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
6.4.4 Deformation of conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
6.4.4.1 Survey techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
6.4.4.2 Measurements with probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .63
6.4.4.3 Permanent installations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
6.4.4.4 Movement of cracks and joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
6.4.4.5 Monitoring deformed shape of conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
6.4.5 Ground deformation (dam near conduit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
6.4.5.1 Surface movements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
6.4.5.2 Subsurface movements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
6.4.6 Other measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .68

Dams and reservoir conduits ix

 6.5 Data acquisition, processing, reporting and interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.5.1 Data acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.5.2 Data processing and reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.5.3 Interpretation and action plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
7 Planning for emergency response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
7.1 Surveillance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
7.2 Access to archive data for reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
7.3 Identification of appropriate ARPE contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
7.3.1 Need for an engineer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
7.3.2 Choice of panel engineer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
7.4 Creating and implementing the emergency plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
7.5 Emergency draw-down . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
7.6 Post-emergency recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
8 Techniques for renovation and repair of conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
8.1 Ranking feasible solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
8.2 Isolation of a conduit from the reservoir head . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
8.2.1 Pipe blocking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
8.2.2 Hot tapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
8.2.3 Line stopping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
8.2.4 Reservoir water level control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
8.2.5 Temporary cofferdams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
8.3 Renovation of pipes buried in embankment fill . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
8.3.1 Internal renovation – general . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
8.3.1.1 Comparing liner solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
8.3.1.2 Isolation, cleaning and inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
8.3.2 Sliplining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
8.3.2.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
8.3.2.2 Technical envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
8.3.2.3 Site suitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
8.3.2.4 Summary of installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
8.3.2.5 Main costs and risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
8.3.3 Close-fit pipe lining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87
8.3.3.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
8.3.3.2 Technical envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
8.3.3.3 Site suitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
8.3.3.4 Summary of installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
8.3.3.5 Main costs and risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
8.3.4 Cured-in-place pipes (CIPP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
8.3.4.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
8.3.4.2 Technical envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
8.3.4.3 Site suitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
8.3.4.4 Summary of installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
8.3.4.5 Main costs and risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
8.3.5 In situ sprayed linings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.3.5.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.3.5.2 Technical envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.3.5.3 Site suitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
8.3.5.4 Summary of installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
8.3.5.5 Main costs and risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
8.3.6 External renovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
8.4 Renovation of pipes in culverts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
8.4.1 External repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
8.4.1.1 Painting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
8.4.1.2 Bolted joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

x CIRIA, C743
 8.4.1.3 Movement accommodation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
8.4.1.4 Restraint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
8.4.1.5 Fittings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
8.4.2 Internal repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
8.5 Renovation and repair of human-entry culverts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
8.5.1 Importance of routine maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
8.5.2 Impact of repairs on leakage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
8.5.3 Localised repair of masonry culverts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
8.5.3.1 Mortar pointing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
8.5.3.2 Patch repairs to masonry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
8.5.3.3 Crack repairs to masonry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
8.5.4 Concrete conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
8.5.4.1 Patch repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
8.5.4.2 Crack sealing and repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
8.5.4.3 Joint sealing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
8.5.5 Invert repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
8.5.6 Filling voids behind culvert walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
8.5.6.1 Grouting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
8.5.6.2 Other fillers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
8.5.6.3 Void filling from the ground surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
8.5.7 Dealing with historic repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
8.5.8 Changes to structural arrangements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
8.5.8.1 Design and construction considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
8.5.9 Culvert lining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
8.5.9.1 Sprayed concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
8.5.9.2 Reinforced concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
8.5.9.3 Plate-bonding (steel, carbon fibre or GRP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
9 Techniques for replacement of conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
9.1 Replacement conduit design (general) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
9.1.1 Determining location requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
9.1.1.1 Interaction with membrane liner materials . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
9.1.1.2 Replacement on the same alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117
9.1.2 Pumps and siphons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
9.1.2.1 Pumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
9.1.2.2 Siphons to provide draw-down capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
9.1.2.3 Siphon constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
9.1.3 Determining hydraulic requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
9.1.3.1 Control position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
9.1.3.2 Flow capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
9.1.4 Determining operational requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
9.1.4.1 Sedimentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
9.1.4.2 Zebra mussels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
9.1.4.3 Debris . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
9.1.4.4 Vandalism, security, public safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
9.1.5 Determining isolation options and structural requirements . . . . . . . . . . . . . . . . . . . . . . 125
9.2 Replacement of pipes buried in embankment fill . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
9.2.1 Typical structural details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
9.2.1.1 Pipe material options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
9.2.1.2 Precast concrete pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
9.2.1.3 DI pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
9.2.1.4 PE pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
9.2.1.5 Details to avoid low pressure zones in fill . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
9.2.1.6 Anti-seepage collars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
9.2.1.7 Filter diaphragms, collars and drains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
9.2.2 Construction process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Dams and reservoir conduits xi

 9.2.2.1 Control of reservoir water level during construction works . . . . . . . . . . . . . . 135
9.2.2.2 Removal of the existing conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
9.2.2.3 Placement of new conduit and compaction . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
9.3 Replacement of pipes within culverts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
9.3.1 Replacement planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
9.3.2 Replacement feasibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
9.3.2.1 Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
9.3.2.2 Isolation from the reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
9.3.2.3 Temporary function replacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
9.3.2.4 Pipe material and jointing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
10 Techniques for abandonment of conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
10.1 Abandonment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
10.1.1 Abandonment process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
10.1.2 Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
10.1.3 Planning – access and isolation requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
10.1.4 Specification of fill material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
10.1.5 Specification of filter collar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
10.1.6 Other design considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
10.1.7 Preparation of conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
10.1.8 Filling process and recording . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
10.1.9 Post-construction monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
10.2 Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
10.2.1 Removal process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
11 Optimising the remedial works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
11.1 Identifying problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
11.2 Engaging appropriate specialist advice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
11.2.1 Emergency actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
11.2.2 Gathering further information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
11.3 Identification of deterioration mechanism and appropriate solutions . . . . . . . . . . . . . . . . . . . . 149
11.3.1 Identification of appropriate solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
11.3.2 Summary flowcharts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
11.4 Ranking feasible solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
11.4.1 Cost and risk summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
11.4.1.1 Construction cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
11.4.1.2 Construction risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
11.4.1.3 Operational cost (and benefit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
11.4.1.4 Operational risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
12 During and after works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
12.1 As-built records . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
12.2 Monitoring and surveillance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
12.3 Lessons learned . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Statutes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166

A1 Case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Boxes
Box 8.1 Application of mortar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Box 8.2 Completing patch repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Box 8.3 Guidance on crack repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Box 8.4 Objectives and considerations of concrete conduit repair . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Box 8.5 Selection of repair materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

xii CIRIA, C743

Box 8.6 Sealing and repairs of cracks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Box 9.1 Pipe material classifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

Case studies
Case study A1.1 Birkenburn reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Case study A1.2 Brent reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Case study A1.3 Brunclough reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Case study A1.4 Cofton reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Case study A1.5 Daventry reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Case study A1.6 Elton reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Case study A1.7 Harthill reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Case study A1.8 Knight and Bessborough reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Case study A1.9 Loch Lebgei and Dubh Loch reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Case study A1.10 Napton reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Case study A1.11 Pebley reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Case study A1.12 Redbrook reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Case study A1.13 Rudyard reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Case study A1.14 Shon Sheffrey reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Case study A1.15 Tardebigge reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Case study A1.16 Torside reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Case study A1.17 Un-named new (raw water storage) reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Case study A1.18 Upper Rivington reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Case study A1.19 Warmwithens reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Case study A1.20 Yeoman Hey reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Case study A1.21 Redmires reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217

Figures
Figure 2.1 Cast in situ concrete conduit constructed in the 1960s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Figure 2.2 Conduit constructed from precast concrete pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Figure 2.3 Flexible couplings on steel pipeline in conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Figure 2.4 Example of a CMP conduit in a 1970s flood storage reservoir, Lamorbey Park . . . . . . . . . . . 9
Figure 2.5 Stone masonry lined conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Figure 2.6 Brick masonry lined conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Figure 3.1 Ingress of fines into a non-pressurised conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Figure 3.3 Carbonation damage to concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Figure 3.4 AAR in a concrete dam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Figure 4.1 Deposits at drain outlet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Figure 4.2 Flail used to clean the inside of a metal conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Figure 5.1 Panoramo ROV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Figure 5.3 TITSCIT ROV for combined CCTV and sonar scans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42
Figure 5.4 The Schmidt hammer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Figure 5.5 Principle of ultrasound thickness measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Figure 5.6 Ultrasonic equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Figure 5.7 Impact-echo method – steel balls mounted on spring-steel rods . . . . . . . . . . . . . . . . . . . . . 44
Figure 5.8 Impact-echo NDT method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Figure 5.9 Sub-surface crack diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Figure 5.10 Example of pipeline inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Figure 5.11 A typical example of GWT data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Figure 5.12 Magnetic flux leakage equipment inside a large diameter mortar-lined steel pipe . . . . . . . 46
Figure 5.13 P-wave RFEC TC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Figure 5.14 32-channel line-cat crawler system using low frequency electromagnetic techniques
to scan pipes at up to 3 m/min . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Figure 5.15 Internal BEM equipment used for scanning ferrous pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Figure 5.16 Taking a 100 mm core through a brickwork tunnel lining . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Figure 5.17 Coring through a cast-iron underground tunnel lining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

Dams and reservoir conduits xiii

Figure 5.18 Compressive strength testing of a 300 mm diameter brick core while simultaneously
measuring strain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Figure 5.19 Installation of an array of temperature probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Figure 5.20 Temperature distribution resulting from a survey of an array of temperature probes . . . . . 51
Figure 5.21 Electro-scanning equipment, sonde and cable (a), principle of operation (b) and scan
showing variation of scan current with position and demonstrating repeatability (c) . . . . . 52
Figure 5.22 Diagram illustrating the AquaTrack™ system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Figure 5.23 Portable Willowstick mapping equipment combining magnetic sensors, GPS and laptop . 53
Figure 5.24 Summary interpretation, Shon Sheffrey dam, Wales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53
Figure 5.25 Robotic equipment combining CCTV, PPR and LiDAR for internal inspection of pipes,
sewers and conduits up to 900 mm dia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Figure 6.1 V-notch weir for monitoring flow rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Figure 6.2 Typical plot of piezometer readings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Figure 6.4 Arrangements of targets and of electronic distance measurement (EDM) measuring
stations along the tunnel axis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Figure 6.5 Horizontal magnet extensometer installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Figure 6.6 Liquid level transducer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Figure 6.7 In place inclinometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Figure 6.8 SAA installed inside sewer tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Figure 6.9 Tell-tales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Figure 6.10 Leica TM30 robotic total station . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Figure 6.11 Pillar monument with target prisms used for monitoring dam deformations . . . . . . . . . . . . 67
Figure 6.12 InSAR measurements using TerraSAR-X images, La Pedrera dam, Spain, July 2008 to
June 2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Figure 6.13 Crest deformation, La Pedrera dam, Spain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Figure 8.1 Hot tapping system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Figure 8.2 Pumping arrangement for reservoir draw-down . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Figure 8.3 Sliplined pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Figure 8.4 Spacers and grout pipes on pipe before insertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Figure 8.5 Damaged HDPE pipe due to high external grouting pressures . . . . . . . . . . . . . . . . . . . . . . . . 86
Figure 8.6 Viking-Johnson pipe connection on a PE pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Figure 8.7 Extensive rusting on a bolted pipework connection in a UK embankment dam . . . . . . . . . . 99
Figure 8.8 Typical patch repair to two courses of brickwork (a) and with pinning detail (b) . . . . . . . . . 104
Figure 8.9 Installation of stitching bars along a crack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Figure 9.1 Design process and main considerations for conduit replacement . . . . . . . . . . . . . . . . . . . 115
Figure 9.2 Sub-atmospheric pressure in siphon operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Figure 9.3 Siphon arrangement for reservoir lowering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Figure 9.4 Zebra mussel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Figure 9.5 Concrete pipes at a small UK embankment dam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Figure 9.6 Rigid pipe cradle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Figure 9.7 Concrete pipe on full width cradle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Figure 9.8 Full reinforced concrete encasement for flexible pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Figure 9.9 Stresses causing settlement and movement in conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
Figure 9.10 Shear forces following conduit construction within a trench . . . . . . . . . . . . . . . . . . . . . . . . 132
Figure 9.11 Recommended measures to reduce arching of fill over conduits . . . . . . . . . . . . . . . . . . . . 133
Figure 9.12 Typical anti-seepage collars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Figure 9.13 Typical conduit filter detail in a homogenous embankment dam . . . . . . . . . . . . . . . . . . . . . 134
Figure 9.14 Restricted access in conduit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Figure 9.15 Wachs cutter in operation before pipe replacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Figure 11.1 Process for optimum solution identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Figure 11.2 Decision flowchart for buried conduit normally containing water under pressure . . . . . . . 151
Figure 11.3 Decision flowchart for non-pressurised conduit (eg human-entry culvert, spillway pipe) . 152
Figure A1.1 Reservoir outlet pipe downstream valve, fitted with end plate and three glands, ready
for grouting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Figure A1.2 Site team preparing to start grouting of reservoir outlet pipe . . . . . . . . . . . . . . . . . . . . . . . 172

xiv CIRIA, C743

Figure A1.3 Cross-section at draw-off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Figure A1.4 Seepage at the head of the tunnel, and collection trough . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Figure A1.5 Installing temperature probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Figure A1.6 GRP liners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Figure A1.7 Bulging GRP liner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .176
Figure A1.8 Cross-section through draw-off works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Figure A1.9 Valve pre-works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Figure A1.10 CIP liner emerging from upstream end under water pressure . . . . . . . . . . . . . . . . . . . . . . . 178
Figure A1.11 Liner cured and trimmed, downstream end . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Figure A1.12 New valve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Figure A1.13 Cross-section through draw-off works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Figure A1.14 Leakage of defective joint three metres from the valve in the lower draw-off pipe . . . . . . 180
Figure A1.15 Joint seepage in draw-off pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Figure A1.16 Teapot lid valve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Figure A1.17 Seepage along the annulus between the liner and the CIP pipe . . . . . . . . . . . . . . . . . . . . . 182
Figure A1.18 Cross-section through draw-off works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Figure A1.19 Clay intrusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Figure A1.20 Attempts to monitor clay intrusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Figure A1.21 Repointed culvert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Figure A1.22 Cross-section through draw-off works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Figure A1.23 Cross-section through draw-off works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Figure A1.24 ROV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Figure A1.25 Use of ROV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Figure A1.26 ROV control position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Figure A1.27 Two existing 1.37 m riveted steel inlet pipes from the Walton treatment works into
Knight and Bessborough reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Figure A1.28 The constructed scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Figure A1.29 Completed trapezoidal channel cut through peat dam at Dubh Loch reservoir . . . . . . . . . 193
Figure A1.30 Completed trapezoidal channel cut through peat dam at Loch Lebgei reservoir . . . . . . . . 194
Figure A1.31 Final stage of work to cut trapezoidal channel through peat dam at Dubh Loch reservoir . . . 194
Figure A1.32 Cross-section through draw-off works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Figure A1.33 Valve tower . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Figure A1.34 Guard and service valves in valve tower . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Figure A1.35 Original valve in gunited culvert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Figure A1.36 Cross-section through draw-off works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Figure A1.37 Plug valve in the Neptune shaft in 1998 and 2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Figure A1.38 Augering machine in thrust pit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
Figure A1.39 Cross-section through draw-off works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Figure A1.40 Original upstream valve operating mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Figure A1.41 Dry well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
Figure A1.42 Cross-section through pre-1987 draw-off works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Figure A1.43 Cross-section through present day draw-off works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Figure A1.44 Access to the upstream end of the conduit during construction . . . . . . . . . . . . . . . . . . . . . 207
Figure A1.45 A 750 mm diameter outer concrete outlet pipe, laid on dry-mix concrete haunching . . . . 209
Figure A1.46 Concrete anti-seepage collars along the conduit length . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Figure A1.47 Breach of the embankment along the line of the recently constructed draw-off tunnel . . 213
Figure A1.48 Reservoir leakage through the cracked brickwork culvert . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Figure A1.49 Ochre deposited on invert of culvert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Figure A1.50 Upstream end of scour pipes in access shaft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Figure A1.51 Remedial works under way – note the black HPPE pipe inside the original CI pipe . . . . . . 218
Figure A1.52 Damage to structural lining pipe (note bulge in top right of pipe) caused by excessive
grout pressures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Figure A1.53 Link seal blown out of upper end of pipe by high grout pressures . . . . . . . . . . . . . . . . . . . . 219

Dams and reservoir conduits xv

Tables
Table 2.1 Properties of some old bricks used in bridge and tunnel construction . . . . . . . . . . . . . . . . . 13
Table 2.2 Statistical analysis of properties of brick samples from old railway structures . . . . . . . . . . 13
Table 2.3 Comparison of typical strength and density of some common UK
building stones with other construction materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Table 2.4 Mortar mixes and compressive strengths used in the UK and corresponding strengths
of masonry using different bricks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Table 2.5 Advantages and disadvantages of upstream control facilities . . . . . . . . . . . . . . . . . . . . . . . . 17
Table 3.1 Failure mode identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19
Table 4.1 Failure mode detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Table 5.1 Degradation of conduit materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Table 6.1 Instrumentation for monitoring conduits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Table 8.1 Sliplining – technical envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Table 8.2 Sliplining – site specific issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Table 8.3 Sliplining – summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Table 8.4 Close fit pipe lining – technical envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Table 8.5 Close fit pipe lining – site specific issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Table 8.6 Close fit pipe lining – summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Table 8.7 CIPP lining – technical envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Table 8.8 CIPP lining – site specific issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Table 8.9 CIPP lining – summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Table 8.10 In situ sprayed lining – technical envelope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Table 8.11 In situ sprayed lining – site specific issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Table 8.12 In situ sprayed lining – summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Table 8.13 Material types used in concrete patch repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Table 8.14 Grout types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Table 8.15 Impact of dam movement on structural modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Table 9.1 Conduit replacement on same alignment – summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117
Table 9.2 Conduit replacement with pumps – summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Table 9.3 Conduit replacement with siphons – summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Table 9.4 Structural classification of common pipe types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Table 10.1 Conduit abandonment – summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Table 11.1 Cost and risk – comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Table A1.1 List of case studies by topic(s) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

xvi CIRIA, C743

Glossary

The formation of vapour bubbles in water when the static pressure falls
Cavitation 
below the vapour pressure of water (which can occur inside certain types
of pumps and ejectors). When the bubbles move to areas of higher pressure
they may implode, causing shockwaves, which can damage the internal
components of pumps and ejectors.

A means to convey water from a reservoir through, under, or around an

Conduit 
embankment dam in a controlled manner or give access to the draw-off
works and monitoring instrumentation.

A closed conduit carrying a watercourse beneath an obstruction such as a

Culvert 
road, railway or canal. The term ‘closed’ implies that a culvert has a hard
soffit and invert. The term conduit implies the conveyance of water some
or all of the time, but excluding tunnels and underpasses for vehicles,
pedestrians and animals.

Glipp joints The Glipp system is a proprietary system for joining lengths of concrete pipe.

A fracture in a formation initiated by the injection of water or grout under a

Hydraulic fracture 
pressure in excess of the local formation tensile strengths. The propagation
of a hydraulic fracture causes the formation to move away from the water or
grout-filled fracture.

The movement of soil particles as a result of chemical actions and/or

Internal erosion 
unbalanced seepage forces produced by percolating water.

A small ‘open caisson’ shaped to fit a dock wall or other concrete structure,
Limpet dam 
generally the draw-off tower in the case of a reservoir. It is lowered by crane
into the water in front of the conduit inlet and in contact with the concrete
structure. It is then and pumped out so that the external water pressure
forces it into close contact with the concrete at the edges and holds it in place.
Access is through the open top.

The work of a mason, strictly referring to work in stone, but commonly used
Masonry 
to refer generally to work in either brick or building stone, as it is here.

The procedure of measuring effluent characteristics such as flows or

Monitoring 
pollutants by means of instruments.

A device inserted into a drill hole that acts to prevent return of the fluid to
Packer 
be injected around the injection pipe and through which the injection pipe
passes. Usually an expandable device activated mechanically, hydraulically
or pneumatically.

This term has often been used generically in literature to describe various
Piping (FEMA) 
erosion processes, not all of which hold to the classic definition of the term
piping. Piping in the classic sense is characterised by the formation of an
open tunnel that starts at a downstream seepage exit point and progresses
back upstream toward the reservoir. This classic type of piping is often
known as ‘backward erosion piping’ and in this document, the term ‘piping’
will used for this type of piping.

Standard dimensional The ratio of the pipe diameter to wall thickness.

ratio (SDR)

Dams and reservoir conduits xvii

 The migration of soil particles through the soil matrix driven by flow through
Suffosion/suffusion 
the soil. Suffusion is a contributor to the manifestation of internal erosion.

The recording of information at a given moment in time eg the number of a

Surveillance 
particular species.

A grout pipe perforated with rings of small holes at specified intervals,

Tube-à-Manchette 
which is annulus (sleeve) grouted into a drill hole. Each ring of perforations
is enclosed by a short rubber sleeve fitting tightly around the pipe so as to act
as a one-way valve when used with a dold ended packer with perforated mid-
section that isolates a stage for injection of the grout.

The Victualic system is a propriety mechanical system for joining lengths of

Victualic joints 
steel pipe.

xviii CIRIA, C743

Abbreviations and acronyms

AAR Alkali-aggregate reaction

ABS Acrylonitrile butadiene styrene
AC Alternating current
ACOP Approved code of practice
AONB Area of Outstanding Natural Beauty
ARPE All Reservoirs Panel Engineer
ASCE American Society of Civil Engineers
ASSI Area of Special Scientific Interest (Northern Ireland)
BDS British Dam Society
BEM Broadband electromagnetic system
BRE Building Research Establishment
C&C Cut-and-cover
CaO Calcium oxide
CBDG Concrete Bridge Development Group
CCTV Closed circuit television
CDM Construction (Design and Management) Regulations 2015
CEATI Centre for Energy Advancement through Technological Innovation
CI Cast iron
CIP Cured-in-place
CIPP Cured-in-place pipe
CMP Corrugated metal pipe
COSHH Control of substances hazardous to health
CRT Canal & River Trust
Defra Department for Environment, Food and Rural Affairs
DI Ductile iron
DTM Digital terrain model
DWI Drinking Water Inspectorate
EDM Electronic distance measurement
FEMA Federal Emergency Management Agency (USA)
FERC Federal Energy Regulatory Commission (USA)
FoS Factor of Safety
GCL Geosynthetic clay liners
GPa Gigapascal
GPR Ground penetrating radar
GPRS General packet radio service
GPS Global positioning system
GRP Glass reinforced plastic
GWUT Guided wave ultrasonic testing
HDS High-definition surveying
HDPE High density polyethylene

Dams and reservoir conduits xix

HSE Health and Safety Executive
ICE Institution of Civil Engineers
ICOLD International Commission On Large Dams
ID Internal diameter
IEEE Institute of Electrical and Electronic Engineers
IGN Information and guidance notes
InSAR Interferometric synthetic aperture radar
kPa Kilopascal
LFET Low frequency electromagnetic technique
LiDAR Light Detection And Ranging
LNR Local nature reserve
LVDT Linear voltage differential transformer
MDPE Medium density polyethylene
MEMS Micro-electro-mechanical systems
MFL Magnetic flux leakage
MPa Megapascal
NCE New Civil Engineer (journal)
NDT Non-destructive testing
NIEA Northern Ireland Environment Agency
NRW Natural Resources Wales
O&M Operation and maintenance
OD Outside diameter
OPC Ordinary Portland cement
PCCP Prestressed cylinder concrete pipe
PE Polyethylene
PFA Pulverised fuel ash
PFR Prescribed form of record
PIV Pipe inspection vehicle
PPE Personal protective equipment
PPG Pollution prevention guidelines
PPR Pipe penetrating radar
PSG Project steering group
PVC Polyvinyl chloride
QCE Qualified civil engineer
RFEC Remote field eddy current testing
ROV Remotely operated vehicle
RTK Real-time kinetic GPS
SAA Shape accelerator array
SAR Synthetic aperture radar
SASW Spectral analysis of surface waves
SBR Styrene butadiene rubber
SCADA Supervisory Control And Data Acquisition
SCUBA Self-contained underwater breathing apparatus
SDR Standard dimensional ratio
SEPA Scottish Environment Protection Agency

xx CIRIA, C743
SPA Special Protection Area
SRC Sulphate resisting cement
SSSI Site of Special Scientific Interest
TISCIT Totally integrated sonar and camera inspection technique
TWL Top water level
uPVC Unplasticised polyvinyl chloride
USACE United States Army Corps of Engineers
USBR United States Bureau of Reclamation
USEPA United States Environmental Protection Agency
USSD United States Society on Dams
UU United Utilities
UV Ultraviolet
WIMES Water Industry Mechanical and Electrical Specification
WIS Water Industry Specifications
WRc Water Research Centre

Dams and reservoir conduits xxi

xxii CIRIA, C743
1 Introduction
1
This chapter reviews the need for a guidance
document for reservoir owners regarding
Note
The term ‘owner’ is used throughout this document to include
2
problems associated with conduits beneath or reservoir undertakers, which is the legal term under the

3
within earth embankment dams, which could lead Reservoirs Act 1975, and operators or agents who may have
control of the dam on behalf of the owner.
to severe leakage and dam failure. The scope of
the guide is described, types of conduit are considered and issues associated with conduits are discussed.

1.1 BACKGROUND
The Reservoir Safety Research and Development Strategy (Atkins, 2009) identified priority areas
4
for research projects aimed at drawing together best practice, operational experience and recent
developments in the management of existing UK dams. One important research area identified was the
development of guidance for the inspection, monitoring, maintenance and repair of tunnels in dams and
reservoirs. The Environment Agency commissioned a scoping study to identify whether there was a need
5
for additional research. After consultation, the study’s scope was extended to include all conduits (Pether

6
and Rowland, 2009).

Defects in and associated with, conduits through dams have resulted in dam failure. The most dangerous
mechanism of failure is internal erosion of an embankment dam’s fill by water passing along the contact
between the conduit and the fill. This has long been recognised as a potential failure mode, yet early
detection of defects can be difficult. Structural defects in conduits can lead to an increased risk of
internal erosion by a number of mechanisms, can put at risk the operational efficiency of the reservoir 7
and affect the availability of emergency draw-down facilities.

Numerous documents are already available giving specific guidance on the safety assessment of
embankments dams (eg BRE, 2009), concrete and masonry dams (eg Kennard et al, 1996), and of valves
and pipework (eg Reader et al, 1997). These documents contain some reference to safety assessment of
8
conduits, but there is no document that provides comprehensive guidance on the safety assessment of

9
conduits through dams.

1.2 SCOPE
This guide is aimed primarily at the assessment of, and remedial works to, existing conduits. It is focused
on existing conduits in UK conditions and covers design issues in so far as they influence the behaviour
10
of an existing conduit, or as part of remedial works. Some overseas information has been included where
it was considered applicable to UK conditions.

A conduit in this context is considered to be a means to convey water from a reservoir through, under,
or around an embankment dam in a controlled manner or to give access the draw-off works and
11
monitoring instrumentation.

This guide has been prepared to provide guidance on reservoir specific issues and covers the following:

‰‰ description of types of conduit commonly associated with UK reservoirs

12
‰‰ description of failure modes for reservoir conduits
‰‰ visual inspection techniques
‰‰ monitoring techniques
A1
Dams and reservoir conduits 1
‰‰ non-intrusive investigations
‰‰ intrusive investigations
‰‰ maintenance
‰‰ replacement
‰‰ abandonment
‰‰ record keeping.

1.3 CONDUIT TYPE AND USES

The conduits considered here are primarily those associated with embankment dams. Conduits have
varying forms and uses and vary from site to site because each dam is unique.

The materials used to construct an embankment vary markedly from site to site as does the size, form
and location of conduits within the embankment, foundation, abutment.

Stonework conduits and, in some cases, composite brickwork and stonework construction were popular in
Victorian times. Later reinforced concrete placed in situ and even precast concrete was used. There are
also cases where clay pipes and timber have been used for conduits as well as steel and cast iron.

Some conduits are backed with or surrounded by concrete and in many cases the conduit is surrounded
by puddle clay.

Pipelines are also a type of conduit and can be laid within the fill of the dam, around the dam or in the
foundation. Pipes come in a variety of sizes and materials including timber, cast iron (CI), ductile iron
(DI), steel, stainless steel, plastic, precast concrete etc. They can also be laid within a culvert or tunnel.

The principal types of conduit in contact with the natural ground or the dam fill are:

‰‰ tunnels driven through natural ground or fill

‰‰ cut-and-cover (C&C) culverts in the foundation of the dam
‰‰ culverts located in the dam foundation and covered with dam fill
‰‰ pipelines in trenches in the dam foundation
‰‰ pipelines in the dam fill
‰‰ pipelines in a culvert or tunnel.

The arrangement where the conduit is incorporated in a tunnel through one of the abutments is not
included by this guide.

The principal uses of conduits are to provide:

‰‰ outlets from overflows – these usually flow part full, but under some conditions may flow full and
under pressure
‰‰ scour or bottom outlets – these may flow under pressure in pipes to discharge water downstream
of the dam, but may also discharge into larger tunnels or culverts passing through the dam or its
foundation
‰‰ draw-off pipes – these normally flow under pressure. They may be laid directly within a dam or its
foundation, or may be located within a dry culvert or tunnel
‰‰ flood regulation and control – to release waters temporarily stored
‰‰ stream diversion during construction
‰‰ power penstocks conveying water to turbines
‰‰ access.

2 CIRIA, C743
Conduits generally can be classified into two types:

1 Pressure flow: where there is pressurised flow throughout the conduit length to the point of
regulation, control or the end of the culvert. 1
2 Non-pressurised flow: where flow is at atmospheric pressure for all or part of the conduit length.

1.4 ISSUES ASSOCIATED WITH CONDUITS 2

1.4.1 Types of failure
Conduits have been constructed through dams for centuries. Defects in and associated with, conduits
through dams have resulted in dam failure. If conduits are not designed and constructed, inspected and
3
maintained correctly, there can be an increased probability of failure.

The most dangerous mechanism of failure is internal erosion of an embankment dam’s fill by water
passing along the contact between conduit and the fill. This has long been recognised as a significant
4
potential failure mode, but early detection can be difficult. However, there are other failure modes

5
worthy of consideration.

Fracture of a conduit could release water under pressure or provide a seepage path into the conduit
causing internal erosion or piping failure. Structural failure of a conduit, particularly a culvert, could
result in displacement of part of the structure and form a passage for water along the conduit/fill interface.
Conduit failure could damage scour outlet facilities and prevent them operating in an emergency.

The main issues associated with problems with culverts include:

6
‰‰ differential settlement and embankment spreading, causing cracking and joint opening allowing
the ingress or egress of water and fill
‰‰ differential settlement leading to reduced soil stresses and potential for hydraulic fracture
7
‰‰ seepage/erosion along the culvert

8
‰‰ deterioration of the materials.

1.4.2 Legislative aspects

9
Many conduits will be associated with reservoirs that come under reservoir safety legislation, the
Reservoirs Act 1975, applied to reservoirs designed to hold or capable of holding more than 25 000 m3
above the land adjoining the dam. The Floods and Water Management Act 2010 proposed changes to
the Reservoirs Act 1975 including the reduction of the threshold volume that the Act applies to 10 000
m3. The reduction in reservoir volume, at the time of writing, has not yet been implemented. Where
works on such reservoirs might affect their safety, the owner should normally appoint a qualified civil
engineer (QCE) to supervise the design and construction of such works.
10
The Reservoirs Act 1975 (as amended by the Water Act 2003 and the Flood & Water Management
Act 2010) applies to England but does not apply to Northern Ireland, which is introducing its own
legislation. Wales will be introducing its own legislation, and at the time of writing, the un-amended
Reservoirs Act, 1975 still applies as it does in Scotlland where the Reservoirs (Scotland) Act 2011 has
11
not yet been implemented.

In respect of planning law, reservoir safety works relating to conduits fall under ‘permitted
development’, as there is usually limited visual impact of the works. However, if the reservoir falls within
12
environmentally protected areas, such as National Parks, Sites of Special Scientific Interest (SSSI)/

A1
Areas of Special Scientific Interest (ASSI) or Special Protection Areas (SPA), Areas of Outstanding
Natural Beauty (AONB), Ramsar sites (wetlands of international importance) and local nature reserves
(LNR), close liaison may be required with the relevant stakeholders during planning of the works, and

Dams and reservoir conduits 3

constraints may apply to the works. Listed Building status and fishing interests can also be constraints in
some cases.

Where a reservoir impounds a watercourse, which contributes to a known flooding problem within the
region, the planning authority is likely to be particularly interested in the works to be completed on a
spillway conduit, and in any proposed reservoir draw-down. Consent from the Environment Agency to
discharge into the downstream watercourse is required, despite permitted development exclusions.

In all regions of the UK, legislation pertaining to protected species and invasive species may constrain
work on a reservoir site. Common issues include:

‰‰ birds’ nests
‰‰ protection of bats
‰‰ protection of badgers
‰‰ protection of other species (great crested newts, water voles, white-clawed crayfish etc).

Particular legislation that could effect on works to conduits includes:

‰‰ The Conservation (Natural Habitats &c) Regulations (amendment) 2007.

‰‰ The Wildlife and Countryside Act 1981.
‰‰ The Countryside and Rights of Way Act 2000.
‰‰ Natural Environment and Rural Communities Act 2006.
‰‰ The Protection of Badgers Act 1992.
‰‰ The Wild Mammals (Protection) Act 1996.
‰‰ The Animal Welfare Act 2006.
‰‰ Pollution prevention guidelines (PPG) published jointly by the Environment Agency, Scottish
Environment Protection Agency (SEPA) and the Northern Ireland Environment Agency (NIEA).

Currently there is no hierarchy of legislation, so the Reservoirs Act 1975 cannot overrule any of the
legislation listed here. It is up to the reservoir owner and other parties to negotiate an arrangement,
which attempts to comply with all legislation.

When planning works to conduits reservoir owners should also take into account guidance from PPG5
(EA/NIEA/SEPA, 2007) and PPG6 (EA/NIEA/SEPA, 2012).

Such works may also come under the definition of ‘construction works’ under the Construction Design
and Management (CDM) Regulations 2015, and so these regulations and other health and safety
regulations may also apply.

1.4.3 Downstream discharges

Where functional and available for use, scour and draw-off pipework often discharge water stored in
the lowest reaches of the reservoir. It is important to note that the beds of most old reservoirs in the
UK are likely to retain a depth of silt. The amount and location of the silt within the reservoir depends
on site conditions, but some is likely to be deposited at the upstream toe of the dam and around any
draw-off facilities.

Once flow starts through the draw-off pipe, high velocities local to the inlet of these pipes can cause
the suspension of silt near to the inlet. Where the inlet has been buried by the silt, large volumes
can be mobilised through the pipe. Such silty flow can cause significant environmental damage in
watercourses downstream, so discharge of silty flow from the reservoir through the scour or draw-off
pipework require consent from the applicable regulating authority (the Environment Agency, Natural
Resources Wales [NRW], SEPA, and NIEA). Such consent can require extensive planning to mitigate
the impact of any released silt, and at some sites it may not be possible to use the scour pipework to
draw down the reservoir.

4 CIRIA, C743
New guidance is currently being developed by the Environment Agency on drawdown capacity for

1
reservoir safety and emergency planning, which is due to be published at the end of 2015. Section 2.4 of
the guide will give guidance on environmental permitting requirements associated with operating low
level outlets. It will advise when a permit is required and outline the process to obtain one depending
on whether the reservoir is in England, Scotland or Wales. It will also recommend good practice for
operating outlets to minimise environmental impacts. For more information refer to EA (2013).

It is noted that the Water Industry Act 1991 requires water companies to obtain consent before 2
discharging into the watercourse downstream. However, the Act does not apply to other organisations
such as the Canal & River Trust (CRT). The Environment Agency has the power to prosecute for
polluting discharges. CRT uses its discretion as to when to discuss discharges with the Environment
Agency and other organisations would be expected to apply the same approach. 3
4
5
6
7
8
9
10
11
12
A1
Dams and reservoir conduits 5
2 Conduits

A review of the historic development of conduits has been included in Section 1.3. This chapter considers
the shape of conduits through dams, the materials from which they are constructed, the associated outlet
and control structures, and the influence of design on performance.

2.1 CONDUIT FORMS

Conduits through dams have been designed and constructed using a range of cross-sectional
shapes, including:

‰‰ circular, eg pipes
‰‰ ‘D’ shape, eg early masonry culverts
‰‰ modified ‘D’ where the straight side walls slope inwards to provide a narrower invert
‰‰ horseshoe-shaped that optimises the cross-sectional areas
‰‰ square or rectangular, eg precast concrete culverts
‰‰ egg-shaped culverts, eg early brick culverts.

Many culverts are straight, but bends were sometimes incorporated to align the conduit with an
original watercourse.

Most culverts incorporate a slight fall from the upstream to downstream at a uniform gradient to ensure
that they are free-draining to avoid sediment deposition.

2.2 CONDUIT MATERIALS

Various materials have been used for the construction of conduits through dams and a summary of the
historic use of materials has been included in Section 1.3. The selection of the materials for conduits
would have been influenced by cost, availability, maintenance and buildability. The most common
materials used in the construction of conduits are:

‰‰ concrete – reinforced, cast in situ and precast

‰‰ metal – steel, DI, CI, and corrugated metal pipe, eg ‘Armco’ culverts.
‰‰ plastic – thermoplastic and thermoset glass reinforced plastic (GRP)
‰‰ masonry – stone and brickwork.

The strength and performance of these conduits is influenced by the chemical make-up of their
constituent materials. Each material has certain design, construction and maintenance advantages and
disadvantages, which can depend on a conduit’s particular situation and use.

2.2.1 Concrete
Concrete conduits can be constructed from either reinforced, cast in situ or precast concrete. Portland
cement was invented in the 1820s, but took some time to be produced with a consistent quality and for
widespread application. Reinforced concrete started to replace masonry as the material of choice for
tunnels in about 1900 onwards (McKibbins et al, 2009). In dam and reservoir conduits, the transition
from masonry to reinforced concrete appeared to take some time with reinforced concrete used as
early as 1911, but masonry still used up to about 1940. The choice of materials was probably governed

6 CIRIA, C743
by the availability and cost of materials in conjunction with labour costs. Unreinforced concrete or a

1
combination of masonry and concrete has also been used.

2.2.1.1 Reinforced cast in situ concrete

2
Reinforced cast in situ concrete
conduits have been used extensively
in dams. They are versatile and can
be designed to satisfy specific project
and site conditions, allowing for
different conduit sizes and shapes. A
properly designed and constructed 3
reinforced cast in situ concrete
conduit has a service life of at least
100 years. It should be remembered
that dams can have lives longer than
this and it might be appropriate to
4
apply techniques that can extend the

5
design life.

Figure 2.1 Cast in situ concrete conduit constructed in the 1960s

(courtesy Severn Trent Water)

2.2.1.2 Precast concrete pipes

6
Precast concrete pipes are manufactured throughout the UK and are extensively used for sewerage and
stormwater drainage. They are manufactured in a wide range of diameters (typically 300 mm to 2400
mm), are often lower cost than alternatives for these diameters, and so are suitable for use as spillway 7
pipes at small UK embankment dams. They are not suitable for the long-term transfer of pressurised
flow, although they have been used to contain a pressurised pipe inside in case of bursting of the internal
pipe. They are not designed for connection to valves, but penstocks can be fitted to a headwall at the end
of a length of pipe (and within manholes where required). 8
Precast concrete pipes typically

9
use spigot and socket joints,
which provide flexibility to the
pipeline. These are typically sealed
using a flexible sealing ring of

10
rubber or similar material. Some
manufacturers integrate the ring
into the pipe socket, making the joint
less susceptible to bad workmanship
during installation. This is known as
a ‘Glipp’ system, is readily available
from major manufacturers and 11
promises enhanced sealing.

12
Precast concrete pipes have very long
life expectancy, with manufacturers
Figure 2.2 Conduit constructed from precast concrete pipes (courtesy claiming a 120-year design life.
D A Bruggemann)

A1
Dams and reservoir conduits 7
2.2.2 Metal
Types of metal pipes used in the construction of conduits are:

‰‰ cast iron (CI)

‰‰ ductile iron (DI)
‰‰ steel
‰‰ corrugated metal pipe.

Most pipes in UK dams are made from CI, DI or steel.

2.2.2.1 Grey CI and DI

The difference between grey CI and DI is the difference in the type or structure of the carbon (graphite)
present in them. They are prepared by adding carbon in the hot beds where they are liquefied, but in
the case of DI, magnesium is also added. The graphite in CI has a rod-like structure or a random flaky
pattern, whereas the graphite in DI is shaped like nodules or spheres. Due to the spherical shape of the
graphite in DI, its molecules have less discontinuities and give rise to its ductility.

Vertically cast grey CI pipes were manufactured from the early 1800s until the early 1950s. Many of the
early pipes were of questionable material quality, had variable cross-section and contained inclusions and
laminations. However, there are many CI pipes still functioning satisfactorily more than 150 years after
their installation. The pipes either had flanges cast onto them or were of the spigot and socket type with
lead-run joints.

In the early 20th century spun CI pipes began to be manufactured by centrifugal casting. Molten iron is
introduced into a rotating mould, which generates a centrifugal force that holds the iron in place against
the mould until it solidifies. This process gave a much more consistent product. The first standardisation
of sizes of CI water pipes in Britain occurred through the publication of BS 78-2:1965. Spun DI, which
has greater strength and resilience than grey iron, was introduced in the 1960s, and is now the only iron
product available. From the 1970s onwards pipes and flanges have been produced in metric sizes.

In the 19th century, bends, junctions and tapers were often produced specifically for a particular dam,
but with the introduction of DI, fittings became standardised. DI pipes are manufactured in metric sizes
from 80 mm to 1600 mm diameter as standard. The pipes are either flanged or have spigot and socket
joints with rubber sealing rings and can be anchored for greater security. DI pipes can also be jointed
together with steel sleeve fittings such as Viking-Johnson couplings and can be fitted to valves using cast
flanges or Viking-Johnson flange adapters. These steel fittings can provide additional flexibility to the
pipework system, but care needs to be taken over corrosion protection and pipe movement.

Early CI pipes were dipped in proprietary tar or bitumen solutions, but their coating could become
damaged. DI pipes are generally provided with a zinc spray coating before coating in bitumen and are
often cement lined with ordinary Portland cement (OPC) or sulphate resisting cement (SRC) mortar.
Polythene sleeves can also be applied at the works or can be fitted on installation in the ground.

Grey CI generally exhibits good corrosion resistance and has been used extensively in British dams until
the late 1960s.

DI pipes have been used extensively in potable water transfer in the UK and are common at many
embankment dams. As with precast concrete pipes, they have rigid walls. DI pipes can be used for the
transfer of pressurised flows, and pipes can be easily fitted to DI fittings (eg valves) using flanged joints.
DI fittings of standard sizes are very common and are used as standard in the water industry – the ability
of DI pipework to connect to DI fittings without the use of flange adaptors is of significant benefit.

DI is not as brittle as CI and DI pipes have very long life expectancy – over 100 years if overloading and
very corrosive ground are avoided.

8 CIRIA, C743
DI pipes are used for pressurised and unpressurised flow in embankment dams. They can be

1
manufactured in any lengths, but are commonly made in lengths of around 6 m. DI pipes can be cut by
a circular saw on site to suit changing conditions. However, this could lead to a defective joint if poorly
completed and should be avoided through careful survey before design.

2.2.2.2 Steel
Since the 1940s steel pipes have been
2
used for most pipelines with a diameter
larger than about 1200 mm (see Figure
2.3). Where CI pipework requires
replacement within towers and tunnels, 3
steel pipe is often used as it is lighter
than DI and special fittings can easily be
replicated in this material.

Steel pipes are manufactured from steel

4
ingots that are rolled into plate and

5
are shaped and welded longitudinally
and circumferentially or long strips are Figure 2.3 Flexible couplings on steel pipeline in conduit (courtesy
Severn Trent Water)
spirally welded to form a pipe. In the UK
steel pipe is manufactured in diameters from 80 mm to 4000 mm in standard lengths from 6 m to 13 m,

6
although lengths up to 18 m are available. Joints are flanged, spigot and socket, ‘Victaulic’ joints, or Viking-
Johnson fittings and adapters. Bends, tapers, junctions etc are made from cut and welded pipes.

The early steel pipes were coated in bitumen for corrosion protection and, as with cast-iron pipe, cement
lining was later developed. Modern protection systems include galvanising, proprietary paint systems,
epoxy, polymer or powder coating and externally steel pipes can be coated or wrapped with bitumen
or coal tar or sleeved with plastics. Steel pipes are also often used to form a conduit, which is then
surrounded in reinforced concrete. Cathodic protection is also commonly used to help prevent corrosion
7
of steel pipework.

In the last 20 years stainless steel pipework has become competitively priced and is now often used
instead of coated steel.
8
2.2.2.3 Corrugated metal pipes (CMP) and culverts
CMP (also known as Armco pipe in the UK) is manufactured from sheet steel that is corrugated for a
9
greater bending resistance and generally galvanised or polymer coated and rolled into pipe lengths.
The pipes are produced in diameters from 150 mm to about 7 m and in lengths typically between
7 m and 14 m. Jointing is by metal
banding and non-destructive testing 10
(NDT) is required for inspection of
the pipes. Particular attention needs
to be paid to the quality of welding.
The pipes are usually round, but can
be formed into other shapes. This
11
material is generally not used for

12
pressures pipes. The design life is
quoted as 50 years however, because of
its association with many dam failures,
the US Bureau of Reclamation (USBR)

A1
does not permit the use of CMP in its
embankment dams (FEMA (2005).
Figure 2.4 Example of a CMP conduit in a 1970s flood storage
reservoir, Lamorbey Park (courtesy D A Bruggemann)

Dams and reservoir conduits 9

This type of pipe was extensively used for land drainage and flood alleviation schemes in the 1970s
due to their hydraulic profile and ease of construction. Figure 2.4 shows an example at a flood storage
reservoir in the UK constructed in the 1970s.

2.2.3 Plastic
For the construction of new conduits in significant and high hazard embankment dams, plastic pipes
are encased in mass or reinforced cast in situ concrete to assure the quality of compaction against the
conduit. The service life of plastic pipe is about 50 to 100 years, which is less than concrete. Plastic pipes
might be preferred in aggressive soil conditions where water or soil could attack concrete. If used without
concrete surround, special measures should be taken to ensure good compaction all around the pipe and
prevention of erosion along the soil/pipe interface.

Plastic pipes used in construction and renovation of conduits are constructed from thermoplastic and
thermoset plastic, which consists of resins composed of polymerised molecules mixed with lubricants,
stabilisers, fillers and pigments. Types of plastic materials are discussed in the following sections.

Existing conduit linings can be renovated using plastic pipe and this is discussed in more detail in Chapter 8.

2.2.3.1 Thermoplastic
Polyethylene (PE) is a type of thermoplastic, meaning that it can be softened through the application
of heat, and then hardened by cooling, without causing deterioration to the material. Other types (less
commonly used in embankment dams) include polypropylene and polybutylene. Polyvinyl chloride
(PVC) is also a thermoplastic, which is commonly used within dam drainage systems.

PE pipes are extensively used in the UK water industry for water supply and sewerage. Pipes used for
these purposes have solid walls, and typically come in sizes ranging from less than 100 mm up to around
1200 mm (outside diameter, OD). There are various strengths and densities of PE available, commonly
referred to as low, medium or high density polyethylene pipes (MDPE/HDPE). HDPE is most commonly
used where high structural strength is required.

In all except the smallest sizes of PE pipe (less than 25 mm) the ratio between wall thickness and
outside diameter remains constant for a given pressure rating of the pipe. This relationship is called the
standard dimensional ratio (SDR). As the SDR reduces, the pipe wall thickness increases.

Solid wall PE pipes have been used extensively at UK reservoirs to slipline existing iron pipes or masonry
brickwork/culverts that have deteriorated. They have been used much less as stand-alone conduits
through the impermeable zone of the dam. This may be due to a preference of designers for more rigid
pipes, which have a longer track record of use in UK dams, and concern relating to the interaction of a
flexible pipe system with the surrounding fill material, particularly through the impermeable zone.

PVC pipe sections can be connected using socket and spigot joints, which can be a major cause of
water leaks for these types of pipes. A socket and spigot joint has the potential for leakage due to poor
installation or due to separation as the embankment settles and spreads laterally. Joints need to be tested
carefully when using these types of pipes to ensure that the gasket has not rolled out during installation.
Joints can also be formed with cemented (glued) connections. On straight pipes, double socket couplers
are required to join a length of pipe, and fittings (branches, bends etc) are provided with a socket to
receive the connecting pipe. The appropriate cement (solvent) and good quality workmanship are
required to ensure watertight joints are formed. Testing as for socket and spigot joints should be carried
out when cemented joints are used.

Solid HDPE and MDPE pipe sections can be joined using the electrofusion method, which is a rapid
jointing technique mainly used on site for an efficient method of assembly for pipes, fittings and
prefabricated sections. The system incorporates couplers to join sections together and an electrofusion

10 CIRIA, C743
control box, which carries out the welding process. HDPE and MDPE pipes can also be butt-welded,

1
which is an economical jointing technique. Butt-welding is well suited to prefabricating pipe sections and
for making special fittings.

2.2.3.2 Twin wall PE

In addition to solid wall pipes, structured wall HDPE pipes are now commonly used to transfer and store
stormwater. Such structured wall systems allow lightweight pipes of large diameters to be used as the
2
pipes exhibit good bending resistance to external loading due the structured wall. The use of such pipes
for reservoir conduits needs careful detailing to deal with the corrugated profile on the outside face of
the pipe.
3
2.2.3.3 Thermoset plastic
Thermoset plastics cannot be reformed because of their rigid nature after manufacturing and curing.
Thermoset plastics commonly used for lining of non-pressurised conduits are cured-in-place (CIP). A
cured-in-place pipe (CIPP) is also called an ‘elastic sock’. CIPP is usually formed of polyester needles-
4
felt or glass fibre-felt reinforcement pre-impregnated with a suitable polyester resin. This process is

5
usually done at the factory for quality control purposes. The inner surface of the CIPP liner is generally
a coated membrane of polyester, PE, surlyn, or polyurethane, depending on the type of application. The
membrane provides a hydraulically efficient inner surface and low friction to the CIPP liner.

2.2.3.4 Glass reinforced plastic (GRP)

The pipes are formed by a technique in which glass fibre, thermosetting resins and aggregate are
6
introduced in layers into a rotating mould or mandrell. Pipes are available in diameters between 800
mm and 4000 mm and in lengths up to 12 m. The main advantage of GRP pipe is that it is durable,
corrosion resistant, lightweight and has a low friction coefficient. The joints are either socket-and-spigot,
double socket or flanged. A full range of fittings are manufactured and some ‘specials’ can be produced
7
to order. Due to its light weight the pipes can be relatively easily handled in confined spaces. It can also
be used to slipline existing pipes. Design life is quoted at around 50 years, but there have been many
problems with its use and care in its adoption is advised.
8
2.2.4 Masonry
In this guide the term masonry is
used to cover both brick and stone
construction (see Figure 2.5).
9
10
During the rapid development of
canal, rail and sewerage systems in
the 19th century, brick and stone
were the predominant material

11
used in the construction of linings
for tunnels. The techniques for
constructing conduits in C&C
conditions or on the dam foundation
for burial within the embankment
are likely to have followed the
practice for lining tunnels. 12
Most of the linings were of brick Figure 2.5 Stone masonry lined conduit (courtesy Dŵr Cymru)
because in most parts of the UK
easily-won stone was rare, while the availability of clays for making bricks meant that brick was generally
the material of choice. However, in some parts of the country such as the Pennines, Scotland, Wales and
A1
Dams and reservoir conduits 11
south-west England, there were plentiful supplies of stone and sandstone, so stone linings tended to be
used in these areas. Composite stone and brick masonry linings were also constructed. Often the clay
spoil from the excavations was used for making the bricks. Also, bricks were small, easily transported
and handled, which made them ideal for construction inside confined spaces. Generally, it was necessary
to construct several skins of brick in sidewalls and multiple rings in arches to achieve the necessary lining
strength and watertightness.

When stone was used as the lining material, the linings were typically evenly dressed stonework with 3 mm
joints. The form of construction was generally a single ring of relatively large and heavy, regularly cut and
shaped stone blocks with thin mortar joints. Composite or hybrid linings were developed because it was
difficult to handle and place the large stone blocks, especially in the higher parts of the tunnel arch. The
hybrid linings tended to comprise thick stone sidewalls supporting brickwork arches. Not all masonry was
ashlar and there are examples where random rubble was used in the construction of the conduit.

Masonry linings for tunnels were typically constructed as arches. There was development of the shape
from vertical sidewalls supporting an arch (typically early canal tunnels and rail tunnels) to the more
efficient horseshoe shape with curved sidewalls, frequently adopted for rail tunnels. Sewerage systems
tended to be circular or egg shaped. Conduits associated with dams are predominantly circular or
vertical sidewalls supporting an arch, which provides a flat invert for more space to install pipes.
Often an empirical approach to the design of the arches was adopted at the time owing to the lack of
knowledge or understanding of soil or rock mechanics.

Open joints are common in the crown of masonry conduits. This is due to the difficulties of filling joints
when closing the arch. Sometimes a ‘key-stone’ was used. The best bricklayers were not always used for
conduits and when pointed after the centring was struck, hence the worked appeared satisfactory, but in
reality on occasion was not up to standard.

McKibbins et al (2009) includes useful information on lining profiles, thickness and quality, and
descriptions of lining construction, inverts and drainage, brickwork bonds and construction joints in
brickwork. This guide includes a summary of the causes of masonry deterioration and reference should
be made to it for details.

2.2.4.1 Brick
During the middle of the 19th century there
was a great demand for bricks, which led to
improvements in the efficiency of brick making
plant and processes. Where large quantities
of bricks were required for tunnel lining (see
Figure 2.6), the bricks were often manufactured
on site using materials excavated locally. The
original composition of the clays used to make
bricks as well as the temperature and duration
of the firing process controlled the composition,
characteristics and quality of the fired brick.

The position of bricks within the kiln during the

firing process also influenced the quality of the
bricks. The fired bricks were graded according to
their quality so that they were used with the best
quality bricks reserved for the inside face. The
poorer quality bricks were often used in the rings
behind the facing or as random rubble or fill. So
there could be significant variability of the quality
of bricks within a single structure. Figure 2.6 Brick masonry lined conduit (courtesy Dŵr
Cymru)

12 CIRIA, C743
The process of aging and deterioration of the bricks in conjunction with often chemically harsh

1
environments such as acid moorland water and the original variability will have contributed to the
current condition of masonry in conduits.

Results of tests carried out on bricks from a variety of railway structures to discern the physical
properties of bricks from the period 1840 to 1910 are shown in Table 2.1.

Table 2.1 Properties of some old bricks used in bridge and tunnel construction (after McKibbins et al, 2009) 2
Manufacture Compressive Elastic
Source Description
date strength (MPa) modulus (GPa)
Sugar Loaf Tunnel (north of
Llandovery)
Alfreton Tunnel
Large regular handmade red brick

Large regular handmade blue brick

1850–1900

1850–1890
34.1

55.2
6.2

9.1
3
Watford Tunnel Long thin distorted handmade red brick 1840–1850 22.9 3.9

Bridge 251 near Grantham Small rough handmade red brick

Bridge 31 near Windermere Small rough handmade red brick

1852

1840–1850
26.3

19.6
8.7

1.5
4
Bridge 49 near Wing Regular handmade red brick 1850–1900 28.8 6.8

5
Bridge 65 on Preston to
Large handmade red brick 1850–1910 25.6 9.4
Lancaster line
Harringworth Viaduct Small rough handmade red brick 1840–1850 17.9 12.9

Notes

6
MPa = Megapascal, GPa = Gigapascal
The cross-sectional area for a small brick is 210 mm2 and for a large brick 250 mm2 to 300 mm2.

Table 2.1 illustrates the variability of the compressive strength and especially the elastic modulus of
old bricks. British Rail (Temple and Kennedy, 1989) carried out an extensive testing programme to
investigate the variability of compressive strength and elastic properties of brickwork from masonry
structures of different ages (mainly between 1840 and 1910) across the UK. Statistical methods were 7
applied to arrive at general conclusions that could be applied for tunnel assessments. The results are
summarised in Table 2.2.

Table 2.2 Statistical analysis of properties of brick samples from old railway structures (after McKibbins et al, 2009)
8
Brick type Characteristic strength (MPa)a Modulus of elasticity (GPa) Poisson’s ratio

9
Red and yellow bricks 16.5 5.2 0.11

Blue bricks 70b 15.6 0.16

Notes
a
Value exceeded by 90 per cent of the bricks tested in a large sample.

10
b
Value is based on typical results from a small sample size.

Table 2.2 shows the considerable difference in strength and modulus of old blue bricks (engineering
bricks) and old red or yellow bricks (probably various non-engineering class bricks, eg stock bricks and

11
gault bricks) found in the structures included in the study. It should be emphasised that the results
given in the table are typical values only and that there are many different types of brick with properties
differing from those presented.

A good reference on the historic production of bricks and their characteristics is provided by
Hammond (1981).
12
2.2.4.2 Stone
Construction material quarried from natural rock resources is termed ‘stone’. It is one of the oldest
building materials known and has been the preferred material for the construction of buildings since
the earliest times of civilisation. The exceptional durability was a notable feature of stone in addition to
A1
Dams and reservoir conduits 13
its aesthetic appeal. However, the stone used for tunnel linings was often not chosen for its durability,
but was a by-product of the excavation process perhaps stone left over from that used in other parts
of the structure. There are existing tunnels that have experienced deterioration of the stone linings,
either because of poor material selection and use in the first place or the subsequent application of
inappropriate repairs. This deterioration can be hastened by harsh conditions, such as acid moorland
water, permanent dampness/wetness and freezing and thawing in the tunnel environment.

Although a wide range of rocks have been used as building stone (see Table 2.3), the most commonly used
stone in the UK were the sedimentary rocks, limestone and sandstone. In the north and west of England,
Scotland and Northern Ireland igneous rocks such as granite occur and these have been used as building
stone. Some of the principal characteristics of building stone in the UK are summarised in Table 2.3.

Table 2.3 Comparison of typical strength and density of some common UK

building stones with other construction materials (after Smith, 1999)

Typical compressive
Masonry unit material Typical density (kg/m3)
strength (MPa)
Stokeground Bath – base bed 22.5 2126

Stokeground Bath – top bed 13.8 1988

Portland Roach 52 2100

Portland Whitbed 36 2200

Welsh Blue Pennant 158 2630–2850

Clipsham limestone 32 1826

Woodkirk Yorkstone 54 2400

Granites 100–350 2500–3200

Concrete (typical) 40 2240

Bricks (typical commons) 20 1800

Bricks (engineering class A) 70 Up to 2800

2.2.4.3 Mortar
The functions of mortar in masonry are to:

‰‰ Avoid excessive local stresses that might develop at points of contact and promote even load transfer
between masonry units (brick or stone) by providing an even contact surface between units.
‰‰ Allow the masonry to function as a composite material by physically binding the masonry units
together to form the masonry fabric. The inclusion of mortar influences important physical
characteristics of the masonry fabric, such as compressive strength and modulus of elasticity (some
degree of articulation was conferred upon masonry arches by the use of weak and flexible lime-
based mortars and allowed them to respond plastically to stresses).
‰‰ Allow the structure to ‘breathe’ by providing preferential pathways for the movement of moisture
through a masonry structure and to concentrate deterioration by providing a sacrificial component
rather than the masonry units themselves.
‰‰ Provide a degree of watertightness by filling joints in water bearing strata.

The thickness of the mortar will have an influence on the behaviour of the masonry structure. Well-
cut stone masonry ashlar typically rest directly on one another or on a very thin bed of mortar as a
lubrication layer during construction. In such cases, behaviour of the masonry is dependent principally
upon the properties of the stone. When thicker mortar beds were required for less well-dressed stone, the
properties of the mortar gains importance. When mortar is present in high proportions such as in rubble
masonry and mortared rubble fills, the mortar properties will have significant influence on the structural
performance of the masonry. The strength and structural behaviour of brickwork will also be influenced to
some degree by the properties of the mortar depending on the thickness of the mortar beds.

14 CIRIA, C743
Thin mortar beds between well-dressed stone are likely to be in a state of uniform compression, where its

1
strength is less significant. In contrast, the mortar beds between uncoursed stone are likely to be thicker
and less even, and more likely to be subject to non-uniform stresses leading to greater potential for
compression and direct load transfer between masonry units.

The mortars used for masonry construction were generally produced from lime cements. The lime
(calcium oxide, CaO) was usually derived from natural limestone, including chalk. The nature of the raw
materials as well as the presence of impurities or additions in conjunction with the firing conditions in 2
the kiln had a significant influence on the characteristics of the lime mortar. Lime cements can exhibit
different properties:

‰‰ pure limes (‘fat limes’) are produced from pure limestone or similar materials. They harden slowly
by reaction with atmospheric CO2, known as carbonation or air setting. They are non-hydraulic
3
cements, ie they do not set in water.

4
‰‰ hydraulic limes do not rely entirely on reaction with CO2 from the atmosphere to set, but include
components that set by chemical reaction with water. The inclusion of clay impurities in the
limestone raw material or the direct addition of hydraulic material would affect the setting action
when mixed with water.

The characteristics of mortars based on pure lime and hydraulic limes can exhibit significant
variability, principally related to their degree of hydraulicity (ie set by chemical reaction with water).
At one end, a very pure lime that relies entirely on air-setting produces a mortar that sets very slowly
5
over long periods and remains weak and plastic, generally with crumbly texture. At the other end,
mortars based on strong hydraulic lime reach a set and achieve higher strength more rapidly when in
the presence of sufficient moisture. 6
McKibbins et al (2009) includes a detailed discussion on the following aspects of lime mortars:

7
‰‰ hydraulic limes, natural cement and ‘Roman’ cement
‰‰ modern cements and gauging of mortar.

Table 2.4 gives indicative ranges of compressive strengths of mortars for different brick strengths.

Table 2.4 Mortar mixes and compressive strengths used in the UK and corresponding strengths of masonry using
different bricks (after McKibbins et al, 2009) 8
Type of mortar (by volume) Brick strength (MPa)
Mortar

9
Mortar 7 20 35 50b 70c
Cement: Masonry Cement:sand strength
designationa
lime:sand cement:sand +plasticiser range (MPa) Characteristic compressive
strength of brickwork (MPa)
1 1.0:0–0.25:3 – 11–16 3.5 7.5 11 15 19

10
2 1:0.5:4.5 1:2.5–3.5 1.3–4 4.5–6.5 3.5 6.5 9.5 12 15

3 1:1:5–6 1:4–5 1:5–6 2.5–3.6 3.5 6 8.5 11 13

4 1:2:8–9 1:5.5–6.5 1:7–8 1–1.5 3 5 7 9 11

11
5 1:3:10–12 1:6.5–7 1:8 0.5–1 2 4 6 7.5 8.5

6 0:1:2–3d – 0.5–1 2 4 6 7.5 8.5

7 0:1:2–3 e
– 0.5–1 2 3 3.5 4.5 5

12
Notes
a From BS 5628-1:1992.
b Class B engineering brick.
c Class A engineering brick.
d Hydraulic lime.

A1
e Pure lime.

Dams and reservoir conduits 15

2.2.5 Timber
There are a number of cases where timber
has been used to construct reservoir
conduits. The amount of information on
these conduits is limited, but it is known that
timber can last a long time in permanently
submerged conditions. However, the timber
can be prone to attack by invertebrates and
deterioration through changes in water
chemistry. For example, at Coombe Park in
Gloucestershire a timber culvert controlled
by a timber plug valve was used to provide a
discharge facility through an earthfill dam
(see Figure 2.7).

For further information see Case studies A1.9

Figure 2.7 Timber control works, which could incorporate a
and A1.11.
timber conduit (courtesy National Trust)

2.2.6 Vitrified clay pipes (ceramic)

In the UK these pipes tended to be used for drainage purposes, eg toe drains, but they have been used
for conduits in pre-Victorian dams and reservoirs. Currently, the common size range is from 100 mm to
300 mm diameter, but larger diameters might be available from some manufacturers. BS EN 295-1:2013
and BS EN 295-2:2013 to BS EN 295-7:2013 cover this type of pipe. In modern dams their use might be
expected to be very limited due to their brittleness and susceptibility to accidental damage.

2.3 CONDUIT GENERAL ARRANGEMENTS

2.3.1 Outlet works

Location of the control valves in the outlet works is critical in establishing the level of risk posed by a
conduit during its life. The control point determines the typical water pressure in the conduit, at any
given point, and so the likely consequences of any leak in the conduit wall.

Control can be provided at upstream, central or downstream locations. Note that for uncontrolled
conduits (for spillways or open conduits at flood storage reservoirs) control is always upstream unless
discharge is constrained downstream.

As discussed in Chapter 3, potential failure modes are present irrespective of the position of the control.
However, the provision of control at the downstream toe of an embankment is inherently more risky for
pressurised single wall pipes and is discouraged for all except very low risk reservoirs.

2.3.1.1 Upstream control

The provision of upstream control is generally more expensive than downstream control. It is the
preferred control location for dam safety. An inlet structure is required with a control valve or penstock,
preferably with a separate guard valve. Three options are normally available:

1 Using sluices/valves with inclined spindles resting on the embankment shoulder.

2 Using hydraulically actuated submerged valves.
3 Providing a tower with an access bridge to a point directly over the valves or an access tunnel
or culvert.

16 CIRIA, C743
The relative advantages and disadvantages of each arrangement are noted in Table 2.5.

Table 2.5

Arrangement
Advantages and disadvantages of upstream control facilities

Advantages Disadvantages
1
‰‰ can provide abstraction at ‰‰ higher construction costs

Intake tower
‰‰
various levels
less likely to have bent
spindles if operated
‰‰

‰‰
connection between tower and conduit can be prone to
differential settlement
more structure to maintain
2
correctly ‰‰ possible safety issues related to access
easier to provide guard and

3
‰‰ ‰‰ vulnerable to seismic activity (not usually a problem in
operation valves/penstocks. the UK).
‰‰ spindle length is limited by ability to impart torque
without fracture
‰‰ prone to settlement of the dam bending spindles – more

4
likely after conduit replacement
‰‰ normally much cheaper ‰‰ more difficult to have guard and operational valve
than provision of a tower, arrangements
Sluice valves with cheaper than submerged
inclined spindles resting ‰‰ more difficult to provide abstraction at more than one
hydraulically actuated valves level
on the dam face

5
‰‰ man access to control point ‰‰ valves difficult to access for maintenance/repair
at dam crest.
‰‰ inclined spindles can be bent by ice/flotsam/vandalism
‰‰ timber draw bars can deteriorate
‰‰ buckling can be a problem with push-pull bars

6
‰‰ prone to vandalism.
‰‰ normally cheaper than the ‰‰ requirement for a hydraulic pump
provision of a tower
‰‰ dependent on long hydraulic fluid lines, prone to
‰‰ not constrained by length blockage by air pockets and failure
Submerged hydraulically
or settlement of upstream

7
actuated valves ‰‰ valves difficult to access for maintenance/repair
shoulder
‰‰ normal operation has least ‰‰ risk to water quality from burst hoses (vegetable oil could
safety hazards. be used) joints are a problem.

2.3.1.2 Middle control

Control can also be provided by valves positioned along the pipe so they are below the dam crest. Where
8
the dam has a central impermeable core and the valves are sealed within it, this places less significance

9
on the watertightness of the conduit wall upstream of the core.

However, where the impermeable fill layer is towards the upstream face or is provided by an upstream
liner, central control would place the valves downstream of the impermeable barrier, and would
introduce a length of pressurised pipe through fill downstream of the impermeable layer. This should be
avoided in the interests of dam safety.

The two most significant problems with central control are the operation of the valves within the conduit
10
at this position, and access for maintenance. For the latter, man access needs to be provided to the valves

11
from either end of the conduit or from the crest. This then requires either a large conduit internal area,
or a tower through the centre of the dam. Both contribute to costs, and the latter interacts with the core
and will impart shear forces into the core increasing the risk of hydraulic fracture, particularly where the
core is narrow.

Again, the provision of separate guard and duty valves is strongly preferred. Valves can be operated by
hand or hydraulically or electrically actuated.
12
2.3.1.3 Downstream control
This should be avoided with single walled conduits, as it introduces a length of pressurised pipe A1
downstream of the main impermeable zone of the dam (irrespective of where this is). Any leak or

Dams and reservoir conduits 17

burst in the pipe is likely to wash out some of the dam fill, possibly leading to collapse. Downstream
control is an inherently unsatisfactory measure. However, it does provide significant advantages
relating to construction cost and access to the valve(s) for operation and maintenance (O&M), but these
considerations should not take precedence over reservoir safety considerations.

At some UK dams, downstream control is provided on a pipe housed within a culvert, which extends to
the central core. At such sites, leakage from the pipe can safely exit the culvert without interacting with
the dam fill. This facility can also be provided by placing the pressurised pipe within another, retaining
a gap in the annulus for passage of leakage water.

Where downstream control is found in existing reservoirs, owners should review the risks, taking expert
advice and consider mitigating these risks by moving the control position, lining the pipes or replacing
the system entirely.

2.3.2 Spillways
When conduits are used as spillways, they are usually associated with shaft or drop inlet type and
generally comprise a crest (eg bellmouth), a conduit and a stilling basin or similar structure. The
crest may or may not be provided with control facilities such as gates. The system generally should be
designed to operate with crest control over the full range of expected discharges such that the weir does
not drown out. The conduit should be designed to run full or such that the flow never reaches the roof.
When the flow touches the roof of the culvert, pressure fluctuations may occur that the conduit was not
designed for. In such situations provision of venting facilities could be considered.

Cavitation effects need to be considered carefully during the design of drop type spillways as there are a
number of cases where severe damage has been caused by this phenomenon.

Where additional spillway capacity is required at a reservoir the most cost-effective solution is usually
to construct an auxiliary spillway over the crest of an embankment dam in the form of an open channel
with a bridge to maintain access along the crest. These structures, although not strictly conduits, will
pass through the dam core and provide a preferential seepage path at high reservoir levels that can
lead to similar failure modes as may affect conduits. Care should be taken that the structure is able to
articulate as the embankment settles and that the joints are waterproof.

2.3.3 Design details

The designer of new conduits needs to consider any particular structural design requirements that might
apply to a conduit passing through an embankment and take account of the failure modes described in
Chapter 3. Key issues include:

‰‰ avoiding the effects of internal erosion and piping

‰‰ cracking should be minimised and special attention needs to be given to conduits on soft
foundations where settlement and spreading of the embankment can be significant and lead to
opening of joints
‰‰ avoiding cavitation or erosion from high velocity flow
‰‰ the range of flows that might occur over the year
‰‰ the difficulty in shutting down for maintenance or repair, owing to compensation flow
requirements downstream
‰‰ ensuring that gates and valves remain functional, tight tolerances might be required
‰‰ allowing for inspection and maintenance requirements (provision of space to manoeuvre).

18 CIRIA, C743
3 Failure modes
1
This chapter considers how conduits may, by their very presence, cause a dam to fail by initiating
internal erosion within the surrounding embankment, causing a local disturbance leading to hydraulic
2
fracture or may deteriorate in use or even collapse entirely causing a loss of surrounding soil.

3.1 MAIN MODES OF FAILURE 3

Many failures of embankment dams have been associated with the presence of a conduit located in the
fill itself or in the foundation beneath. Brown and Gosden (2004) and others before them estimated
that in the UK about 60 per cent of incidents appear to be related to internal erosion, although not all
of these would have been associated with the conduit. The mere presence of the conduit can change the
4
degree of compaction or the pressure in the surrounding soil, which can lead to erosion. In addition,
defects in the conduit or in the badly compacted fill where the conduit acts as a discontinuity through
the fill itself can cause erosion either by seepage into an unpressurised conduit or by seepage out of a
pressurised conduit into the embankment.
5
6
The five main failure modes are:

1 Erosion of soil into a non-pressurised conduit.

2 Erosion of soil by water escaping from a pressurised conduit.
3 Erosion of soil by water flowing along outside of a conduit.
4
5
Erosion of soil near to a conduit.
Collapse of the conduit.
7
Table 3.1 shows failure modes that could be associated with conduits and the how these might be
manifested in an inspection.

Table 3.1 Failure mode identification (after Brown and Gosden, 2004)
8
9
Caused by
Hydraulic fracture of
uneven compaction
of conduit, joints or

watertight element
inadequate design

of backfill or other
Defect in conduit

foundation/anti-
seepage collars
(Section 3.2) or

support defect

Threat from conduit in embankment1

Inadequate or

(Section 3.3)
Non-uniform

10
support

Erosion of soil into a non-pressurised

Piping failure or sink holes leading to dam instability
conduit
Erosion of soil by water escaping from a
pressurised conduit
Erosion of soil by water flowing along
Piping failure leading back to the conduit defect

Piping failure or sink holes leading to dam instability

Piping failure or sink holes
leading to dam instability
11
outside of a reservoir or conduit water

12
Erosion of soil near to a conduit Piping failure or sink holes leading to dam instability

Collapse of the conduit Piping failure or sink holes leading to dam instability
Note
1 Can be caused by either ‘piping’ or ‘internal erosion’.

The illustration of erosion into a conduit that follows relates to a homogeneous conduit and is intended
to show the fundamental mechanism. It is recognised that most dams in the UK have clay cores and A1
there is a tendency for a sinkhole to appear at or close to the crest potentially reducing freeboard.

Dams and reservoir conduits 19

3.1.1 Erosion of soil into a non-pressurised conduit
If a defect develops in the conduit due to some deterioration mechanism, or if a joint in the conduit
opens due to movement and seepage occurs through the surrounding embankment, soil particles may be
carried into the conduit (see Figure 99 in FEMA, 2005). For this failure mode, the conduit needs to have
an interior pressure lower than the seepage pressures in the surrounding soil.

When the reservoir fills, a phreatic surface will

develop within the embankment and there will be
hydraulic pressure around the conduit. Upstream of
any impermeable core this pressure will be the head
in the reservoir. If the surrounding soil is susceptible
to piping, particles may enter the conduit through a
defect. Depending on the soil type a piping failure
may develop or a sinkhole may be formed linking the
defect in the conduit to the reservoir. If the defect
becomes big enough, large flows into the conduit can
occur leading to failure of the conduit or sink holes
in the face of the embankment, either of which may
cause instability and possible failure of the dam.

If an hydraulic fracture or some other preferential

flow path has formed near to the defect in the conduit,
internal erosion will occur as soil particles are washed
into the conduit.

For further information see Case studies A1.2, A1.16

and A1.18.

Figure 3.1 shows the ingress of fines into a non-

pressurised conduit, which is also discussed in
Figure 3.1 Ingress of fines into a non-pressurised Case study A1.14.
conduit (courtesy Dŵr Cymru)

3.1.2 Erosion of soil by water escaping from a

pressurised conduit
If there is a defect in a conduit acting as a supply or scour pipe or as a reservoir overflow under pressure,
water is likely to flow into the surrounding embankment (see Figure 100 in FEMA, 2005). Conduits that
were not designed to be pressurised may become so if a blockage occurs or if a downstream gate or valve
is fitted after construction, and this also may cause water to pass into the dam.

Water flowing from the conduit defect will eventually flow from the downstream face of the embankment
or other exit, such as a drain. If there is no protective filter, soil particles may start to be washed out and
piping failure may start. The piping failure will typically progress upstream, to link back to the conduit
defect. If the defect becomes big enough or if no remedial action is taken, sufficient erosion may occur to
breach the embankment. Soils not susceptible to piping require a concentrated flow path for significant
erosion to occur, unless there is an hydraulic fracture or some other preferential flow path has formed
near to the defect, in which case piping may occur (see Case study A1.8).

3.1.3 Erosion of soil by water flowing outside a conduit

This failure mode is usually associated with embankment seepage from the reservoir through the soils
surrounding the conduit. The seepage along the interface between the conduit and surrounding soil may
be concentrated enough to result in piping, and if the soils are susceptible, a connection to the reservoir
may lead to a breach in the embankment (see Figure 101 in FEMA, 2005).

20 CIRIA, C743
If soils surrounding the conduit are resistant to piping, but cracks or preferential flow paths occur

1
from poor compaction techniques or later develop from hydraulic fracture, continued flow through
the preferential flow paths will result in internal erosion. Erosion similar to that caused by piping, may
then develop.

3.1.4 Erosion of soil near to a conduit

A conduit passing through the core of an embankment dam is one of the main causes of differential
2
settlement, which can lead to hydraulic fracture within the embankment (see Figure 103 in FEMA,
2005). When an earthfill dam experiences an hydraulic fracture, a pathway is created along which water
from the reservoir can flow easily and erode the soil on the sides of the crack. This can progress to the
point where breaching can occur if the crack is not intercepted by a filter zone. Even high plasticity clays
3
that are not dispersive can erode over time.

When a crack forms in the soil surrounding the conduit, if the soil is highly erodible, then the crack is likely
to rapidly enlarge. The water discharging at the downstream face of the embankment dam will typically
be muddy, and a vortex may form at the entry point on the upstream slope. The erosion tunnel can then
4
enlarge to the point that the reservoir is emptied and the breaching process is completed. A tunnel-shaped

5
hole will exist after the failure, if the soil can support the roof of the tunnel. If there is lack of support, the
tunnel roof is likely to collapse and the failure will have the appearance of an open breach in the dam.

3.1.5 Collapse of the conduit

If the conduit suffers serious structural failure it is likely that the surrounding soil will fall into the gap
and the resulting void will migrate upwards and form a sink hole. Note that flow from the reservoir into
6
the tunnel will occur if the hole is below water level. Otherwise the failure will be via erosion of soil into a

7
non-pressurised conduit or if the conduit is pressurised erosion of soil by water escaping.

3.2 CAUSES OF DEFECTS IN CONDUITS

Deterioration of the conduit can produce gaps or holes, which can lead to water flowing into or out of a
conduit. These may be formed by any of the following processes (Sections 3.2.1 to 3.2.3) and will initiate 8
the failure modes described in Sections 3.1.1 and 3.1.2.

9
Repair of any defects discovered during inspections should be carried out in a timely manner to
halt or slow down deterioration. More costly repair or replacement of the conduit may be required if
deterioration is allowed to progress unhindered.

3.2.1 Abrasion
Where flows in the conduit are rapid and the water contains sands, and gravels, abrasion of the parent
10
material and of any protective coatings can occur. Concrete pipes are particularly susceptible to abrasion,

11
especially where construction practice has been poor, as can brick and masonry conduits. CI pipes can
also become abraded with section loss, particularly at the invert.

3.2.2 Corrosion of metals

Corrosion is the destruction of materials by chemical reaction with its environment. In metals this
process is brought about by oxidation or by galvanic action with a dissimilar metal.
12
Metal pipes and metal bolted culvert sections are susceptible to corrosion especially when they are in
contact with aggressive soils or are laid inside wet or damp culverts. This can lead to loss of mechanical
strength and cross-section. Corrosion may be general or local and pitting may occur where a protective A1
coating has been damaged. Pitting can lead to perforation if protective coatings are not maintained.

Dams and reservoir conduits 21

When CI corrodes, the iron becomes a product of graphite flakes and phosphide residues, which
occupies the dame space and disguises the deterioration as there is no obvious change in the surface
appearance. This graphite corrosion product has some strength and may prevent leakage even if it
penetrates the full wall thickness. This is an important ability of cast iron. Some weaker corrosion
products can be identified by surface crazing and scored with a knife, but the harder forms bond with
the metal and are difficult to scratch. The weaker forms can fall out to leave obvious pits in the metal
surface (Reader et al, 1997).

DI usually corrodes to a weaker product than CI, often causing leakage as soon as the corrosion
penetrates the thinner pipe walls (Reader et al, 1997).

Steel corrosion causes a gradual reduction in the effective thickness of the pipe wall, which reduces the
strength of the pipe and its ability to resist internal and external pressure.

The rate of corrosion is accelerated by increasing soil acidity, resistivity, moisture content, the presence of
soluble salts and oxygen content.

Corrosion can also occur in the steel reinforcement within reinforced concrete conduits. Inadequate cover,
poor concrete design and compaction and structural under-design can lead to cracking and spalling of the
concrete, which exposes the reinforcement to rusting. The rusting reinforcement then expands causing
more spalling and cracking of the concrete and further corrosion. This mechanism, if allowed to continue,
can lead to loss of structural strength and the conduit may ultimately fail or even collapse.

3.2.3 Concrete deterioration

There are several mechanisms that can cause a concrete conduit to deteriorate, discussed in the
following sub-sections.

3.2.3.1 Freeze-thaw
This is not normally a problem in the UK. However, in northern and highland areas of Scotland where
an overflow culvert beneath a dam can be subject to freezing winds and damp or wet conditions, the
concrete may deteriorate. In new construction air entrainment in the concrete mix can reduce the
damage caused by freeze-thaw and in existing structures surface treatments and specialist repair
materials and techniques are available.

3.2.3.2 Sulphate attack

Sodium, magnesium, and calcium sulphates existing in soils and groundwaters react chemically with
the hydrated lime and hydrated aluminates in the cement paste in concrete. The volume of the reaction
by-products is greater than the volume of the cement paste from which they are formed, resulting in
disruption of the concrete due to expansion.
Concrete experiencing sulphate attack (see
Figure 3.2) is characterised by a disintegrated
appearance. In new construction, a sulphate
resistant Portland cement or a combination
of suitable cement and pozzolan should be
specified, when it is recognised that concrete
will be exposed to soil and groundwater
containing sulphates. The application of a thin
polymer concrete overlay or sealing compounds
may be beneficial for existing structures
experiencing sulphate attack. Otherwise,
removal and replacement of concrete with
Figure 3.2 Sulphate attack of concrete (courtesy Oz Coasts) sulphate resistant cement should be considered.

22 CIRIA, C743
3.2.3.3 Carbonation
The carbonation of concrete is one of the
main reasons for corrosion of reinforcement 1
(see Figure 3.3). Carbon dioxide from the
air penetrates into concrete and reacts with
calcium hydroxide to form calcium carbonates
resulting in shrinkage. In the presence of
moisture, CO2 changes into dilute carbonic
2
acid, which attacks the concrete and also

3
reduces its alkalinity, accelerating corrosion
of the steel. In a conduit, especially if it is not
ventilated, the concentration of CO2 in the air
can be much higher than usual. However, given

4
good quality concrete and sufficient depth of
cover, carbonation should not occur. Figure 3.3 Carbonation damage to concrete (courtesy
Wikipedia)

3.2.3.4 Chloride attack

Steel reinforcement that comes into contact with chlorine ions from a marine environment (although
unlikely in a dam conduit), or where calcium chloride has been used in the concrete mix as an accelerant,
5
may be subject to chloride damage. Chlorides attack the reinforcing steel through electrolytic action
producing black coloured rust. In dense high strength concrete, a significant loss of the reinforcing bar
can take place without the normal visible signs of disruption to the concrete cover, such as spalling. In
poor quality concrete cracking and spalling will occur. Specialist repair techniques are needed to combat 6
chloride attack. In new concrete control of the aggregate source and admixtures should prevent attack.

3.2.3.5 Alkali-aggregate reaction (AAR)

AAR occurs when certain types of sand and aggregate (eg opal, chert, flint, or volcanic material with
7
high silica content) are exposed to sodium and potassium hydroxide alkalis in Portland cement (see

8
Figure 3.4). In a moist environment, a gel is formed around the reactive aggregate, creating tension
cracks and extensive expansion and fracturing of the concrete. This expansion, cracking, and loss of
concrete strength can lead to pathways
for seepage or localised collapse of the

9
conduit. Concrete containing alkali-
reactive aggregate may show immediate
expansion and deterioration, or it
may remain undisturbed for many
years. Concrete experiencing AAR is
characterised by ‘map’ cracking on the
surface. In new construction, aggregate 10
sources containing negligible potentially
alkali-reactive materials, low alkali
cements, and pozzolan replacement of a
portion of cement, should be used. 11
There is no proven method for

12
eliminating AAR in existing structures.
Figure 3.4 AAR in a concrete dam (courtesy Atkins Ltd) The process is slowed if water is excluded.

3.2.4 Defects in precast concrete conduits

Precast concrete units are generally produced in controlled conditions to high quality. However, as well
as being susceptible to the same mechanisms of deterioration as cast in situ conduits (although to a lesser
A1
Dams and reservoir conduits 23
extent) other problems may occur with culverts or tunnels constructed from precast concrete sections or
pipes. These have been used to form new culverts in existing dams, and it is vital that there are no voids
formed or left behind the segments. This is done by ensuring that adequate back-grouting is carried out
to fill voids and re-establish adequate earth pressure in the surrounding soil (see Case study A1.19).

3.2.5 DEFECTS IN BRICK AND MASONRY CONDUITS

In the UK conduits large enough for human-entry have historically been constructed of brick or
masonry or a combination of both. These are only rarely built as pressurised culverts. The integrity of
such culverts depends upon the structure being in relatively uniform compression. Loss of horizontal
support from the soil can reduce the compressive forces and allow individual bricks or blocks to fall from
the soffit or cause the structure to distort or become unstable.

Individual bricks and blocks can delaminate or otherwise deteriorate to the extent that they may not be
sufficiently strong to withstand the compressive forces. Also the mortar between bricks and blocks can
deteriorate or fall out. If this occurs the culvert should be repointed to prevent more serious damage,
although it should be noted that this will affect the groundwater flow and pressures around the culvert
(see Section 8.4.2).

Often weep-holes are formed by leaving out bricks and blocks, possibly to relieve the water pressure on
the culvert. These can provide a route for seepage and leaks leading to erosion of the surrounding soil
(see Case study A1.14).

3.2.6 Defects in CI
CI pipes and bolted segments, especially those over 100 years old, can be particularly troublesome.
This is because they can become delaminated and contain imperfections, inclusions and cracks, which
may not be noticeable, even if they can be inspected, and may only be revealed by NDT. Such pipes are
susceptible to failure if the internal or external pressures acting on the pipes are changed significantly or
if any mechanical damage occurs.

CI bolted segment conduits can become unstable if the bolts corrode and there is loss of compression in
the ring. These should be protected or replaced as necessary.

3.2.7 Poor conduit design

The conduit can be damaged if the forces acting upon it have not been adequately considered at the
design stage.

Differential settlement due to foundation discontinuity can cause offsetting of joints. Differential
settlement of the embankment dam can create loading greater than the conduit can accommodate,
resulting in cracking and excessive deformation.

Spreading of the embankment dam can cause separations in the conduit joints. As compressible soils under the
embankment dam consolidate, the vertical movement is frequently accompanied by some lateral spreading. As
soils spread laterally, sections of the conduit may separate if provision has not been made (eg by pre-cambering
the conduit or special joint design), leaving joint openings through which water can then move.

The vertical load on a conduit can be in excess of overburden pressure (Penman and Charles, 1975).

3.2.8 Cavitation
Cavitation can occur where there is high velocity flow combined with discontinuities, irregularities or
misalignments. These can cause negative pressures leading to bubbles of vapour forming within the
fluid. When these bubbles collapse, a local shockwave occurs that can damage the conduit surface or

24 CIRIA, C743
joints. This damage produces irregularities that can lead to further cavitation damage downstream.

1
Cavitation is not likely to occur where the flow velocity is below 10 m/s.

Cavitation can also occur just downstream of gates and valves where there is a rapid change in pressure.

3.3 CAUSES OF HYDRAULIC FRACTURE AND

CRACKING
2
3
Hydraulic fracture is common near conduits, because the conduits can create differential stresses and
strains in the surrounding embankment soils. Hydraulic fracture of embankment dams can occur when
the pore water pressure is greater than the minimum total stress in the earthfill at a point within a dam.

Physical cracking may also occur without hydraulic fracture.

There are many causes of hydraulic fracture or cracking within embankment dams, some of which
may be sufficiently close to the conduit to be associated with internal erosion failure (see Section 3.1).
4
Features that may lead to hydraulic fracture are:

‰‰ a relatively thin clay core

‰‰ cross-section settlement
5
‰‰ cross valley settlement

6
‰‰ valley arching
‰‰ differential foundation settlement or settlement within the embankment
‰‰ staged construction
‰‰ poorly compacted layers.

Causes of cracking and hydraulic fracture associated with conduits are discussed in subsequent sections. 7
3.3.1 Uneven foundation
If conduits are located on foundations that are not uniform, differential settlement can lead to cracking
within the fill and joint problems in the conduit.
8
9
A hard rock foundation is the ideal, but even here slush grouting or dental concrete may be required to
provide a uniform foundation. On soft rock the formation may need protection from the elements before
constructing the conduit.

Where a conduit is constructed on a soil foundation, differential settlement is likely to occur. Even
if the support is uniform, the loading will vary along the line of the conduit. The soils may vary and
give different levels of support. In the worst case the conduit may be founded partly on soil and partly 10
on rock. When this is expected, efforts should be made to provide a degree of articulation while
maintaining a watertight conduit.

Some designers have supported conduits on piles to provide a uniform foundation. However, this is
generally not a good solution as a poorly compacted or low stress region can be formed beneath the
11
conduit leading to a preferential leakage path.

3.3.2 Inadequate or uneven compaction and backfilling 12

The type of fill around a conduit, its moisture content, permeability and the plant used to compact it,

A1
all affect the degree of compaction and ultimately erodibility of the soil. All of these aspects can cause
differential settlement leading to hydraulic fracture, cracking and hydraulic fracture, all of which can
vary along the line of the culvert.

Dams and reservoir conduits 25

If the conduit has been laid in a trench, the angle of the sides and its width has an effect on the stress
regime around the conduit when the embankment is subsequently placed over it. A wide trench with
side slopes flatter than 1:1 is preferred and the worst case would be where the conduit is placed in a deep
narrow vertically-sided trench backfilled with soil. In the latter case, arching across the trench will occur
leading to low earth pressure within the trench and the possibility of fracture and cracking, which can
vary along the seepage/leakage points.

3.3.3 Inadequate structural capacity

If the conduit has not been adequately designed structurally, gross deformation of the conduit section
can cause cracking and fracture in the soil, as can differential and longitudinal movement across poorly
designed joints. The conduit may even collapse causing large deformations in the surrounding fill with
the potential for embankment failure.

3.3.4 Cut-off collars

In the past, particularly in the USA, it was considered good practice to form anti-seepage or cut-off
collars around a conduit, especially within the core of an embankment dam. However, numerous failures
occurred even where the anti-seepage collars were considered to have been properly designed and
installed. It was found that hydraulic fracture associated with differential movements in compacted fills
was the primary failure mechanism, creating cracks through which scour could occur. Seepage collars
were seen to be ineffective in preventing many failures associated with conduits, and were even thought
to have contributed to some failures by inducing additional differential settlement and impeding proper
compaction around the conduits, so that cracking of the surrounding earthfill could more easily occur.

However, some practitioners consider that they do have a role to play in some low-head situations,
provided that care is taken in their implementation. (Note that poor construction of any detail relating to
conduits can lead to an increased risk of failure). Further information is provided in Section 9.2.1.

26 CIRIA, C743
4 Conduit examination and
cleaning
1
2
This chapter deals with the examination of the embankment both surrounding and within the conduit
using human-entry or remote equipment, identifying the defects that can lead to dam failure. It also
gives details of the cleaning techniques that can be used to facilitate such examination. 3
4.1 PREPARATION
Entry into a conduit may be a confined space entry and, in keeping with modern health and safety
4
management, the first consideration should be whether the inspection can be carried out without

5
human-entry, eg by closed circuit television (CCTV) camera or similar. However, this requirement needs
to be balanced against the quality of information that will be obtained, and it is generally recognised that
direct examination by the human eye provides the most reliable information.

The success of a conduit examination depends upon good planning and preparation.

In preparing for the examination of a conduit through a dam it is necessary to consider: 6

‰‰ purpose of the examination

7
‰‰ requirement for/feasibility of human-entry
‰‰ equipment needed including safety equipment and training
‰‰ preparation of an examination plan
‰‰ data available to inform the examination
‰‰ personnel required to carry out and support the examination.
8
4.1.1 The purpose of the examination
Good quality regular surveillance of a conduit reduces the probability of failure of the dam. Any
examination of a conduit, whether it is a routine examination by the reservoir keeper or it is part of the
statutory inspection by an inspecting engineer, should be planned in advance. In particular the reason
9
for the examination needs to be borne in mind when deciding the method, extent and timing of the
examination.

Defects in a conduit need to be noted and photographed so that they can be compared with the findings
10
of subsequent and previous examinations and any deterioration of the defect logged so that a decision

11
can be made as to the need for remedial works to be carried out. It is also important to a reservoir owner
that defects are remedied at the appropriate time and do not progress to the stage where they become
inordinately expensive or threaten safety. For example, if the bolts in the CI segments in a culvert
beneath the dam are rusting, wire brushing and painting at the appropriate time will be much less costly
than replacement of all the nuts and bolts. In the case of grey CI segments, the emphasis should be on
the crown bolts that secure the key.

It is also very important that the modes of dam failure that are associated with conduits beneath an
12
embankment are recognised so that any indications that such a failure is taking place can be looked for
and reported upon (see Chapter 3, Section 4.4 and Table 4.1).

The failure modes that might be detected in an examination are summarised in Table 4.1.
A1
Dams and reservoir conduits 27
Table 4.1 Failure mode detection

Evidence of potential failure

Seepage from the

downstream dam
Defects observed

dam upstream or
accompanied by

in a pressurised
soil particles
Possible failure mode

downstream
Sinkholes or
instability in
into conduit
Seepage

conduit

face
Erosion of soil into a non- pressurised conduit Y N/A N Y

Erosion of soil by water escaping from a non-pressurised

N/A Y Y Y
conduit

Erosion of soil by water flowing along outside of a conduit N Y Y Y

Erosion of soil downstream on line of conduit N Y Y Y

4.1.2 The requirement for/feasibility of human entry

As mentioned in Section 4.1.1, the first consideration should be whether the information can be obtained
without human-entry. Once the decision has been taken to carry out a human-entry examination, a
safe method of working should be developed with advice from appropriately trained health and safety
professionals. Note that the information provided here should not be considered as a comprehensive
procedure for confined space working, but rather an overview of issues that need to be addressed. The
judgement and experience of those responsible for the entry will need to be exercised at all times.

Note that only personnel with specific training in working in confined spaces can form part of any entry
team. Untrained personnel can be taken in as long as a trained person is assigned solely to look after
their well-being.

It is not possible to give a conduit size or condition, which it is safe to enter. There are many factors that
will influence their decision. The main consideration is the ability to rescue personnel in the event of an
accident. This will depend on the:

‰‰ size and inclination of the conduit

‰‰ standard of ventilation and lighting
‰‰ ease of access, the presence of pipework or other plant
‰‰ material the conduit has been carrying, if any
‰‰ presence of sediment or other deposits
‰‰ likelihood of gas or oxygen deficiency
‰‰ ability to lock-off valves, pumps etc.

If the conduit has been entered regularly, and there has been a risk assessment developed by suitably
qualified personnel, there is likely to be a safe system of work in place and a record of hazards likely
to be encountered, and of previous experience during examinations such as gas alarms. Under these
circumstances, it will usually be safe for the reservoir owner and their staff and other inspectors to enter,
provided that they have had suitable training. However, it needs to be remembered that conditions may
have changed since the previous examination, eg a tunnel that before had a satisfactory oxygen level may
now be dangerous to enter if there has been a change of air pressure or wind direction.

If the conduit is entered only rarely, a specialist team may be required to carry out the initial
examination or accompany the reservoir owner’s staff or inspecting engineer to ensure safety. Following
this, a safe system of work can be established for future examinations.

A notice should be posted in a convenient place at the conduit entrances giving the name of the access
road, the grid reference and the postcode for the dam to allow easy location by the emergency services.

28 CIRIA, C743
An approved code of practice (ACOP) on safe working in confined spaces is available (HSE, 2014).

1
Reference to this ACOP does not remove the responsibility of those involved to get advice from trained
health and safety professionals when planning and/or carrying out a confined space entry.

4.1.3 Equipment and precautions

The selection of the equipment needed for a confined space entry should be determined on the advice
of appropriately trained health and safety professionals. The following list of equipment is not definitive
2
as each confined space entry needs to be considered on a case-by-case basis. However, this list gives an
overview of the range of safety equipment that could be needed for a confined space entry:

‰‰ gas detectors
‰‰ escape breathing apparatus
3
‰‰ safety helmets
‰‰ cap lamps and torches (even if the conduit has lighting)
‰‰ safety harness
4
‰‰ fall arrest equipment if entry is by ladder

5
‰‰ personal protective equipment (PPE)
‰‰ some means of communicating with/alerting the ‘top person’ – voice relay, walkie-talkie, air horn etc
‰‰ communication device for the top person to alert the emergency services (consider alerting the
emergency services in advance if works are to be carried out).

If water can be discharged into the conduit or from pipework within the conduit, all valves and gates etc
controlling this discharge should be locked off to avoid accidental or malicious operation. If butterfly
6
valves are the only control a secondary means of control may need to be fitted if discharge could be a

7
serious problem. A permit to work system might be appropriate in these circumstances.

If the conduit is full or partially (half) full of water, divers will need to be employed and will need to
follow the appropriate regulations. A diving plan or diving hazard assessment should be prepared
before any dive examination in line with the Diving at Work Regulations 1997. Most commercial diving
companies have their own plans. The use of remotely operated vehicles (ROVs) can also be considered.

If the conduit is physically too small for human-entry or is considered too dangerous, a CCTV or other
8
remote survey method will be necessary (see Chapter 5).

4.1.4 The preparation of an examination plan 9

A detailed examination plan should be prepared to identify all features to be inspected, areas of known

10
problems or defects, and areas of potential problems. The examination plan should also identify any
special access, or equipment requirements. An examination checklist is typically prepared as part of an
examination plan. The checklist is used to identify specific examination objectives and is also useful in
developing the final examination report.

Where regular examination is to be undertaken the examination plan and checklist should be
incorporated into a safe system of work to be followed on each occasion.
11
Distance from the conduit entrance or bulkhead wall should be marked on the wall to facilitate the
identification and the position of important elements, such as the clay core. ‘Clock’ notification can be
used to identify positions around the conduit at a particular chainage. 12
4.1.5 The data available to inform the examination
Typical documents that might be reviewed before an examination include: A1
Dams and reservoir conduits 29
1 The health and safety files and operating manual.
2 Design summary and calculations for the conduit.
3 Geotechnical reports on the dam.
4 Geological and seismicity reports at the dam site or in the region.
5 Hydraulic reports on the conduit or any pipework within it.
6 Plans and specifications.
7 As-built drawings.
8 Construction progress reports.
9 O&M records.
10 Reports of previous conduit examinations.
11 Inspecting engineers’ reports (under Section 10 of the Reservoirs Act 1975).
12 Supervising engineers’ statements (under Section 12 of the Reservoirs Act 1975).
13 The prescribed form of record.
14 Owner’s operating procedures.
15 Any risk assessment reports for the dam and the conduit.

4.1.6 The personnel required on the examination

For routine examination for which there needs to be a safe system of work, the examination team
should include:

‰‰ a reservoir owner’s representative who is familiar with the conduit (preferably the reservoir keeper
or the headwork’s controller)
‰‰ the inspecting engineer (if it a statutory examination under the Reservoirs Act 1975)
‰‰ the supervising engineer (if it is a statutory reservoir), or an experienced civil engineer or
geotechnical engineer.

Unless the examination is routine, it should be carried out or managed by a specialist team.

Consideration should be given to include a structural engineer in the team if the integrity of a culvert is
under review.

The minimum number of people in the examination team for human-entry is likely to be governed
by company procedures. However, some authorities suggest a minimum of three – two to carry out the
examination in case one is injured or incapacitated and one to act as top person to communicate with
the inspectors and to summon assistance if required. Where the risk assessment determines, the team
membership might be more.

4.2 FREQUENCY OF EXAMINATION OF CONDUITS

The frequency of examination will depend upon the examination type. It is assumed in this section that
human-entry to the conduit will be possible. Where it is not, the suggested frequency will be the same,
but specialist examination equipment will be needed (see Section 4.6).

4.2.1 Initial examination on first filling

The first time that an embankment dam impounds water is critical. The embankment will be under
hydraulic load for the first time and will adjust to this loading. The conduit will also be tested for the
first time. So, during this period the examination frequency of the conduit and the embankment should
be high. The construction engineer, if it is a statutory reservoir, and the design engineer if not, will give
instructions as to the rate at which the reservoir should be filled and these may include holding the

30 CIRIA, C743
reservoir at particular levels for a specified period. During first filling the conduit should be inspected

1
daily or more often especially when the level is being raised. This frequent surveillance will also be
necessary where the overflow has been raised or there have been remedial works to the conduit.

The construction engineer is in place for a minimum of three years. This is as a means of recognising
that ‘faults’ develop in the early life of a dam and within this period examination frequency is likely to
be high until there is confidence that the dam is behaving as designed. The frequency of examination
during this period generally would be agreed with the construction engineer. 2
4.2.2 Periodic examination
Periodic examination is generally controlled by reservoir safety legislation appropriate to the reservoir 3
location in the UK. For a statutory reservoir in England, after the final certificate has been issued by
the construction engineer, the reservoir will be inspected by the inspecting engineer at intervals of
up to 10 years and at least yearly by the supervising engineer as specified by the Reservoirs Act 1975.
The first statutory examination is within two years of the final certificate. The supervising engineer
should examine the conduit at least once a year both externally and internally where human-entry is
4
possible. Where inspection frequencies are recommended in the inspection engineer’s report these

5
should be complied with as minimum. Where the reservoir is not covered by the Reservoirs Act 1975 the
examination should be carried out by an experienced reservoir engineer at similar intervals.

The reservoir safety legislation in other parts of the UK also sets out the requirements for periodic

6
examination and reference should be made to the relevant legislation for specific information.

4.2.3 Routine examination

7
Routine examinations should be carried out by the reservoir owners or their staff. This may take the
form of a weekly or monthly check on the quality and quantity of any leakage or on any movement of the
dam associated with the conduit. It should also include a visual inspection from the ends of the conduit
where possible, and it is suggested that an internal examination of the conduit is carried out a least once
a year where human-entry is possible.

4.2.4 Special examination

8
A special examination will be carried out when a unique opportunity exists for examination such as
when low water conditions exist in a reservoir exposing a structure that is normally submerged or the
reservoir is at a record high level. 9
4.2.5 Supplementary or emergency examination
The conduit should be inspected following any unusual event such as an earthquake, scour discharge, 10
flood event, the observation of any local defect in the dam, unexplained changes in flows or where a
known problem is getting worse and special monitoring is required.

4.3 EXTERNAL EXAMINATION

11
12
Dam failure associated with conduits is generally caused by collapse of the conduit, loss of soil into the
conduit, erosion by water flowing out of the conduit if it is pressurised, or into the conduit if it is not, or
by seepage along the line of the conduit. The external examination should search for surface signs of
these failure mechanisms such as cracks, localised settlement, depressions and/or sinkholes.

A1
Dams and reservoir conduits 31
4.3.1 Ground movement
The embankment should be examined locally for the presence of slips, cracks, depressions or sink holes
that may have been caused by the loss of soil into or along the line of the conduit. In extreme cases a
whirlpool may appear in the reservoir. Any such features that cannot be readily explained should be
investigated by geotechnical investigation of the embankment and by internal examination of the conduit
where possible. Localised settlement along the line of the conduit could indicate internal erosion into the
conduits (Charles et al, 2014).

Geotechnical examination could be by careful excavation, by conventional boreholes, window sampling,

or by a geophysical survey to assess the size of any underground void.

4.3.2 Seepage flow

Any seepage near to a conduit should be channelled or piped such that it can be monitored for quality
and quantity. Any flows being discharged into the conduit, should be reduced or if possible stopped
and the conduit emptied. The effect of this on the seepage should be observed and noted and the
source of the seepage ascertained by internal examination of the conduit, if possible. If there is no
effect on the seep, the reservoir could be lowered slowly by one or two metres, where feasible, and the
flow plotted against rainfall and reservoir level. Deposits of sand, silt or precipitates of dissolved iron or
manganese may form at the exit point of seepage (see Figure 4.1). Water escaping from intergranular
seepage in granular soils may create sand boils, and the flow may not be turbid. If the flow is seen to
react to changes in reservoir level or if the flow increases at a steady reservoir level or if the flow should
become cloudy or discoloured,
the reservoir could be lowered
further until the flow ceases.
Expert advice should then be
sought and the source of the leak
identified (see Section 5.3).

If the embankment is seen to be

wet in the area of the conduit,
the wet area should be marked
and watched to see if it increases
with time or reservoir level and
rainfall. An area of wetness
may coincide with a change of
vegetation cover. The flow in
the conduit should be collected
Figure 4.1 Deposits at drain outlet (courtesy D A Bruggemann)
and stopped and the conduit
emptied. The effect on the wet
area should be observed and noted. A drainage system should be constructed in an attempt to dry out
the area and any flow should be measured and checked for discolouration. If the area does not dry out,
the reservoir should be lowered where feasible, and the effect monitored. Expert advice should then be
sought and the source of the leak identified (see Section 5.3).

4.3.3 Normal flow from the conduit

Normal flows from the conduit, such as the compensation flow, should be regularly observed and
checked for any unusual changes in quantity and quality. In particular if the flow should become
discoloured or carry sediment an immediate internal examination of the conduit should be carried out
(see Case study A1.18).

32 CIRIA, C743
4.3.4 Additional concerns
During operation of the conduit, additional items of concern can include:

‰‰ unusual noises. Any unusual noises such as popping, banging, or noise of vibrations should be
1
investigated. Vibrations may occur if the conduit is not properly supported. Vibrations could

2
adversely affect the conduit and surrounding backfill
‰‰ pulsating or unstable flow
‰‰ unexplained reductions in discharge capacity
‰‰ unexplained increases in discharge capacity.

4.4 INTERNAL EXAMINATION

3
4
Where possible a conduit should be regularly (see Section 4.3) examined internally either by human-
entry or using CCTV or ROVs.

The interior examination should look for structural defects, surface defects and deterioration, leakage of
water into or out of a conduit, or the deposition of eroded material.

4.4.1 Masonry culverts

5
Conduits of brick or stone construction are normally large enough for human-entry in which case
they are referred to as culverts. Such culverts should be checked for signs of structural distress such as
cracking, spalling and de-lamination, loose or missing bricks and stone and missing mortar. The cross- 6
sectional dimensions of the culvert should be checked for deformation and the profile of the invert
should be surveyed to determine if there has been any settlement. The presence of ponding in the invert
can give an indication of settlement especially if the culvert was constructed at a constant slope. Large
cracks or steps may occur where the culvert is on an uneven or compressible foundation or where slips
have occurred during construction.
7
8
The locations of cracks should be documented using a crack map or similar reporting method. By
comparing with the details of pre-existing cracks it is possible to determine if they have changed in width
or length. Tapping the brickwork or blockwork may reveal hollow sections behind the culvert or that
layers of the construction have become de-bonded.

Mortar between the bricks or blocks should be checked for softening by chemical action or falling out.
Leaching of the lime can also lead to the weakening of the mortar and sometimes precipitates in drains
9
causing a blockage. Culverts can also deteriorate by the action of running water eroding or plucking
elements from the invert and sidewalls, especially at the point where pipes discharge onto the surface of
the culvert. Chemical or frost action within the bricks or stones can cause loss of the face, which can be
severe and lead to complete loss of a brick or block particularly at the downstream end of culverts. 10
Water leaking into a culvert can cause erosion of the mortar and the bricks and stones, and can also

11
erode the soil behind the wall causing lack of support and structural collapse. Where weep holes have
been left in the culvert walls these can form an exit point for ‘pipes’ formed in the embankment above by
internal erosion and material can be carried into the culvert (see Case study A1.18).

12
If the culvert normally contains running water or is pressurised and is de-watered for examination,
water may be seen running back into the culvert from the surrounding ground showing that seepage out
of the culvert is likely to be occurring under working conditions.

Stalactites and stalagmites can form within culverts if they are constructed through limestone rock or
backfill. There have also been instances of the grout used to remedy a leak in the dam above being
deposited within the culvert. The leaching out of the grout from the dam may render the grouting
ineffective.
A1
Dams and reservoir conduits 33
4.4.2 Concrete conduits
These may be cast in situ or be precast as pipe sections or segments. Some problems may have an
underlying cause similar to those found with brick or masonry conduits, causing cracking and spalling.
In addition, the widening or crushing of joints and discontinuities and misalignments across them may
also occur where there are movement or construction joints in in situ concrete, or at the joints between
adjacent precast units. Again crack mapping and measurement should be carried out. Tapping or
dragging a heavy chain along the invert may reveal voids behind a concrete culvert. This can occur
when concrete segments have been used in the construction and back-grouting is needed to prevent
flow along the outside of the culvert or to fill voids. Lack of back-grouting may have caused the failure
at Warmwithens dam in 1970 (see Case study A1.19). In addition the bolts holding concrete segments or
sections together may corrode leading to failure.

In certain cases protective coatings are applied to concrete conduits. The examination should include
a survey of such a coating, by a concrete specialist if necessary, to prevent further deterioration of
the coating and prevent enhanced deterioration in the area of damage. Deterioration caused by
sulphate or chloride attack may be more severe in concrete than in brick and masonry conduits, by
sulphate or chloride attack, and AAR. Concrete can also be abraded and eroded by flowing water and
cavitation can remove material from the surface (Reader et al, 1997). Rusting or other deterioration
of reinforcement within concrete conduits may also cause spalling and a loss of the section’s ability to
withstand the design loading.

Loss of material by leakage into or out of the culvert creating voids may also occur with concrete
conduits, leading to lack of support. This may take much longer to become apparent with concrete
conduits than with brick or masonry.

Common faults in both in situ and precast concrete conduits that can lead to leakage are associated with
the sealing element between conduit sections.

In in situ reinforced concrete conduits the waterbar, which is a flat section of rubber or unplasticised
polyvinyl chloride (uPVC), needs to be secured in position across the joint to prevent displacement
during pouring of the concrete. The waterbar is often flattened by the weight of concrete. Also the
presence of the waterbar, even if it is not displaced, can lead to voids being formed or honeycombing of
the concrete surrounding it, which can also lead to leaks. Often the problem is not detected until a leak
occurs during service.

In precast concrete conduits the positioning of the flexible gasket between the sections is vital in
preventing leaks. The gasket can be displaced or twisted during fitting or jointing of the conduit
sections, which is also often not detected until the leak occurs.

4.4.3 CI and steel conduits

Large steel or CI pipes, some large enough for human-entry, have been used as culverts beneath dams
and CI segments (tubbing) have been used at the point where a brick, masonry or concrete tunnel
crosses the clay core of a dam. Structural problems similar to those found in concrete conduits can occur.
However, CI can de-laminate or may contain inclusions that weaken the material. Steel and CI pipes can
also become severely corroded leading to loss of cross-section and structural instability. Also the bolts
between CI segments may corrode leading to lack of continuity.

Steel pipes in particular are likely to have a protective coating to prevent corrosion and withstand
abrasion. Any damage to such a coating should be repaired immediately if accelerated deterioration in
the area of the damage is to be avoided.

CI and steel conduits also are joined with flexible gaskets that can be misplaced during jointing.

34 CIRIA, C743
4.4.4 Corrugated metal pipes (CMP)
CMP have particular problems when used as conduits in an embankment dam. The main problem is the
quality of the joints between the pipe sections. The joints can separate due to differential settlement and/ 1
or embankment dam spreading, and this can result in leaking joints. The joints are often incorrectly
assembled on site, also resulting in leaking joints. CMP is considered a particularly flexible material,
but its design requires that the earthfill surrounding the pipe provides structural stability to the pipe.
If the surrounding backfill does not provide adequate support, the pipes are subject to distortion and
deflection. Also the circular shape and corrugations on the exterior surface makes compaction of the
2
earthfill against the conduit difficult to achieve. So, examination for distortion and faults in the pipe
joints in CMPs is particularly important.

CMPs have also been found to be more than usually susceptible to corrosion because the protective
3
coating is often damaged and the pipe wall is usually comparatively thin.

4.4.5 Pipes 4
Occasionally, pipes are laid through an embankment without protection and are usually made from

5
CI. Such pipes are susceptible to corrosion and are normally lined with concrete or mortar, sliplined
with plastic pipe, or are fitted with some form of cast in situ liner. Pipes laid in fill with no upstream
control are considered inherently unsafe because if there is failure of the pipe, there is no ready means
of stopping the flow of water until the reservoir empties. Pipes are not usually suitable for human-entry
and examination needs to be via CCTV and/or ROV if the pipe cannot be dewatered.

Many culverts or tunnels will contain pipework. The condition of this pipework should be assessed and 6
remedial works considered. If this pipework should fail it is possible that damage to the tunnel may
occur or in the case of a catastrophic failure there will be danger to operatives or examination staff from
flying debris or inundation of the culvert.

Examination should include the pipework fittings and the means of anchorage.
7
4.4.6 De-watering
The de-watering of conduits and pipes beneath a dam may be needed for a proper inspection, however,
8
care needs to be taken when attempting this. The possibility exists of external pressures being high

9
enough to damage the de-watered conduit or vents becoming plugged, causing negative internal
pressures to develop and collapse the conduit.

4.5 SPECIALISED INSPECTIONS

10
4.5.1 Underwater

11
All diving operations must be carried out in accordance with the Diving at Work Regulations 1997.

Underwater examinations can be carried out by either scuba (self-contained underwater breathing
apparatus) operations or surface-supplied air diving operations. The scuba approach has more
flexibility than the surface-supplied air system, but communication with the diver is limited, whereas
a communication line can be included with the air lines. The potential for longer underwater stays is
greater with the surface-supplied air system. 12
Diving for an examination inside a conduit where the diver does not have a direct route to the surface is

A1
a specialised area of diving. There are no specific criteria for defining conduits, which can or cannot be
inspected by divers and conduits large enough for divers to enter may include factors that preclude diver
examination. The requirements of the Regulations should be met for all diving examinations.

Dams and reservoir conduits 35

Generally dive examinations are costly and are significantly influenced by the water temperature, the
depth of the dive and the altitude at which the dive is to take place. FEMA (2005) suggests that the cost
of dive examinations can be about three to five times higher than ROV examinations. This comparison is
based on USA conditions, but is probably similar in the UK.

The advantage of a dive examination is that instruments for testing the structural integrity of the
conduits can be applied to assess the general condition of the conduit.

Instruments could include the following:

‰‰ rebound (Schmidt) hammer

‰‰ magnetic reinforcing steel locator
‰‰ ultrasonic pulse velocity meter.

The potential for hands-on, tactile examination of features in poor visibility or those covered with
shallow layers of sediment is another advantage of diving surveys. Often there are gratings and grilles to
be removed from the conduit before entry is possible.

In deciding whether or not to carry out a diving survey, the principle of seeking alternative methods to
eliminate risks should be applied.

4.5.2 ROV examinations

ROVs were originally developed for industrial purposes to inspect oil and gas pipelines, and offshore
platforms. They are now used for underwater examinations of entrance and terminal structures and
conduits. Normally the unit is linked to the surface by power and communication links, although
untethered vehicles are also available. Generally, ROVs comprise a video unit, a power source for
propulsion, vehicle controllers and a display monitor. Additional equipment can be fitted to expand the
vehicles’ capabilities and include a manipulator or cutting arm and water samplers as well as instruments
to measure water clarity, light penetration and temperature.

For underwater surveys of conduits, the ROV is usually inserted from the upstream end of the conduit
and might require assistance from a diver to manoeuvre the unit into position.

The following limitations of ROV examinations were identified in USACE (1995):

‰‰ a 2D view produced and the full extent of any defects may not be projected
‰‰ visibility may be impaired by turbid water conditions
‰‰ orientation or position of the ROV may be difficult to determine
‰‰ manoeuvrability can be constrained in ‘tight’ areas and strong water currents might cause the
tether to become entangled.

Further discussion on ROVs is given in Section 5.1.

4.5.3 CCTV examinations

CCTV systems are available for inspecting the inside of conduits with 25 mm to 2 m diameter. The
system has been widely used for the examination of pipework for many years and is generally available
in most parts of the UK. Crawler-mounted camera units are available as well as underwater applications.
Colour CCTV examination systems incorporate ‘forward looking’ cameras and optical mechanisms with
microprocessor controls that allow scanning of almost a complete hemisphere with no moving parts.

Detailed examination of joints, junctions and defects can be carried out using sideways-looking (pan and
tilt) cameras with automatic side illumination. Such equipment can provide full hemispherical viewing
with a tilt angle of 240o and rotation angle of 360o.

36 CIRIA, C743
The success of a CCTV examination depends on the quality of the equipment and experience of the

1
operator. The issues to consider in perforaming a successful CCTV survey are as follows:

‰‰ light – the amount of light is critical to the success of the examination. Lack of clarity hinders
assessment of the integrity of the conduit
‰‰ camera – the video camera should be able to pan and tilt as well as looking straight ahead
‰‰ distance meter – should be superimposed on the videotape/digital record
‰‰ compass – a compass unit will provide azimuth and inclination readings superimposed on the
2
videotape/digital record

3
‰‰ narration – all examination videotapes/digital records should include narration by the operator
‰‰ drawings and photographs – copies of all available design and/or as-built records of the conduit
and dam should be on site during the survey
‰‰ measurements and data collection – the collection of data at the time of the CCTV survey on the

4
reservoir water level, nearby piezometric levels, other historical operation records with date/time
will enhance the technical evaluation of the survey
‰‰ videotape/digital data library – previous CCTV surveys should be reviewed by the operator and
other examination personnel before the start of the CCTV survey.

4.6 CLEANING 5
The cleaning of conduits beneath dams is not a process that is carried out at UK reservoirs in normal
circumstances. At most reservoirs, valves are regularly operated and the conduit is normally cleaned as
a result. 6
However, FEMA (2005) notes the following regarding cleaning:

“Small, inaccessible conduits are especially vulnerable to plugging issues. Cleaning is usually
only an issue where man-entry is not possible. If a conduit requires cleaning, it should only
7
be done after careful consideration of the potential effects on known or suspected deterioration
within the conduit. The basic philosophy used in the cleaning of conduits should be to
“do no harm”, this means a very cautious approach is required.”
8
4.6.1 Reasons for cleaning
Cleaning of a conduit may be necessary for examination for defects, before any renovation or re-lining
or to improve the flow capacity of the conduit. If a conduit has not been periodically operated, certain
mechanisms may lead to deposits or growths within the conduit.
9
4.6.1.1 Sediments and debris
Low level draw-off and scour pipes may become filled with silt over time if the valves are only operated
10
for a short period or if either the scour valve or the control valve is kept closed in sequence during the

11
operation. Also, if a conduit does not discharge water completely out of the system or if the discharge
channel is adversely sloped, water may partially or completely submerge the exit portal. If this
occurs, sediments and debris can back up into the conduit, resulting in sediment deposits or debris
accumulation. Culverts carrying pipes may need to be cleaned if they become blocked with sediment or

12
ochre to allow human-entry.

4.6.1.2 Hard deposits

Conduits experiencing inward seepage through a joint may develop solid deposits where the seepage
water evaporates. These deposits often contain calcium carbonate, which precipitates out of solution as
the mineral calcite. Calcite will form deposits when the calcium ion and bicarbonate ion concentrations A1
in the water increase to the point where they exceed the capacity to dissolve in water. Stalactites and

Dams and reservoir conduits 37

stalagmites may form. Hard deposits of calcium carbonate precipitate may also develop when the
seepage water evaporates.

4.6.1.3 Bacterial growths

Often, bacterial growths are soft and easy to remove, but in some situations these growths can become
hard and mineralised. Aerobic bacterial growth can also create hazardous conditions by depleting the
oxygen in the air of a confined space.

4.6.1.4 Zebra mussels

The Zebra mussel has become a particular problem in the UK in recent years. They can grow on the
inside of conduits and are particularly difficult to dislodge. See Section 9.1.4.2 for further information.

4.6.2 Cleaning methods

If the intake pipe becomes blocked with silt upstream of any valve or gate, it is best cleared from within
the reservoir. If the reservoir can be emptied the silt can be removed from the entrance to the conduit
by an excavator and the pipe cleared by pressure jetting or air-lifting. If emptying the reservoir is not
practical the silt can be dredged from the mouth of the conduit or dispersed by divers using air or water
jets or air lifters if the silt is not too deep. This will be made easier if the reservoir level can be lowered.

It may be possible to remove the silt from the conduit from the downstream end by jetting. However,
once the pipe is cleared and the water flows again removal of the jetting equipment may be difficult and
also impractical if the head in the reservoir is high.

The improper use or the selection of incorrect cleaning equipment may cause additional damage to
a deteriorating conduit and further degrade its structural integrity. The type of conduit material (ie
concrete, plastic, or metal) needs to be considered when selecting the appropriate cleaning method.
Cleaning of inaccessible conduits should only be considered after following a thorough CCTV inspection
of the conduit. If a deteriorating conduit is cleaned without the benefit of CCTV examination, the
conduit may become unknowingly damaged. Indications of obstructions within the conduit may include
reduced outlet flow capacity. If obstructions are found during the CCTV examination, the method
of cleaning can be evaluated and a preferred method selected. Sometimes, CCTV examination and
cleaning are done on the same day. Some cleaning services have limited CCTV examination equipment.
Any cleaning should be attempted only in the presence of qualified and experienced staff representing
the owner of the embankment dam. Complete documentation (including photographs) of all activities at
the site is highly recommended. The success of any conduit cleaning depends upon accessibility, type of
cleaning required, and the cleaning method used.

A variety of cleaning methods are available:

‰‰ Flushing – if debris and sediments are not significant, adequate cleaning may be obtained by
merely flushing the conduit with water. Flushing can be accomplished by opening a gate or valve
and allowing water to flow through the conduit or by inserting a flexible hose and pumping water
into the conduit. In many cases, volume and low pressure is all that is needed to adequately clean
the conduit.
‰‰ Pressure jetting or washing – pressure jetting or washing involves the use of a flexible hose
attached to a metal nozzle that directs jets of water out in front of it to loosen debris and sediments
in the conduit. The nozzle is propelled forward by reverse angle jets. The reverse angle jets also
push debris and sediments backwards toward the end of the conduit, where the flexible hose exits.
Pressure washing is best suited where biomasses or mineral encrustation are to be removed. The
pressure selected for cleaning should fully consider the condition of the types of conduit material,
age, and type of joints.

38 CIRIA, C743
 The lowest possible pressure that effectively cleans the conduits should be used. The jets on the

1
nozzle should be angled no more than about 30 degrees, so the jets are not aimed directly at the
conduit wall. The nozzle should be kept rotating and moving and should not be allowed to remain
in one spot during jetting.
‰‰ Mechanical – mechanical cleaning uses rotating brushes or a metal flail (see Figure 4.2), such

2
techniques have been used in the UK for cleaning the inside of conduits.
‰‰ Cleaning pig – these have wire brushes to scrape the walls of the conduit. A variety of brushes are
available, depending on the type of cleaning required and the existence of any coatings on the
interior surface of the conduit. Cleaning pigs are generally available in diameters up to 1200 mm.

In some situations, minor cracks

or joints experiencing seepage may
3
eventually seal themselves by calcite
deposition. This process occurs when
calcite precipitates out of solution and
forms a deposit in the crack. Calcite 4
deposits typically mineralise and harden
over time. Where the examination
shows locations within a conduit where
this has occurred, cleaning with high
pressure jets could remove enough of
5
the calcite deposition to cause seepage

6
to begin again. This possibility needs
to be carefully considered before
performing any cleaning operations
within the conduit.

7
Figure 4.2 Flail used to clean the inside of a metal conduit (courtesy
AECOM)

8
9
10
11
12
A1
Dams and reservoir conduits 39
5 Conduit assessments –
surveys, integrity testing and
leakage investigations

This chapter reviews the equipment that is available to survey, measure and carry out testing within the
conduit to provide additional quantitative information to supplement the examination.

For the purposes of this chapter, conduits through embankment dams can be divided into the following
three categories:

1 A culvert or pipe large enough to permit internal access.

2 A pipe within a culvert that is large enough to permit access to the exterior of the pipe and the
interior of the culvert.
3 A smaller pipe or culvert buried within the embankment fill.

The assessment of a conduit will generally include and start with a visual inspection (see Chapter 4),
although this may not always be possible.

This chapter outlines various techniques available for more detailed investigations that may follow the
initial visual inspection, including surveys, scans, invasive and non-invasive testing as may be required.
Such surveys and testing of the pipe or culvert may be undertaken to:

‰‰ investigate its shape including distortions arising from settlement and loss of circularity
‰‰ establish its structural condition
‰‰ investigate leakage into or out of the pipe or culvert
‰‰ investigate the presence of voids in the surrounding ground.

While some testing involves direct measurement, most NDT needs to be carried out and interpreted by
a specialist. Many of the techniques described here have been used to assess the condition of tunnels,
sewers or pipelines, but few have been used specifically for the assessment of conduits through dams.
Only those methods that have found application in a civil engineering context are described here, but
it is recognised that other NDT methods are being developed. Further information about the various
techniques is referenced in each section.

5.1 SURVEYS
In many cases conventional surveying techniques may be sufficient to survey the interior of a conduit.
Where greater detail is required there are several technologies available that have been developed over
the last 10 years to survey pipes sewers and tunnels, some of which may be adapted for use in conduits
through embankment dams. Where access is straightforward a variety of methods may be used, while
conduits with difficult access and those that are fully-submerged may require remotely operated
equipment. Such equipment may be mounted on a self-propelled tractor that runs along the base of the
conduit. Where the conduit consists of one or more pipes within a masonry culvert, a complete detailed
survey of the culvert may be difficult because pipes may impede conventional surveys and will partially
obscure scans. In addition, some internal inspection may also be required for the pipes.

40 CIRIA, C743
5.1.1 Conventional surveys
The interior of a large culvert may be surveyed using conventional techniques with total stations and
laser distance measurements referenced back to an external datum. If permanent instrument mountings 1
are positioned at several points along the axis of the conduit and targets are positioned on the conduit,
it may be possible to combine an initial survey with establishing a datum set of readings for subsequent
movement monitoring (see Section 6.4.4).
2
5.1.2 Imaging
Most imaging technologies are camera-based methods, for example CCTV with a pan, tilt and zoom
camera. Often mounted on a ROV, these can view conduit or pipe surfaces above the water-line. Where
clear water is present, it may also be
3
possible to obtain useful images though
water although reflections from the water
surface may cause glare, and movement of
the ROV may stir up any sediment present.
4
While CCTV is the primary option, if a

5
detailed high-quality evaluation of the
interior is required, other techniques
should also be considered.

Digital scanning equipment developed

in Germany (the Panoramo 3D optical
pipeline scanner) uses two digital cameras 6
with 185 degree wide angle lenses
mounted on the ROV (see Figure 5.1). A
succession of forward and backward facing Figure 5.1
photographs using strobe illumination are
Panoramo ROV (courtesy IBAK)

taken as the ROV travels through the conduit or pipe at up to 20 m/m. Digital processing allows seamless
7
perspective and unfolded views as well as creation of 3D geometric model.

In the UK the SETT digital scanning system uses a digital, high-

resolution scanner to produce a forward image, as well as a 360 degree
Further information 8
SETT digital scanning system:
image of the interior wall of the culvert, normally a sewer. A 360 http://tinyurl.com/l47fxfg

9
degree scanning camera is mounted on a robotic, remote-controlled,
wheeled tractor which travels through the pipe at a constant speed of about 13 m/m. The camera
continuously scans the pipe’s inner surface creating a series of adjacent section views covering the pipe’s
circumference. Specially developed computer software processes these scanned sections and stores them
for further analysis as a single complete record of the survey run. This makes it possible to view 100 per
cent of the pipe wall at the same angle and light intensity thereby minimising operator error that could
result from image skewness, light reflections, and shadowing. When the forward view is combined with 10
the scan a three dimensional perspective can be produced.

It is sometimes appropriate to use additional sensors on the ROV such as laser scanners, sonar and
thermal imaging camera. For example pipelines can be inspected using remotely operated tracked
vehicles tethered by a fiber optic cable. The inspection system consists of a modular, long-range, multi-
11
sensor vehicle that is capable of providing a wide variety of high quality data, including:

‰‰ high definition digital pan-tilt-zoom CCTV

‰‰ high speed profiling with sonar designed specifically for pipeline inspection
12
‰‰ laser profiling, eg Light Detection And Ranging (LiDAR) with 3D reporting capabilities

A1
‰‰ electromagnetic sensors for assessing the structural integrity of prestressed concrete cylinder
pipe (PCCP)
‰‰ additional sensors as required.

Dams and reservoir conduits 41

5.1.3 Laser scanning (LiDAR)
Where access permits 3D surveys
may be undertaken using high
speed laser scanners such as the
Leica High-definition surveying
(HDS) system. 3D laser scanning
or LiDAR is the process of
shining a reflectorless laser line
over a surface in order to collect
3D data. Data is captured by a
camera sensor mounted in the
laser scanner, which scans and
records thousands of 3D points
in space each second. Using
measurement of time of flight Figure 5.2 3D survey of bowling rail tunnel (courtesy Severn Partnership)
or phase change, each point is
surveyed to an accuracy of a few millimetres. The 3D survey points collected are then combined to form
a point cloud that represents the surface surveyed, and the processing is carried out using proprietary
software. Scans would be undertaken from several positions along the conduit and referenced to the
external survey grid. These survey techniques have been used for surveying numerous rail tunnels, for
example see Figure 5.2. Similar techniques have been used by subsurface laser scanning for surveying
sewer tunnels (Hudson and Hodgson, 2009).

5.1.4 Sonar scans

Sonar scans may be used to image the interior surface of a fully submerged conduit including those
made of masonry, concrete and metal. Sonar technology involves the emission of an acoustic pulse from
a transducer and the subsequent receipt of an echo. The time delay from the instant of transmission to
the moment of reception is used to determine the distance to the reflecting surface.

Typically a rotating narrow-beam sensor is mounted on a ROV or floating platform, and moved steadily
through the conduit. The acoustic reflections from the conduit surface are recorded in a series of cross-
track slices. When stitched together each slice forms an image of the conduit, and the resulting images
can be combined and overlaid on a 3D wire mesh surface. At high frequency (eg 2 MHz) good resolution
may be obtained enabling identification of cracks and other surface features. Lower frequencies are
required for penetrating the conduit wall to examine wall thickness, delamination and other planar
defects (Andrews, 1998, and Duran et al, 2002), but slow scanning speeds may be required and the
resolution may be poor.

Conduits are most likely to be either

empty or fully submerged, but for
partially submerged conduits an ROV
with both sonar equipment and CCTV
(known as totally integrated sonar and
camera inspection technique, TISCIT)
might be used. This combination of
equipment can be either trolley or
crawler mounted (see Figure 5.3) or
float mounted dependent on conduit
diameter and flow conditions, although
it would be difficult to use such
equipment if the conduit contains an
inner pipe (see Case study A1.7).
Figure 5.3 TITSCIT ROV for combined CCTV and sonar scans (courtesy
OnSite Ltd)

42 CIRIA, C743
5.2 CONDUIT INTEGRITY

5.2.1 Introduction
1
The wide variety of conduits and materials encountered in practice necessitates several approaches
to integrity assessment. The common materials used are listed in Table 5.1 together with various
types of degradation. 2
Table 5.1 Degradation of conduit materials

Material

Masonry
Typical degradation

Spalling, softening, crumbling, flaking, blistering, cracking and mortar loss from joints
3
Concrete Cracking and spalling, reinforcement corrosion, sulphate attack, freezing and thawing

PCCP

CI, DI, steel

Delamination, loss of prestress, wire breakage, concrete degradation

Corrosion, rust, delamination, graphitisation corrosion

4
CMP Corrosion, rust

HDPE Should not corrode, joints and welds may fail

Where it is possible to access the inside of a culvert, NDTs may be used to supplement visual inspections
5
in assessing its condition and that of any pipes contained within it. Subsequently invasive techniques may

6
be used to obtain samples for laboratory testing. Where a pipe is surrounded by the embankment fill
various NDT techniques may be used to assess the condition of the pipe.

5.2.2 Non-destructive testing (NDT)

5.2.2.1 Introduction 7
NDT techniques are increasingly used to measure

8
the thickness and quality of buried pipelines, and
similar techniques may be used to assess the state
of conduits. When considering the use of NDTs
it is important that the objectives of the testing

9
are clearly specified, and to ensure that chosen
methods are well-suited. Some of the commonly
used NDTs are outlined in subsequent sections,
but the technology is continually developing.
Useful reviews of the techniques are given by
Lillie et al (2006) and Liu and Kleiner (2013). It
should be noted that these techniques are normally 10
undertaken by a specialist contractor who will also
interpret the tests.

5.2.2.2 Surface hardness

11
The surface hardness of concrete is often assessed

12
using the Schmidt hammer rebound test (BS EN
12504-1:2009). This widely used NDT (see Figure
5.4) measures the rebound of a spring-loaded
hammer when hit against a prepared concrete

A1
surface. The rebound depends on the energy
absorbed by the concrete, which may be loosely
correlated with its strength. In testing conduits the Figure 5.4 The Schmidt hammer (courtesy Cemex)

Dams and reservoir conduits 43

relative hardness gives an indication of the relative quality of the concrete at different locations, rather
than an absolute indicator of concrete strength. Voids can also be detected at times.

5.2.2.3 Ultrasonic test methods

Pulse techniques
Ultrasonic pulse techniques may be used to measure the thickness and quality of metal conduits. Point
measurements of the thickness are based on the time taken by the ultrasound wave to pass through the
pipe wall. While one-way measurements may be made where there is access to both sides, for buried
culverts and pipes a single probe incorporating both transmitter and receiver transducer may be used
and the time for the reflected wave to return to the surface is measured (see Figures 5.5 and 5.6). The
technique can be deployed readily in the field and is reported to give good accuracy. Careful surface
preparation is essential to obtain reliable measurements. The coating of any cleared areas should be
carefully reinstated.

Figure 5.5 Principle of ultrasound thickness Figure 5.6 Ultrasonic equipment (after Wright, 2011)
measurement (courtesy Advanced NDT Ltd)

Impact-echo technique
More sophisticated ultrasonic techniques may be required to determine the quality and thickness of
a concrete culvert because ultrasound disperses rapidly in concrete. The impact-echo technique was
invented at the US National Bureau of Standards and developed at Cornell University (Sansalone
and Streett, 1998). It has subsequently
been accepted into mainstream practice for
health monitoring of structures and further
descriptions are given by Carino (2001),
Malhotra and Carino (2004).

The impact-echo technique is based on the use

of transient stress waves generated by a small
impact. A diagram and photographs of the
method are shown in Figures 5.7 and 5.8. A
short-duration mechanical impact, produced
by tapping a small steel sphere against a
concrete or masonry surface, produces low-
frequency stress waves (up to about 80 kHz)
that propagate into the structure and are Figure 5.7 Impact-echo method – steel balls mounted on
reflected by flaws and/or external surfaces. spring-steel rods (after Carino, 2001)

44 CIRIA, C743
Multiple reflections of these waves within the structure

1
excite local modes of vibration, and the resulting
surface displacements are recorded by a transducer
located near to the impact (see Figure 5.9). The
fundamental equation of impact-echo is d = C/(2f),
where d is the depth from which the stress waves are
reflected (the depth of a flaw or the thickness of a solid
structure), C is the wave speed, and f is the dominant
2
frequency of the signal (obtained from the results of
a test). To determine thickness or depth of a flaw, the
wave speed C needs to be known. It can be measured by
observing the travel time of a stress wave between two
transducers held a fixed distance apart on the concrete
3
surface or by performing a test on a solid slab of known
thickness and observing the dominant frequency.

Figure 5.8 Impact-echo NDT method (courtesy

4
Impact-Echo)

Sansalone and Carino (1989) described

the first use of such tests to detect
5
delamination in concrete slabs. More
recently Marshall and Fisk (2010)
describe its use in assessing the
structural condition of a PCCP. They 6
state that the average strength of the
concrete pipe core, the presence of
wire breaks and areas of delamination,
can all be determined from the
measured resonant frequency and
7
wave attenuation. Gorzelańczyk et al
(2012) describe the successful use of
ultrasonic tomography and the impact-
echo method in determining the
8
thickness of a concrete tunnel lining.
Figure 5.9 Sub-surface crack diagram (courtesy Impact-Echo)

Guided wave ultrasonic testing (GWUT)

This technique uses the pipe wall as a conduit for the
9
ultrasonic energy. An externally mounted guided
wave propagation collar (see Figures 5.10 and 5.11)
is used to generate low-frequency ultrasonic Lamb
waves. This transmits ultrasonic pulses around the
10
full circumference of the pipe, which propagate along

11
the pipe for a distance of 30 m. Discontinuities such as
corrosion and cracks give a differing reflection back to
a receiving transducer at the collar, from which the size
and position of the defect may be determined. Losses of

12
wall thickness can also be identified and mapped.

Figure 5.10 Example of pipeline inspection (courtesy

A1
Spiralboy)

Dams and reservoir conduits 45

 Note
The green band indicates the
position of the transducer array.

Figure 5.11 A typical example of GWT data (courtesy Spiralboy)

5.2.2.4 Electromagnetic test methods

Several technologies are based on electromagnetic principles, but all are variations on two, which are
applicable to ferrous pipes, or pipes with a ferrous component:

‰‰ magnetic flux leakage (MFL)

‰‰ eddy currents.

Magnetic flux leakage (MFL)

This electromagnetic method of the NDT is used to detect corrosion, pitting and wall loss in metallic
pipes. With this technology, permanent magnets are used to temporarily magnetise the steel pipe and
the magnetic field changes are recorded and analysed. The magnetic flux is uniform if there are no
flaws in the wall of the pipe. If internal
or external flaws are present, such as
pitting, corrosion or other forms of
damage, the magnetic flux is distorted
outside the wall of the pipe, and this
distortion or ‘leakage’ is measured by
Hall Effect sensor devices whose outputs
vary in response to a changing magnetic
field. Thomson and Battelle (2009)
report that MFL is the most commonly
used inspection technology for oil and
gas pipeline inspection. However, MFL
tools cannot yet detect all metal loss
or reliably detect other defects such as
axial cracking. MFL technology can be
used inside a pipe (see Figure 5.12) and Figure 5.12 Magnetic flux leakage equipment inside a large diameter
also externally. mortar-lined steel pipe (courtesy Pure Technologies)

Remote field eddy current testing (RFEC)

RFEC is another electromagnetic field inspection technology that can be used inside metal and pre-
stressed concrete pipes, and can also assess their integrity (Atherton et al, 1989, and Atherton, 2000). A

46 CIRIA, C743
low frequency alternating current is used to energise a solenoid coil placed inside the pipe. This induces

1
eddy currents within the pipe wall and part of the field propagates inside the pipe and part passes
through the pipe wall and travels along its outside. The field inside is attenuated rapidly, and that on
the outside is detected by sensitive receiver coils placed just inside the pipe at least 2.5 pipe diameters
from the source. The travelling external field is affected by the thickness of the pipe wall, and changes
to the phase and magnitude of the field are used to detect defects such as corrosion and cracks. By using
equipment of nearly the same size as the internal diameter of the pipe with numerous receivers, it is
possible to survey the whole of the inside of the pipe. However, this equipment does not appear to be
2
widely used in the UK at present.

Electromagnetic testing/transformer coupling (RFEC/TC)

A similar RFEC electromagnetic inspection
3
technology may be used for condition assessment
of PCCP and in particular to detect broken
pre-stressing wires. Corrosion of the steel wires
can lead to wire breakage that could ultimately 4
result in bursting of the pipe under pressure.
In this technique a trolley-mounted transmitter
coil positioned close to the inside of the pipe
produces an electromagnetic field. This field
interacts with the pre-stressing wires in a
5
manner similar to a transformer. A second coil is
typically mounted coplanar with the first except
diametrically opposite to it, and acts as a pickup.
The pre-stressing wires in the pipe amplify
6
Figure 5.13 P-wave RFEC TC (courtesy Pure Technologies)
the signal that is recorded by the receiver. If

7
there are broken wires, the signal is distorted.
Figure 5.13 illustrates equipment that is operated in a dewatered pipe or, if necessary, robotically driven
through a depressurised pipe.

Low frequency electromagnetic technique (LFET)

The technique is potentially useful for monitoring
the degradation of metal pipes of a wide range of
8
sizes. It is based on electromagnetics, and consists

9
of a set of driver coils and corresponding pickup
coils. A low frequency alternating current (AC) in
the driver coils induces an AC magnetic field in a
metal pipeline, which is detected by the pickup coils

10
positioned close to the metal surface. If there is
pipe wall loss and pitting, a measurable distortion is
introduced to the induced field that can be detected
using the pickup coils. By using low frequencies (eg
5 Hz to 30 kHz), the penetration of the magnetic
field is fairly uniform throughout the pipe wall and
defects on both the inside and outside surfaces can 11
readily be seen on a display. Equipment may be
operated manually or using a remote controlled

Figure 5.14 32-channel line-cat crawler system using

crawler system (see Figure 5.14). With the latter
low frequency electromagnetic techniques to the scanner can be moved steadily along a pipe,
scan pipes at up to 3 m/min (courtesy TesTex) scanning speeds of 3 m/min have been reported.
12
Broadband electromagnetic system (BEM)
Whereas conventional eddy current techniques use a single frequency for testing, the BEM technique A1
transmits a signal that covers a broad frequency spectrum typically ranging from 50 Hz to 50 kHz (Liu

Dams and reservoir conduits 47

 et al, 2012, and Liu and Kleiner, 2013).
The recorded signal from a broadband
transmission contains more information,
and allows detection and quantification
of various wall thicknesses as well as the
effective conductivity of the complex
through-wall components of a pipe.
Changes in conductivity reflect changes
in material properties. The pipe wall
thickness is estimated indirectly by
measuring signal attenuation and phase
delay of the secondary magnetic field.

BEM scanning of pipes may be external

or internal, the former with a hand-
held scanner, the latter with an in-line
Figure 5.15 Internal BEM equipment used for scanning ferrous pipes
(courtesy Rock Solid Group) inspection pig, which is driven or drawn
through the pipe (see Figure 5.15). The
pipe needs to be out of service, emptied, and cleaned of loose deposits in order to run the pig. In CI
pipes, BEM can identify and locate metal loss, graphitisation and cracks, and identify wire breakage in
PCCP. It does not require direct contact with bare metal to detect pits and other metal loss.

5.2.3 Other test methods

A number of other innovative methods that are still in the research stage have been considered for integrity
testing of pipes and sewers. These include spectral analysis of surface waves (SASW), micro-deflection,
thermal imaging and fibre optics (see Tuccillo et al, 2010, Lui et al, 2012, and Liu and Kleiner, 2013).

5.2.4 Invasive coring and testing

Where the size of the conduit permits access, its wall thickness and condition may be investigated by
drilling core holes using rotary coring equipment. Such equipment may be bolted to the conduit surface
(see Figures 5.16 and 5.17) or if space permits a larger rig may be introduced. Carefully executed drilling
of coherent materials should yield good core or sample recovery, and subsequently these may be sent to
a laboratory for testing. In planning such coring, provision should be made for preventing the possible
ingress of water, and if necessary a stuffing box should be used. Where the presence of an external void is
suspected, these may be investigated through full-depth core holes using an endoscope. Each hole needs
to be thoroughly sealed to avoid future problems. Before undertaking such coring a thorough assessment
should be made of the conditions behind the wall including the presence of water under pressure.

Figure 5.16 Taking a 100 mm core through a brickwork tunnel lining Figure 5.17 Coring through a cast-iron
(after McKibbins et al, 2009) underground tunnel lining
(after Wright, 2011)

48 CIRIA, C743
Laboratory testing may then be undertaken to:

‰‰ identify the materials

‰‰ assess their physical and chemical 1
characteristics
‰‰ assess their current condition
‰‰ assess the presence extent and causes of
material degradation.
2
Testing may include physical testing of cores (eg
hardness, strength and stiffness, see Figure 5.18),
microscopic/petrographic examination of samples to
identify material types and assess their degradation,
3
and chemical analyses where appropriate. A detailed
description of the range of laboratory investigation
techniques is beyond the scope of this document,
however detailed guidance of the following can be
4
found elsewhere:

‰‰ brickwork and masonry in general (see Edgell

et al, 2005) 5
‰‰ metallic materials (see Bussell 1997, Rajani,
2000, and Makar and Rajani, 2000)

6
Figure 5.18 Compressive strength testing of a 300 mm
‰‰ concrete (Concrete Society, 2000 and 2014, diameter brick core while simultaneously
CBDG, 2002, and BRE, 2000). measuring strain (courtesy CERAM)

5.3 LEAKAGE INVESTIGATIONS

Concentrated uncontrolled leakage of water through a dam may result in dangerous internal erosion or
7
a build-up of pore water pressures within the downstream slope. As was discussed in Chapter 3, failure

8
modes linked to water leakage include erosion of soil by:

‰‰ water flow into a non-pressurised conduit

‰‰ water escaping from a pressurised conduit
‰‰ water flowing along outside a conduit
‰‰ instability due to water flow through the dam near to a conduit.
9
Visual inspection of the embankment and conduit and leakage/seepage monitoring are likely to give the
first warnings of problems. Initial investigations will focus on establishing the level of risk to the dam,
identifying the source of the water, and whether it is carrying sediment in suspension.

The first step in any investigation should be to develop a detailed understanding of the dam cross-
10
section and its construction, the underlying geology, and hydrogeology. Further detailed investigations
may be required to determine the water origin, preferential flow paths, recharge zones, and transit time
as discussed by Charles et al (1996) and may include: 11
‰‰ flow rate and pore water pressure monitoring
‰‰ chemical analyses
‰‰ flow studies with natural and artificial tracers
‰‰ temperature measurements
12
‰‰ geophysical tests.

There are numerous geophysical test methods that may be considered, and several new technologies
have been developed since the study by Charles et al (1996). Of these the most promising techniques
A1
Dams and reservoir conduits 49
for use in the context of conduits are described here. The interpretation of geophysical tests should be
undertaken by the specialists who design and carry out the test programme.

5.3.1 Initial investigations

5.3.1.1 Flow measurements
The initial investigations should include the establishment of flow monitoring system such as an
automated V-notch or weir (see Section 6.4.2). There should be provision for trapping suspended soil
and measuring turbidity. Measurements should be taken over a period of time and plotted against the
reservoir level. If there is a single defect such as a crack, soil pipe or sink hole, leakage from upstream
may be linearly related to reservoir level and fall to zero below the level of the level of the defect. Multiple
leaks will result in more complicated relationships.

5.3.1.2 Water analyses

The investigations should attempt to establish whether the source of water is the reservoir itself, or
whether much of the flow derives from rainfall, and whether it is carrying sediment in suspension
(Charles et al, 1996). This will necessitate sampling the water both from the reservoir and where the
water emerges, and then comparing both the chemical analyses and the presence of natural tracers. The
geochemical signature of the water will be influenced by the ground or rocks through which it flows, but
valuable information may be obtained from such analyses.

Chemical analyses can be supplemented by the use of tracers (Charles et al, 1996, Contreras and
Hernández, 2010). Water generally contains natural tracers such as the stable heavy isotopes,
Hydrogen-2 (deuterium) and Oxygen-18. Comparisons of their relative abundance may be used for
identification of the source of the water. The point of inflow may be explored through the use of tracers
such as salt and fluorescein.

5.3.1.3 Pore water pressure measurements

While the water analyses will be useful, readings from piezometers can also play a part in identifying
the leakage path. Piezometric levels should be correlated with reservoir level and rainfall to identify
anomalies. Piezometers installed immediately upstream of a central core of a rockfill dam may show
a local depression of piezometric level near to a leak through the core. Elevated piezometric levels
downstream of the core also indicate localised leakages, although it is not likely that piezometer readings
alone will be sufficient to pinpoint the defects because of their limited zone of influence.

5.3.2 Temperature measurements

Temperature measurements have been used since the 1950s for detecting leakage in embankment dams.
Where there is a direct connection to the reservoir through a defect the water emerging downstream will
be at the same temperature as the reservoir when the transit time is small.

Seepage carries heat through the embankment in an advective process that gradually alters the
temperature of the ground through which it passes. An investigation of ground temperatures across
the dam may reveal anomalies that indicate developing problems, and longer term monitoring of the
temperature of concentrated flows in or around conduits may also be useful. The interpretation of such
measurements is difficult if there is not a significant seasonal variation of reservoir temperature. It can
be useful to time the survey to take advantage of the maximum temperature differentials (ie in summer
or winter). The use of temperature probes for investigations of concentrated leakage is described here,
while routine monitoring of dams using fibre-optic systems is described in Section 6.4.2.

50 CIRIA, C743
5.3.2.1 Profile of temperatures
In this technique temperature measurements
are made at regular intervals along a section 1
typically aligned with the axis of the dam
(Dornstädter et al, 2006). Metal tubes are driven
or drilled into the ground at about 5 m to 10
m intervals (see Figure 5.19) and chains of
temperature sensors generally placed at one
2
metre intervals are inserted in the tubes. The in

3
situ ground temperatures at different depths are
taken after the tube’s temperature has equalised
with the ground temperature. The measured
temperatures are immediately mapped on a

4
computer in the field (see Figure 5.20), and
additional temperature probes can be installed
Figure 5.19 Installation of an array of temperature probes
where temperature anomalies are detected. So, (from Dornstädter, 2013)
vertical and horizontal boundaries of seepage
zones are localised on site. The system was developed in Germany, and has been used for surveys at a
number of UK dams including Cam Loch (Dornstädter et al, 1997), see Case study A1.2.
5
6
7
8
Figure 5.20 Temperature distribution resulting from a survey of an array of temperature probes
(from Dornstädter, 2013)
9
5.3.2.2 Temperature measurements at specific locations
Under normal circumstances the variation of temperature of the seepage water lags behind that of
10
the reservoir due to the elapsed seepage time, but where the temperatures are found to be in phase

11
this indicates a direct connection. Measurements may focus on points where water enters the conduit
and/or at its downstream end, or perhaps water in any seepage zone near to the conduit as at Brent
Reservoir (see Case study A1.2), and/or at its downstream end. Temperature measurements made at
various positions along a submerged conduit could also help to establish the points of water entry. Such

12
measurements require the installation of temperature probes as described in Section 6.4.2.

5.3.3 Electrical resistance measurement (electro-

scanning)
Defects in conduits are not always observable unless water is flowing through them and leakage into or
from a fully submerged conduit may be difficult to detect and pinpoint. Electro-scanning technology
A1
Dams and reservoir conduits 51
(ASTM F2550-06, Harris and Dobson 2006, and Tuccillo et al, 2010) developed for locating leaks from
sewers could also be used for locating leaks in fully submerged conduits. Conduit materials such as
masonry, concrete and plastic are poor conductors of electrical current. However, a defect in the conduit
wall that leaks water will also leak electrical current, whether or not water infiltration or exfiltration is
occurring at the time of the test. Measurements are made of the current flowing from an electrode at
the ground surface to a special sonde (see Figures 5.21a and 5.21b) as it is drawn or pushed through a
flooded pipe. The electrical resistance of the current path between the sonde and the surface electrode
is very low except through the conduit wall. The sonde is designed to focus the measured current flow so
as to pinpoint the location of leaks, and repeat tests have shown good repeatability (see Figure 5.21c).

The technology was evaluated in field demonstrations in the USA and reported by Martell et al (2011) where it
compared favourably with CCTV and multi-scanning technologies. Electro-scans were first demonstrated in
the UK in a sewer survey in Topsham for South West Water in 2012. It is not known whether this technology
has been used for the assessment of submerged conduits in embankment dams, but the indication is that it is
likely to be more effective than CCTV surveys at pinpointing submerged leaks.

a c
Figure 5.21 Electro-scanning equipment, sonde and cable (a), principle of operation (b) and scan showing variation of
scan current with position and demonstrating repeatability (c) (courtesy Electro Scan Inc.)

5.3.4 Electro-magnetic
There are non-intrusive geophysical methods available that are specifically designed for mapping
groundwater (Willowstick, 2014). This low frequency (380 Hz) electro-magnetic method involves the
establishment of an electric current through the groundwater by applying an alternating voltage between
an upstream electrode positioned in the reservoir and a second electrode placed in the seepage flow
downstream. An alternating current in a conductor generates an alternating magnetic field around the
conductor, and in this application the conductor is the water flowing through the dam. The system is
shown diagrammatically in Figure 5.22. A portable unit containing three magnetic coil sensors (X Y
and Z axes) is used to map the resulting magnetic field across the surface of the dam (see Figure 5.23)
and to track patterns that help characterise groundwater flow. On processing of the data, 2D maps and
3D models are generated and interpreted, in conjunction with other hydrogeologic data, to provide
an enhanced definition of preferential groundwater flow paths, enabling individual flow paths to be
identified and plotted both on plan and elevation. The technology has been used at numerous dams
around the world and several in the UK including Torside Reservoir (Kofoed et al, 2008) and Bartley
Dam (Kofoed et al, 2014).

The technology has been used at numerous dams around the world including those in the UK. An
example of the output is given in Figure 5.24.

52 CIRIA, C743
 1
2
3
4
Figure 5.22 Diagram illustrating the AquaTrack™ system (courtesy Willowstick Technologies LLC)

5
6
7
8
Figure 5.23 Portable Willowstick mapping equipment combining magnetic

9
sensors, GPS and laptop (courtesy Willowstick Technologies LLC)

10
11
12

Figure 5.24 Summary interpretation, Shon Sheffrey dam, Wales (courtesy Willowstick
A1
Technologies LLC)

Dams and reservoir conduits 53

5.3.5 Other geophysical methods
A number of other geophysical techniques have been developed for leak detection and tracing in
embankment dams. These include acoustic methods, resistivity, and self-potential methods. The
latter two are discussed by Johansson (2004), but appear to be less straightforward to interpret than
temperature and Willowstick testing.

5.4 VOID DETECTION

Where there has been significant leakage into or out of a conduit or through the surrounding ground,
it is possible that erosion of soil has resulted in the development of voids near to the conduit. The
principal technology to evaluate the material surrounding the conduit and identify possible voids is
ground penetrating radar (GPR) undertaken from within the conduit. Where anomalies are found this
is usually followed up by invasive drilling, inspection through core holes using endoscopes and then
grouting if necessary.

5.4.1 Ground penetrating radar (GPR)

GPR is commonly used to locate buried services and voids beneath the ground. It operates by
transmission of electro-magnetic waves at frequencies between 50 MHz and 2.5 GHz. The waves are
reflected back when they encounter a change of surface with contrasting dielectric properties. The
choice of antenna frequency is important and depends on the depth of interest and materials involved.
Clayey or saturated soils may attenuate the signal and limit the depth/range to which this technology is
effective to a few metres. Most GPR systems use wave pulses, and are interpreted in the time domain,
and the distance to the reflective boundary is inferred from the time delay between transmission and
receipt. Currently researchers are exploring the use of frequency and spatial domain techniques (Hao et
al, 2012). All GPR data are difficult to interpret and this requires highly skilled operators.

GPR equipment has been developed specifically for investigations through non-metallic pipelines, sewers
and tunnel linings (Parkinson and Ékes, 2008). Ékes and Neducza (2012) describe a project where such
equipment was successfully used to assess the condition of a 0.7 m diameter reinforced concrete pipe.
Sometimes referred to as pipe-penetrating radar (PPR) this may also be used in evaluating the condition
of conduits, conduit bedding and the presence of voids in the surrounding ground. It can be operated
from within a tunnel either using a hand-held scanner or mounted on a ROV in conjunction with other
equipment such as CCTV and laser scanning (see Figure 5.25). The equipment may be used to make a
series of longitudinal and transverse
scans to provide information on
the thickness and condition of the
conduit and the presence of voids
behind.

Voids that have already been

identified are extremely useful in
calibrating a GPR survey, because
GPR data profiles can be difficult
to interpret. On identifying other
anomalies it is then necessary to
follow up with intrusive investigation.

Such equipment has recently been

used for investigations of a reinforced
concrete tunnel lining at Glendoe
Reservoir near Loch Ness where
Figure 5.25 Robotic equipment combining CCTV, PPR and LiDAR for
voids were suspected. A grid of GPR internal inspection of pipes, sewers and conduits up to 900
mm dia. (courtesy SewerVUE Technology Corp)

54 CIRIA, C743
profiles was collected along the 200 m tunnel, with supplementary low frequency ultrasonic testing in

1
targeted locations. The survey identified 23 voids with sufficient accuracy to enable targeted repairs
(Fugro, 2013).

5.4.2 Other test methods

A number of other innovative methods that are still in the research stage have been proposed for
investigating voids and bedding around pipes and sewers. These include seismic tomography, gamma-
2
gamma logging, and infrared thermography. FEMA (2005) gives details of a site where zones of low
compaction around a corrugated metal pipe were successfully identified using seismic tomography.
These techniques are discussed in the reports for US Environmental Protection Agency (USEPA) by
Tuccillo et al (2010), Liu et al (2012) and Liu and Kleiner (2013).
3
4
5
6
7
8
9
10
11
12
A1
Dams and reservoir conduits 55
6 Instrumentation and
monitoring

This chapter is concerned with the equipment and techniques that can be used to detect and record
changes within and around the conduit, and predict future behaviour and performance.

6.1 PURPOSES OF MONITORING

Regular visual inspections of dams and conduits are central to assuring dam safety as discussed in
Chapter 4. Indeed, in discussing observation and instrumentation Peck (1972) wrote that:

“An instrument too often overlooked in our technical world is a human eye connected to the brain
of an intelligent human being. It can detect most of what we need to know about subsurface
construction. Only when the eye cannot directly obtain the necessary data is there a need to
supplement it by more specialized instruments. Few are the instances in which measurements
themselves furnish a sufficiently complete picture to warrant useful conclusions.”

So instrumentation may play an important role in providing supplementary data for use by owners
and engineers in the assessment of the current condition and performance of the dam and associated
conduits. Although it is not common UK practice to install instrumentation specifically to monitor
conduits, typically instrumentation will be used:

‰‰ during construction of a new dam to confirm that its geotechnical performance is within
predicted limits
‰‰ for routine in service general health monitoring to give warning of possible developing problems
(USSD, 2013)
‰‰ in exceptional circumstances as part of investigations of specific problems.

Before planning the instrumentation in detail, the geotechnical questions to be answered should be
considered carefully. Dunnicliff (1988) wrote that “every instrument on a project should be selected and
placed to assist with answering a specific question; if there is no question, there should be no instrumentation”. It is
suggested that instrumentation should be planned to assist with evaluation of a specific possible failure
mode as discussed in Section 6.2.

There is an extensive literature on the use of geotechnical instrumentation for monitoring field
performance (eg Dunnicliff 1993, 2012, ASCE, 2000, and Bassett, 2011). Rather than consider the wider
uses of instrumentation in dams, this chapter focuses on instrumentation relevant to conduits.

6.2 IDENTIFICATION OF KEY PARAMETERS

Before planning a programme of monitoring there should be a thorough assessment of possible modes of
failure linked to the conduit (see Chapter 3) and the associated geotechnical uncertainties. The common
failure modes discussed previously were erosion of soil by water:

‰‰ flow into a non-pressurised conduit

‰‰ escaping from a pressurised conduit
‰‰ flowing along outside of a conduit
or instability resulting from water flow through the dam near to a conduit, or collapse of a conduit.

56 CIRIA, C743
Key visual indicators should be identified that will give reassurance, or alternatively give early warning

1
of developing problems. However, there may be little visual indication of some problems initially (for
example egress of water from a pressurised conduit), so quantities to be measured with instrumentation
need to be identified carefully. Instrumental monitoring may give sufficient warning that dam failure
can be averted, or at least enable timely evacuation to avoid loss of life.

The primary quantities to be assessed are likely to be one or more of:

‰‰ water ingress or egress and suspended sediment

2
‰‰ seepage conditions near to the conduit
‰‰ conduit movement, alignment and distortion
‰‰ conduit condition 3
‰‰ the presence of voids outside the conduit.

Other quantities may also be relevant to the developing condition of the conduit such as surface
deformations and subsurface movements of the dam near to the conduit. 4
6.2.1 Prediction of values and planned response
An important part of the design of an effective instrumentation scheme is the prediction of the observed
parameters and their expected variation with time and reservoir level for all conceivable circumstances.
5
The values at which further actions on the part of the dam owner and engineer are triggered should

6
be identified. Also, the actions that should be taken in response to deteriorating conditions should be
planned in some detail, and on-site emergency plans should be drawn up. Originally developed by
Terzaghi and now known as the Observational Method (Peck, 1969, and Nicholson et al, 1999), this
approach to the use of instrumentation provides a robust means of managing safety.

6.3 SPECIFICATION OF INSTRUMENTATION 7

In specifying any instrumentation it is essential to ensure that it is suitable so that it will give reliable
readings over the required period of observations. Dunnicliff (1988) provides detailed advice on the
selection of appropriate instrumentation. Some key factors to consider are: 8
‰‰ purpose of the instrumentation and parameters to be monitored

9
‰‰ locations at which readings are required
‰‰ frequency of monitoring required
‰‰ simplicity of operation and the need for special readout units
‰‰ manual reading vs. requirements for data-logging and remote reading
‰‰ range of expected values, precision, accuracy, repeatability
‰‰ the operational environment that may affect durability
10
‰‰ design life, reliability, robustness, possibility of replacing faulty equipment

11
‰‰ requirement and feasibility of recalibration in service
‰‰ need for instrumentation redundancy to enable validation of unexpected data
‰‰ need for electrical power
‰‰ means of installation and possible associated risks
‰‰ effect of instrument and its installation on measured parameters.
12
A specialist instrumentation supplier should be involved in the specification at an early stage, as well
as the subsequent equipment installation and commissioning. While simple instrumentation can be
read and the readings recorded and processed by the dam owner, more complicated installations may
necessitate regular visits by a specialist technician. A1
Dams and reservoir conduits 57
6.4 TYPES OF INSTRUMENTATION

6.4.1 Introduction
When a conduit is constructed as part of a new dam there is the opportunity to install instrumentation
to monitor its performance. Instrumentation can be installed outside the conduit that can be used for
routine monitoring. For dams already in service any instrumentation will probably have to be installed
inside or through the walls of the conduit, the instrumentation and any associated pipes or cables will
require careful protection, and the conduit will have to be sealed.

There are numerous types of geotechnical instruments that could be used in monitoring conduits and
the surrounding ground. Many of these are described in some detail by Dunnicliff (1988 and 2012) and
so detailed descriptions will not be given here. Over the last 10 years numerous instrumentation systems
have been developed that can be data-logged and read remotely, and new systems are being developed all
the time. Bearing in mind that instrumentation for dams should be as simple and robust as possible – the
focus here is on those systems that have proved reliable over many years. Some of the instruments of most
relevance to monitoring conduits are listed in Table 6.1, but clearly other systems could be included.

Table 6.1 Instrumentation for monitoring conduits

Quantity measured Possible locations Instruments

‰‰ calibrated container, weir, flume, V-notch
Water ingress Defects, conduit exit ‰‰ sediment trap, jar samples, turbidity meter
‰‰ water temperature.
‰‰ weir, flume , V-notch
Seepage from around
Near to exit ‰‰ sediment trap, jar samples, turbidity meter
conduit: rate and quality
‰‰ water temperature.
Pore-water pressures Outside conduit ‰‰ piezometers.

Total stresses Outside conduit ‰‰ total stress cells.

‰‰ staff gauge
Reservoir level Reservoir
‰‰ piezometer.
‰‰ internal survey, hydrostatic profile gauge, liquid level gauge,
horizontal inclinometer, Shape accelerator array (SAA)
Conduit settlement, horizontal magnetic extensometer
Crown, invert, haunches ‰‰
elongation and distortion
‰‰ laser distance measurement
‰‰ electrolevels.
‰‰ strain gauges, demec gauge, crack meter (manual and VW),
Conduit condition Joints and defects
electrolevels.
‰‰ inclinometers, survey targets
Dam body movements Dam body at end of conduit
‰‰ extensometers.
‰‰ survey targets, LiDAR, interferometric synthetic aperture
Surface movements Dam surface above conduit
radar (InSAR).

6.4.2 Water measurements

Several possible modes of dam failure are linked to internal erosion resulting from ingress of water
into a conduit through open cracks or joints, leakage from a pressurised conduit, or other unplanned
seepage/leakage through the ground near to a conduit. A breach could develop if a flow of water erodes
and transports embankment or foundation material either into the conduit or from the dam near the
downstream end of the conduit.

A combination of regular visual inspections of the embankment and conduit together with regularly
monitoring seepage flows and pore pressures are considered to be best practice to give warning of
developing problems (FEMA, 2005 and USSD, 2013). Any sudden change of seepage/leakage flow rate
without apparent cause could indicate a developing problem.

58 CIRIA, C743
Instrumentation may be used to monitor:

‰‰ rate of water ingress through cracks or joints

‰‰ rate of seepage from around the conduit and nearby 1
‰‰ water pressures outside the conduit

2
‰‰ water temperature measurements at key locations
‰‰ water level in the reservoir and its temperature.

6.4.2.1 Water ingress into conduit

Water ingress through individual cracks or joints may be measured using calibrated containers and
a stopwatch. The total inflow is best monitored using a weir, V-notch or flume constructed at the
3
downstream end of the conduit. Continuous measurements may be obtained using a level gauge linked
to a data-logger, but precautions should be taken to ensure that the level gauge cannot be restricted by
debris. The readings may then
be related to reservoir water level 4
and rainfall, and will indicate
whether flow rates are constant
or increasing with time. An
example of a V-notch is shown
in Figure 6.1. V-notches should
5
be constructed to ISO 1438:2008

6
and can have different angles
of notch according to the flow
being measured. Measurement
of localised flows from defects in

7
conduits may necessitate use of
pipes or gutters to channel the
flow to the point of measurement.
Such measurements were
undertaken recently at Brent
Reservoir (see Case study A1.2).
Figure 6.1 V-notch weir for monitoring flow rate (courtesy Severn Trent
Water)
8
There should also be provision for monitoring sediment carried with the water. Regular visual
inspections will show discoloured or turbid water at the point of water entry, sediment deposited along
the base of the conduit and behind any weirs, V-notches etc. Such examinations may be supplemented by
regularly sampling the water or using a turbidity meter. 9
6.4.2.2 Seepage flows
If water is emerging from around the conduit at its downstream end (for example if the conduit is
surrounded by a granular layer), or from the dam nearby, the flow may be measured using a weir,
10
V-notch or flume preferably sited upstream of the toe drainage system. Continuous readings may be

11
obtained using a level gauge linked to a data-logger – note that precautions should be taken to ensure
that the level gauge cannot be restricted by debris or the notch/flume accumulate debris. The readings
may then be related to reservoir water level, and will indicate whether flow rates are constant or
changing with time and/or reservoir level, or any otherwise anomalous data.

Again there should be provision for monitoring sediment carried with the water. Regular visual
inspections will show discoloured or turbid water at the point of water entry, sediment deposited along 12
the base of the conduit and behind any weirs, V-notches etc. Such inspections may be supplemented by
regularly sampling the water and testing it (or just letting it stand) or using a turbidity meter.

If chemical dissolution is suspected, for example where the dam is in an area of carbonates or gypsum,
chemical tests on the seepage water may be appropriate.
A1
Dams and reservoir conduits 59
6.4.2.3 Pore water pressures
Piezometers can give useful information on the patterns of seepage within a dam and uplift pressures in
its foundations although discrete measurements can only ever give part of the picture – almost inevitably
the piezometer will not be located at the point where a leakage path is opening up. Piezometric levels
should be plotted against time alongside reservoir levels as shown in Figure 6.2.

Figure 6.2 Typical plot of piezometer readings (courtesy Canal & River Trust)

These piezometers are located along the dam axis (see Andrews and Dornstädter, 2000). The
distribution of pore-water pressures along the outside of a conduit may provide useful information about
the presence of leakage paths through the dam, or out from pressurised conduit. Standpipe piezometers
have been installed on either side of the conduit at Shon Sheffrey in Wales (see Case study A1.14).
Where piezometers are installed within a permeable collar placed around the whole circumference
of the conduit, any seepage along the outside of the conduit will be intercepted and will influence the
piezometer readings. Further information may be gained from temperature measurements (Andrews
and Dornstädter, 2000, Dornstädter et al, 2006 and Sjödahl and Johansson, 2012).

Piezometers are instruments that measure the pore water pressures at a specific point in the ground,
and they need to be installed carefully to ensure readings are reliable. In new dams piezometers are
often buried with the cables or tubing is run in trenches through the body of the dam to a gauge-
house. In existing dams piezometers will often be installed at the base of a borehole, which is then
sealed with cement-bentonite grout to ensure it does not create an unwanted water path. Great care
needs to be taken to ensure the installation process does not unwittingly damage the core of the dam
by hydraulic fracture.

There are several types of piezometer used in dams, including standpipe (or Casagrande) piezometers,
hydraulic piezometer, pneumatic piezometers, fibre-optic piezometers and vibrating wire piezometers.
The systems differ both in the reading system and, depending on the permeability of the surrounding
soil, in their response time. A detailed discussion is given by Dunnicliff (1988). The terminology is
sometimes confusing. Piezometers installed in a borehole have seals above and below and measure water
levels at a specific point. Observation wells or standpipes have no seals, but are surrounded by gravel and
measure average water levels in the overlying ground.

For measurements in permeable soils, standpipe piezometers may be satisfactory, and readings may be
taken manually or logged using a pressure transducer. Where the piezometric level drops below the
piezometer tip the water will drain out, and when it subsequently rises, air in the tip will readily vent
into the atmosphere. Hydraulic piezometers were commonly installed in UK dams during construction.

60 CIRIA, C743
These may be de-aired periodically giving confidence that the readings are reliable, but they are less

1
suited to data-logging. Also the pipework should not rise more than five metres above the piezometric
level being measured.

For measurements in low permeability soils and in situations where data-logging is needed, rapid
response piezometers (eg vibrating wire piezometers) should be used. If not installed during dam
construction, they may be installed either in a borehole or through a core hole from the conduit,
however there are cases where they are known to drift. Zero drift should be checked before installation. 2
Vibrating wire piezometers are known to be reliable and are probably the most widely used remote
reading piezometer for that reason. However, consideration should be given to using a second system (eg
hydraulic piezometers) to corroborate the readings. In situations where the piezometric level drops below
the piezometer tip, the tip should be made from low air-entry ceramic, so that when water levels recover,
the ceramic tip will de-air itself and the correct water pressure is registered by the instrument.
3
6.4.2.4 Water temperature measurements
If the temperatures of the reservoir water, water in the toe drain and the conduit water are recorded
4
and plotted against time, this can yield useful information about the water path from the reservoir to

5
the point of water exit. Where the temperature of exit water is out of phase with the reservoir water
temperature, this indicates that the time taken for the water to pass through the dam is significant.
Conversely if the water is of a similar temperature to the reservoir water, there is probably a direct
connection. Temperature measurements may be obtained using a variety of instruments. Vibrating

6
wire piezometers routinely have thermistors in the tip that may be used for temperature measurements.
Recently fibre-optic systems have been installed in a number of larger dams to obtain distributed
temperature measurements (Beck et al, 2010, Dornstädter 2013, Johansson and Sjödahl, 2009, Almog et
al, 2011, and Sjödahl and Johansson, 2012) as part of routine monitoring.

6.4.2.5 Reservoir level and toe drain water 7

The water level in the reservoir should be monitored so that the flow/leakage rates may be related to
reservoir level. While manual reading using a staff gauge may be sufficient, if other instrumentation
is being logged, it is straightforward to monitor the water level using piezometer equipment such as a
vibrating wire piezometer fitted with a thermistor for measuring temperature. Seepage emerging from
the dam into the toe drainage system should be routinely monitored at one or more locations using a
8
weir, V-notch or flume with continuous measurements obtained using a level gauge linked to a data-

9
logger. During an emergency it is important to monitor levels and flows regularly.

6.4.3 Total stress measurements

10
The stresses exerted by fill on the top and sides of a conduit at the end of construction are strongly
influenced by the relative stiffness of the conduit, the ground supporting it and the surrounding fill.
If the conduit is on a stiff foundation and settles less than the surrounding fill it will attract stress,
whereas if it is on a relatively compressible foundation the fill may arch over the top of it. If the pore

11
water pressure is of a similar magnitude to the total stress there is then a risk that hydraulic fracture
may develop.

Conduits in new dams may be instrumented using total stress cells on the surface of the conduit to
measure total stress. Typically these are constructed from two stainless steel plates, welded around their
periphery with the narrow gap between the plates filled with hydraulic fluid. The pressure developed in
the fluid is measured using a pressure transducer. The measured total stress may then be compared with 12
pore water pressures measured with nearby piezometers.

A1
These instruments have not been found to be reliable after any significant period of time and
interpretation is very difficult because readings are very sensitive to installation and contact stresses.

Dams and reservoir conduits 61

6.4.4 Deformation of conduit
Movement of the conduit could open cracks or joints that might eventually lead to escape or ingress of
water or even to collapse of the conduit if it became severely overstressed and/or distorted. So the main
purpose of monitoring conduit deformations is to quantify distortions. Where dams and conduits are
founded on compressible strata they are very likely to experience differential settlements, and potentially
longitudinal expansion. Flexible conduits may also experience distortion of their cross-section. Where
rigid pipes are enclosed within a culvert, they may be subject to distortion arising from deformation
of the culvert that could eventually lead to fracture. While the greatest movements are likely to occur
during the early years of a dam’s life, movements of highly stressed foundations may continue for a long
time due to creep.

There is an extensive general literature on soil-structure interaction and the relationship between
damage and distortional strain for different types of structure (see Kastner et al, 2003). It is likely that
instrumentation will focus on settlement and elongation and will probably not give complete information
about distortions.

The simplest instrumentation will involve visual observations. Where this is not possible, tubing may be
installed along the full length of the conduit for the insertion of probes, or if continuous readings are
required transducers may be installed at discrete locations. Some of these systems are outlined as follows.

6.4.4.1 Survey techniques

The deformations of large conduits that can be accessed
easily may be measured by undertaking periodic surveys.
Permanent monitoring stations should be set up that can
be surveyed repeatedly over many years.

Carefully planned and executed precise level surveys with

a precise level and invar staff (albeit with illumination) can
detect movements to better than ±1 mm. If the conduit
is reasonably straight, sighting from the conduit exit onto
scales permanently mounted on the conduit may be possible,
although lesser accuracy may be achieved. Alternatively a
laser level or piper laser on a permanent mounting at the
exit to the conduit may be used in the measurement of offsets
from the laser beam to the crown and sides. Longitudinal
distances from the same mounting can be measured using a
laser distance meter reflecting off targets. The instruments
Figure 6.3 Typical dimensions to be determined should be mounted in a completely stable position that can be
accurately (courtesy R F T Nash) confirmed by external surveys.

Diametral and diagonal lengths

may also be measured with a tape
extensometer or a laser distance
meter supported on mountings on
the sides of the conduit (see Figure
6.3). With secure mountings
for the instruments and targets,
distance readings may be
obtained with an accuracy of a few
millimetres or better.

In tunnel construction total

Figure 6.4 Arrangements of targets and of electronic distance
stations placed on several
measurement (EDM) measuring stations along the tunnel axis
measuring stations (see Figure 6.4) (from Kontogianni and Stiros, 2003)

62 CIRIA, C743
along the tunnel axis are used to determine the 3D co-ordinates of targets fixed on the primary lining

1
with an accuracy of a few millimetres (Kontogianni and Stiros, 2003), but this is not likely to be feasible
in a small conduit.

6.4.4.2 Measurements with probes

The use of probes inserted into tubing is well established in geotechnical practice. Horizontal tubing
has been used in a number of dams for measurement of vertical and horizontal deformations. Several
2
instrumentation systems are available in which the probe is drawn through the tubing by a cord with a

3
return pulley and readings taken at regular intervals.

Settlements may be measured using a hydrostatic profile gauge drawn through profile tubing fixed
to the conduit. The measuring system consists of a probe containing a sensitive pressure transducer
connected by small-bore fluid filled tubing to a reference tank that is normally housed in the tubing reel.
As the probe is pulled through the profile tube, it is positioned using physical markers on the connecting
tubing and readings are taken of the pressure differential at that point between the reel and the probe.
From this, the change in elevation of each point along the profile tube can be measured and movements
4
calculated by comparison with the base readings. The profile tubing should be sufficiently flexible to
accommodate movement of the conduit. The reel is positioned on a stable reference datum near the exit
to the conduit. The repeatability of such systems is not particularly good and where accurate readings
are required and inclinometer or a permanent installation (see Section 6.4.4.3) would be preferable.
5
6
Settlements may also be measured using a horizontal inclinometer. The probe contains an accelerometer
to measure its inclination to the horizontal and is mounted on spring-loaded wheels normally 0.5 m
apart that run in grooves in special tubing. The PVC or Acrylonitrile butadiene styrene (ABS) tubing
should be fixed to the conduit at regular intervals with one pair of grooves oriented vertical. It should
also have telescopic joints that are located carefully in relation to joints in the conduit. The probe
is drawn through the tubing manually and readings taken at intervals equal to the wheel spacing.
Successive readings of tilt are stored in the readout unit and the vertical position of the casing relative to 7
the end of the tubing is obtained by integration. The vertical deflections of the casing are obtained by
comparison with the base set of readings.

Horizontal movements can be measured using a magnet extensometer system in conjunction with either
of the these systems. Here a probe containing a reed switch is drawn through the profile tubing, which
8
passes through a number of ring magnets that are fixed to the tube (see Figure 6.5). As the probe passes

9
through each ring magnet, the reed switch closes and completes a circuit sounding a buzzer at the cable
drum. The cable is graduated allowing measurement of offsets from a reference point at the conduit exit
to be determined.

10
11
12
A1
Figure 6.5 Horizontal magnet extensometer installation (courtesy Sisgeo)

Dams and reservoir conduits 63

6.4.4.3 Permanent installations
If it is necessary to monitor settlement of a conduit at discrete points without gaining access regularly
there are numerous instrumentation systems that might be used. In selecting instrumentation the
designer should consider the points given in Section 6.3.

There are several systems based on the liquid level principle. The hydraulic overflow gauge is
conceptually the simplest, and uses the principle that the levels of a liquid in a U-tube will be at the same
elevation in each limb. Overflow cells fixed at the points of interest are linked by tubing to a readout
installation at a similar elevation at the exit of the conduit. De-aired water is introduced to fill the tubing
and allowed to overflow into the measurement cell. The water levels at each end of the tube then stabilise
at the same elevation as the overflow and readings are taken manually.

Other liquid level systems use a pressure transducer at the measurement position. This is used to sense
the hydrostatic pressure in a fluid-filled tube lead from a header tank mounted on a stable reference
point at a higher elevation. Changes of pressure with time are used to determine settlement. There
is long experience of using vibrating wire pressure gauges in installations for this purpose in dams.
Recently, compact systems using solid state pressure transmitters have been developed for use during
tunnelling and deep excavations in urban areas and are reported to be extremely stable (see Figure 6.6).
However, there is no experience of the use in long-term measurements in dams.

Figure 6.6 Liquid level transducer (courtesy Getec UK)

An alternative measurement system uses a set of inclinometers or electrolevels permanently installed within
suitably fixed inclinometer tubing (see Figure 6.7). The measured inclinations are converted to offsets,
and integrated to determine the profile. Changes relative to the initial readings are used to determine
settlement since installation. A disadvantage of electrolevels is that they are sensitive to temperature.

Figure 6.7 In place inclinometer (courtesy Geokon)

64 CIRIA, C743
SAA is a similar system that

1
has been developed recently.
Arrays up to 100 m in length
are constructed in inter-
connected 0.3 m or 0.5 m long
segments each incorporating a
microelectromechanical system
(MEMS) linked to an interface
2
by wireless network. The arrays
may be clipped to a conduit or
installed within simple 25 mm
plastic tubing (see Figure 6.8).
The arrays are waterproof
3
and may be used to monitor
movements of conduits that are
generally submerged. A benefit
of the system is that it is suitable
4
for data-logging and remote Figure 6.8 SAA installed inside sewer tunnel (courtesy Geo-Observations)

5
reading. The software supplied
makes the interpretation straightforward and, if necessary, data from any individual inclinometer may
be excluded reducing the system vulnerability to individual failure or drift.

6.4.4.4 Movement of cracks and joints

To monitor movements across joints or cracks it is common practice to use Demec gauges or Vernier
6
callipers, and reliable readings may be obtained over many years. Using installations with three studs,
displacements may be determined accurately in two orthogonal directions. Sometimes plastic tell-
tales are sufficient, and while simple devices such as the Avongard tell-tale may suffice, more accurate
readings can be obtained using a Tell-Tale (see Figure 6.9). 7
8
9
10
11
12

Figure 6.9 Tell-tales (courtesy Avongard and Moiré Tell Tales)

A1
Dams and reservoir conduits 65
For conduits with limited access, vibrating wire or linear voltage differential transducers (LVDT) crack
gauges may be permanently mounted across a crack or joint. At each location a pair of gauges is needed
to determine displacements in two orthogonal directions. These are claimed to be very stable and may
be read manually, remotely or data-logged.

6.4.4.5 Monitoring deformed shape of conduits

Where survey techniques as previously discussed are not appropriate or feasible, the changing shape of
a conduit may be monitored using transducers similar to those used for monitoring cracks and joints
mounted to measure diametral distances. Such instruments are claimed to be very stable and may be
read manually, remotely or data-logged.

If the complete cross-section of the conduit is to be monitored a series of tiltmeters mounted on

articulated bars fixed to the conduit lining may be used. Using a series of electro-levels, the Bassett
Convergence System (Bassett et al, 1999) was specifically developed for measuring the movement of
tunnel linings – it is similar in principle to the more recent SAA (see Section 6.4.4.3).

6.4.5 Ground deformation (dam near conduit)

Understanding the deformation of the dam near to a conduit may be helpful in giving warning of
developing problems. While surveys alone will not necessarily reveal the formation of a sinkhole, such as
when combined with visual inspection and other instrumentation, surveys are often useful in the routine
monitoring of dams.

Ground deformations near to a conduit may be monitored using SAA. As part of a recent European-
funded urban flood project, the use of SAA sensors within flood embankments to support an online
early warning system was tested at a site at Boston, Lincs (Abdoun et al, 2013)

6.4.5.1 Surface movements

It is common practice to install
a number of survey monuments
(see Figures 6.10 and 6.11) that
may be used in monitoring
surface movements of a dam, for
example along the crest and on the
downstream slope above the conduit.
These may be surveyed regularly
using accurate levels and (depending
on the resolution required) total
stations, the latter mounted on
permanent pillars (Almog et al, 2011).
A more sophisticated system could
use a number of robotic total stations
(eg Duffy et al, 2001). Such automated
Figure 6.10 Leica TM30 robotic total station (courtesy Leica Geosystems)
survey systems have been used
extensively during recent underground construction in London (Beth and Obre, 2005).

The rapid development of remote sensing techniques suggests that these may be useful for asset
management although at present they are not used much for monitoring dams. A recent study Long et
al (2012) has reviewed their role in the monitoring of flood embankments. The study concluded that
portable GPS equipment (real time Kinematic, RTK-GPS) may be used to determine elevations to an
accuracy of a few centimetres. Other remote sensing techniques such as terrestrial or airborne LiDAR, or
satellite InSAR could be also be useful in routine monitoring of the surface deformations of dams near to
a conduit although at present the data processing requirements are considerable.

66 CIRIA, C743
Absolute positions of the ground surface may be

1
obtained using laser scanning techniques such as
LiDAR, which is a technique in which a narrow laser
beam can be used to map physical features with
very high resolution. The scanner may be mounted
on a tripod or a moving platform, eg an aircraft
or helicopter, and the data processed to produce a
cloud of thousands of data points. The data may then
2
be interpolated to produce a digital terrain model
(DTM) on a grid as fine as 0.25 m. The helicopter-
based system, Fast Laser Imaging and Mapping
Airborne Platform (FLI-MAP™) was evaluated for
monitoring flood defence assets by work carried out
3
Figure 6.11 Pillar monument with target prisms used for
be Burgess (2004), and Auriau et al (2013) – with a

4
monitoring dam deformations (after Duffy et
helicopter flying at a height of 150 m and scanning al, 2001)
a width of 105 m, the vertical accuracy is currently
considered to be around ±30 mm (see work on the Cheshunt North Reservoir in Almog et al, 2011).

5
Movement of the ground surface may be monitored using satellite-based synthetic aperture radar (SAR).
On each pass of the satellite, a radar scan is undertaken and the difference in phase of the outgoing and
return waves is recorded. The data is processed to produce a very narrow effective radar beam. While
the total distance from the satellite to the target (ie the number of whole wavelengths) is not known,
the extra fraction of a wavelength can be measured extremely accurately. By comparing two or more
images obtained from reflections from the same ground target it is possible to resolve movements along
the line of sight to millimetre accuracy. Reflectors may be placed specifically for this purpose, but it is 6
also possible to use persistent scatterers (ie use reflections from the same point that appear in multiple
images). Such images have been collected at regular intervals by the European Space Agency for over
20 years, so this technique is potentially well suited to long-term asset monitoring. It has been used in
several studies of long-term ground subsidence (Raucoules et al, 2007), movement of landslides, and
more recently included in several published studies on dams (Grenerczy and Wegmüller, 2011, Voege
7
et al, 2012, and Tomás et al, 2013). In a study of movements of the La Pedrera dam in Spain, Tomás et al
(2013) present longitudinal and transverse profiles of settlements that steadily increase to a maximum
of 16 mm over 2.5 years (see Figure 6.12 and 6.13). While such measurements appear to give a useful
indication of general movements of the dam, the resolution in this study is at least 10 m by 10 m,
8
considerably larger than reputed sinkholes.

9
10
11
12
Figure 6.12 InSAR measurements using TerraSAR-X images, La Pedrera dam, Spain, July 2008 to June 2010 (courtesy
R T Jover)
A1
Dams and reservoir conduits 67
Figure 6.13 Crest deformation, La Pedrera dam, Spain (courtesy R T Jover)

At the present time it appears that no remote sensing techniques have sufficient resolution to completely
replace geodetic survey techniques. This is likely to change as the technology develops and new satellite
systems are established.

6.4.5.2 Subsurface movements

Subsurface movements near to a conduit may be monitored using standard geotechnical instrumentation,
such as inclinometers (eg SAA) and magnet extensometers. Care should be taken so that the drilling of
boreholes and grouting of the tubing does not adversely affect the dam, for example by accidentally causing
hydraulic fracture.

6.4.6 Other measurements

There is renewed interest in monitoring acoustic emissions from deforming soil masses. Koerner et
al (1978) describe field studies at various sites including earth dams, fills, and embankments. They
conclude that deforming soil masses generate emissions that exceed background noise levels, and that
those generating a high level of acoustic emissions are actively unstable. The system has also been used in
detecting leaks from pipes. While the technique may prove to be useful in detecting developing internal
erosion and in detecting progressive strain of the conduit structure more research is needed.

68 CIRIA, C743
6.5 DATA ACQUISITION, PROCESSING,
REPORTING AND INTERPRETATION 1
6.5.1 Data acquisition
The requirements for monitoring vary widely depending on the project, and the choice of data-logger
will also depend on the type, number, precision and speed of measurements required. Single and
2
multi-channel data-loggers can be used to monitor all types of sensors including vibrating wire gauges,

3
strain gauges, MEMS (analogue and digital), thermistors and linear potentiometers. They may be
tailored to suit the available power supply (preferably mains as well as stand-alone wind/solar sources)
and communication requirements (wireless link to sensors, direct link to laptop, cellular general packet
radio service [GPRS] modem). The chosen system should have low power consumption, be of rugged
construction, and have a proven track record.

6.5.2 Data processing and reporting

4
Documentation relating to the instrumentation should be prepared immediately after installation, and
would include:

‰‰ information about visual surveillance requirements and procedures

5
‰‰ details of the instruments installed and their purpose
‰‰ instrumentation monitoring procedures
‰‰ reporting requirements 6
‰‰ expected ranges, alerts and trigger values
‰‰ procedures to be used if results fall outside those expected ranges.

It will be necessary to decide on the frequency and procedures used to acquire and confirm the validity 7
of data, and resolve irregular or atypical readings, evaluate data and decide on any action needed.
Clarity is needed on:

‰‰ When are instruments read, how, and by whom?

‰‰ How fast is the quantity being monitored likely to change and what is the basis for the reading
8
frequency? Should the frequency be changed during floods?

9
‰‰ Are there any factors such as submergence of the conduit that may restrict the ability to acquire
data? What procedures are to be followed to assure dam safety if the dam or instrumentation is
inaccessible for a portion of the year?
‰‰ How will the raw data be confirmed to be valid? What quality control procedures are needed to

10
ensure the reliability of the data collected?
‰‰ How will those reading the instrumentation be trained?

Initial data processing and recording should be undertaken as soon as instrumental data is acquired.

11
Simple manual records and graphs will highlight unexpected trends, and rogue data should be
investigated immediately. For example piezometer data should be plotted alongside reservoir levels to
show whether there are any unusual trends.

Staff who undertake regular visual inspections and take and collate the readings from the instrumentation
should be properly trained. Clear procedures should be established to ensure that the supervising engineer
and possibly the inspecting engineer and dam owner are alerted to unexpected behaviour quickly. 12
6.5.3 Interpretation and action plan
The data from instrumentation should be critically reviewed and evaluated by an experienced engineer
on a regular basis, as well as in response to unusual behaviour observed on site.
A1
Dams and reservoir conduits 69
The following questions are examples of the items that should be considered when compiling the
evaluation (FERC, 2012):

‰‰ How do the readings relate to the expected performance of the dam and conduit?
‰‰ The use of the terms ‘normal’, ‘within historical range’ and ‘no adverse trends’ should include an
interpretation. Why is it normal? What is the historical range?
‰‰ How do the readings relate to the trigger values and action levels?
‰‰ How do the readings relate to the design or installation of the instrument?
‰‰ Have any calibrations changed? What is the effect on previous readings?
‰‰ Are electronic sensors, semi-automated or fully automated systems being confirmed by manual readings?
‰‰ Does the instrument still fulfil its intended purpose?
‰‰ Is the data supported by the visual observations of the behaviour of the dam and conduit?
‰‰ Is there any trend to the data?
‰‰ Are all similar instruments in similar parts of the dam responding similarly?
‰‰ Can the movement/trend of the readings of an instrument be correlated to any movement/trend of
the readings of another type of instrument? (eg is a rise in the water level in a piezometer reflected
by an increase in flow entering the conduit? Are piezometers responding to changes in headwater/
tailwater levels?)
‰‰ Has enough data been plotted to assess long-term trends?

As discussed in Chapter 7 the actions that should be taken in response to deteriorating conditions should
also be planned in some detail, and on-site emergency plans should be drawn up.

70 CIRIA, C743
7 Planning for emergency
response
1
2
This chapter deals with the production of plans to be followed in an emergency.

Embankments both in the UK and around the world are owned by a variety of types of owners, which
can vary from a private owner with little experience and resources, to large water companies and
3
government organisations with considerable experience and resources. The responsibilities of the owner

4
may attract criminal liability if subject to the Reservoirs Act 1975 in the UK, but in all cases they will
attract liabilities under common law (eg Rylands v Fletcher 1868, Charles et al, 2014).

The failure of an embankment not only has severe consequences for persons and property downstream,
but can also be serious for an owner resulting in reputational damage and even loss of operating licence
in the case of a water company, or bankruptcy in the case of a private owner. The potential for failure
always exists, but the probability of failure can be reduced by surveillance and monitoring, which seeks
5
to identify the early stages of a failure process before it develops further.

However, surveillance and monitoring are of little benefit if no action is taken to prevent a failure
from developing. Preventative actions are always more effective if planned in advance as part of a well-
developed emergency action strategy.
6
7
Problems with conduits can lead to problems with the whole embankment dam, such as settlement and
sinkholes that can ultimately threaten the stability of embankment dams. So, pre-planned emergency
actions should be considered for the failure of conduits.

7.1 SURVEILLANCE 8
Any form of inspection, monitoring or surveillance may assist in detecting a developing problem before
it can affect the operation and safety of the dam. Early intervention with sufficient time to detect a
problem and then carry out mitigation measures can often ‘save’ the dam and prevent failure. Mitigation
measures as described in later chapters, including renovation and repair works, may well be required,
but one of the most important elements of trying to avoid a release of water from the reservoir will
9
be its removal from the reservoir basin by drawing down the water level. Other mitigations may be to

10
divert water away from the reservoir, and perhaps increase draw-off rates to treatment works and scour
discharge. Emergencies and incidents involving the conduit may prevent the use of that conduit to draw
the reservoir down.

11
Visual inspections and examinations are carried out periodically and usually include not only the dam,
but also the appurtenant structures and the related conduits.

Surveillance by trained personnel – usually a visual inspection, but sometimes supplemented by

measurements – seeks to find and reveal features or trends that are indicative of a more serious problem
developing. Structural defects and deterioration typically develop over a period of time and can be
observed and watched. Experienced engineers can often identify these defects and potential problems
before the conditions either within the embankment or conduit become serious. Timely intervention can
12
identify an issue before it becomes serious, allowing sufficient time to react and to reduce the probability
of failure by removing some or all of the water from the reservoir. Regular and appropriate surveillance
is known to reduce the probability of failure. A1
Dams and reservoir conduits 71
However, some situations can suddenly arise and cause serious damage to the culvert in a short period
of time. Examples might include the planned full scour discharges through conduits, which might occur
every six months, or during/after an emergency draw-down operation as a reaction to a problem at the
site, or during/after a seismic event where damage might have occurred. In these types of instances it
would be important to inspect/examine the conduit.

Surveillance regimes will usually include a walk through of the conduit, if accessible for a ‘high risk’ or
‘high consequence’ structure, at least once a month. Other forms of surveillance might include a visual
examination from the downstream end without entering the conduit and looking for flows of water at
the downstream end of a culvert and/or listening for water falling into the conduit.

7.2 ACCESS TO ARCHIVE DATA FOR RESERVOIR

When an incident or emergency happens, archive data is all important for those trying to interpret what
is or has happened, and to establish:

‰‰ how the embankment will react

‰‰ how water inflowing to the reservoir might be diverted
‰‰ how water levels with the reservoir might be reduced
‰‰ what resources, facilities and assistance is at hand and needs to be employed.

In this respect there are many sources of data, including drawings, the prescribed form of record (PFR),
valve schematics, catchment plans, emergency plans, and on-site plans. Much of this information would
be of great value in trying to manage an incident or emergency. However, information is more likely to
be available for dams subject to legislation and may be non-existent for privately owned structures or
those not subject to legislation.

For reservoirs subject to legislation it would be likely that the supervising engineer would hold much of
this data, but for reservoirs not subject to legislation this material may not have been collated.

For ‘high consequence’ or ‘high risk’ dams it is likely that an on-site plan as part of All Reservoir Panel
Engineers’ (APRE) emergency action plan would have been developed and be written in such a way that
it would include much of this information.

7.3 IDENTIFICATION OF APPROPRIATE ARPE

CONTACTS
The responsibility for maintaining a safe embankment rests with the owner. Safely maintaining and operating
a reservoir in a safe condition is a key element of preventing a failure and limiting the liability that an owner
might face. An important part of trying to ensure safety is to employ a qualified dam engineer, whether they
are a supervising engineer, an inspecting engineer, or a QCE within the meaning of the Reservoirs Act 1975,
to periodically examine, inspect or give advice on works and in all cases to assess the possibility of problems
developing that could lead to failure. These engineers develop an understanding of the structure and how it
has behaved and will behave and are invaluable in the case of an emergency or incident.

7.3.1 Need for an engineer

Embankment dams and/or appurtenant structures will deteriorate with time. Changes in loading
conditions could also affect the structure. Failure of an embankment dam, whether due to conduit
deterioration or any other reason, could leave the owner liable for loss of life and property damage
downstream. For this reason, and often required by law, the owner should undertake regular and
periodic inspections, maintenance and occasional repair and rehabilitation of the structure or parts of
the structure to try to ensure the safety of the structure.

72 CIRIA, C743
The owner will require the services of a QCE from one of the panels, constituted under the Reservoirs

1
Act 1975, Flood and Water Management Act 2000 and/or the Reservoirs (Scotland) Act 2011, for those
reservoirs subject to legislation and it would be sensible to employ the same engineers for advice on
reservoirs not subject to the Act.

Choosing a panel engineer with the competency and experience will be important in trying to achieve
a successful outcome. If no drawings or records exist the engineer may be able to develop basic plans
and calculations to help better understand the structure, evaluate its stability and understand the 2
possible failure modes, the speed of the likely development of those modes and the consequence of
failures. Accordingly, the engineer may be able to assist the owner with the management of the incident/
emergency. They may also provide the owner with assistance in selecting a contractor to conduct the
repairs or remediation work if necessary, and can provide site supervision if required. 3
7.3.2 Choice of panel engineer
It is important to appoint an engineer who is from the appropriate
panel constituted under the Reservoirs Act 1975 or the Reservoirs
APREs
4
A list of APREs is given on the Gov.uk
(Scotland) Act 2011 with the relevant experience of the type of dam, website: http://tinyurl.com/pz4b3q7

5
and experience of dealing with remedial works, engineering works,
and repairs on that sort of dam (see APREs box).

7.4 CREATING AND IMPLEMENTING THE

EMERGENCY PLAN 6
The Water Act 2003 recommends flood plans are produced for dams that might threaten persons and
property downstream of the dam.

The flood plan, for a dam that could inundate an area downstream where there is a potential for loss of
7
life or significant damage, includes an ‘on site’ element – the work that an owner could carry out on site

8
to try to prevent failure of the structure. The other elements of the emergency action plan or flood plan
include an inundation map, and an ‘offsite plan’ – the latter being the plan that the local authority and
emergency services would use to try to evacuate people and to protect persons and property.

The on-site plan will consider potential failure modes and specify pre-planned actions to be followed
to attempt to prevent failure and to minimise risk to people and property downstream. The plan will
contain information and procedures to assist the owner in issuing warning and notification messages
as well as acts to mitigate or alleviate a potential dam failure. This might include diversion of outflows,
9
means of scouring the reservoir, the places to position pumps with their associated suction and delivery
pipes, the sourcing of materials such as sandbags, clay and timber.

The plan would also give details on access routes, the places where control rooms and welfare facilities
10
could be placed, which mobile phones work in the area, and contact phone numbers for those to be

11
notified of an impending failure.

The inundation map will delineate the area that would be flooded as a result of the embankment
failure for normal and flood conditions or uncontrolled release. The maps are used by the owner and
the emergency services to facilitate the timely notification and evacuation of areas affected by failure
and/or flooding.

The maps should provide information on depths and velocities of water and speed of travel of a flood
12
wave to assist in evacuation, particularly where critical infrastructure is located within the affected area.
However, the maps should only be used as guidance as there are many assumptions in the analysis used
to produce them. A1
Dams and reservoir conduits 73
Any plan needs to be maintained and exercised to ensure it remains
Flood risk management
effective. When the plans are exercised those involved will then be aware
Sample plans from:
of their roles and responsibilities. Details of sample plans can be found on http://tinyurl.com/ovfz75q
the Defra website (see Flood risk management box). These can be used to
inform the process of writing a plan.

When an accident or incident occurs the event may progress to ‘total’ failure and release of the reservoir
contents, but the process may not go that far. It is often the panel engineer on site who will be asked to
use their judgement to estimate whether failure is going to occur and the timescale.

The rate of propagation of the incident/failure will vary depending on the mode of failure. The panel
engineer should be able to assist in a view as to the speed of failure, and the response of the owner
and support staff will vary. If the speed of failure is fast, it may be that people cannot or will not enter
the area in which preventative works might be carried out. In the case of imminent failure it may be
untenable for even the emergency services to send staff into the affected area, even to evacuate people.

However, where some preventative works can be carried out, immediate actions might include
monitoring the quality of any water discharging – and its rate of flow, searching for sink holes in the
upstream face, reduction of water levels, plugging of sink holes, or the placement of a protective filter of
sand and gravel over exit points.

If the failure of the conduit is involved, with piping of material through the culvert, the closure of any
valve or penstock to protect the embankment from further erosion may be essential. However, if the
piping is downstream of any valve, then it may be very difficult to control the situation.

Monitoring of the conduit for settlement and inflows might also be carried out, but again in an
emergency it may not be possible. In the event of an emergency, access into the conduit may be
prevented.

If the situation is declared as an emergency, is not stable and is deteriorating, then it will be necessary
to implement the emergency action plan and start reservoir draw-down (see Section 7.5). This may not
necessarily be through the existing conduit and other facilities may be required.

If a sinkhole forms it might be filled with appropriate materials that might be transported directly to
the defect in the conduit and help block any exit paths for eroded material. A careful assessment of the
situation will be required.

7.5 EMERGENCY DRAW-DOWN

One of the most frequent actions taken to reduce the risks in an emergency situation is to undertake an
emergency draw-down.

Alternative means of reservoir draw-down may be required if the outlet works are compromised or
inoperable. Releases through the conduit might not be possible.

The selection of an appropriate means depends on the size of the reservoir, the physical features of the
dam site access arrangements, the availability of equipment and materials, the volume of water to be
released and the required rate of release.

Ideally the dam will have adequate and operable facilities to draw the reservoir down – usually at a
rate of at least one metre per day. There are many features to allow inflows to be diverted or to control
inflows, but this is not always the case.

Other means to reduce reservoir levels might include the following.

Siphons
A single or a series of siphons can be used to partially drain a reservoir, in particular a small reservoir.
Siphons are usually fitted with a downstream valve and a means of priming the siphon.

74 CIRIA, C743
Siphons are often placed over the crest of the dam and should discharge onto a properly prepared area

1
to prevent erosion.

Siphons can be constructed of PVC, HDPE, or steel pipe and because they usually have to be transported
by hand, they generally do not exceed 300 mm in diameter. Because of negative pressures the pipe
should be sufficiently rigid to withstand the collapsing forces. Pipe joints need to be watertight.

Siphons have the advantage of being able to be installed relatively quickly, can be automatic, can be
installed with the reservoir at least partially full, and are relatively cheap.
2
Pumps
Pumps can be used either singularly or as a series of pumps to assist in the draining of the reservoir. 3
Pumps are usually self-contained and are either wheeled, trailer or skid mounted.

They can be positioned on the crest and discharge to the stream or to a spillway channel. Pumps can be
petrol/diesel or electrically powered. Floating pumps on pontoons are also available. 4
The area in which pumps can be positioned is often limited. All the suction and delivery pipes have to

5
be carefully positioned so that they can safely discharge the waters to the stream below the reservoir.
This is often a logistical problem and there may be a restriction on what equipment can be deployed. It is
sensible to lay the pipes on plastic sheeting in case there is leakage from the pipe.

In an emergency situation, for example during prolonged storms affecting a wide area, sufficient pumps
may not be available. The supplier may not be able to provide pumps and equipment outside normal
working hours or perhaps on a bank holiday. Some owners have decided to purchase their own pumps 6
and temporary siphons to overcome the reliance on pump suppliers.

A key element of an emergency plan is to establish where pumps will be obtained from, how they
will be brought to the site, and how access to the pump set-up area will be maintained, even in an
emergency situation.
7
8
Other techniques to lower the reservoir could include:

‰‰ removal or lowering of the top of a drop shaft spillway structure or a side channel spillway
‰‰ the crest might be removed in stages
‰‰ removal of the control structure of the spillway
‰‰ excavation of a trench through a carefully considered location beyond the abutment to provide an
additional spillway facility
9
‰‰ care should be taken to ensure that erosion of the channel formed does not result in an

10
uncontrolled release larger than the planned release. The channel could be protected with erosion-
resistant materials placed to prevent the material from being washed away
‰‰ controlled breach of the embankment.

The breach position should be carefully selected at a point where it can be controlled. Ideally the breach
should be formed in a staged and controlled way discharging to an erosion-resistant channel.
11
7.6 POST-EMERGENCY RECOVERY
An investigation would follow the immediate emergency actions when the situation allows. This
investigation would have to be carefully planned and executed. The investigation often involves careful
12
excavation and the gathering of information, and may often be combined with the actual repair. Also,
the conduit may need repair as well as the embankment above it.

A post-incident report must be written and lodged with the enforcement authority. A1
Dams and reservoir conduits 75
8 Techniques for renovation
and repair of conduits

This chapter discusses techniques for the renovation of buried pipes, the internal and external repair of
pipes in culverts, and the renovation and repair of human-entry culverts.

Like all man-made structures, conduits deteriorate over time. As they form a critical element of the dam
structure, deterioration can represent a significant hazard for the dam with associated high risk for
public safety. So dealing with such deterioration effectively is very important.

A reservoir owner/manager will have three main options available when conduit deterioration causes concern:

1 The conduit can be repaired or renovated. Repair would consist of strengthening weakened
sections of the original conduit through the application of new materials, often following the
removal of damaged fabric. This is common with human-entry culverts. Renovation would
generally mean augmenting the original conduit fabric with additional new material, such as
through the provision of a new lining. This is common with pipes.
2 The conduit can be replaced either along the same alignment or in a new position. This may be
preferred if the function provided by the conduit is essential, but unacceptable risks would remain
following any renovation. The replacement conduit can also have increased capacity.
3 If the conduit is in very poor condition and not essential, the most effective way to manage the risk
it poses may be to abandon it entirely by filling or removal.

This chapter details the techniques currently available that relate to point 1, namely the repair and
renovation of conduits. It is split according to the type of conduit, which determines the access available
to complete the works, namely:

‰‰ buried pipes
‰‰ pipes within culverts
‰‰ human-entry culverts and tunnels.

Chapters 9 and 10 of this guide relate points 2 and 3. These are split according to the same three types
of conduit. Chapter 11 discusses the process to be undertaken in identifying which solution is most
suitable for any given situation.

When considering conduit repair or renovation, the options that are available are constrained by
the location, structure, arrangement and capacity of the existing conduit. So the process of solution
identification can be relatively simple compared to conduit replacement, where design options are
unconstrained by the existing conduit. However, once the preferred solution is identified, the planning
and design can be complex, particularly with difficult access and undesirable arrangements, often
making isolation from the reservoir head more difficult.

In all cases the CDM Regulations 2015 will apply to the construction works. It is essential for reservoir
owners to engage suitably competent contractors and designers, in accordance with the regulations.

8.1 RANKING FEASIBLE SOLUTIONS

Throughout Chapters 8, 9 and 10 of this guide, summary tables have been used to help identify the
balance of construction and operation costs and risks associated with common solutions. The impact of

76 CIRIA, C743
each main cost or risk type has been colour coded as follows:

‰‰ red – this solution has a major disadvantage or cost/risk implication

‰‰ orange – medium disadvantage or cost/risk implication 1
‰‰ yellow – minor disadvantage or cost/risk implication

2
‰‰ green – advantageous solution with cost or risk benefit.

These have been based on the consideration of ‘typical’ conduits at UK embankment dams, as follows:

‰‰ buried pipe – 300 mm to 450 mm internal diameter (ID) pressurised CI or DI, or 600 mm ID

3
concrete/brick spillway pipe, buried in a low, old embankment dam, either homogenous or with
central core
‰‰ pipe in culvert – 300 mm to 900 mm ID pressurised CI or DI in a masonry culvert in an old
embankment dam with clay core

4
‰‰ human-entry culvert – more than 2000 mm nominal diameter circular or horseshoe/modified
horseshoe section masonry or concrete culvert containing pipework, with space for human-entry
and access from the downstream end.

5
Many conduits will deviate from these typical arrangements. Some sites will have conduits that, due to
dimensions or materials, fall outside the technical envelope of the solution.

Note that for each solution, the costs and risks considered in the summary tables relate only to the
technical solution, and do not consider the priorities of the reservoir owner/manager, the hazard posed
by the dam, or the quality of information available to the project team. The implication of these on
construction and operation costs and risks are discussed in Section 11.4. 6
8.2 ISOLATION OF A CONDUIT FROM THE
RESERVOIR HEAD 7
All works to reservoir conduits require the working area to be isolated from the reservoir water. The
most reliable method of isolation is the drawing down of the reservoir to below the conduit inlet(s).
However, this is often not possible for a variety of reasons. 8
Where access is available to the upstream end of a pipe via a dry intake tower, the isolation provided by

9
the tower may be sufficient to undertake some types of work, for example lining of the outlet pipe.

Where works are only required to sections of pipe that are downstream of the dam crest or within a
culvert, the dam will provide this isolation in conjunction with valves or other bungs/stoppers, blocking

10
the pipe upstream of the working area.

However, in most other circumstances, temporary isolation is required. This is generally provided by a
combination of reservoir water control and a temporary cofferdam structure.

8.2.1 Pipe blocking 11

Where the upstream end of the pipe is in good condition and works are required downstream only, it
may be possible to isolate the works area from the reservoir head by blocking the pipe. Closed valves,
bungs or stoppers can be used to provide the required blocking mechanism. Following failure of the
blocking mechanism, if a sudden release of water could risk the safety of workers undertaking the
remedial works, or the public, then two blocks within the pipe are usually required. This is known as
12
double isolation. Preferably at least one of the blockages is formed by a trustworthy valve.

Where isolation is obtained through the closure of valves, it is necessary to ensure they are not operated
during the period when isolation is required. Valves need to be locked and labelled. If possible,
A1
Dams and reservoir conduits 77
handwheels should be removed. Automatic valves should not be used for isolation, unless the actuators
are removed or disabled.

It is important that the pipework downstream of the isolating blocks (and upstream of the works site)
is not at risk of failure once emptied, for example if it is subject to reservoir head externally by being
buried in the upstream shoulder of a dam with central core. If there is concern about its condition,
then measures are required to control the external pressures on the pipe, such as lowering the
reservoir water level.

Where access to the upstream end of a pipe for insertion of a bung is difficult (eg due to silt), but the pipe
can be accessed downstream, then it is possible to fit a temporary stopper at the accessible point using
processes called ‘hot tapping’ or ‘line stopping’.

8.2.2 Hot tapping

This is a process whereby a tee and valve are fitted to an existing length of pipe while it remains under
pressure. The process, originally developed in the oil and gas sector, is now commonly used in the UK
water industry on large water main projects. A wide range of pipe materials can be hot tapped, including
CI and DI, steel, plastic and reinforced concrete, and a typical range of diameters is from 50 mm to 900
mm. A number of specialist contractors can provide this service throughout the UK.

The hot tapping process uses a sleeve, valve and drilling rig to make the connection (see Figure 8.1).

Figure 8.1 Hot tapping system (courtesy Mott MacDonald Ltd)

The methodology is as follows:

1 The pipe is exposed and the external surface cleaned where the sleeve is to fit on. Suitable support
for the pipe needs to be kept in place and sufficient space available to allow the drilling rig and
valve to be fitted. Space requirements depend on the size of the new pipe – typically a width of
300 mm on either side of the new pipe is required, and a length of about 3 m perpendicular to the
existing pipe.
2 The existing pipe should be inspected for defects, using internal survey or NDT techniques if
necessary, to ensure the pipe is in reasonable condition.

78 CIRIA, C743
3 A ‘tapping sleeve’ is fitted to the existing pipe. For steel pipes, this is generally welded into position.

1
On old CI pipes (and other materials), the sleeve is clamped tightly to the existing pipe.
4 A valve is fitted to the tee end of the capping sleeve, in the open position.
5 A sealed drill unit is fitted to the end of the valve. This then cuts a hole in the wall of the existing
pipe, typically using a pilot drill and a wider cutting head. The hole is marginally smaller than the

2
new pipe. The drill is then withdrawn through the new valve back into the sealed drill housing.
6 The new valve is closed, allowing the drill to be removed and further pipework or fittings attached
to the valve.

3
The hot tapping process requires a working area for the drill rig to be available beside the pipe to be
tapped. This is dependent on the size of the tap (tee) to be provided, although in general it would be in
the order of 1 m (width along the pipe) by 3 m to 4 m long (perpendicular to pipe). In addition, enough
space needs to be available around the pipe in order to fit the collar. Note that in some circumstances it

4
is possible to tap the existing pipe at an angle other than 90 degrees.

If the hot tapping is undertaken near to an in-line existing valve, a new section of pipe can be fitted to
the new tee, allowing the existing pipe downstream of the in-line valve to be removed. It could then be
replaced, or the line from the new tee used permanently.

If an in-line valve is not present, the new tee can be used to fit a temporary bung into the existing pipe
near to the tee. This process is called ‘line stopping’ (see Section 8.2.3).
5
8.2.3 Line stopping
Once the drill rig is removed, a separate rig can be attached to the gate valve, again forming a seal,
6
allowing the gate valve to be opened. A telescopic arm can then extend to place a folded stopper into the

7
main pipe. This is then unfolded/inflated perpendicular to the flow, to form a seal with the pipe wall.
The water downstream of this stopper can then be drawn-down.

A variety of such stoppers are now available. The pressure that can be resisted by such stoppers is

8
limited, and varies between products. Advice should be sought from a specialist contractor if considering
line stopping.

8.2.4 Reservoir water level control

The water level in a reservoir is most easily controlled by using a weir, which can allow excess inflows to
escape without manual intervention. Where the reservoir is to be artificially drawn-down below the weir
9
level, firstly the storage above this ‘draw-down level’ needs to be removed and secondly sufficient means
to deal with expected inflows needs to be provided.

Full or partial draw-down of reservoirs (see Figure 8.2) can sometimes be achieved by isolating the
reservoir from indirect catchments, turning off inflow pumps, and/or increasing abstractions for use.
10
This is the most economical way of drawing the reservoir down as the value of the stored water is used

11
and draw-down costs are minimised. However, if this is not feasible due to capacity or programme
constraints, draw-downs can be achieved by discharging reservoir water into a nearby watercourse, most
often the one impounded by the dam.

12
Such draw-down is achieved using existing gravity pipework at the site, temporary pumps or siphons.

Using existing gravity pipework in the dam is cheapest if the pipework is available and functional, as
mobilisation and operation costs are small. However, the pipe capacity may be insufficient, the pipe
integrity may be in doubt, and the impact of silt discharges may prevent the required environmental
consents being obtained (see Section 1.4.2).
A1
Dams and reservoir conduits 79
 Over-pumping for long periods
can have considerable fuel
requirements particularly for
larger or upland catchments.
In built-up areas noise
from pumping can also be a
significant issue. The provision
of large pumps will also require
reasonable access to the set-up
location as well as the need
for refuelling if they are diesel
driven. Measures need to be
taken to ensure pump inlets are
raised above silt levels.
Figure 8.2 Pumping arrangement for reservoir draw-down (courtesy K Gardiner)

Siphons can be used to draw down reservoirs. However, their operation is dependent on numerous site
specific factors, as detailed in Section 9.1.2, and the rate of draw-down can be slow. They can though
provide cost-effective draw-down in some cases, particularly when used in conjunction with pumping
capacity to deal with flow variations.

For reservoirs with large direct catchments (relative to reservoir size), particularly those in steep and
wet upstream areas, storm events during any temporary draw-down can cause a rapid change in water
level in the reservoir basin. This can be dealt with by either providing sufficient pumping capacity to
deal with such larger inflows, making use of existing or new flood bypass facilities, or by ensuring that
the cofferdam around the works is high enough to allow temporary storage within the reservoir basin
without it being overtopped.

The design of such temporary arrangements needs to take into account both the cost of providing the means
of water level control, and the cost associated with flooding of the working area. Where the costs of temporary
works are expected to be high, then a financial analysis is required to determine the optimum pumping
provision and cofferdam height. Generally this is undertaken by identifying a design storm event, which
needing to be defended against by the temporary works. The choice of design storm event is based on the:

‰‰ planned duration of the construction works

‰‰ season during which the works are to be undertaken
‰‰ impact of cofferdam overtopping.

The height of the cofferdam is determined by the water level expected due to the design storm event
during the works period.

8.2.5 Temporary cofferdams

Protection of the working area around the inlet to a conduit from retained reservoir water and flood
surcharge is often required for renovation. Such protection is afforded through construction of a
cofferdam around the working area. Typical options include:

‰‰ sheet pile walls (steel or plastic) – note that access for installation plant around the inlet is often a
major constraint, and careful consideration needs to be given to the risk of damage to the inlet or
other vulnerable structures, as well as to the embankments into which the sheet piles are inserted.
Space inside the cofferdam may be constricted by wall props
‰‰ embankments – it is difficult to construct an embankment cofferdam underwater that will be
sufficiently watertight to act as a cofferdam. Stability is another major consideration
‰‰ fabric dams – note that these are limited by the depth of water that can be retained
‰‰ limpet dams – these can be applied to hard structures such as draw-off towers.

80 CIRIA, C743
The optimum cofferdam arrangement will depend on the site specific constraints, particularly the space

1
available, head of water to be retained and nature of the work to be done. The design of a temporary
cofferdam can be difficult until after the reservoir has been drawn-down and the extent of siltation
confirmed (see Case study A1.1).

The most common type of cofferdam is a steel sheet pile system. The sheet piles can be driven into
the reservoir bed or upstream shoulder of the dam, and installed to sufficient depth to prevent flow
through the bed or shoulder material into the cofferdam. Silt lying on the bed need not be removed. 2
The suitability of the sheet pile cofferdam is often determined by access for plant and material deliveries
to the reservoir, and access to the cofferdam location for the piling rig. The latter can be provided by
a barge in the reservoir. Where the dam crest is used, its ability to take the weight of the rig should be
confirmed on the advice of a geotechnical engineer. The design of a sheet pile cofferdam is complex and
needs to be undertaken by competent temporary works designers. Sheet pile cofferdams are not suitable
3
for reservoirs with an upstream impermeable liner or membrane.

There may be value in retaining cofferdams in situ, to hold silt back from the pipe inlet and for any
future repair. If a cofferdam is retained, gaps may be required to enable passage of sufficient water flow
4
to the pipe during reservoir operation. Submerged structures may also require warning signage for the

5
attention of boaters or other recreational users.

8.3 RENOVATION OF PIPES BURIED IN

EMBANKMENT FILL
Many small embankment dams and some larger dams do not have human-entry culverts. Instead, pipes
6
are buried within the embankment fill to provide scour or draw-off facilities or to transfer spill flow from

7
an overflow weir. Even where culverts are present, these often do not extend through the impermeable
zone of the dam, instead ending with a bulkhead through which the pipework extends, surrounded
by fill when passing through the impermeable zone and sometimes the upstream shoulder. As with all
pipes, these will deteriorate over time, and at some stage may require renovation or replacement.

Deterioration can occur due to influences along the inside or outside of the pipe. It can be caused by
flow within the pipe and/or chemical reactions between the pipe and soil or water, which can lead to
changes in chemical structure within the pipe material. Typically the deterioration process occurs over
8
a long timeframe, though this is dependent on the pipe material, the method of its manufacture, the
environment in which it is placed, its use and other factors.

Deterioration is unlikely to be constant over the entire length of the pipe. Localised deterioration can
9
be accelerated due to excessive local stresses within the pipe wall, which may be caused by imperfections

10
from the manufacturing process, by external loads from the embankment fill that may change over time
(as the embankment settles) or by other factors.

Once deterioration of a pipe has been identified, a decision is required as to whether to renovate, replace
or abandon the pipe, or leave it in its current condition. This decision needs to consider the costs,
benefits and risks of each option.

Deteriorated pipes performing a function that need to be retained are either renovated or replaced.
11
For buried pipes, the choice is largely determined by the depth of burial, which is usually the most
significant factor when determining the cost of digging out and replacing. However, an alternative is to
abandon the pipe and provide the function by alternative means, often a replacement conduit. 12
Digging out and replacing carries with it a significant number of risks.

A1
Dams and reservoir conduits 81
8.3.1 Internal renovation – general
Insertion of internal linings has become common at embankment dams in the UK. The most common
systems have a proven track record of halting corrosion, reducing the risk of leakage from the pipe
and providing structural support when this is applicable. However, there are also risks associated with
each type of lining, and many examples exist of ineffective or failed linings. The designer needs to
understand the limits of each type of lining, and potential reasons for failure. The most common types
of lining are:

‰‰ sliplining – a thin walled flexible pipe is pulled into the existing pipe, and the annulus between the
two pipes is filled with grout
‰‰ tight-fitting lining – a form of sliplining where the external diameter of the lining pipe is the same
as the internal diameter of the existing pipe (or slightly larger), and is deformed temporarily to
allow insertion. The original design shape is then restored, forming a tight fit with the existing
pipe without need for grouting the annulus
‰‰ CIPP linings – these linings are inserted into one end of the pipe in the form of a flexible ‘sock’,
impregnated with resins. The sock is expanded within the pipe using air or water so it fits the
existing pipe surface. Following this, the resin within the sock is cured using heat (hot water or
steam) or ultraviolet (UV) light, to form an impermeable, flexible lining on the inside of the pipe
‰‰ in situ sprayed linings – these linings are in the form of resins that are applied to the inside of the
pipe via a robotic spray-gun, which is passed through the existing pipe.

Sliplining and CIPP liners have been used extensively in the UK for many years, including use at
embankment dams. Tight-fitting linings have also been used at UK dams, though less frequently. In situ
sprayed linings are a relatively new technology with limited use to date at UK reservoirs.

Design guidance for lining systems is given in WRc (2001). This guide also provides further detail on all
common lining systems, as well as specialist suppliers for each lining type. Further information is available
for some lining types in the form of Water Industry Specifications (WIS)
and Information and Guidance Notes (IGN), which are published by Water UK
Water UK. These documents are under frequent review so check the Guidance documents and notes are
available from: www.water.org.uk
Water UK website for the latest versions (see Water UK box).

All of the lining types discussed in this guide are Type II (ie they do not require a bond between the
lining and existing pipe) as defined in WRc (2001). Type I linings (ie those that do require a bond
between lining and existing sewer) are only recommended for human-entry culverts where high
confidence annulus grouting can be achieved, or where an in situ coating is considered suitable. These
are discussed briefly in Section 8.4.9 of this guide. In both cases it is essential that any annulus is sealed
to avoid future leakage.

Where the lining is installed in a pipe on which external reservoir head applies, the lining needs to be
able to resist this pressure without collapsing. So in such circumstances there is a maximum depth at
which the lining can be installed below the reservoir flood storage level. The ability of the liner to resist
these pressures is determined by the lining thickness and strength relative to the pipe diameter.

Many providers of lining systems also deliver a system called pipe-bursting, whereby the existing pipe
is fractured in the process of installing a new pipe, allowing greater hydraulic capacity to be achieved.
This technique is not suitable for conduits through embankment dams as the disruption to the interface
between the pipe and the fill may lead to leakage and erosion of the fill material.

For all lining technologies, the pipe needs to first be isolated from the reservoir, emptied, cleaned and
inspected. Inspection by CCTV is not effective if the pipe is full of water and this poses a difficulty at
reservoirs with no form of upstream control where emptying of the reservoir may then be unavoidable.
Inspection is required to determine the most appropriate solution, and to decide on the details of the
installation process and the design of the liner (see Chapters 4 and 5). Some further details on these
preparatory works are given in the following sections, followed by details of each lining type.

82 CIRIA, C743
A number of different materials are available for use in the internal renovation of conduits. Chapter 2 of

1
WRc (2001) offers guidance on the properties of a range of materials suitable for sliplining, CIPP, close-
fit pipes and in situ lining renovations.

8.3.1.1 Comparing liner solutions

The four main types of pipe lining systems are discussed in Sections 8.3.2 to 8.3.6 under the
following headings:
2
1 Technical envelope – technical limits of the solution due to the arrangement and condition of the

3
pipe to be lined.
2 Site suitability – effect of other major issues common at UK reservoir sites.
3 Summary of installation – a brief summary of the installation process.
4 Main costs and risks – a summary table providing a relative comparison of major construction and
operation costs and risks associated with each solution. This has been colour coded, considering a
‘typical’ buried pipe at a UK reservoir (see Section 8.1). 4
8.3.1.2 Isolation, cleaning and inspection
The conduit needs to be isolated from the reservoir to allow its emptying for cleaning and inspection.
Where there is confidence that the existing pipe is in good condition and the lining method to be used
5
has already been identified, the cleaning and inspection can be completed immediately before the works

6
begin, using the same isolation temporary works as that used in the installation.

Most liners do not require a bond to be established with the existing pipe wall. They are generally
designed on the basis that the existing pipe has achieved equilibrium with the loading from the
surrounding soil, and so the liner pipe needs to simply resist the external hydraulic pressure caused by
pore water in the embankment fill. However, pipes need to be cleaned sufficiently to allow installation
without snagging, for the lining to achieve a circular shape (to achieve the required strength), and to 7
minimise the risk of leakage paths between the liner and the existing pipe.

There are many cleaning techniques available (see Section 4.6). Typically swabbing, flushing and/or
high pressure jetting are most economical methods. Cleaning of old and deteriorated pipes carries a
considerable risk of damage, particularly when using high pressure water jetting. Experienced personnel
8
should be used to clean pipes that are to be lined.

Some lining techniques require the pipe to be free of standing water. This is typically achieved through
application of foam swabs through the pipe. This is particularly important where previous settlement has
9
left a sagging profile to the pipe in which water may accumulate at low points.

It is imperative that a visual survey (using CCTV survey equipment where required) is undertaken over the
entire length of the pipe to be lined. This is required to confirm the presence of any obstructions, damage 10
or discontinuities in the existing pipe, as well as to identify leaks that may affect the liner installation. The
risk of unsuccessful installation will usually be unacceptably high if a full survey is not undertaken.

CCTV techniques have advanced considerably in recent years. Some units can now carry out laser profiling
and scanning of the pipe or tunnel. This can be used to determine ovality, pinpoint areas of structural defects
11
or missing/displaced masonry, and provide a highly accurate dimensional survey of the conduit, which can be

12
used to maximise the size of the liner pipe. The survey can also be used to confirm the precise alignment of
the pipe, which may not be evident with small diameter pipes through old embankments. The location and
angle of bends in alignment will influence the remediation assessment (see Section 4.5.3).

A1
Dams and reservoir conduits 83
8.3.2 SLIPLINING

8.3.2.1 General
Sliplining is a common procedure where a flexible pipe
is pulled into an existing pipe, and the space between
is filled with grout (see Figure 8.3). By far the most
common lining material to be used is HDPE, though
GRP or steel can also be used.

Installation can be very rapid and typical liners can

accommodate the reservoir head and length and size
of pipes commonly found at UK embankment dams.
Success is affected by temperature during installation,
and requires care when cleaning the existing pipe and
grouting the annulus between pipe and liner.

Figure 8.3 Sliplined pipe (courtesy Mott

MacDonald Ltd)

8.3.2.2 Technical envelope

The limitations of the liner in relation to pipe size and common structural problems affecting buried
dam conduits are noted in Table 8.1.

Table 8.1 Sliplining – technical envelope

Description Comments

Pipe diameter ‰‰ 100 mm to 2500 mm accommodated.

‰‰ depends on pipe diameter and effectiveness of cleaning
Pipe lining length ‰‰ maximum winching forces on liner will determine friction limits
‰‰ lengths of up to 1 km are feasible.
‰‰ depends on SDR of lining pipe used
Lining thickness ‰‰ can be reduced by using GRP or steel rather than HDPE liner
‰‰ SDR of commonly used pipes range from 11 to 33.
‰‰ not an issue as long as pre-welded liner can be winched through without significant
Pipe slope
bend at change of gradient (within pipe or outside).
Changes in size/shape ‰‰ no transitions in size or shape are possible.

‰‰ long radius bends can be accommodated, tight bends cannot

Bends
‰‰ long radius bend capacity dependent on liner thickness and temperature.
‰‰ solution does not rely on bond with existing pipe – existing pipe material does not
affect performance
Pipe material/preparation
‰‰ existing pipe needs to be well cleaned before installation to reduce the risk of the
liner catching/snagging during installation.
Leaking into pipe during ‰‰ large leaks need to be stemmed before installation
installation ‰‰ water from small leaks can be accommodated by adjusting grout mix.
‰‰ all systems can comfortably bridge gaps and accommodate small steps
Gap bridging (at joints)
‰‰ gaps leading to voids will affect grout take.
‰‰ dependent on sag radius being suitable for liner thickness and temperature during
installation
Deflection along existing pipe
‰‰ pooled water within sagging pipes needs to be removed before installation, and any
sag will be reflected in lined pipe alignment.

84 CIRIA, C743
8.3.2.3 Site suitability
The performance of sliplining in relation to common reservoir construction considerations is noted in
Table 8.2. 1
Table 8.2 Sliplining – site specific issues

Description

Draw-down/isolation
Comments

‰‰ pipe needs to be isolated from reservoir and access available at both ends.
2
‰‰ pipe lengths and welding equipment need to be delivered to site at pipe end where

3
liner is to be inserted. Both can be lifted to final position with tracked plant or crane
if road access not available
Road access
‰‰ grout to be pumped into annulus from one end of pipe – can be pumped up from a
distance if appropriate plant used
‰‰ grouting should be carried out from the downstream end.

4
‰‰ depends on individual pipe lengths and length of lining to be achieved – minimum
of two times pipe lengths plus 3 m (in line of conduit) by 3 m to 5 m wide.
Working area at pipe outlet (end ‰‰ it is advantageous for all pipe welds to be completed before liner is pulled into
where liner is to be inserted) position to avoid pauses in the pulling process, which make the process more
prone to failure

5
‰‰ guidance on the geometry of lead-in trenches is provided in WRc (2001).
Wet weather during installation ‰‰ only impact is in relation to reservoir level control (where required).
‰‰ impacts on installation
Low temperatures during
installation ‰‰ minimum bend radius depends on liner thickness, but that at 0oC is typically 2.5

6
times that at 20oC.

8.3.2.4 Summary of installation

The existing pipe is isolated from reservoir head, allowing access at both ends. Where damage of the
existing pipe causes significant inflow at any point, then the reservoir needs to be drained or other
works undertaken to stem the leakage, otherwise it will affect the annulus grouting. 7
The pipe is then cleaned to remove encrustations. Some cleaning procedures can damage the existing
pipe, so the choice of cleaning methods should be carefully considered and cleaning undertaken by
competent persons. After cleaning, a full CCTV survey of the pipe should be undertaken. Following this
survey, further preparatory works may be required to remove remaining encrustations or stem ingress.
8
Standing water in the pipe is then removed, typically using foam swabs. The survey will also establish the

9
horizontal and vertical alignment of the existing pipe, to enable the proposed liner details (OD, lengths,
flexibility) to be determined.

The liner pipe is brought to site and constituent segments welded. It is important to ensure that the
liner pipe is located as centrally as possible inside the host pipe/culvert to ensure uniformity of the grout
around the annulus. This can be achieved by using standard spacer products, or by welding shaped
guide studs to the outside of the liner pipe (see Figure 8.4). These both help to reduce friction when 10
inserting the liner and to ensure central positioning and circular shape are retained throughout the
grouting process.

Figure 8.4 shows an example of spacer collars on a pipe, accommodating grouting pipes, before
insertion.
11
12
A nose cone is fixed to the end of the liner pipe to ease insertion and allow a central point to connect the
winching cable. The winch is set up at the other end of the pipe, and cable pulled through to connect to the
nose cone. The liner is winched through the existing pipe, preferably in one continuous motion. Guidance
on the calculation of permissible winching forces is provided in Section 3.2 of WRc (2001). Depending on

A1
the site constraints, the liner pipe may need to be welded in sections as it is pulled into position.

Once in place, forms are fitted at either end to prevent escape of the annulus grout. Holes for grout
insertion are created, with a mechanism to fit the grout hoses to prevent release and spillage. Vents to

Dams and reservoir conduits 85

 allow air escape during grouting
need to be installed. It is advisable
that there is a means to allow grout
pressure measurement within the
pipe during installation, as such
data allows early identification
of problems during grouting
to a much greater extent than
pressure monitoring at the pump.
The grouting pressures at various
sections of the pipe/tunnel should
be calculated, accounting for the
head loss in the grouting hoses.
If a long section of annulus is to
be grouted, consideration should
be given to having multiple grout
Figure 8.4 Spacers and grout pipes on pipe before insertion (courtesy
injection hoses to ensure that the
Mott MacDonald Ltd)
grouting process can be continuous.
With larger annulus volumes, there is a risk that the grout viscosity will decrease as it starts to set. This
could increase the pressure required to pump the grout into the annulus, and it may be enough to cause
the maximum external pressure of the pipe to be exceeded.

The grout mixing and pumping equipment is set up and checked for faults and adequate supply of
constituents and fuel before beginning. The grouting process should not start if there is a foreseeable
risk that it may be halted for any length of time before completion.

The grout is normally inserted at one end, though grouting from multiple locations is feasible provided
appropriate means of air evacuation are given. A retarder should be added to the grout mix to maintain
fluidity of the mix. Pre-bagged mix is preferable to ensure a consistent mix. Further information of the
grout mix and specification can be found in WRc (2001).

Potential pitfalls
An important and often overlooked detail when sliplining is selection of an appropriate SDR rating
of pipe. Many incorrectly assume that the pressure rating of the pipe (normally used for distribution
purposes) is the same externally as internal. In fact the permissible external pressure is far less than the
internal pressure ratings. Further information on the permissible external pressures of a range of pipes
is given in Chapter 5 of WRc (2001).

The maximum external head of water should generally be taken as the reservoir flood water level, as
it is quite conceivable that the pipe could be
subjected to this. If the pipe is in any way
deformed during the installation process, the
allowable external head significantly reduces.
This might mean that a pipe with lower SDR
value (and thicker wall) may be required.

Grouting is a specialist operation and should be

undertaken by competent specialists. Grouting
pressures should be strictly monitored during
the grouting process. If a blockage occurs in
the annulus (ie if a section of grout starts to
set), high pressures can develop, which can
easily cause the pipe to be crushed (see Figure
8.5 and Case study A1.21). Figure 8.5 Damaged HDPE pipe due to high external grouting
pressures (courtesy Mott MacDonald Ltd)

86 CIRIA, C743
Chapter 8 of WRc (2001) provides further information. Grout volumes should be carefully calculated

1
before the operation, and measured accurately during the pumping. The volume of water required and
number of grout bags in each grout pan are simple ways to do this.

8.3.2.5 Main costs and risks

The main costs and risks relating to both construction and operation are noted in Table 8.3.

Table 8.3 Sliplining – summary

2
Cost or risk Project element

Mobilisation and
labour/plant costs
Comments

‰‰

‰‰
depends on isolation cost
plant and labour costs are typically low.
3
Construction cost
Typical material

4
‰‰ depends on diameter and length, but generally low.
costs
‰‰ risk of damaging pipe during cleaning process, depends on process
Damage to existing used
structure ‰‰ risk of damaging pipe during pull-in process at steps or discontinuities

5
‰‰ excessive local grout pressure could rupture existing pipe if very weak.

Construction risk ‰‰ lining pipe can snag during installation

‰‰ poor grout mixing/installation can lead to local high pressure deforming
Other construction liner, or voids to be left enabling leakage path formation
risks ‰‰ if damage is done it is very difficult to rectify
‰‰

‰‰
measures required to prevent floatation of liner pipe (see Section 5.2.1
and 5.2.2 of WRc, 2001).
depends on lining thickness, pipe diameter and roughness of existing pipe
6
Capacity reduction ‰‰ the cross-sectional area of the pipe is reduced

7
‰‰ the lining will provide a smoother surface mitigating the impact on capacity.
Operational cost
Surveillance/ ‰‰ increased confidence in pipe integrity may allow surveillance to be
maintenance/ reduced
running costs ‰‰ running costs may increase where capacity reduced.

8
‰‰ risk reduced by halting of deterioration
Good installation
‰‰ grouting the annulus can seal leaks associated with the original pipe.

Operational risk ‰‰ leakage between liner and pipe if pipe is not well cleaned or grouting
unsuccessful
Poor installation
‰‰ damage to pipe during poor cleaning can lead to weakening and

Note
potential collapse.
9
See Section 8.1 for guidance on applicability and colour coding.

8.3.3 Close-fit pipe lining 10

8.3.3.1 General
Close-fit pipe lining is an alternative to conventional sliplining, where the lining pipe is temporarily
reduced in cross-section to allow insertion into the existing pipe. There are several system options 11
available to temporarily change the pipe shape, including rolled (‘rolldown’), die drawn (‘swagelining’)
(see Case study A1.12) and folded systems. All systems have been developed based on the use of PE pipes.

Once the lining is in position it is expanded to full circular cross-section. Dependent upon the method
chosen, the lining may expand naturally. By choosing a lining pipe external diameter that is the same or
12
slightly larger than the existing pipe internal diameter, the lining should fit very tightly with the existing

A1
pipe, which negates the need to grout the annulus. This also has the advantage that the hydraulic
capacity is maximised, and may be increased over the original capacity depending on the roughness of
the existing pipe.

Dams and reservoir conduits 87

Installation can be very rapid and systems can typically accommodate the reservoir head and length
and size of pipes commonly found at UK embankment dams. Success is affected by temperature during
installation, and care is required when cleaning the existing pipe.

8.3.3.2 Technical envelope

The limitations of the liner in relation to pipe size and common structural problems affecting buried
dam conduits are noted in Table 8.4. The technical envelope is largely the same as sliplining.

Table 8.4 Close fit pipe lining – technical envelope

Pipe diameter ‰‰ 100 mm to 2300 mm accommodated.

‰‰ depends on pipe diameter and effectiveness of cleaning

‰‰ maximum winching forces on liner will determine friction limits
Pipe lining length
‰‰ typically lengths of 300 m are feasible, although lengths of up to 1500 m have been
achieved in a single pull within water supply systems.
‰‰ depends on SDR of lining pipe used
‰‰ reductions in lining thickness for structural solutions are limited by the requirement
Lining thickness
to use PE
‰‰ thin walled non-structural solutions are feasible.
‰‰ not an issue as long as pre-welded liner can be winched through without significant
Pipe slope
bend at change of gradient (within pipe or outside).
‰‰ no transitions in size or shape are possible
Changes in size/shape ‰‰ close fit lining is not suitable where the original pipe is segmented due to the risk of
the new pipe snagging at the joints.

‰‰ long radius bends can be accommodated, tight bends cannot

Bends ‰‰ radius bend capacity dependent on liner thickness, temperature and the system
adopted to reduce the pipe cross-section.

‰‰ most solutions are fully structural and do not rely on bond with existing pipe so
existing pipe material does not affect performance
Pipe material/preparation ‰‰ non-structural solutions exist for minor repairs that rely on the structural strength of
the host pipe
‰‰ existing pipe needs to be thoroughly cleaned before installation in all cases.

Leaking into pipe during ‰‰ large leaks need to be stemmed before installation
installation ‰‰ installation with minor leaks is feasible.

Gap bridging (at joints) ‰‰ all structural systems can comfortably bridge gaps unless these include a step.

‰‰ sagging can be accommodated up to the maximum allowable bend for each system
Deflection along existing pipe
‰‰ any sag will be reflected in lined pipe alignment.

Options
The cross-section of the pipe includes:

‰‰ Rolldown – process achieves a reduction in liner diameter of up to 10 per cent by passing a PE

pipe through a rig typically comprising of two rollers. The change in liner diameter is permanent
until pressure is applied to revert the liner to its original diameter meaning that the rolldown
process can be completed near or off-site and the rolled liner transported to the pipe. Rolldown is
suitable for small diameter liners (typically 100 mm to 500 mm) and can generally accommodate
bends of up to 11.25o. 300 m of lining can be fitted in a single typical installation. On one site as the
reservoir water cooled in winter the pipe is known to shrink and a slight flow in the annular space
that forms occurs.
‰‰ Die drawn – achieves a reduction in liner diameter by pulling a PE pipe through a die of smaller
diameter located at the inlet of the pipe. Die drawing is suitable for a range of liner diameters from
100 mm to 1000 mm and can generally accommodate pipe bends of up to 11.25o.
‰‰ Folded – liners are generally folded into ‘C’, ‘U’ or ‘H’ shapes during manufacture or on-site before

88 CIRIA, C743
 installation. When folded during manufacture the liners are delivered to site on coils. Folded liner

1
installation is suitable for liner diameters in the range of 75 mm to 1500 mm and can generally
accommodate bends of up to 22.5o.
‰‰ Non-structural – feasible using any of the methods previously mentioned with thin-walled liners.
This option is only suitable where the existing pipe has a high structural strength, where there

2
is minimal risk of water entering between the liner and culvert wall, and only suitable for fixing
minor leaks. The liner itself shall have no structural capability once installed.

8.3.3.3 Site suitability

The performance of close fit pipe lining in relation to common reservoir construction considerations is
noted in Table 8.5.
3
Table 8.5 Close fit pipe lining – site specific issues

Description

Draw-down/isolation
Comments

‰‰ pipe needs to be isolated from reservoir and access available at both ends.
4
‰‰ pipe lengths, welding equipment and any rolling, die or folding rig needs to

5
be delivered to site at pipe end where liner is to be inserted. Most rigs can be
Road access delivered to site on trailers
‰‰ a winch is required at the pipe outlet. These can be lifted to final position with
tracked plant or crane if road access not available.
‰‰ all lining pipe units need to be welded together before passage through a rolling,

6
die or folding rig and insertion into pipe
‰‰ the length of the working area needs to exceed liner length
‰‰ small lengths of liner are not advisable as the welding process is specialised and
Working area at pipe outlet (end
expensive
where liner is to be inserted)

7
‰‰ a working area of about 8 m length is required at the end of the pipe at the same
level as the existing pipe to ensure the new liner can be installed easily
‰‰ this working area needs to be larger for die drawing systems to accommodate the
die rig at the pipe entrance.

Wet weather during installation ‰‰ only impact is in relation to reservoir level control (where required).
Low temperatures during
installation
‰‰ liners become difficult to manipulate through rolling, die drawing or folding rigs at
temperatures below 0oC. 8
8.3.3.4 Summary of installation
Pipe preparation
9
Large leaks into the pipe need to be stemmed before installation. Small ingresses can be accommodated.

Rolldown
The pipe rolldown rig can be located close to the end of the pipe where the liner is to be inserted, but
10
as the change in pipe diameter is permanent, this does not need to be located at the entrance to the

11
pipe. Pipe lengths can be welded following rolldown and before insertion, however, this is generally not
economic for all lengths of pipe as the welding process is specialised and relatively expensive.

To install, the rolled down liner is pulled through the existing pipe using a winch located at the pipe
inlet or outlet.

To revert the pipe to its original diameter following installation the pipe is filled with cold water and
held at a pressure of up to twice its working pressure for 12 hours. Pressure testing of the pipe can also
12
be achieved during the reformation process. The water used to revert the pipe size is not contaminated

A1
and can be easily and safely discharged into a nearby watercourse.

Dams and reservoir conduits 89

Die drawn
The die rig needs to be located at the end of the pipe through which the liner is to be inserted. The liner
is pulled through the die and into the pipe using a large winch located at the pipe outlet. The change
in diameter is not permanent, so the liner needs to be kept under tension until installed meaning the
process of die drawing cannot be completed offsite. Winching loads are high, and large winches are
required for this method with purpose-built restraints mounted on a reinforced concrete slab at the pipe
outlet. Winch loads of up to 70t are not uncommon. The liner is reformed to its original diameter when
the tension is no longer applied.

Folded
Liners that are folded during the manufacturing process are installed by pulling the liner through the
pipe using a winch located at the pipe outlet. These liners are reformed by applying heat and pressure
(usually steam).

On site systems are available to achieve deformation of liners to a ‘U’ shape by passing the pipe through
a rig comprising of a plough that deforms the liner into shape. The rig needs to be carefully designed
to avoid overstressing the liner before installation. Bands need to be fitted around the deformed liner
to prevent it beginning to revert and rubbing against the host pipe during installation. The rig can be
located close to the end of the pipe where the liner is to be inserted, but does not need to be located at
the entrance to the pipe. It is possible to weld liners before they pass through the rig.

The liner is pulled into the pipe using a winch and draw line. The liner is reformed to its original shape
using cold water held at pressure as for rolled down pipes. Pressure testing of the pipe can also be
achieved during the reformation process. The bands around the liner shape break during this process.

Non-structural
Non-structural solutions are feasible using any method with thin walled liners. This option is only
suitable where the existing pipe has a high structural strength, where there is a low risk of water getting
between the liner and pipe and only suitable for fixing minor leaks. The liner itself shall have no
structural capability once installed.

Joints
Standard PE connections can be used, such as Viking-Johnson joints (see Figure 8.6) however, the liner
needs to be reverted to its original diameter using a stainless steel insert before joining.

Figure 8.6 Viking-Johnson pipe connection on a PE pipe (courtesy Mott

MacDonald Ltd)

90 CIRIA, C743
8.3.3.5 Main costs and risks
The main costs and risks relating to both construction and operation are noted in Table 8.6.

Table 8.6 Close fit pipe lining – summary

1
2
Cost or risk Project element Comments

‰‰ depends on isolation cost

Mobilisation and labour/ ‰‰ plant and labour costs are typically low (slightly more than sliplining)
plant costs ‰‰ costs for die drawn lining is more expensive due to winching
Construction cost
arrangements.

Typical material costs

‰‰

‰‰
depends on diameter and length, but generally low (less than
sliplining due to lack of grouting).
risk of damaging pipe during cleaning process, depends on
3
Damage to existing process used

4
structure ‰‰ risk of damaging pipe during pull-in process at steps or
discontinuities.
‰‰ lining pipe can snag during installation
Construction risk
‰‰ inappropriate pipe size or distortion in existing pipe can lead to
liner not expanding correctly, leaving voids in annulus

5
Other construction risks
‰‰ if installation is not done correctly it is very difficult to rectify
‰‰ health and safety issues for die drawing relating to winch cable
under tension if winch is located away from pipe exit.
‰‰ depends on lining thickness, pipe diameter and roughness of
existing pipe

6
Capacity reduction
‰‰ capacity reduced less than sliplining, may increase in some
Operational cost circumstances due to low friction of new lining.
Surveillance/ ‰‰ increased confidence in pipe integrity may allow surveillance to be
maintenance/running reduced

7
costs ‰‰ running costs may reduce where capacity increased.
Good installation ‰‰ risk reduced by halting of deterioration.
‰‰ leakage between liner and pipe if pipe is not well cleaned or
Operational risk installation unsuccessful
Poor installation

8
‰‰ damage to pipe during poor cleaning can lead to weakening and
potential collapse.

Note
See Section 8.1 for guidance on applicability and colour coding.

8.3.4 Cured-in-place pipes (CIPP) 9

8.3.4.1 General
CIPP lining is an established technology for arresting the deterioration of conduits. A resin-impregnated
flexible liner is pulled or pushed into the pipe and inflated, then cured.
10
Currently no CIPP liner system is approved by the Drinking Water Inspectorate (DWI) for use in pipes
transferring water for potable supply. However development of new lining systems is ongoing and it is
expected that a DWI approved lining system will be available in the near future. 11
8.3.4.2 Technical envelope
The limitations of the liner in relation to pipe size and common structural problems affecting buried
dam conduits are noted in Table 8.7.
12
A1
Dams and reservoir conduits 91
Table 8.7 CIPP lining – technical envelope

Description Comments
150 mm to 2200 mm accommodated, depending on curing method:
‰‰ water: diameters up to 2200 mm
Pipe diameter
‰‰ air/steam: diameters up to 1200 mm
‰‰ UV: diameters up to 1500 mm.
Depends on pipe diameter:
Pipe lining length ‰‰ small pipes (150 mm to 225 mm), maximum length is 200 m due to friction
‰‰ large pipes longer lengths are feasible up to 800 m.
‰‰ 3 mm to 55 mm
Lining thickness
‰‰ can be reduced through using carbon fibre reinforced liners, which are more expensive.

Pipe slope ‰‰ maximum drop over pipe to be lined is around 2 m (with insertion at lower end).

Changes in size/shape ‰‰ one transition in shape or size is feasible within the lining (depending on manufacturer).

Bends ‰‰ most systems can accommodate bends of up to 45o.

‰‰ solution does not rely on bond with existing pipe – existing pipe material does not affect
Pipe material/ performance
preparation
‰‰ existing pipe needs to be well cleaned before installation.
‰‰ large leaks need to be stemmed before installation to avoid resin being washed off liner
Leaking into pipe before curing, and to avoid lining deformation
during installation ‰‰ small leaks can be countered by fitting a pre-liner to prevent water contacting the CIPP liner
before curing.
‰‰ all systems can comfortably bridge gaps and accommodate small steps
Gap bridging (at joints) ‰‰ note that step accommodation may require insertion from higher (upstream) pipe end and
lack of external support at gaps may allow liner to stretch, leading to a locally weaker zone.

Deflection along ‰‰ no problem except pooled water within sagging pipes need to be removed before installation
existing pipe ‰‰ any sag will be reflected in lined pipe alignment.

Options
Different solutions are available for pressurised and non-pressurised pipes. Solutions for non-pressurised
pipes can be used where small pressures of up to 0.5bar are encountered without relying on the existing pipe.

There are also a number of resin types available. Standard resins are based on polyester. Epoxy resins
are stronger and are used where a thinner liner is necessary. They are also used where chemical
resistance is important. It is possible to introduce a ‘patch’ to a polyester impregnated liner where epoxy
is used instead. This can strengthen a short length of the liner where it is expected to resist additional
load due to a defect in the existing pipe.

An alternative to standard linings is a lining incorporating carbon fibre, which can increase strength and
allow a thinner lining section to be used. ‘This is more cost effective with larger pipe diameters (ie >600
mm) because of the high cost of carbon fibre components.’

Air/steam curing of liners is generally faster than water curing, both in terms of setting up of the
equipment and time taken to cure the liner (see Section 8.3.4.4). However, air lining is only suitable for
pipe diameters up to 1200 mm.

The hot water curing process allows for a greater control over the curing temperature, reducing the
risk of overheating the liner, which may result in a reduction in strength of the finished liner during air/
steam curing. It is suitable for all pipe sizes up to 2200 mm.

UV cured liners include a layer of chopped glass, which achieves two to three times the strength of water
or steam-cured liners. This leads to savings in weight and maximises retained pipe capacity. UV cured
lining systems should be designed in accordance with ATV-M 127-2:2000 (please note this is a German
code (see Case study A1.4).

92 CIRIA, C743
If the conduits are perforated, ie for land/toe drainage purposes, failed pipes can be remediated by CIPP

1
systems and a robot used to cut holes/slots through the lined pipe and host pipe after curing. The CIPP liner
can be designed as a fully-structural liner negating the need for dig up and replacement of the conduit.

It is currently possible to create CIPP liners that can accommodate a change of shape or size at one point
along a conduit.

8.3.4.3 Site suitability

2
The performance of CIPP lining in relation to common reservoir construction considerations is noted in
Table 8.8.

Table 8.8 CIPP lining – site specific issues

3
Description Comments

Draw-down/isolation
‰‰

‰‰
pipe needs to be isolated from reservoir and access available at both ends
some techniques involving access only from one end are available, but are high risk
4
and not recommended for use at reservoirs.
water curing – requires access for scaffold delivery crane, boiler trailer, water

5
‰‰
deliveries (if required)
‰‰ scaffolding be placed at pipe outlet, water deliveries and boiler can be up to 30 m
from this point
‰‰ tankers shall be required to remove water afterwards

6
Road access ‰‰ air and steam curing (only applicable for smaller pipes, access for liner delivery and
steam plant required) less onerous access requirements than watercuring
‰‰ UV curing, access road not required
‰‰ curing plant consists of computer controlled winch and light train
‰‰ plant is generally located in ready built trucks requiring road access, but there are

7
separate systems that can be transported by tractor/tracked vehicles.
‰‰ water curing (scaffold only) about 5 m x 5 m (requires access for deliveries nearby)
Working area at pipe outlet ‰‰ water curing – scaffolding + crane + deliveries – about 20 m x 30 m
(end where liner is to be
inserted) ‰‰ air and steam curing – about 10 m x 5 m

8
‰‰ UV curing – 3 m x 8 m.

Wet weather during installation ‰‰ only impact is in relation to reservoir level control (where required).

Low temperatures during

‰‰ only impact is access for delivery to site.
installation

8.3.4.4 Summary of installation

9
The pipe needs to first be isolated from the reservoir, cleaned and inspected as noted in Table 8.8.

Pipe preparation
For CIPP liners, all large leaks into the pipe needs to be stemmed before installation. Small ingresses can
10
be accommodated through using a pre-liner (plastic/nylon sheeting) to ensure water does not wash out

11
resin from lining before curing takes place.

Access
Suitable access needs to be available for the installation plant and material deliveries. The extent of

12
access requirements is determined by the lining type and curing process as follows:

‰‰ CIPP lining with water curing – requires a scaffold to be constructed immediately beside the pipe
end from which the liner is to be inserted. A mobile crane is required to assemble the scaffold and
inversion ring, preferably as close as possible to the insertion point. Water tankers (where used)
and boiler trailers need to be able to access within 30 m of the insertion point – it is preferable if
they can be situated beside the scaffold, but access along the dam crest is often suitable. For pipes
of 450 mm and smaller, an alternative to the scaffold called a ‘chip unit’ can be used, which allows
A1
Dams and reservoir conduits 93
 installation with less space required at the pipe.
‰‰ CIPP liner with air/steam curing – does not require such large plant. Plant is required only for
delivery of the liner, and air/steam blower. This needs to be located within 10 m to 20 m of the
insertion point.
‰‰ CIPP liner with UV curing – plant consists of computer controlled winch and light train. Plant is
generally located in ready-built trucks requiring road access, but there are separate systems that
can be transported by tractor/tracked vehicles. Location of plant is dependent upon length of lined
pipe and winch length.

For all types of CIPP lining, human-entry needs to be available to both ends of the pipe that is to
be lined. The access to the end not used for insertion is required to trim the end of the liner, which
overshoots the pipe exit by one to two metres. There are some methods available to install a CIPP lining
with access to only one end of the pipe. However, such methods are considered at high risk of failure
(blocking the pipe) and so are not recommended for use at reservoirs.

Installation and curing

‰‰ Water cured lining (all pipe sizes) – the lining sock is hung from the scaffold and filled with cold
water, which drives the sock along the pipe. The water does not have to be specially treated, and
reservoir water is typically suitable. Sufficient pressure is exerted for the lining to take the shape of
the existing conduit along its entire length. When this is achieved, the water is circulated through
a boiler so it heats up. When above 75 degrees centigrade, the liner will start to cure. The curing
process typically takes 4 to 20 hours to achieve depending on the lining length and pipe diameter.
When this is completed the water is gradually cooled down, and then removed from the site for
treatment and disposal. The curing water contains contaminants (such as styrene) and should not
be released into the environment without treatment. Sufficient measures need to be in place to
prevent accidental contamination of nearby watercourses. Water curing systems are generally the
cheapest where good access is available to the insertion point for all plant (without having to spend
large sums on access improvements), and where a supply of water is available.
‰‰ Air/steam curing (pipes up to 1200 mm diameter) – the lining is blown into the pipe with
compressed air and inflated. Steam is then blown in to cure the lining. This can avoid using water
and has less onerous access requirements than water curing, with no scaffolding, large boilers or
water tankers required. Curing times vary from 4 to 10 hours.
‰‰ UV curing (pipes up to 1500 mm diameter) – the lining is brought to the site in a flat, folded form
and is winched through the pipe. Air is then pumped into the lining to expand it to fit the shape of
the pipe. A robot with camera and lights is then inserted through a gland. The camera can confirm
that the liner has been successfully installed, and then the lights are turned on. These cure the
pipe rapidly. UV cured liners can be installed and cured within four to five hours, minimising
conduit down time and the period over which reservoir draw-down is required.

In all instances, the lining sock will extend just beyond the end of the pipe. This end piece is then cut off
and removed.

Further guidance on the installation and curing of CIPP is provided in Section 3.4 of WRc, 2001.

8.3.4.5 Main costs and risks

The main costs and risks relating to both construction and operation are noted in Table 8.9.

94 CIRIA, C743
Table 8.9 CIPP lining – summary

Cost or risk Project element Comments

‰‰ for water/steam curing, generally around 80 per cent of

1
Mobilisation and labour/plant project costs (excluding any cofferdam or major pumping
costs to hold down reservoir water level). This is reduced for
Construction cost

2
UV curing.
‰‰ around 20 per cent of a typical project. Higher for epoxy
Typical material costs
resin, UV curing, reinforced composite liners.
‰‰ only relates to damaging pipe during cleaning process –
Damage to existing structure
depends on process used.
Construction risk
Other construction risks
‰‰

‰‰
liner snags during installation.
stop/starts during installation can cause liner to stick
(mainly occurs due to breaks in water supply).
3
‰‰ depends on lining thickness, pipe diameter and

4
Capacity reduction roughness of existing pipe
‰‰ capacity can increase or decrease.
Operational cost
‰‰ increased confidence in pipe integrity may allow
Surveillance/maintenance/ surveillance to be reduced
running costs

5
‰‰ running costs may reduce where capacity increased.

Good installation ‰‰ risk reduced by halting of deterioration.

Operational risk ‰‰ leakage between liner and pipe if pipe is not well cleaned
Poor installation ‰‰ damage to pipe during poor cleaning can lead to

6
weakening and potential collapse.

Note
See Section 8.1 for guidance on applicability and colour coding.

8.3.5 In situ sprayed linings 7

8.3.5.1 General
In situ sprayed lining is a new technology for arresting the deterioration of conduits. A resin
(polyurethane or epoxy) is sprayed onto the inside of the existing pipe from a spinning delivery head,
which is winched through the pipe.
8
9
Various alternative systems are available to cure the resin. In some cases UV lights are passed through
the pipe, which subsequently cures the resin lining after installation. Alternative systems are available
in which a computed controlled delivery head mixes and applies resin and hardener, and eliminates the
need for UV curing. However, these systems have been developed primarily to improve drinking water

10
quality in heavily corroded pipes and only offer remediation of small leakages in pipes.

The thickness (strength) of applied linings and the pipes that can be accommodated are restrictive (see
Section 8.3.5.2). DWI approved products are available.

8.3.5.2 Technical envelope

The limitations of the liner in relation to pipe size and common structural problems affecting buried
11
dam conduits are noted in Table 8.10.

12
A1
Dams and reservoir conduits 95
Table 8.10 In situ sprayed lining – technical envelope

Description Comments

Pipe diameter ‰‰ 100 mm to 1000 mm.

Pipe lining length ‰‰ up to 150 m.

Lining thickness ‰‰ typically 3 mm to 20 mm.

Pipe slope ‰‰ depends on capacity of winching system.

Changes in size/shape ‰‰ cannot be accommodated.

Bends ‰‰ sharp bends cannot be accommodated.

Pipe material/preparation ‰‰ iron, steel, cement and concrete pipes.

Leaking into pipe during ‰‰ needs to be avoided

installation ‰‰ lining applied through water will not be effective and will require remediation.
‰‰ small gaps (5 mm) can be bridged, but can result in local thinning of lining
Gap bridging (at joints)
‰‰ steps risk snagging the resin hoses and unequal spray application in the vicinity.

‰‰ no problem except pooled water within sagging pipes needs to be removed before
Deflection along existing pipe installation
‰‰ any sag will be reflected in lined pipe alignment.

Note
Due to relatively thin applications, in situ sprayed liners have limited capacity to resist external heads compared to other lining systems.

8.3.5.3 Site suitability

The performance of in situ sprayed liners in relation to common reservoir construction considerations is
noted in Table 8.11.

Table 8.11 In situ sprayed lining – site specific issues

Description Comments

Draw-down/isolation ‰‰ pipe needs to be isolated from reservoir and access available at both ends.
‰‰ requires delivery of mixing rig (small trailer) and resin materials at one
pipe end – can be located at a distance from the outlet (eg on dam crest)
Road access depending on total hose length available and the pipe length to be lined
‰‰ access for CCTV camera and spray head rig insertion at one pipe end, winch at
the other (both can be manually handled).
Working area at pipe outlet (end
‰‰ very small, sufficient to insert spray head rig (1 m x 2 m).
where liner is to be inserted)
Wet weather during installation ‰‰ only impact is in relation to reservoir level control (where required).

Low temperatures during installation ‰‰ lining should not be undertaken in temperatures (of pipe wall) less than 3°C.

8.3.5.4 Summary of installation

The pipe needs to be isolated from the reservoir, cleaned and inspected as noted in Table 8.11.

Existing pipe condition

In situ sprayed liners can only be used when significant ingress flows are not present. Excessive inflow
may dilute the resin application and reduce effectiveness.

Access
Access requirements for in situ resin lining are less onerous than for the other lining systems. The
distance from which the mixing rig (including resin materials), air compressor and generator needs to be
situated from the pipe end is dependent on the length of hose available. Hose lengths are usually of the
order of 200 m.

96 CIRIA, C743
Installation process

1
A mixing rig is set up close to the pipe outlet, and sufficient quantities of resin to undertake the lining
process are heated and mixed in the rig. The resin is typically brought to site as two separate components
(base and activator), which react when mixed.

After the pipe has been cleaned and swabbed dry, it is inspected over its length using a CCTV system.
This should also prove the feasibility of the spray head rig passing along the entire length of the pipe
without snagging/entrapment during the lining process.
2
The two component materials are pump-batched by volume, mixed and delivered under pressure to
the spray head, which is winched along the pipe. Fast curing systems deliver the resin and hardener
separately to the delivery head, where they are mixed on delivery. A smooth, resilient coating of the
required thickness is applied by the spray head. The entire pipe is lined in one operation.
3
4
Once the lining is in place it cures automatically. Cure times vary, with epoxy resins requiring about 16
hours to cure and polyurethane resins offering curing within 60 minutes.

After sufficient curing, the lining is inspected again to ensure its adequate installation and to identify
any faults. Providing the pipe has been lined satisfactorily, it is then flushed as required for potable
supplies, and brought into operation.
5
It is possible to return treated pipes into service within one day using spray linings with polyurethane resins.

8.3.5.5 Main costs and risks

The main costs and risks relating to both construction and operation are noted in Table 8.12.
6
7
Table 8.12 In situ sprayed lining – summary

Cost or risk Project element Comments

Mobilisation and labour/ ‰‰ relatively low cost

Construction cost plant costs many suppliers deliver a single unit application rig to site.

8
‰‰

Typical material costs ‰‰ relatively low cost as replacement pipe is not required.
Damage to existing ‰‰ only relates to damaging pipe during cleaning process – depends
structure on process used.
deposits not removed during cleaning can be lined over, causing

9
‰‰
premature failure
Construction risk ‰‰ ridges in the lining can be formed due to jerking of the spray
Other construction risks head through hoses snagging or winch gear malfunction
‰‰ protruding obstacles or jerking action of the lining head may
result in sections of the main with no coating

Capacity reduction
‰‰

‰‰
steps and joints in pipe can result in uneven application of lining.
liner is very thin and friction reduced so capacity generally
10
increases.
Operational cost ‰‰ increased confidence in pipe integrity may allow surveillance to

11
Surveillance/maintenance/ be reduced
running costs
‰‰ running costs may reduce where capacity is increased.
‰‰ risk reduced by halting of deterioration, but no structural benefit
Good installation
of liner and unable to remediate large leaks.

12
‰‰ damage to pipe during poor cleaning can lead to weakening and
Operational risk
potential collapse
Poor installation
‰‰ thin liner can easily deform due to underestimated external
pressures.

Note
See Section 8.1 for guidance on applicability and colour coding.
A1
Dams and reservoir conduits 97
8.3.6 External renovation
External renovation of buried pipes is rarely carried out, as common pipe materials are inert and not
susceptible to significant corrosion on the outside of the pipe. Also, it requires exposure of the pipe,
and if this is feasible, then the optimum solution in reducing future risks associated with the pipe would
normally be to replace the damaged section. However, there may be situations where the pipe cannot
easily be taken out of service to facilitate internal renovation or replacement.

If a section of buried pipe has suffered external deterioration, and replacement of that section is not
preferred, then it could be exposed and a protective layer applied to the outside face of the pipe. Specialist
advice should be sought when specifying the protective layer, as ground conditions aggressive enough to
cause significant pipe deterioration would require a specialist product. Such situations do arise with respect
to pipes laid through contaminated ground, but this is not normally a consideration at dam sites.

8.4 RENOVATION OF PIPES IN CULVERTS

Human-entry culverts in embankment dams generally relate to spillways or draw-off works (including
scour). In the latter instance, the culvert will usually have at least one pipe running along its length. As
with all pipes, these will deteriorate with time, and at some stage may require renovation or replacement.

Deterioration can occur along the inside or the outside of the pipe. Deterioration of the inside face of the
pipe can be caused by abrasion or chemical interaction with carried water, and halted by the provision
of a resistant lining within the pipe. An internal survey should be undertaken to ensure that pipe wall
thickness has not reduced to an unacceptably low figure.

Deterioration on the outside of the pipe will most likely be caused by corrosion, impact due to works in
the culvert, or movement of the pipe relative to its surroundings.

The following sections discuss options relating to repairing the existing pipes. Alternatively a part of or
the entire pipe may be replaced (see Section 9.2).

8.4.1 External repair

8.4.1.1 Painting
A wide range of painting systems can be applied to pipes externally to inhibit corrosion. These generally
require surface preparation and the application of a primer, top coat and sealer. Water companies
generally stipulate that all painting for anti-corrosion protection meets Water Industry Mechanical and
Electrical Specification (WIMES) 4.01 (ESR Technology, 2005). This specification has been designed to
meet the requirements of paint systems at a wide range of water site installations, including steel, CI and
DI pipes, valves and other fittings.

All systems depend on thorough surface preparation and suitable ambient conditions, though some
coatings are more tolerant to surface moisture than others. WIMES 4.01 covers the standard of substrate
preparation for each material type with reference to applicable British and harmonised standards.

It is important to note that the cleaning method used needs to take account of the pipe material. For
example, abrasion systems used to clean steel could cause significant damage to softer CI pipes. Abrasion
of CI can remove any mill scale, which otherwise acts to protect against corrosion.

Coatings used for repair need to be compatible with substrates or original coatings.

Painting fumes may cause a dangerous atmosphere to build and additional ventilation may be required in
enclosed or confined spaces to deal with this. In addition, heaters or dehumidifiers may be required (with
careful consideration of exhaust locations). The ambient temperature is critical for the curing of some coatings.

98 CIRIA, C743
8.4.1.2 Bolted joints
Pipework in UK embankment dam culverts often has
flanged connections. These are points of weakness within the 1
pipeline system and are prone to corrosion, with substantial
flaking on the bolt material occurring during rusting (see
Figure 8.7). The associated loss of section can often look
considerably worse than what is found, however such bolts
should be replaced before application of tools to the bolt
2
heads or nuts becomes impossible. Such replacement should

3
be carefully planned with bolts replaced in a sequence
considering the loads on the remaining bolts during the
replacement operation.

4
Such bolted connections may also be subject to unbalanced
loads caused by movement of the pipe or variation in the
flanges or gasket. This can cause the loads to be distributed
unequally between the bolts. Such joints can be reinforced
with additional support, anchor blocks or tie bars.

5
Figure 8.7 Extensive rusting on a bolted
pipework connection in a UK
embankment dam (courtesy
8.4.1.3 Movement accommodation Canal & River Trust)

Pipes in embankment dam culverts are subject to movement

caused by a variety of reasons:

‰‰ settlement or spreading of the dam foundation, causing the support and restraint positions to move
6
forcing the pipe to move or changing its support conditions

7
‰‰ thermal expansion or contraction of the pipe
‰‰ pressure surges
‰‰ seismic events.

8
To safely deal with these movements, the pipeline should be designed to be adequately restrained, but
also to allow flexibility where required.

Lack of flexibility is a particular problem for flanged pipes. This is commonly overcome by using spigot
and socket joints in places, or introducing flexible flange adaptors. These are often prone to enhanced
corrosion compared to the rest of the pipe as they are often of thin sections.

Where a flange adaptor is subject to dripping water from above, it should be protected by guttering
9
covering with a plastic or rubber sheet or wrapped in a proprietary waterproof tape.

Badly corroded flange adaptors should be refurbished or replaced. 10

8.4.1.4 Restraint
Pipes in culverts can be encased within the culvert invert, or supported at regular intervals on masonry,
brickwork or concrete saddles. Such pipework should be restrained from excessive movement from all causes.
11
Where provided (or retrofitted), such restraint is commonly in the form of metal straps that arch over the
top of the pipe and are bolted to each supporting saddle. Such straps have often been flimsy in relation
to the thermal movements of the pipe, and prone to cutting into the coating of the pipe. Alternatively the 12
connection to the saddle may have come loose. Straps should be painted or replaced where damaged,
and enhanced where their sufficiency is questionable.

Restraint may be provided by encasing the pipe in concrete at each support. However, the ability of the
pipe to move longitudinally at each support point should be considered.
A1
Dams and reservoir conduits 99
Note that where pipe coatings have not been stripped back before pouring the concrete a poor bond
is likely, allowing some movement. If the pipe is within a service reservoir and can be dewatered, the
arrangement may have to be designed to resist flotation forces.

Lateral pipe restraints are sometimes required to resist hydrodynamic forces induced by bends in the
conduit alignment.

8.4.1.5 Fittings
Due to the range and complexity of valves and their deterioration, managing valves at dams is outside
the scope of this guide. For more detail on the management of valves at UK dams see Reader et al (1997),
which also provides advice on the condition assessment of pipework in dams.

8.4.2 Internal repair

See Section 8.3.

8.5 RENOVATION AND REPAIR OF HUMAN-ENTRY

CULVERTS
The challenges associated with human-entry culverts in embankment dams are considerably different to
those associated with pipes. Failure modes and materials are different and often more variable – shape,
size and alignment change over length, and the size of the structure and access all lead to a wide range of
problems and solutions. Routine maintenance of the culvert fabric is also possible, while replacement of a
human-entry culvert is usually unfeasible due to the likely costs.

This section discusses the renovation and repair of human-entry conduits in UK embankment
dams. The majority of these are of masonry construction while modern dam culverts are
predominantly formed of reinforced concrete. So, guidance provided here is based on culverts
formed of these materials.

8.5.1 Importance of routine maintenance

Access to most human-entry reservoir culverts is restricted because of their enclosed or confined nature,
and the need for adequate security arrangements to protect vulnerable pipework from vandalism or
terrorism. Occasionally, routine maintenance and even examination/inspection is sometimes delayed for
several years because of the greater costs and planning complexities associated with working in culverts.

Regular inspection of conduits is an important aspect of ensuring the dam does not progress towards
failure. However, regular basic maintenance is also important.

The main reason for completing cyclic maintenance within a culvert is to slow the deterioration of the
culvert itself or the assets within it, particularly deterioration caused by water and humidity. This is
largely achieved by cleaning, controlling leakage flows and completing timely small repairs. Such actions
will have benefits for:

‰‰ Pipework and valves within the culvert. Cyclic wetting and drying accelerates rust, particularly in
crevices such as those around flanges, nuts and bolts, requiring higher frequency replacement. As
noted elsewhere within this guide, the replacement of pipework and valves can be a very expensive
exercise, and extension of life due to reduced atmospheric corrosion can be very variable.
‰‰ Electrical systems within the culvert such as for lighting or for monitoring devices.
‰‰ The culvert wall fabric. Minor and minimally disruptive repairs can reduce the requirement for
more significant remedial works over time.

100 CIRIA, C743

Dam owners should recognise the value of routine surveillance and maintenance for dam management,

1
and should budget for sufficient funds and resources. These activities should be viewed as necessary and
unavoidable to ensure the safe and efficient operation of a reservoir.

Owners should encourage contractors to find ways of improving efficiencies in their methods of working.
Consideration should also be given by the owner to co-ordination of work activities. The completion of
several maintenance activities at one dam at one time can provide savings in procurement, particularly
relating to mobilisation. Alternatively, owners managing a portfolio of dams could consider the 2
packaging of maintenance works for all dam culverts within a geographic region to be undertaken by the
same maintenance team.

Records of maintenance and repair works to culverts, including before and after photographs and
measurements where applicable are invaluable especially when establishing the cause and effect of any
3
new defect as well as the cost and programme for remediation.

Cleaning of culverts and drainage maintenance are discussed in Section 4.6.2. Typical repairs to
control leakage or improve wall fabric are discussed in the following sections. It is the asset manager’s
4
responsibility to determine which specific works are required, how often they are to be carried out and
to allocate the necessary resources and budgets to complete the works.

8.5.2 Impact of repairs on leakage

5
The impact of repairs on leakage paths should be seriously considered before starting such work. In spite
of great efforts to ensure that the culverts were water-tight, many culverts in embankment dams have
developed small leaks of water entering from the surrounding saturated fill, which are long established 6
and show no sign of carrying fill material. Such small leaks may accelerate the deterioration of pipelines,
valves or electrical systems within the culvert. Reservoir managers or engineers may be tempted to seal
such leaks to alleviate this deterioration, even though the leaks do not cause concern relating to the
culvert or to dam safety. 7
In general, it is not advisable to seal such leaks. The leaks act to reduce pore water pressure in the

8
surrounding fill, and the resulting pressure increase caused by sealing the leak may unintentionally
lead to more serious or hidden damage such as hydraulic fracture of the dam core or instability of the
downstream shoulder. In addition, the local increase in pore water pressure may cause more leakage in
a different location, at the next weakest area of the culvert wall. Finally, the removal of points at which

9
leakage can be monitored for long-term trends means that it will be more difficult to link past leaks with
future performance.

Such low risk leakage flows should instead be managed through the creation of a passive drainage system.
The principal of a passive drainage system is the management of ingress water rather than prevention.
This means transferring flows away from assets susceptible to damage within the culvert by provision of an
adequate water control system that may include guttering or placing a protective cover over particular assets. 10
The one disadvantage with passive measures is that there is usually a residual maintenance responsibility

11
to ensure that the drainage paths and installed guttering are kept clean and free of debris. The cost of
this maintenance is often underestimated and if neglected can be more costly and problematic.

Where a small leak enters the culvert through a defect in the culvert fabric, it may be prudent to provide

12
a small weep pipe to allow leakage measurement in conjunction with any re-pointing or patch repair.
However, if the leak is carrying soil material into the culvert then there may be a serious problem, which
would warrant a more thorough investigation and probably substantial remedial works.

Flow from a series of leaks or weep pipes may be carried along the culvert wall in a guttering system
fixed to the culvert wall or a channel along the invert. Such improved drainage within the culvert should
not obstruct visual inspection of the culvert, and measurement devices should be located for ease of
measurement. For further details on systems for recording drainage flows see Section 6.4.
A1
Dams and reservoir conduits 101
8.5.3 Localised repair of masonry culverts
The most common material that has been used for the construction of human-entry conduits in UK
embankment dams is masonry, using either bricks or stone blocks. These materials are generally very
durable, requiring little maintenance. However, in some instances, poor construction quality may require
localised repairs where loss of the wall strength has occurred. Otherwise deterioration can be accelerated
by severe conditions, particularly around downstream portals, or along spillway culverts that are subject
to the passage of air, water and detritus. The most likely types of repair to brickwork are discussed in the
following sub-sections, namely re-pointing, patch repairs and reconnecting leaves of brick arches.

The most common repairs to masonry culverts are mortar pointing, patch repairs and crack repairs are
also discussed. The most common repairs to concrete culverts are patch repairs, crack sealant and joint
repair, which are discussed in Section 8.5.4. Other works that are also applicable to masonry structures
are discussed in subsequent sections. Considerable thought should be given to the effect of such repairs
on established leakage paths. The following guidance is applicable to parts of culverts where leakage
paths will be unaffected or formalised through insertion of drains through the culvert wall.

8.5.3.1 Mortar pointing

Loss of pointing and jointing mortar loosens masonry units (potentially leading to their dropping from an
arch), and also reduces the ability of the masonry to transmit and evenly distribute external forces. This
can cause the concentration of stresses in localised areas, potentially leading to cracking and distortion.

Assessing need
Repointing may be required to prevent progressive deterioration of the masonry fabric. Repointing is
most frequently required close to portals, particularly where subject to severe weather conditions. Frost
damage can affect the entire culvert, particularly if it is ventilated at both ends. Areas subject to flowing
water can also require repointing from time to time. The extent of mortar loss, which determines that
re-pointing is required, is subject to opinion, and may vary between owners. McKibbins et al (2009)
recommends the following when assessing the need for repointing:

“Waiting until the majority of the pointing in a lining has completely deteriorated or fallen out
before carrying out repointing is not advisable, because other damage may have already occurred
to the structure. For some infrastructure tunnel owners, a loss of mortar to a depth of over 20 mm
may trigger repointing. In cases where mortar shows signs of deterioration and the decision is made
to repoint, the joint should be cleaned out to a depth at least twice the width of the joint, or to a
maximum depth of 18 mm to 25 mm from the finished face” (Railtrack, 1996).

Mortar should not be removed by high pressure water hoses as this can lead to structural damage unless
the process is very carefully controlled.

The removal of the mortar before repointing should be undertaken carefully to prevent damage, loosening
and/or displacement of facing brickwork or masonry units. The depth to which loose and soft mortar
should be removed is dependant to some degree on the size of these units. In general, the depth should not
exceed 70 mm for brickwork, but may be increased from stone work with larger masonry units.

Programme considerations
If repointing is recommended in open-ended culverts that are rarely accessed, a survey should be
undertaken to confirm if protected species such as bats are present. The presence of bats may require
an alternative roost to be created and enhanced monitoring during the works, significantly adding
to cost and programme. Alternatively they may require timing restrictions to avoid working during
hibernation periods.

Pointing is affected by ambient conditions. Ideally repointing should be programmed to avoid freezing
temperatures and, where lime-based mortars are used, wet weather.

102 CIRIA, C743

Choice of mortar

1
The mortar used when re-pointing needs to be chosen carefully considering the mortar and masonry
present in the structure. Masonry work at old dams was generally completed using lime mortars, which
are generally much more flexible than mortars formed with Portland cement. These allow the culvert
to follow the cyclic movement of the surrounding fill without cracking. Where lime mortar is present, a

2
comparable mix should be used in repairs. If a stiffer Portland cement mortar is used then this may lead
to cracking and rapid deterioration of the masonry units and mortar.

Where lime mortars are to be used, specialist advice should be sought on the appropriate mix and any
additional (eg environmental) constraints on application.

Application of mortar
3
Lime mortar should typically be applied with pointing keys to ensure joints are adequately filled (see
Box 8.1).

Box 8.1 Application of mortar (from McKibbins et al, 2009) 4

When cleaning out for repointing, care should be taken to avoid damaging the surrounding masonry units. Hand tools

5
(quirks and long necked jointing chisels with parallel faces) are normally adequate where the old mortar is weak.
Where joints are thin and dense mortar has been used, it can sometimes be difficult to remove. The use of mechanical
tools should only be considered when necessitated by the scale of the work. Where such methods are necessary,
appropriate equipment should be used by skilled operatives to prevent damage to the masonry units. Cutting out using
angle grinders is not advised as the risk of damage to the masonry units is great.

6
Where the deterioration of the jointing mortar is extensive, resulting in voids and friable mortar deep in the joint, to the
extent that it is considered beyond repair by repointing techniques, then pressurised compressed air or mechanical
repointing may be necessary. The loose and very soft mortar should be removed back to more solid material
A suitable mortar can then be injected to fill the joints under pressure using compressed air or mechanical pointing
equipment to pressurise the mortar and force it through a hose to a gun nozzle. The operator should then build up mortar

7
in layers from the back of the joint to the front in one continuous operation to avoid cold joints.

8.5.3.2 Patch repairs to masonry

8
The replacement of a number of stone units or bricks may be required where the strength of a culvert
lining is undermined due to localised deterioration of mortar and/or the stone or brick.

In order to ensure a continuous stress distribution within a culvert, it is preferable complete a patch
repair using the old bricks/masonry units with new mortar. Where the units have deteriorated to an
unacceptable level, replacement with recycled or new units will be necessary. Care should be taken to
ensure the new units are similar in size and property to the existing units to prevent an overly-strong or
9
overly-stiff repair. This can result in a hard spot in the tunnel lining leading to local stress redistribution,
potentially causing premature failure or distress of the surrounding area.

Where the decision is made to use new units, McKibbins et al (2009) recommends that: 10
“Where new bricks are used, types with low water absorption characteristics should be specified
to avoid expansion causing compressive stresses in surrounding bricks. This can also be avoided
with the use of lime mortars. Over the years, a wide variation in brick sizes have been manufactured,
and bricks in old culverts may not have the same dimensions as new bricks. When renewing 11
brickwork, consideration should be given to matching the existing brick size if possible as
the bed joints may have to be deeper or thinner than those in the original brickwork.”

Manufactured and natural stone masonry units should comply with BS EN 771-5:2011 and BS EN 771-
6:2011 respectively. Where possible they should be selected based on proven durability and resistance to
12
weathering in a similar climate and exposure condition to the masonry to be repaired.

Design of masonry structures should be undertaken in accordance with BS EN 1996-1-1:2005. A

guidance note for the application of this standard in the UK is also available, namely PD 6697:2010.
A1
Dams and reservoir conduits 103
Whether new or existing units are used it is generally good practice to ensure that the new jointing
mortar is weaker than the unit being used in the repair. New mortar should also have adequate
permeability to allow the brickwork to breathe and for moisture to evaporate through the joints rather
than through the masonry units.

For further information on the selection of materials for masonry repair see McKibbins et al (2009) and
the Concrete Society (2005) (see Box 8.2).

Where individual bricks/units in the crown or small areas in a wall require replacement, support
measures are generally not required. However, where a larger area, comprising several bricks, is to be
renewed over the crown of the tunnel, some form of temporary formwork may be required to support
the brickwork until the mortar has attained sufficient strength.

Following the breaking out of the deteriorated area, the exposed surface should be inspected and
the planned works should be reviewed against any further repair works that may be identified by the
inspection.

Box 8.2 Completing patch repairs (from McKibbins et al, 2009)

Where possible, when carrying out a repair, the bond of the original brickwork should be maintained and the new
brickwork keyed into the surrounding brickwork. No extra stiffening of the lining should be introduced through further
bonding between successive rings forming the arch of the lining. If this is done the structural behaviour of the lining may
be altered.
If the brickwork is to be tied between successive rings on a repair of more than one brick ring thickness, the use of
stainless steel brick fishtail ties is one suitable method of tying back new brickwork, installed on a diamond pattern with a
nominal pitch of 400 mm × 400 mm.
Repairs to stonework require similar considerations to those discussed for brickwork, but employ slightly different
techniques and labour skills, so are likely to require the use of a specialist contractor with suitably skilled and experienced
masons. Local repairs, which involve the replacement of a small number of masonry units or damaged parts of units
only, can be achieved either by replacement with new stone or by piecing in to repair the damaged areas only. Stonework
repairs are pinned back to the original fabric of the lining in a similar way to brickwork repairs.
The use of plastic repairs where mortar is used to replace original stonework, or the use of bricks to replace stone should
be avoided wherever possible, because repairs may fail prematurely or damage adjacent masonry fabric.

a b
Note the use of a centering rib to support the repair.

Figure 8.8 Typical patch repair to two courses of brickwork (a) and with pinning detail (b)

8.5.3.3 Crack repairs to masonry

Crack repairs to masonry culverts in embankment dams should only be carried out once the cause of
the crack has been identified and can be dealt with. Particular attention should be paid to ascertaining
the cause of new or widening cracks. Unless this cause of a crack is identified and dealt with, the crack
is likely to reoccur. It is noted that cracks are most likely to develop shortly after construction, or on first
impounding of the dam (see Box 8.3).

104 CIRIA, C743

Box 8.3 Guidance on crack repair (from McKibbins et al, 2009)

Superficial repairs to cracking involve sealing the surface of the crack to prevent the ingress of moisture and deterioration
of the adjacent materials, but do not restore structural connection between the masonry either side of the crack.
Longstanding inactive cracks can be repaired using mortar
1
materials that should not be too hard or brittle, or else

2
small movements are likely to result in a recurrence of the
cracking and failure of the repair. Cracks that are expected
to experience further movement, for example, through cyclic
moisture or thermal variations, can be treated as joints
and sealed with a flexible material that will accommodate
the anticipated range of movement. If a crack is acting as

3
a drainage path then a permanent pipe or other means of
drainage may be incorporated into the repair. The pipe may be
connected to a water collection system if dripping or flowing
water cannot be tolerated.
Where cracks have confined themselves to the mortar joint

4
lines they can be repaired using normal pointing methods.
However, cracks that pass through the masonry units
themselves are more difficult to treat, and patch repairs may
be necessary. Alternately the crack may be stitched using
a stitching bars installed diagonally through the crack at
suitable intervals or, in the case of brick linings, along the

5
brick courses over the length of the crack. Several proprietary
systems are available on the market that can be used.
The advantage of stitching is that it offers an element of
reinforcement to the cracked masonry that repointing does
not achieve. Figure 8.9 Installation of stitching bars along a crack

It is recommended that the ends of any stitching bars are staggered to avoid cracks forming at either end. 6
8.5.4 Concrete conduits
Concrete conduits have been extensively used in modern dams with high standards of construction 7
supervision, so major structural deterioration is rare. However, redistribution of ground stresses caused
by differential settlement can cause cracking, as can thermal effects. Settlement is most apparent at
joints, where steps can form in the wall and invert, with joint material lost and concrete on either side
of the joint dislodged. Concrete is most likely to deteriorate in places where the cover to reinforcement
was insufficient or poor construction led to honeycomb patches or other defects, which may only become
8
apparent after time.

The most common repairs to concrete culverts are patch repairs, crack sealant and joint repair, which
are discussed in this section. Other works that are also applicable to masonry structures are discussed in
9
subsequent sections.

Suitability and compatibility of materials and their influence on the structural behaviour of the culvert
are important considerations (see Box 8.4). Reference should be made to BS EN 1504, the comprehensive 10
BS for products and systems for the protection and repair of concrete structures. This standard includes
10 parts – Parts 9 (BS EN 1504-9:2008) and 10 (BS EN 1504-10:2003) are of particular relevance to those
determining a strategy for and executing repairs on concrete structures.

Box 8.4 Objectives and considerations of concrete conduit repair (from McKibbins et al, 2009)
11
12
In carrying out any repair the objective is to provide adequate protection to the existing lining components to enable the
lining to function as a load bearing structure. Unless load is removed from structural elements before repair, the repair
will only contribute to the resistance of extra loads. The ability of the repair to resist loading will depend on a range of
characteristics including its compressive strength, elastic modulus and bond strength with existing concrete. Differential
shrinkage and creep should also be considered. Suitability and compatibility of materials and their influence on the
structural behaviour of the culvert are important considerations.

For those determining a strategy for and executing repairs on concrete structures, see BS EN 1504-
9:2008 and BS EN 1504-10:2003.
A1
Dams and reservoir conduits 105
8.5.4.1 Patch repairs
Similar to masonry culverts, localised degradation of concrete conduits may lead to the need to carry out
patch repairs which require the removal and replacement of cracked, spalling or otherwise defective areas
of concrete. Removal can be achieved using mechanical breakers or a high pressure water jet and the
boundary of the repair should extend a short distance in sound concrete beyond the extent of the degraded
material. Well defined edges should be achieved by making saw-cuts at the boundary of the repair.

If reinforcement is present within the boundary, existing material should be removed from behind
the reinforcement (to allow the reinforcement to be entirely surrounded by the repair material) and to
provide full bond lengths. The reinforcement itself should be surface cleaned and treated. Consideration
of replacement of reinforcement should be made where it is heavily corroded. Cathodic protection
systems may also be introduced to prevent further rusting following the patch repair.

Any decision to carry out a patch repair needs to be made recognising that local patch repairs will not
protect the area beyond repair, and the incipient anode effect may even accelerate the deterioration of
adjoining areas. As a result, the dam owner should accept the need for future patch repairs, which can
be reduced by combining patch repairs with cathodic protection (see Box 8.5).

Box 8.5 Selection of repair materials (from McKibbins et al, 2009)

There are a wide variety of generic and proprietary concrete repair materials available and selection of appropriate materials
is an important factor in achieving a successful repair. This should be based on a consideration of the function of the repair
(for example, whether structural or non-structural, cosmetic, or to protect reinforcement from chlorides or carbonation), and
any constraints (for example, on the size and shape of the repair, its location, or on the period available for application or
hardening/curing). Compatibility between the physical and chemical properties of the repair material and its substrate is
an important factor, in particular shrinkage/expansion, strength and stiffness and coefficient of thermal expansion. Where
repairs are extensive or performance particularly critical, trials and testing should be considered.
As well as plain cementitious materials (concretes and mortars) a wide range of polymer modified cementitious materials
and resin-based materials are available, which can be used to meet specific repair requirements. A consideration of the
characteristics of these materials and their suitability for differing applications is beyond the scope of this publication,
and specialist advice should be sought when selecting and specifying repair materials.

Further guidance on mortars and concrete is given in BS EN 1504-3:2005.

Materials are placed either by hand or using a flowable mix with formwork. More detail is provided in
Table 8.13.

Table 8.13 Material types used in concrete patch repairs

Method Comments
‰‰ repair mortars are most commonly used for patch repairs with limited size, less than 1 m area and
up to about 30 mm in thickness
‰‰ for thicker repairs, and for use in larger areas, proprietary repair concrete or design mix repair
concrete may be used
‰‰ generic Portland-cement-based materials can be used, but need to be compensated for shrinkage
‰‰ bond strength may be limited (although this may be improved by making mechanical connections
Hand placed with sound concrete or by cutting back beyond reinforcement). Depending on the repair location
mortar or grout and dimensions it may be difficult to place and compact
‰‰ more expensive, proprietary polymer-modified cementitious materials are frequently used
where there are particular requirements for high bond strength, good chemical resistance, low
permeability, rapid setting and curing and limits on the thickness of repair
‰‰ these materials can also be applied to vertical or overhead surfaces without formwork and built up
in thin layers to form a thicker repair, although this can introduce discontinuities that potentially
reduce durability.
‰‰ for large repair areas flowable grout or self-compacting concrete can be used and does not require
Flowable grout or compaction/vibration in situ
concrete ‰‰ materials are carefully poured into formwork to produce the required repair shape and are useful
for rebuilding badly damaged elements.

Further guidance is provided by the Concrete Society (1991).

106 CIRIA, C743

8.5.4.2 Crack sealing and repair
As for masonry culverts, crack repairs to concrete conduits in embankment dams should only be carried
out once the cause of the crack has been identified and can be dealt with. Unless this action is completed, 1
the crack is likely to reoccur. Unlike their masonry equivalents, reinforced concrete culverts have little
ability to accommodate settlement, which can lead to excessive stresses building up within the culvert
structure. Where the culvert has cracked due to continued overstressing then strengthening of the
culvert structure may be required. Where regular periodic inspection shows that cracks are no longer
propagating, these may be sealed to preserve durability and/or prevent water seepage. Where cracking
2
is the result of reinforcement corrosion it may be appropriate to breakout defective concrete and repair

3
using the patch repair techniques discussed earlier (see Box 8.6).

Box 8.6 Sealing and repairs of cracks (from McKibbins et al, 2009)

Crack sealing and structural repairs are generally carried out by injection of a flowable repair material through a series

4
of injection ports inserted along the crack’s length. Typically the injection is carried out under pressure. A wide variety of
materials, including SBR (styrene-butadiene rubber), acrylic and co-polymer emulsions, as well as various types of low-
viscosity resins are potentially suitable for crack sealing and repairs. Resin-injection is a specialist activity and requires
careful selection of materials and experienced specialist contractors for good results.
Selection of a suitable material will depend on a variety of factors, including crack width, the requirement for

5
accommodation of future movement, the need for restoring load transfer, and whether any moisture is present in
the substrate. Where considerable extra movements are anticipated, crack-sealing materials are likely to fail and
consideration should be given to chasing out a groove at the surface of the crack and using a flexible sealant capable of
accommodating the movement.

6
If the crack is a source of water ingress from fill upstream or within the impermeable zone, the crack
should be sealed by injection of a suitable resin-based grout.

Further guidance on the principles and methods for crack injection of concrete structures are included
in BS EN 1504-9:2008, under principles 1 and 4. Specifications for injection products and systems are
included in BS EN 1504-5:2013.
7
8.5.4.3 Joint sealing (where the joint has opened due to settlement
or pulling apart)
Construction joints within reinforced concrete structures (such as large modern dam conduits in the
8
UK) generally have at least partial steel running through the joint, which limits the risk of displacement
at joints. However, differential settlement may cause changes in culvert alignment to form at such joints,
which can then lead to compression and spalling of concrete on either side of the joint and squeezing out
of any flexible joint sealant. Note that such damage could occur on the outside face of the culvert, which 9
is invisible to internal visual inspection. Where this is suspected, appropriate NDT techniques should be
used to check the thickness and condition of the structure.

Any part of the joint in tension can encourage a leak to form along the joint, with an accompanying
build-up of calcareous or ocherous deposits. If the joint opens due to severe differential settlement,
10
material could be washed into the culvert. Where the joint runs along the culvert invert and it is subject

11
to spill or scour flows, the sealant material may be dislodged by the flow.

Spalled material on either side of the joint should be replaced as detailed under patch repair (see Section
8.5.4.1). If further movement is expected, suitable compressible filler can be added along the joint to
accommodate this.

Where precast concrete pipes or box culverts are used in embankment dams, their joints are typically
sufficiently flexible to allow the required amount of flexibility to deal with differential settlement.
12
However, where an embankment spreads as it settles, such joints may be pulled apart. Where this is likely
to happen to the extent that surrounding material may be eroded, replacement of part of the culvert or
conduit is probably the optimum solution. Where this is not possible, lining of the entire culvert should
be considered, or a plate bonded over the problematic joint(s).
A1
Dams and reservoir conduits 107
8.5.5 Invert repairs
The inverts of culverts through embankment dams are an integral part of the culvert structure. Older
masonry culverts have inverts of masonry or mass concrete. The invert is subject to uplift forces and
loads from internal pipework, and is subject to redistribution of ground stresses due to differential
settlement along the culvert. Stresses imparted by these loads may cause cracking of an invert. Many
inverts are also subject to sediment laden, abrasive flows during operation of scour valves, which can
cause erosion, particularly where concentrated by pipe supports or changes in alignment.

Where a culvert invert is extensively cracked or is showing signs of excessive deterioration or distress, either
strengthening or replacement repair works will need to be carried out. Where the invert of a culvert passing
through soft ground has to be replaced it may be necessary to do this work in short sections to ensure the culvert
stability is not adversely affected. Open joints in the invert can lead to flow underneath, necessitating repair.

Invert repair work is likely to be disruptive and may remove the ability to operate the reservoir scour
valve during the works, potentially requiring an alternative draw-down facility to be put in place.

Options available include patch repair, replacement of part of the invert and/or re-lining the existing
invert with a reinforced concrete overslab, providing there is sufficient headroom clearance to
accommodate the new invert. Where scour forces are particularly high, the invert can alternatively be
protected by locally bonding a plastic or metal plate.

8.5.6 Filling voids behind culvert walls

8.5.6.1 Grouting
The most commonly used method of filling behind culvert linings is grouting. However, grouting in
isolation may not provide the whole solution to any particular problem, and may also be attended by
certain risks, as previously discussed. So, it is prudent to consider what the full consequences of grouting
will be and to mitigate any negative potential effects. For example, if increased flows are anticipated in
ungrouted zones, then some form of drainage, such as a filter collar, may be required.

It should also be appreciated that grouting is not always a permanent solution and it may need to be
supplemented, or periodically re-applied (Gibson, 1992).

Void grouting behind linings

In areas where voids have been identified remedial grouting can be considered. The selection of a particular
grouting technique will depend on the size of the voids and the extent of the void network, together with the
properties of the lining, the properties of the ground to be grouted, and the purpose of grouting.

A number of documents give guidance on grouts and grouting methods. Chapter 8 of WRc, 2001 gives
detailed guidance on grout materials, grout selection, admixtures, mechanical properties, workability
and durability, as well as placement techniques, supervision and quality control. Although the application
is slightly different much of this guidance is relevant to work on and in culverts. Rawlings et al (2000)
also provides a range of information and guidance of grouting in the ground, although the specific
application of filling voids behind tunnel or culvert linings is not addressed.

Grouting solutions frequently require a high degree of confidence in the filling of the annulus being
grouted up. Consequently, shrinkage, bleeding and segregation of the grout needs to be minimised.
Grout mixes without special admixtures may exhibit the following shrinkage characteristics:

‰‰ plastic shrinkage during the period up to the initial set

‰‰ bleeding as the solid particles of the grout settle
‰‰ thermal expansion during setting and hardening of the grout due to heat generated by hydration
‰‰ drying shrinkage as water is removed by hydration, evaporation and absorption.

108 CIRIA, C743

So, it is critical to control the water/solids ratio and the cement content of the grout, and admixtures are

1
often useful in this regard. WRc (2001) provides guidance on such admixtures.

It is also necessary to consider the size of the annulus or voids to be filled. McKibbins et al (2009)
provides information on the use of grouts where grouting is to be carried out to prevent water ingress.
Cementitious grouts are particulate and so have limitations on their use in sealing up very fine
apertures. However, these limitations have been overcome to some extent by the development of finer
cement products. Chemical (resin based) grouts have traditionally been used for the finest apertures. 2
Table 8.14 gives an indication of the most appropriate grout type for certain conditions.

3
Table 8.14 Grout types (from McKibbins et al, 2009)

Grout type Particle size of grout Practical joint aperture range

Ordinary Portland cement 80–100 microns > 400 microns

4
High early strength cement 40–60 microns > 200 microns

Microfine cement 10–12 microns > 50 microns

Silicate gels, chemical/resin grouts N/A < 50 microns

Cementitious grouts are frequently used where the voids are of significant size. Cementitious, particulate
grouts can be used with or without sand aggregates, but they have a high dead weight when wet and
their influence on structural stability should be taken into account in both the short- and long-term.
5
However, lightweight aggregates and fillers are available if necessary. Where there are very narrow voids,
the use of micro-fine cement grouts may be better to ensure successful penetration of the cracks, fissures
or voids. For the finest cracks it may be necessary to use chemical grouts. 6
Chemical grouts include sodium silicate, acrylates, polyurethanes, lignins, and resins. Of these,

7
polyurethane resins are the most common for repair of leaking joints and cracks in conduits.
Polyurethane is primarily used as a void-filler and can displace water that causes a load on the structure.
The single-component polyurethanes are very water reactive and can expand up to many times their
original volume. However, this leaves a material that is of low density and strength. The two-component
polyurethanes can be controlled to produce a more modest expansion, with a considerably higher
strength. The other main type of resin used in injection is acrylic-based. These are two-component
materials that can be used in sealing earth structures, dams or embankments, but are more commonly 8
used to inject into cracks or fissures in rock, concrete or masonry. The resin can be made to react within
a few seconds or up to half an hour or more, as required. Acrylic resin has an extremely low viscosity,
comparable to that of water, and so can be injected safely at very low pressures while still being capable of
penetrating the very fine cracks. Both types require specialist pumping equipment. 9
In practice, a combination of grouts and grouting technique may be required. For example, in large
voids the primary grouting may be with cementitious grouts, with or without fillers, followed by
secondary injection of a chemical grout to tighten up the ground and prevent minor seepages. Ideally
trials should be first carried out to determine the most appropriate approach to be adopted. 10
Consideration should be given to the grouting sequencing. It is often appropriate to sequence the

11
grouting so that the grout front progresses in the direction of increasing water pressure to maximise
grout penetration into the voids. Where it safe to do so, the installation of grout holes in advance of the
grouting operations can be advantageous as it allows the progress of grouting to be observed, and the
assessment of the size, location and interconnectivity of voids. However, consideration should also be
given to the location and number of grout holes, and how they are to be filled, to minimise the risk of
them becoming seepage pathways in the future. Care should be taken when first penetrating the lining
to avoid a sudden flush of soil and water under pressure as the voids are tapped. 12
Tube-à-manchette grouting
A technique that can be used for grouting in soil and can be carried out from the crest of a dam. Details
of the method can be found in Rawlings et al (2000). This technique was adopted at Shon Sheffrey Dam in
Wales (see Case study A1.14) and the grouting works are described in Bruggemann and Francis (2014).
A1
Dams and reservoir conduits 109
8.5.6.2 Other fillers
Apart from cementitious grouts and chemical grouts, many branded specialist grout products exist for
specialist applications. For very large voids, other bulk filling materials may be considered.

Concrete
Concrete, which is occasionally used in the filling of large voids behind tunnel linings, may also be used.
It is fluid enough to be pumped and produces a very high strength final product. However, the high
weight of wet concrete may lead to high stresses during placement. Its poor penetration properties and
shrinkage characteristics also mean that the use of concrete will, in most cases, have to be followed by
subsequent grouting to seal the residual fine cracks.

Foamed concrete
Foamed concrete is produced by injecting a foaming agent into concrete to produce a low density
concrete. Foamed concrete has been used to fill relatively small voids, such as the void between a new
insert pipe and the original culvert in sewer re-lining projects, and for filling much larger voids, most
notably recorded for the Heathrow airport tunnel collapse (NCE, 1994).

Foamed concrete may be an alternative technique if larger voids are found behind culvert walls and
require filling. Correctly applied, it can have several advantages, because of its placement characteristics.
Foamed concrete:

‰‰ does not settle, so requires no compaction

‰‰ is lightweight, so does not impose large loadings
‰‰ is free flowing, spreading to fill all except the smallest voids
‰‰ has excellent load spreading characteristics
‰‰ requires no maintenance.

However, careful application by experienced designers and operators needs to be employed to avoid the
mix collapsing or shrinking significantly. Some form of grouting is also most likely to be required after
the application of foam concrete.

Smaller applications are considered feasible, and cases have been recorded of sewer repairs.

Other materials
Large underground voids have also been filled with other bulk filler materials, such as mixtures of water
and rock paste, pulverised fuel ash (PFA) etc. However, the limited volume of most culvert voids makes it
unlikely that such applications would be practicable for culvert void filling.

8.5.6.3 Void filling from the ground surface

For conduits that are not accessible to human-entry, voids around the conduit have to be filled externally.
This is normally achieved by drilling or installing injection points from the surface of the embankment
dam. As the depth to the conduit below the installation point increases, so does the degree of difficulty
of injecting grout at the precise location required along the conduit. The risk of damage to the conduit
from drilling errors or from excessive injection pressures also increases.

Grout is applied to the subject area through the drill hole under pressure. Unless carefully controlled,
this method has the potential not only to cause damage to the conduit, but also to cause hydraulic
fracture within the embankment dam itself. Drilling from the surface of the embankment dam is not
advisable for situations where the reservoir water surface cannot be lowered. If the reservoir cannot be
lowered, grouting along the upstream portion of the conduit is not practicable.

FEMA (2005) provides some case studies are recorded of grout being placed through boreholes using

110 CIRIA, C743

only the pressure from gravity (ie with no pumping). Holes with large grout ‘takes’ should be re-drilled

1
and grouted again to ensure that the voids are filled, possibly with a modified grout mix.

Care should be taken in the design of the grout hole locations to maximise the potential for filling the
voids. In addition, to minimise the potential for grout ingress into the culvert, any repairs required to
the culvert lining, such as the filling of cracks or other defects, should be undertaken before grouting
takes place. Unless conditions require otherwise, grout holes should generally be drilled and grouted
in an appropriate pattern beginning at the downstream end of the culvert and progressing towards the 2
upstream end (ie progressing against gravity).

8.5.7 Dealing with historic repairs 3

Historic repairs are common in old culverts that have been subject to excessive movement, structural
deterioration or leakage, or where the original design was not suitable. For example:

‰‰ partial or full lining of the conduit with concrete, gunite or metal sheets
‰‰ patch repairs of walls or inverts
4
‰‰ void filling with various materials
‰‰ structural elements such as concrete or steel beams and struts.

These repairs may be problematic as they can:

5
‰‰ introduce new load paths within the structure that are inappropriate
‰‰ restrict flow of water along the culvert
‰‰ conceal potential deterioration of the original fabric
6
‰‰ deteriorate quickly, or cause accelerated deterioration of areas near to the culvert

7
‰‰ restrict access into the culvert.

In dealing with historic repairs, the engineer needs to try to ascertain their original function, and
determine the need for action. There is considerable risk associated with interfering with the existing
structure, which may have reached a state of equilibrium under normal loading.

Where inappropriate materials have been used, or the repair was carried out without understanding the
underlying issue, further damage around the repair may be apparent. However, consideration should
8
be given as to whether this damage is historic or ongoing, and whether it has been caused by the nearby
repair or is more likely to be due to changes elsewhere in the dam.

Where structural supports were added and are subsequently deteriorated to a dangerous condition,
9
a full structural analysis of the conduit and surround is recommended rather than simply replacing

10
like for like. Such structures may have been added in response to some movement in the original
conduit, which may indicate a leakage path or overloaded zone outside the conduit, and should be
targeted for improvement.

Historical plate repairs can be particularly problematic, especially where a plate has been fixed to the
conduit wall concealing the structure behind it. The function of the repair may have been to stem a
significant leak into the culvert, and there is a risk that such a leak would recommence if the plate is 11
removed. Alternatively the surrounding fill may have further consolidated since the repair, sealing up
the leakage path leading to the conduit (if one was present).

The primary function of the plate could also have been to support an area of masonry that was prone
to movement due to insufficient thickness or poor pointing. Removing the plate may remove necessary
12
support and lead to further damage or collapse.

A1
Dams and reservoir conduits 111
8.5.8 Changes to structural arrangements
Major changes to dam culverts are typically undertaken for the following reasons:

‰‰ change of control position – where a conduit buried in the embankment fill carries pressurised
flow. Valves can be introduced at the upstream end to improve safety. This often requires an inlet
structure to provide access to the valve, such as a new tower at the upstream toe or shaft through
the upstream shoulder
‰‰ access improvements – parts of the structure may be removed to improve access, typically in
relation to refurbishment of pipes or valves within the culvert
‰‰ structural enhancement – structural members or lining systems may be added to improve the
structural capacity of the conduit
‰‰ installation of new pipework – this may impose loadings on the conduit
‰‰ changing an upstream wet culvert to a dry culvert.

A cautious approach needs to be taken when considering the amendment of an existing culvert.
This is primarily related to the impact of fill movement and settlement on the conduit structure.
An understanding of the existing and new load paths within the structure is also crucial. These are
discussed in Table 8.15.

Table 8.15 Impact of dam movement on structural modifications

Application Comments
‰‰ where a new structure such as a tower or shaft is added, the structure changes the load on
the embankment fill and foundation in its vicinity. This will affect the future movement of the
surrounding fill
‰‰ any additional vertical load will increase settlement of underlying compressible layers, and
New structures also any shaft may restrict ongoing horizontal movement due to ‘spread’ of the settling
embankment, leading to horizontal or rotational movement and additional stresses in
both old and new structures. These would normally be most prominent at the connection
of old and new structures, although could have an effect anywhere on the structure. Such
movement may also stress pipework or bend valve spindles.
‰‰ internal structural elements in shafts or culverts may act as struts, resisting compressive
forces from the surrounding water or fill following some structural movement, even where
Access improvements clearly not designed for this function
(structural element ‰‰ examples include intermediate flooring and part-height walls used to support pipework
removal) ‰‰ removal of such elements for access, and replacement with similar structures that do not
provide equivalent resistance (such as simply supported flooring) could substantially weaken
the structure.
‰‰ most dam culverts are simple structures where all loads are transferred within the walls
without secondary or tertiary load paths provided by beams or columns. In such structures
the wall acts as a panel, redistributing variations in the external load. This redistribution
may lead to overloading of some parts of the culvert, with visible signs of distress such as
cracking or displacement of blocks. An engineer may be tempted to reinforce the distressed
New structural area with supporting beams or struts
members or lining ‰‰ the addition of such structural members may transfer the load to form point loads at
other places in the structure that are not able to distribute the load to the surrounding
fill effectively. Further movement of the conduit may also lead to overloading of the new
structural members
‰‰ new linings can provide support to the conduit fabric without adjusting the load paths.
However, they can adversely affect leakage around the conduit (see Section 8.4.2).

No works should be undertaken until the capacity of the modified structure is analysed considering
possible future movement. This may be a major undertaking in itself, requiring physical investigations
to determine key properties and dimensions of the culvert, and the creation of a complex model to
understand the interaction between the culvert and the surrounding fill. The loads to be resisted include
the loads relating to the design storm event as well as applicable seismic loads.

112 CIRIA, C743

8.5.8.1 Design and construction considerations
Structural designs should consider the safety of operatives at all stages of any proposed demolition and
reconstruction, and proceeding incrementally in multiple stages where access for work is restricted. This 1
may require careful planning and co-ordination. Human-entry culverts within embankment dams are
generally confined spaces, and access is often constrained by large pipes within the culvert. Also, due to
operational requirements, the available time for repairs or renovations may be very short. Due to these
factors, a relatively high proportion of available time may be taken up in activities such as: 2
‰‰ safety procedures

3
‰‰ gaining access to and from the work site
‰‰ transporting plant and materials
‰‰ setting up and dismantling temporary access equipment
‰‰ making safe any incomplete works at the end of each session.

In such situations the time required to complete the work, and its cost, may be significantly increased
compared to similar work at other sites to which the owner may be accustomed.
4
8.5.9 Culvert lining
Methods and materials are available to line entire human-entry culverts that are deemed to be
5
structurally inadequate. These can provide structural support either independently of the existing

6
culvert material, or to work as a composite section with the existing material by sufficient bonding
between the old and new materials. The design process for either type of solution is detailed in WRc
(2001). This refers to structural solutions (which require bonding between old and new materials to
transfer shear forces) as Type I, and non-structural solutions (which do not require such bonds) as Type II.

Culverts, which have deteriorated to the extent that lining is considered ineffective, will often have
developed leaks, transferring water from the surrounding fill. As discussed in Section 8.5.2, the 7
treatment of such leakage paths should be carefully considered where a liner system is to be introduced.

8
The most common lining types and materials suitable for human-entry culverts are discussed in
subsequent sections.

8.5.9.1 Sprayed concrete

Sprayed concrete is frequently used in the tunnelling industry where it is necessary to replace large
areas of defective concrete or to thicken and strengthen the existing tunnel linings. It has the advantage
9
that it can be applied rapidly over large areas in a single operation. The concrete can either be dry mix

10
or, for reduced rebound, wet mix and can be modified (eg by the incorporation of polymers) to meet
performance requirements. It is generally considered good practice to include a light steel fabric within
the repair, or alternatively a fibre reinforced concrete mix may be applied. Skilled and experienced
operatives are required and it is not feasible to produce very thin (less than 100 mm) layers or to treat
small areas. Practical restrictions apply in confined spaces.

8.5.9.2 Reinforced concrete

11
Reinforced concrete linings involve the fixing of steel, placement of shutters and pouring of concrete
within the culvert. Given the restrictive space available, and the significant reduction in culvert size
following construction, this type of lining is unlikely to be used for UK embankment dams except in
lining unlined rock tunnels (eg spillway tunnels) within natural ground under or around the dam.
12
A1
Dams and reservoir conduits 113
8.5.9.3 Plate-bonding (steel, carbon fibre or GRP)
Plate-bonding techniques, using either steel or fibre-reinforced composite materials, can be used to
strengthen structural elements against bending or shear forces. They may be used in combination
with conventional concrete repairs to restore capacity to damaged culverts, or improve the capacity to
resist future loading. These techniques rely on bond strength with the original concrete or repairs, so
careful inspection of the substrate concrete to identify weakened or delaminating areas is vital. Careful
preparation of the substrate surface should be carried out before applying adhesives and bonding.
Further guidance is given in by the Concrete Society (2004). The principles and methods for plate-
bonding of concrete structures are included in BS EN 1504-9:2008, and specifications for products and
systems are included in BS EN 1504-4:2004. In culverts plate-bonding techniques have the potential
advantage that they can increase strength and resistance to certain forces without significantly reducing
clearances within the culvert.

114 CIRIA, C743

9 Techniques for replacement
of conduits
1
2
On some occasions, replacement of all or part of a conduit may prove the optimum solution for dealing
with the risks posed by it. The replacement conduit needs to be designed so that it has the required
hydraulic capacity, resists blocking, is buildable and structurally sufficient and poses no significant threat
to reservoir safety.
3
4
In comparison to the rehabilitation of an existing conduit, there are many more options available with
replacement. For example, the location and capacity of a new pipe is not usually constrained by the
existing pipework, and the new pipe can have a different control arrangement. So the designer of a
replacement conduit needs to follow a process to identify the optimum solution. An example of such a

5
process is shown in Figure 9.1.

Determine location options (see Section 9.1.1)

‰‰ if function not required, consider abandonment

6
‰‰ avoid excavation in dam if possible
‰‰ consider high level pipe, possibly pumps or siphon(s).

Determine hydraulic requirements (see Section 9.1.3)

7
‰‰ determine flow capacity required
‰‰ determine design of control systems
‰‰ consider energy dissipation requirements at the outfall.

8
Determine operational concerns (see Section 9.1.4)
‰‰ need for screening/variable draw-off levels at inlet
‰‰ zebra mussels or other fouling concerns.

9
Determine isolation options (see Section 9.1.5)
‰‰ upstream cofferdam
‰‰ upstream valve closure (part replacement of pipe)
‰‰ hot tapping/line stopping (part replacement of pipe).

Determine structural requirements

10
Buried pipes (see Section 9.2) Pipes in culverts (see Section 9.3)

11
‰‰ earth and water loadings ‰‰ construction access
‰‰ deflection requirements. ‰‰ pipe connections and restraint.

Detailing

12
‰‰ constructability
‰‰ cost
‰‰ programme.

Figure 9.1 Design process and main considerations for conduit replacement

A1
Dams and reservoir conduits 115
Very few buried conduits at UK embankment dams are replaced on the same alignment, due to the high
construction costs and risk of damaging the impermeable zone of the dam. However, such replacement
is generally limited to shallow buried pipes, typically at small or low risk reservoirs. Section 9.1 provides
guidance on the replacement design common to all conduits, while those relating to buried pipes only
are given in Section 9.2.

The replacement of pipes in culverts is generally made difficult through restricted access. Connections
to existing pipework and restraint of the new pipe are critical aspects of the design. These issues are
discussed in Section 9.3.

Human-entry culverts are seldom replaced during the operational life of an embankment dam in the
UK. This is due to the high costs and risks associated with such work compared to repair or renovation.
In addition, most human-entry culverts in UK embankment dams are constructed of masonry or
reinforced concrete, which are inert materials less prone to severe deterioration than pipework.

It is unlikely that the balance of replacement and renovation costs will change, and so the replacement
of many such conduits is not expected in future. This guide does not cover the replacement of
human-entry conduits. However, many aspects discussed in Section 9.2 would be relevant to anyone
considering such works.

There are significant consequences associated with the failure of a conduit through an embankment
dam, so the designer of the replacement conduit needs to consider the design of every detail, many
of which may have little consequence in pipeline design for other sites. In all cases, if the new conduit
is through or over the crest of an embankment dam, the designer should have experience with
embankment dams. Where works are undertaken at reservoirs subject to the Reservoirs Act 1975 (in
the UK), and those works have an effect on the safety of the reservoir, the design and construction of
the work should be completed under the supervision of a QCE. For more details on reservoir safety
legislation in the UK, including the definition of a statutory reservoir, see Section 1.4.2. Where siphons
or pumps are involved then specialist experience of this type of work is also required.

9.1 REPLACEMENT CONDUIT DESIGN (GENERAL)

9.1.1 Determining location requirements

Is the conduit required at all?
Some conduits were constructed to serve a purpose that is no longer required, or has little value. At a
site with more than one conduit, others may have sufficient capacity to draw the reservoir down, even
in an emergency. Alternatively it may be possible to rely on the supply of temporary pumps or siphons
to provide this capacity. If the conduit function is not required, the existing conduit can be removed or
abandoned (see Chapter 10).

Can construction through the dam be avoided?

Construction of a new pressurised pipe or tunnel, which is remote from an embankment dam, carries
much reduced consequences as a result of failure compared to the equivalent through the dam’s
impermeable zone. In addition, the pipe or tunnel is less (or not at all) susceptible to deformation during
the remaining lifetime of the dam. Due to these reasons, it is strongly recommended that where such
alternatives are available, the new conduit should not be located within the dam fill.

Can a high-level conduit be used?

Where the new conduit is to be buried through the embankment, it may be possible to replicate the
required functionality of the existing conduit at a higher level. This will minimise the footprint of
the excavation and disturbance of the impermeable zone, avoid disturbance of the deeper areas of
the core (which are subject to higher hydraulic pressures during operation) and save excavation

116 CIRIA, C743

costs. Siphons or pumps may be suitable at the site and allow significant excavation of the dam to be

1
avoided (see Section 9.1.2).

Note that spillway conduits can be replaced by the construction of a surface spillway chute elsewhere at
the site. This is not covered within this guide.

9.1.1.1 Interaction with membrane liner materials 2

Reservoirs have been constructed in the UK in recent decades that are lined with an impermeable
membrane, typically welded plastic sheeting (PE or PVC) or geosynthetic clay liners (GCL). These
are normally fixed to the upstream face. These reservoirs are often small farm reservoirs and do not
impound a watercourse, and instead have pumped inflows. 3
Where at all economically possible, conduits should not be placed through a plastic membrane. It is

4
difficult to form an adequately durable seal between the membrane and the conduit, primarily because
the membrane will move in response to changing water levels that may rupture or damage it. The
conduit will not move with the membrane and so is likely to stress it to the point of damage, particularly
if a concrete inlet structure is fixed to the pipe where it passes through the membrane. However,

5
specialist contractors can provide effective seals for conduits that need to penetrate the lining.

Normally conduits can be avoided at such reservoirs through using pumped inflow and outflow
arrangements. However, spillway pipes passing through the membrane are often the best way to
discharge excess inflows safely. Some lined reservoirs feature a dual-purpose outlet and spillway pipe,
which minimises the number of penetrations through the liner.

Where it is necessary to pass a pipe through a membrane, the pipe should be at as high a level as possible
6
where passing through the membrane. An economic and proven sealing system is to band the membrane

7
to the outside of the pipe using tightened rings to form an adequate seal. Enough slack should remain
in the membrane near to the pipe to accommodate some movement. Note that where the pipe inlet is
close to the reservoir perimeter (particularly where an up-stand on the end of the pipe forms a spillway
bellmouth), ice forming between the pipe and bank can force the pipe to move. The inlet should be a

8
sufficient distance into the reservoir to prevent this.

9.1.1.2 Replacement on the same alignment

Where bypassing the dam with a new conduit is not possible, like-for-like replacement is an obvious
option. The full draw-down range of the original conduit can be replicated, so the same trench or culvert
is used during construction, and the flow can remain gravity fed, and operational simplicity can be
9
maintained. However, this solution retains a conduit in the embankment fill, which may deteriorate over
time to form a major hazard again in future.

The main costs and risks for replacing conduits on the same alignment are noted in Table 9.1. 10
Table 9.1 Conduit replacement on same alignment – summary

Cost or risk Project element

Mobilisation and labour/

Comments

‰‰ potentially very large water control and earthworks costs

11
plant costs ‰‰ most likely effective for conduits less than 3 m below dam crest
Construction cost and where draw-down is easily achievable.

Typical material costs

Damage to existing
‰‰

‰‰
relatively cheap – many options available.
significant risk of damaging core, connecting to retained inlet or
12
structure outlet structures.
Construction risk ‰‰ flood overtopping when trench excavated through dam can
Other construction risks
‰‰
cause breach
overloading during construction can damage new conduit. A1
Dams and reservoir conduits 117
 Cost or risk Project element Comments

Capacity reduction ‰‰ N/A (should not be a problem).

‰‰ high level of surveillance recommended immediately after

works that disturb the core. Long-term surveillance could
Operational cost Surveillance/maintenance/ probably be reduced
running costs
‰‰ works can improve access to valves
‰‰ running costs same or reduced.

Good installation ‰‰ risk reduced by improved structural integrity and control.

Operational risk ‰‰ leakage along conduit, conduit pulled apart at joints or

Poor installation leaks due to installation damage – requires at least partial
replacement.

Note
See Section 8.1 for guidance on applicability and colour coding.

The construction risks for standard conduit replacement are generally high, though dependent on the
dam catchment hydrology, original construction details, and depth of excavation. In retaining a conduit
through the dam fill or foundation, a potential hazard is retained as an operational risk. Costs associated
with rectifying damage during its operational life may be high.

9.1.2 Pumps and siphons

Any new pipe through an embankment will become a hazard to dam safety, although with appropriate
detailing and construction quality the level of risk may be acceptable. However, this risk can be further
reduced or avoided altogether by not placing the conduit through the impervious water retaining
element of the dam such as the core. This is possible by replacing the pipe function by the use of pumped
over-pipes or siphons. These solutions can also prove significantly cheaper than the construction of a
new low level pipe through the embankment, but can pose a number of new risks.

9.1.2.1 Pumps
Pumped abstractions can offer optimum solutions at some sites where replacement on the same
alignment is uneconomic. Pumping can avoid deep excavations into the embankment fill, which reduces
the risk of damage to the core and water control/isolation costs. The existing conduit can be abandoned
and the pump line or siphon constructed at the dam crest or elsewhere on the reservoir rim.

The main costs and risks for replacing conduits with pumped flow are noted in Table 9.2. The risks will
vary depending on whether the pumping plant is temporary (mobile) or permanent.

Table 9.2 Conduit replacement with pumps – summary

Cost or risk Project element Comments

Mobilisation and labour/plant costs ‰‰ low for small – medium pump installation.
Construction cost
Typical material costs ‰‰ depends on pump specification and capacity.

Damage to existing structure ‰‰ interaction with surface works only.

Construction risk ‰‰ construction programme dependent on power supply
Other construction risks
connection.
Capacity reduction ‰‰ N/A (should not be a problem).

Operational cost ‰‰ high running (power) cost for operation – impact

Surveillance/maintenance/running depends on planned operation regime
costs
‰‰ maintenance costs could also be high.

118 CIRIA, C743

 Cost or risk Project element Comments

Operational risk
Good installation
‰‰

‰‰
risk reduced by removing operating/empty conduit
from dam
pumps dependent on power connection.
1
‰‰ maintenance costs could be high to achieve necessary
Poor installation

2
reliability.

Note
See Section 8.1 for guidance on applicability and colour coding.

Pumps are commonly used for operational abstractions and emergency reservoir draw-downs. The
biggest disadvantage of using pumps for operational abstractions is the running costs (power and
maintenance), although installation costs associated with obtaining a local power source and protecting
3
the pumps from adverse weather conditions can also be considerable in remote locations. However, they

4
are commonly relied upon for emergency draw-down of small reservoirs in the UK, given their flexibility
and deploy-ability at short notice.

Use of temporary pumps to draw down large reservoirs in an emergency is constrained by the availability

5
of large capacity pumps at short notice, the mobilisation time and the space required for setting up
around the dam area.

The application of pumps for short-term or emergency abstractions can be assisted by careful planning
and some preparatory works, particularly relating to inlet and outlet positions. Outlets should be
positioned not only to avoid erosion of unprotected slopes, but also to ensure adequate protection of any
downstream watercourse from silt pollution. 6
Areas for the deployment of pumps for emergency draw-down purposes should be carefully located to

7
ensure that suitable access to critical areas of the dam remains viable during adverse weather and flood
conditions. Often outlets are laid in plastic sheeting to protect the embankment should the pipework leak.

9.1.2.2 Siphons to provide draw-down capacity

Siphons can be used to provide operational or emergency draw-down capacity. These are less commonly
used at large reservoirs in the UK, although they are common at small farm reservoirs where small pipe
8
diameters are used. There may be permanent installations or arrangements for emergency mobilisation
in place.

Siphons are conduits through which water flows at sub-atmospheric pressure. This can occur where parts 9
of a pipeline are above the reservoir level at the inlet. Once started, the flow is dictated only by the water
pressure at the upstream and downstream ends and the balance of frictional forces, just as any other
pipeline. So, most of the functionality of a typical pressurised outlet pipe can be provided by a pipe that
does not pass through the dam core at depth, reducing installation costs and normally reducing long-
term risk to dam safety.
10
11
Siphons can provide an optimum solution for the provision of draw-down when implemented correctly.
An example of this is the siphons constructed at Birkenburn Reservoir (see Case study A1.1). However,
where siphons are considered, additional requirements apply and if these are not acknowledged then the
siphon may not function as desired. These are discussed further in Section 9.1.2.3.

The main costs and risks for replacing conduits with siphons are noted in Table 9.3.
12
A1
Dams and reservoir conduits 119
Table 9.3 Conduit replacement with siphons – summary

Cost or risk Project element Comments

‰‰ moderate design and construction costs
Mobilisation and labour/ ‰‰ smaller draw-down needed for siphon installation than like-for-
plant costs like replacement, although large draw-down possibly required
Construction cost
for abandoning existing conduit.
‰‰ largely the same as like-for-like replacement, though possibly
Typical material costs
more/higher spec fittings.
Damage to existing ‰‰ may interact with the top of the dam core, but generally risks
structure are low.
Construction risk ‰‰ may involve excavation of a shallow trench through the dam crest,
Other construction risks with moderate risk of inundation and possible breach formation
‰‰ requires careful installation, particularly pipe jointing.

‰‰ should not be a problem – new siphon can be designed for any

Capacity reduction
required capacity, however, range of draw-down is limited.
Operational cost
Surveillance/maintenance/ ‰‰ operational cost of priming (if appropriate) and costs
running costs associated with vandalism protection.
Good installation ‰‰ risk reduced by removing operating/empty conduit from dam.
‰‰ incorrectly designed siphons will not work
‰‰ siphons are prone to damage by ice
Operational risk ‰‰ reliant on some level of operator skill
Poor installation
‰‰ require reliable means of priming if used in emergency context
‰‰ may produce undesirable vibrations that would be transmitted
to the embankment.

Note
See Section 8.1 for guidance on applicability and colour coding.

Good quality control is required during siphon construction to ensure it is operable. Careful detailing is
required, and some operational costs associated with priming may be necessary. The main restriction relating
to siphon use is the hydraulic limitation in terms of maximum sub-atmospheric pressure in the pipe.

Siphon spillways use siphonic action to greatly increase flow over a weir block above that normally
determined by a given reservoir head. These structures are substantially different to siphon pipes for
abstraction or draw-down. This section of the guide does not relate to siphon spillways.

9.1.2.3 Siphon constraints

The following hydraulic requirements govern the design and performance of siphon outlets.

Maximum sub-atmospheric pressure

The maximum sub-atmospheric pressure (the height of the siphon pipework above the hydraulic grade
line) that can be withstood by the flow (and pipe) is limited. Under ideal conditions at sea level, this
maximum pressure is around 9.5 m. However, a lower value needs to be used for design, which takes
account of the effects of altitude (atmospheric pressure is reduced at higher altitudes), frictional losses
and a factor of safety (FoS). The position of the maximum sub-atmospheric pressure in a typical draw-
down siphon is at the downstream side of the crest (see Figure 9.2).

If the maximum sub-atmospheric pressure is exceeded, cavitation pressures may be reached that will
damage the siphon pipe. In the USA, FEMA (2005) advises that the maximum drop in level from crest to
outlet for a siphon should be limited to around 20 ft. (6 m).

120 CIRIA, C743

 1
2
Figure 9.2 Sub-atmospheric pressure in siphon operation (courtesy Mott MacDonald Ltd) 3
Minimum velocity
To prevent the accumulation of air within the siphon a minimum velocity and of flow needs to be
maintained when the siphon is operated, particularly for extended periods of operation. The minimum
velocity depends on the pipe diameter. As the reservoir level drops, the head difference driving the flow
4
through the pipe reduces, reducing the flow velocity. Once a critical depth is reached the siphon action
will break.
5
Balance of frictional losses
The frictional losses on the inlet pipe need to be less than those of the descending (downstream) pipe.
This is normally achieved by ensuring the downstream leg is longer than the upstream leg.
6
Submergence
The inlet of the siphon needs to be at such a depth to prevent the intake of air (which would break the
siphon action). This can cause a problem at reservoirs with flatter upstream shoulders. Submergence
can be assisted by constructing a hood over the inlet and/or using a bellmouth to reduce local velocities 7
at the inlet.

8
These requirements mean that a siphon cannot always replicate the full functionality of a low level outlet
pipe at a dam. In particular, the height of draw-down, which can be achieved by any siphon, is limited to
around 5 m to 6 m. In addition, a single siphon pipe cannot sustain as wide a range of flows as a low level
pipe. Using two or three siphons with different diameters and lengths may help overcome this constraint,
but such an arrangement has further O&M implications.

Siphon operation 9
For a siphon to operate, it needs to be primed (ie filled with water with all air evacuated). If the upstream
reservoir level is above the pipe crown at its highest point, this will occur automatically. However, this
may mean the pipes pass through the embankment core. If the crown of the pipe lies above the reservoir
water level when operation is required, the pipe can be primed by filling with water from a vent point
at the crest. For this to work, a valve is required at the upstream end of the pipe, introducing additional
10
cost and maintenance issues. Alternatively, priming can be achieved by sucking air out with a vacuum
pump.

Once operating, the flow through a siphon can be throttled by partially closing a valve at the 11
downstream end. This restricts flow through the pipe, but can allow for greater draw-down depths to be
achieved, as long as the minimum velocity is not reached.

Design considerations – pipe material and joints

The siphon pipe needs to be able to resist the sub-atmospheric pressures without collapse, and air-tight
12
joints need to be achieved during construction and throughout the operational life of the siphon. Pipe
materials that can adequately resist negative pressures up to -1 bar (-10 m head) include DI, GRP and
HDPE. Other factors for consideration are typical of any pipeline project and include good hydraulic
performance, ease of handling, readily available fittings and corrosion resistance.
A1
Dams and reservoir conduits 121
Figure 9.3 Siphon arrangement for reservoir lowering (courtesy Mott MacDonald Ltd)

Location
The siphon (see Figure 9.3) should be located to optimise hydraulic performance while not posing
unacceptable risks to dam safety. This often means it will discharge away from the existing downstream
watercourse. Provision needs to be given to transfer the discharged flow back to the watercourse without
damaging the dam mitre. The inlet should be sited not only to avoid air intake, but also silt. The siphons
should be located for good construction and operations access. A siphon over the surface of the crest
will cause considerable disruption to access along the crest. For this reason permanent siphons are often
formed in shallow excavations in the crest to allow crest access to be maintained.

Freezing
If water in a siphon pipe freezes, the siphon is rendered inoperable and may be damaged. This can
be somewhat mitigated by burying the pipe, draining in the winter (where a crest valve is provided)
or maintaining a trickle flow to prevent freezing. When abstraction is not required during freezing
conditions, removal of the inlet pipe from the reservoir for these periods is also an option. Burying the
pipe and setting the valves in chambers also reduces the risk of vandalism.

9.1.3 Determining hydraulic requirements

Once the location and type of the new conduit has been determined, the hydraulic design needs to be
undertaken. The hydraulic design of standard (gravity) conduits is discussed in this section.

When constructing a pipe through an embankment dam, the designer should ensure that the control
and capacity of the new system is optimised to suit operational requirements and minimise the safety
risks relating to operating personnel and dam integrity. Appropriate flow capacity is currently being
considered as part of an Environment Agency research project (Environment Agency, in press).

9.1.3.1 Control position

The choice of control position is critical in establishing the level of risk posed by the conduit during its
operational life. The control point determines the typical water pressure in the pipe at any given point,
and so the likely consequences of any leak in the pipe wall.

Control positions are discussed in detail in Section 2.3.1. As noted, control can be provided at upstream,
central or downstream locations. For spillways or open conduits at flood storage reservoirs control is
always upstream unless discharge is constrained downstream.

As discussed in Chapter 3, failure modes are present irrespective of the position of the control. However,
without adequate measures to mitigate leakage downstream of the impermeable zone, downstream
control on a pressurised conduit is not recommended.

122 CIRIA, C743

9.1.3.2 Flow capacity
The optimum flow capacity depends on the purpose of the conduit.

For conduits used for abstraction, the flow required is determined by the use, and should consider
1
future growth. The conduit may also be required for emergency draw-down. UK reservoir owners with
a large portfolio of dams generally require a draw-down rate of 0.3 m to 1.0 m per day, which could be
substantially different to that previously provided at the dam. Charles et al (2014) provides a summary of
draw-down rates that have been adopted in the past to successfully avert failure during internal erosion
2
emergency incidents in UK embankment dams. Calculations of draw-down rates may take account of

3
other pipework at the dam and typical inflows to the reservoir.

For spillway conduits at impounding dams, the designer should check the outflow based on a flood study
that uses the most recent guidance on rainfall depths, catchment characteristics and analysis methods,
which is appropriate for that category of reservoir. In the UK, ICE (1996) should be referred to when
determining spillway capacity.
4
In all instances, the design should consider the risk of debris or silt blocking the pipe or upstream screen,
or capacity being reduced by downstream conditions.

9.1.4 Determining operational requirements

5
These should be undertaken at the same time as deciding on hydraulic requirements given their interaction.

Reservoir water contains sediment, organic matter and floating debris, all of which can cause operational
problems at conduits. Vandalism is also a concern at many sites. There may also be a requirement to
6
draw water at various depths within the reservoir for quality purposes.

9.1.4.1 Sedimentation 7
Sedimentation rates in UK reservoirs vary, depending largely on catchment characteristics, particularly

8
land use. The larger sediment particles are deposited where the watercourse enters the reservoir, with
gradually smaller particles depositing with distance from the watercourse entrance. Often the sediment
at the dam toe is largely silt deposits, although this depends on the catchment and the reservoir shape.
If sediment builds up over the inlet of a pipe to a sufficient depth, it can act as a plug preventing
abstraction through the pipe. To avoid this, the pipe needs to be regularly operated to flush sediment
around the inlet before it builds up excessively. This operation may be restricted by environmental
concerns (see Section 1.4.3). 9
9.1.4.2 Zebra mussels
In the last two decades, an invasive
species called the zebra mussel (Dreissena
10
Polymorpha) has been found in watercourses

11
and reservoirs throughout the UK (see
Figure 9.4). This species originated in the
rivers of southern Russia.

12
The zebra mussel has had a major impact
on UK aquatic ecosystems by establishing
huge colonies that have the capacity to
filter huge volumes of water in the process
of finding food. This removes nutrients
that would otherwise support other
organisms, and makes the water clearer
Figure 9.4 Zebra mussel (courtesy Paul Beckwith, BWW)
A1
and more susceptible to weed growth.

Dams and reservoir conduits 123

The mussel increases drag on the insides of pipes, and also causes significant operational problems
at intakes from lakes and reservoirs, by blocking screens and conduits and fouling hydro-mechanical
equipment. If the mussels are carried into water treatment facilities they can also severely affect some
treatment processes. The mussels can adhere to most surfaces at high densities and can require large
forces to dislodge them.

A closely related species is the quagga mussel (Dreissena Bugensis). This has invaded the USA and the
Netherlands, and is expected to reach the UK in future years. This acts in the same way as zebra mussels
and causes very similar problems, although it may be more invasive and can survive in conditions where
the zebra mussel cannot. It generally prefers deeper water depths (more than 20 m).

Zebra mussels are expected to become more widespread over time due to climate change, which will
lead to more problems. Their optimum conditions for growth are in slow moving waters (less than 1 m/s)
with high calcium content (more than 28 mg/l) and temperatures in the range of 17oC to 24oC. They
cannot usually form where velocities are higher than 1.5 m/s, water temperature does not exceed 8oC, or
calcium content in the water is less than 11 mg/l (de Lafontaine and Costan, 2000). Defra (2011) provides
guidance on the identification of these and other major invasive species risks to UK watercourses. If the
species are spotted at a reservoir site then they should be reported to Defra.

Given the extent of problems caused by these mussels, it would be prudent to consider measures to
mitigate their impact on works that are to be undertaken at reservoir conduits where mussels are
currently problematic or may be in future.

Such mitigation typically involves:

‰‰ Specifying inlet screens (where required for operational reasons) that are proven to avoid choking
from mussel growth, typically by construction from biocidal materials such as a copper-nickel alloy.
These can be provided as a standard option by some screen manufacturers and DWI approved
products are available.
‰‰ Regular cleaning of raw water pipelines and aqueducts using ‘pigging’ or other systems (see Section 4.6)
to clear off adherent mussels, which can then be removed by flushing or settlement at the pipe outlet.
Multiple parallel mains can be installed to allow regular cleaning without interruption of supply.
‰‰ Killing mussels within the abstracted flow by dosing at the inlet with chlorine (not for potable
applications) or other biocides suitable for potable applications such as the Biobullet system. This is
a DWI approved material, which is ingested by adult zebra mussels and poisons them. Note it may
have impacts on some downstream treatment systems, which need adjusting during dosing periods.
Such dosing is typically used in conjunction with a facility to remove the (dead) mussels from the
flow before entering treatment or process plant that might be damaged by them. More recently
designed water treatment works often include a settlement tank at the inlet to settle out the mussel
carcasses (following dosing).

Two other technologies, which have been installed outside the UK, include filtering systems and anti-
fouling coatings:

1 Filter technology has been used in Europe and the USA to remove zebra mussel larvae from intake
pipes for industrial cooling plants. These are based on the premise that shear forces during very
turbulent flow can kill zebra mussel larvae. A filter medium is set within a chamber where water is
forced into turbulent flow, killing the larvae and allowing their removal.
2 Anti-fouling coatings have been used in the USA, and a wide range of coating types and products
are available. Some of these may become available in the UK in future years. Recently the
USBR has undertaken research into the effectiveness of different coatings that could be used at
reservoir sites. This focused on reservoirs in the south-western region of the USA, where warmer
temperatures have led to particularly acute problems relating to zebra and quagga mussel growth.
This research has indicated that:
a mussels adhere quickly and effectively to steel, stainless steel and standard epoxy coatings at
broadly similar rates, and the force required for removal of the mussels is largely the same

124 CIRIA, C743

 b mussels attached to galvanising coatings as well. This finding was different to that of earlier

1
research by the United States Army Corps of Engineers (USACE), which found that galvanising
was somewhat effective in resisting mussel growth. The authors considered it likely that the
chemical composition of the water may have influenced the effectiveness of the galvanising as
the research was undertaken in different locations

2
c polyurea and polyurethane/polyurea hybrid coatings were also colonised though to a lesser
extent, and mussels were considerably easier to clean off than standard epoxy coatings or steel
d some anti-fouling coatings were tested. These are supposed to work by poisoning the mussels
as they attempt to adhere to the surface. However, these are generally toxic and so are

3
generally not recommended for use in reservoir environments
e silicone-based foul-release coatings performed well in avoiding adherence and proving very
easy to clean. These were likely to be the most suitable method of mitigating the risk from
mussels at most sites. Historically, they have been developed to avoid barnacle and mussel

4
growth on ship hulls, and some are marketed for use in hydropower and water intakes within
the USA. However, no such coatings are currently DWI approved in the UK.

9.1.4.3 Debris
Floating debris can clog conduits and reduce capacity. This can become a dam safety concern if it affects
spillway capacity or emergency draw-down, but in all instances it causes operational problems and
5
rectifying activities can be costly and with high risk.

Where debris is a problem the conduit inlet will probably require screening. Appropriate operational
access for screen clearance and to the end of the pipe needs to be provided. Special screens are available
to avoid blockage by zebra mussels (manufactured with copper alloys), or allow access to the pipe for
6
cleaning equipment.

9.1.4.4 Vandalism, security, public safety 7

Conduits are prone to damage or inappropriate use by vandals. They are also a point of vulnerability

8
within the water supply infrastructure, and so are of particular concern to security professionals.

At all conduits through embankment dams, steps should be taken to avoid unauthorised operation and
potential vandalism. Screens with a clear bar spacing of 100 mm or less are generally considered to be
child-proof. Valve headstocks should be locked off or placed in secure chambers/housing, and detailing
of exposed structural elements should be robust. Cables that are required in the operation of lighting or
valves should be protected against damage and theft. 9
Many reservoirs are used for public recreation. Inlets to conduits can form a significant hazard to
members of the public in or on the water. This may require appropriate signage and barriers to prevent
access to the inlet. 10
9.1.5 Determining isolation options and structural
requirements 11
The means available to isolate the conduit from the reservoir to allow replacement is the same as
those available for conduit remediation (see Section 8.2). The only difference relates to the scale of the
temporary works, which would be greater for a deep conduit replacement due to the larger excavation
footprint, and the greater likelihood of dam failure should any cofferdam be overtopped. In the past, 12
steel sheet pile cofferdams for replacement conduits have been split by bulkheads to manage the risks
associated with overtopping.

The structural design considerations for buried pipes and those in culverts are substantially different so
are discussed separately in Sections 9.2 and 9.3.
A1
Dams and reservoir conduits 125
9.2 REPLACEMENT OF PIPES BURIED IN
EMBANKMENT FILL
This section focuses on the replacement of pipes buried in the embankment fill only. It does not cover
the replacement of larger human-entry conduits as such work would be very rare in the UK and require
bespoke design. It also does not directly relate to conduits in dam foundations or abutments, where the
consequence of failure is lower and so more design options are available. However, much of the content is
applicable to both larger conduits in fill and conduits in abutments.

The replacement of a buried dam conduit requires water level control in the reservoir and significant
earthworks. The costs of these activities rise with pipe depth, and generally preclude the direct
replacement of conduits if they are deeply buried in the embankment fill. The excavation also directly
affects the impervious zone of the dam, risking damage, with further long-term maintenance or
repair implications. As such works affect the safety of the dam their design and construction should be
supervised by a QCE where reservoir safety legislation applies.

For these reasons, replacement of conduits in the UK generally only occurs at small dams where the
conduit in question is shallow, access is reasonable, water control is easily achievable and the conduit,
typically a buried pipe, is small and can be quickly replaced.

As previously noted, a pipe buried in embankment fill can be replaced with a new pipe along the same
alignment, or in a different location altogether, possibly using pumps or siphons. This section primarily
considers the replacement of a pipe at depth to contain water flowing under pressure through the
embankment. The use of pumps and siphons is discussed in Chapter 8.

9.2.1 Typical structural details

Once the location, arrangement, control points and hydraulic capacity of the new conduit have been
identified, the conduit designer needs to complete the detailed structural design of the conduit. In
particular, this includes:

‰‰ identifying optimum pipe material, design of wall thickness, specification of jointing

‰‰ selecting conduit surround or support and trench shape to improve compaction
‰‰ designing any anti-seepage details
‰‰ providing for filter collars or diaphragms, as necessary.

These details are discussed further in the following sub-sections.

9.2.1.1 Pipe material options

New dam conduits can be constructed from a wide range of materials, including cast in situ concrete,
precast concrete, DI, GRP, PE (normally HDPE although other densities are available) and steel. When
correctly specified and designed, all of these materials have the requisite strength to resist loading
typically encountered by conduits through embankment dams.

The structural design of buried pipes in the UK should be carried out in accordance with BS 9295:2010.
Within this Standard, pipe materials are classified by their inherent response to external loads, and their
interaction with the surrounding fill. The three categories used are rigid, semi-rigid, and flexible (see
Box 9.1) and Table 9.4 provides further classifications of the common pipe types.

126 CIRIA, C743

Table 9.4 Structural classification of common Box 9.1 Pipe material classifications (from BS 9295:2010)
pipe types (from BS 9295:2010)

Pipe material Classification Rigid pipe materials have a small deflection on loading, which is
too small to develop any lateral earth pressures. Load is taken by
1
Clay Rigid the pipeline and bending moments are developed in the pipe walls.
Rigid pipes also attract an amplified backfill load upon burial and

2
Concrete Rigid obtain a reaction from their bedding in response. Semi-rigid pipe
materials tend to exhibit a range of behaviour from rigid to flexible.
Reinforced concrete Rigid
Flexible pipe materials deflect towards an oval shape in response
DI Semi-rigid to loading and are able to develop significant lateral earth pressure
from their surrounding embedment.
Thermoplastic Flexible

3
Pipes of different materials are classified in the UK according to
GRP Flexible the strength criterion required to be proven in testing or otherwise
established in design. Where the strength of pipes is established
Thin walled steel Flexible in a crushing test, they are classified as rigid.

BS 9295:2010 describes the calculations and checks, which should be undertaken when designing pipes
of all three classifications. It is particularly important to note that the trench shape has a significant 4
effect on the loading on all pipe types. It is generally recommended that wide trenches with sloping sides
are used when constructing a new pipeline into an existing embankment dam to aid with compaction
and connect to the remaining embankment. However, this leads to higher loads acting on the pipe than
would occur when they are buried in narrow trenches. 5
Within BS 9295:2010, different coefficients and equations are used depending on whether the pipe is

6
laid in a narrow trench or a wide trench, and to take account of ground conditions. The designer should
take care to use the equations that are applicable to the site trench and ground conditions, and not rely
on tables produced by pipe manufacturers.

The general material properties of all materials commonly used in UK embankment dam conduits
are discussed in Section 2.2. However, not all materials are now used for new conduits. As previously
noted, new conduits at existing embankment dams in the UK are typically only considered at small 7
embankments. They are normally short lengths of pipe (less than 100 m), often with a need to
incorporate valves and gates, and they are commonly required to have fast installation. For these
reasons, the most common pipe materials used for such replacement work are precast concrete, DI and
HDPE (structured and solid wall types). The applicability, design and specification of these materials for
new conduits are discussed as follows.
8
9.2.1.2 Precast concrete pipes
As noted in Section 2.2, precast concrete pipes are economical, widely available, and come in a wide
9
range of sizes. They are not designed for connection to valves, though penstocks can be fitted to a

10
headwall at the end of a length of pipe. They are often suitable for use as spillway pipes at small UK
embankment dams. They are not suitable for the long-term transfer of pressurised flow.

Precast concrete pipes typically use spigot and socket joints, which provide flexibility to the
pipeline. These are typically sealed using a flexible sealing ring of rubber or similar material.
Some manufacturers integrate the ring into the pipe socket, making the joint less susceptible to bad
workmanship during installation. This system is readily available from major manufacturers and aims to 11
provide enhanced sealing.

Precast (see Figure 9.5) concrete pipes are heavy for any given diameter when compared to other pipe
systems. Due to their weight, pipe units are about 2500 mm to 3000 mm in length for most common
sizes. So, construction requires heavy plant access to the side of the trench and/or the use of large
12
craneage.

Concrete can be susceptible to attack by sulphates, which may occur in embankment fill or foundation
materials. However, precast concrete pipes are more resistant to sulphate attack than typical in situ
A1
Dams and reservoir conduits 127
concrete due to their manufacturing process.
The sulphate content of the soil should be
checked, and sulphate-resistant cement or PFA
substitute can be used in the pipe manufacture
when greater resistance is required.

Precast pipes are prone to damage during

transport and installation through striking
with other objects. Small damage patches
that may be acceptable for typical stormwater
application will not be tolerable for use
in embankment dams. Careful quality
control during installation is essential. The
pipe manufacturer should be checked for
appropriate independent quality accreditation
such as BS EN ISO 9001:2008. The pipe
should meet the requirements of BS EN
1916:2002 and BS 5911-1:2002.

When specifying a concrete pipe, the designer

needs to consider the likely settlement of the
dam fill above and below the pipe, and the
effect this will have on the joints. Settlement
is often linked with the spreading of the base
of the embankment, and both mechanisms
can cause pipes to be pulled apart at their
joints. Spigot and socket joints are particularly
susceptible to this action. For this reason Figure 9.5 Concrete pipes at a small UK embankment dam
the use of precast concrete pipes should be (courtesy Mott MacDonald Ltd)
very carefully considered in relation to the
anticipated continuing consolidation of the embankment.

Settlement causes particular problems where the pipe connects with other structures such as inlet or
outlet chambers. The settlement of these structures is likely to be different to that of the pipe. Allowance
can be made for this by the introduction of a shorter length of pipe just beside the structure (known as
rocker pipes). The additional joints will accommodate some of the differential settlement.

The load on the pipe should be calculated considering the trench shape and backfill operations, which
are likely to be different to typical pipelaying applications. The structural design of the pipeline should
be completed in accordance with BS 9295:2010, which relates to BS EN 1295-1:1997. As noted previously,
precast concrete pipes are classified as ‘rigid’ in these standards.

Installation should be completed according to the manufacturer’s instructions, particularly the closure of
joints and sealing of lifting holes. Installation and testing should be completed in-line with BS EN 1610:1998.

9.2.1.3 DI pipes
DI pipes have been used extensively in potable water transfer in the UK and are common at many
embankment dams. As with precast concrete pipes, they have rigid walls. DI pipes can be used for the
transfer of pressurised flows, and pipes can be easily fitted to DI fittings (eg valves) using flanged joints.
DI fittings of standard sizes are very common and are typically used in the water industry – the ability of
DI pipework to connect to DI fittings without the use of flange adaptors is of significant benefit.

The pipe manufacturer should be checked for appropriate independent quality accreditation (such as BS
EN ISO 9001:2008). The pipe should meet the requirements of BS EN 545:2010. DI pipes have very long
life expectancy of over 100 years if overloading and very corrosive ground are avoided.

128 CIRIA, C743

DI pipes are used for pressurised and unpressurised flow in embankment dams. They can be

1
manufactured in any lengths, but are commonly made in lengths of about 5 m. DI pipes can also be
cut by a circular saw on site to suit changing conditions. However, this leads to a poorer joint if badly
completed and should be avoided if possible through careful survey before design.

DI pipes typically have spigot and socket (Tyton) joints where buried, to allow for a degree of flexibility
at each joint. Flanges are alternatively added to pipe ends during manufacture, allowing ease of fixture
to valves. 2
Pipes are usually protected externally with a zinc and bitumen-based coating to protect against corrosive
groundwater. Enhanced protection can be provided by using a polymer sleeve, though this is unlikely to
be required in UK embankment dams and may lead to further leakage along the pipe. 3
Internally, pipes are protected from the build-up of corrosion products with a cement mortar lining,

4
although epoxy or polyurethane linings are also available. The choice of internal lining has a significant
effect on the weight of the pipe, and also on the plant required for installation.

As with precast concrete pipes, DI pipes with spigot and socket joints are prone to being pulled apart in
dams where embankment spreading or excessive settlement occur. Many pipe manufacturers allow for
the installation of an anchored gasket to help retain joint cohesion. However, this may not be sufficient at
susceptible locations within dams with a history of significant settlement or spreading. 5
The load on the pipe should be calculated considering the trench shape and backfill operations, which

6
will be different to typical pipe laying applications.

The structural design of the pipeline should be completed in accordance with BS 9295:2010, which relates
to BS EN 1295-1:1997. As noted previously, DI pipes are classified as ‘semi-rigid’ in these standards.

9.2.1.4 PE pipes 7
PE pipes are flexible, and so will deflect when loaded, gaining support from fill on either side. As noted
in Section 2.2.3, several densities and wall thickness are available.

They do not have a long history of use at UK reservoirs as the material is relatively new. However, there 8
is concern that deformation of PE pipes may result in low stresses in the fill above the pipe leading to
the formation of seepage paths. Their use is increasingly common at new reservoirs in the USA, and as
a result, research has been completed by FEMA culminating in an extensive guide to the use of plastic
pipes in embankment dams (FEMA, 2007). 9
Plastic pipes can deform under local high loading without rupture because of their flexibility. This is

10
a good thing when rupture would risk the safety of the dam. However, excessive deformation needs to
be avoided. WRc (2001) offers guidance on the permissible external pressures on PE pipes for different
SDRs. As excessive deformation may go unnoticed with pipes transferring low flows without difficulty,
operators may decide that flexible pipes require a higher frequency of inspection than more rigid pipes.

Flexible pipes have considerably different expansion coefficients to iron pipework, and so care needs
to be taken in the design of connections between the two systems. The structural design of the pipeline 11
should be completed in accordance with BS 9295:2010 and BS EN 1295-1:1997.

9.2.1.5 Details to avoid low pressure zones in fill

Fill around conduits is more prone to zones of low density than elsewhere in a typical embankment. This
12
is because:

‰‰ the haunches of a circular pipe can be difficult to access with compaction equipment
‰‰ heavy compaction equipment cannot be deployed near to the pipe due to the risk of damaging it
A1
Dams and reservoir conduits 129
‰‰ differential settlement can occur along the length of the conduit
‰‰ arching occurs in the fill above or to the sides of the conduit.

Such zones increase the likelihood of leakage paths along the conduit, and so their avoidance is critical in
relation to dam safety. They are discussed along with their mitigation as follows.

Conduit shape
Compaction of fill around conduits in embankment dams has long been a concern of dam engineers in
the UK, and the multitude of dam conduit shapes is in part attributable to this concern. Historically,
haunches under large conduits in masonry or concrete were generally avoided through specifying a
conduit with near vertical walls, so avoiding haunches altogether.

When constructing a replacement conduit with rigid or semi-rigid pipes (eg precast concrete, DI),
particularly through the impermeable zone of the dam, the pipe is often constructed with a cradle of
mass concrete. A typical cradle is shown in Figure 9.6.

Figure 9.6 Rigid pipe cradle (courtesy Mott MacDonald Ltd)

The cradle should extend up to the springline of the pipe and along its entire length. The concrete
used to form the cradle needs to be impermeable, requiring careful consideration of the mix design and
support arrangement when pouring. The mix may also include additives to ensure the rapid setting
of the concrete, to minimise the construction programme. The sides of the cradle should also slope at
around 1(H):10(V) as shown in Figure 9.6, to aid compaction on either side.

Note the cradle can alternatively be widened to the full trench base width (see Figure 9.7).

The cradle will be subject to stresses caused by differential settlement of the pipe. Reasonable settlement
can be accommodated by joints in the pipe. The cradle should have movement joints at similar spacing to
accommodate this. These are typically provided
with a compressible filler board between sections.

During the casting of the cradle, flotation

forces will be applied to the pipe so restraint
is required to maintain the design alignment,
usually in the form of straps.

At some low height dams in the USA, rigid pipe

conduits have been laid on a concrete bedding
supporting the bottom 25 per cent of the pipe
only, rather than a cradle up to the springline,
for reasons of economy (FEMA, 2005). As this
implies a requirement to compact fill within
the pipe haunches, it is not recommended for
Figure 9.7 Concrete pipe on full width cradle (courtesy Mott
use in the UK. MacDonald Ltd)

130 CIRIA, C743

Flexible pipes (eg HDPE) rely on some deformation under load to mobilise resistance from the

1
surrounding fill. Half-pipe cradles should not be used with flexible pipes, as they restrict this necessary
deflection, causing stresses to build up in the pipe wall, which may crack the pipe.

Designers considering use of a flexible pipe need to be satisfied that sufficient compaction under and
around the pipe is feasible where this is critical (through the impermeable zone). Alternatively the pipe
could be fully encased in in situ concrete (forming a rigid pipe). In such circumstances the hoop stress is
transferred from the pipe into the concrete surround. In addition, the concrete needs to be designed to 2
withstand the vertical loads above the pipe. So, concrete surrounds to flexible pipes should be reinforced
for all except very small or shallow pipes. This solution is shown in Figure 9.8. Another solution is
placing the HDPE pipe within a rigid pipe that is cradled.
3
Deflection along length
Deflection along the length of a
conduit in an embankment dam is
likely for a number of reasons (as 4
illustrated in Figure 9.9):

5
1 Load from the embankment
varies with maximum load at
the crest.
2 Load and support conditions

6
vary with each type or zone
of fill through which the
conduit passes.
3 Compressibility of materials Figure 9.8 Full reinforced concrete encasement for flexible pipe (courtesy

7
supporting the conduit can Mott MacDonald Ltd)
vary, particularly if a cut-off
is present or the conduit is well above the foundation.
4 Ongoing settlement and spreading of the embankment will change the shape of the embankment

8
and impart loads onto the conduit.

Settlement along the conduit will cause additional shear and bending stresses within the wall of the
conduit. In addition, spreading of the embankment can cause jointed pipes to be pulled apart.

Pipes should be designed with a fall calculated to be acceptable even after estimated settlement has
occurred. Geotechnical specialists should be engaged to estimate deflections where these are expected to
be an issue particularly if there are concerns relating to previous dam performance.
9
10
Deformation can be minimised by carefully locating the pipe at as high a level as possible and placing it
as late as possible in the construction sequence or placing it along the dam foundation if the pipe needs
to be at depth. During construction the excavated surfaces should be protected from softening due
to precipitation, particularly in the base of the trench. Spreading can be minimised by using suitable
material when rebuilding the embankment on top of the conduit trench backfill. Where existing
materials are to be reused these can be stabilised with additives to provide improved strength.
11
Where deflection does occur, pipe wall stresses are most effectively relieved in rigid or semi-rigid
pipes through the use of flexible joints. Such joints should be located where deflection is expected at
transitions, namely at transitions between zones and close to structural interfaces.

Where a cradle is used with a rigid pipe, joints should be introduced into the cradle to allow flexibility
12
without cracking (for example, at a joint where the pipe bends ‘downwards’). Durable compressible

A1
fillers such as those used in reinforced concrete expansion joints should be used at such joints. Similar
precautions should be used for encased flexible pipes.

Dams and reservoir conduits 131

Figure 9.9 Stresses causing settlement and movement in conduits (courtesy Mott MacDonald Ltd)

Flexible pipes without concrete surround should be able to flex in response to deflection along their
length. However, a realistic estimate of expected deformation is required to ensure the sufficiency of the
chosen pipe. Geotechnical specialists should be consulted.

Preventing arching
Shear stresses occur in fill around conduits, which can lead to pockets of low density material being
created that cause arching. This is typical where a conduit is constructed within a trench in an existing
embankment, as would commonly be the case when replacing a pipe.

After a pipe is replaced, the fill

above it tends to settle more than
the surrounding fill, creating shear
stresses that support the fill over
the conduit. This can lead to low
pressure zones (see Figure 9.10).

Arching can be reduced by

increasing the trench base width
and reducing the angle at which the
sides of the trench are excavated.
As shown in Figure 9.11. The base
width depends on the compaction
plant to be used at either side of the
conduit. Typically a minimum width
of either side of the conduit that is
equal to the width of the compaction
plant plus 10 per cent is required.
Side slopes no steeper than 1(H):1(V)
are recommended, incorporating Figure 9.10 Shear forces following conduit construction within a trench
(courtesy Mott MacDonald Ltd)
benches in deeper cuts for stability.

Compaction around and above conduits should not be completed in vertical sided trenches, unless the
pipe is to be surrounded by granular material. This would only be suitable outside the impermeable
zone (eg within the downstream shoulder). However, individual layers of fill should be benched into the
surrounding fill using shallow, vertical cut benches (see Figure 9.11).

In addition, where only a downstream section of the conduit is replaced, it may be decided to replace the
compacted fill around and above this section of conduit with granular filter or drainage material.

132 CIRIA, C743

 1
2
3

Figure 9.11 Recommended measures to reduce arching of fill over conduits (courtesy Mott MacDonald Ltd)
4
Providing sufficient base width for compaction and reducing trench side slopes may lead to an
excavation much larger than expected by owners or contractors familiar with pipelaying techniques used
for stormwater drainage. Parties should be reminded that these requirements relate to the dam safety 5
and should not be compromised.

9.2.1.6 Anti-seepage collars

Conduits through embankment dams often have anti-seepage details, designed to reduce leakage along
6
the conduit/fill interface by elongating the flow path. Typically the mechanism is the installation of

7
collars or flanges that sit perpendicular to the conduit where it passes through the impervious zone of
the dam. These are commonly known as cut-off or anti-seepage collars. They are generally constructed
from the same material as the conduit. Human-entry conduits or pipes with concrete surrounds may
typically have reinforced concrete collars around 300 mm thick and up to one metre high, while

8
steel or iron pipes may have metal collars. The collars are spaced at 7 to 10 times their height, which
increases the potential leakage path along the interface by 20 to 30 per cent. Another similar detail is the
construction of clay ‘stanks’, which are collars formed of clay.

A series of failures of embankment dams in the USA occurred in the period 1960 to 1980, near to
conduits with cut-off collars. Later investigations concluded that the collars were largely ineffective as
they only influenced water flow 9
in the immediate area around
the conduit. The investigated
failures were found to be caused
by water flowing through cracks
in the fill just beyond the limits
10
of the collars.

The inclusion of collars or stanks

is often still preferred by some
11
dam designers concerned about

12
leakage along the conduit/
fill boundary. From the USA
example, it appears that if anti-
seepage collars are included,
great care is necessary to ensure
that full compaction of the
surrounding fill is carried out A1
right up to the edge of the collar. Figure 9.12 Typical anti-seepage collars (courtesy Mott MacDonald Ltd)

Dams and reservoir conduits 133

This may require the use of different compaction techniques (eg plate compactors) compared with that
used for bulk fill compaction.

Anti-seepage collars (see Figure 9.12) may have some benefits in certain circumstances, particularly in
low-head situations.

If used, collars should be spaced sufficiently to allow good compaction around them. Their sides can also
be sloped to aid compaction.

Whether anti-seepage collars are used or not, it is recommended that a filter diaphragm is constructed
towards the downstream end of the conduit.

9.2.1.7 Filter diaphragms, collars and drains

As noted in Chapter 2, filter layers are commonly used in modern zoned embankment dams on either
side of the core. They can prevent internal erosion from the surrounding less permeable zone, and allow
substantially different materials to be used in the core and dam shoulders.

Leakage paths can form

along the line of conduits
in embankment dams,
either along the conduit/
fill interface, or in cracks in
the surrounding fill caused
by stresses induced by the
conduit presence. This higher
level of risk means that
filters are often appropriate
at conduits, even where not
constructed elsewhere in an
embankment.

The inclusion of a filter zone

where remedial works are
undertaken at an existing Figure 9.13 Typical conduit filter detail in a homogenous embankment dam
conduit, or a new one is (courtesy Mott MacDonald Ltd)
constructed, is normally a
very small proportion of overall construction cost. So, the inclusion of a filter is strongly recommended
where other works are being undertaken at an embankment dam conduit. Such a filter is normally
provided in the form of a filter diaphragm or filter collar. A typical detail is shown in Figure 9.13.

There are three main constituent parts of a conduit filter:

1 Diaphragm or collar.
2 Granular drain.
3 Drainage monitoring point at downstream toe.

The filter diaphragm is a wall, perpendicular to the direction of the conduit, which extends sufficiently
from the conduit to intercept piping flows in the zone of influence of the conduit.

Figure 9.13 shows a typical location for a filter diaphragm in a homogenous dam. In such dams, the
diaphragm should be located downstream of the dam centreline and any cut-off trench. A common
detail is to place it at one-third of the conduit length from the downstream toe. If the dam has an
impermeable core, the filter diaphragm would ideally be placed immediately downstream of the core,
but can also provide benefit if installed further downstream (where only replacing a downstream section
of existing conduit).

134 CIRIA, C743

A rule of thumb for a filter diaphragm is that it should extend to a length of 3D (where D is the conduit

1
diameter) on either side of the conduit, extend to the same distance above the conduit (3D), and 1.5D
below. However, the engineer should use judgement when designing the filter, considering the site layout
and fill characteristics. The filter diaphragm should be at least 900 mm thick.

The grading design of a diaphragm or collar is important if it is to function properly. The principle of
filter design is to produce a material that allows the passage of water while preventing the passage or
movement of soil particles from the soil that the filter is protecting. However, the practice of filter design 2
has developed considerably in recent years and has become more involved than the simple calculation
used in earlier times. Fell et al (2005) chart the development of methods of filter design from Terzaghi’s
early rules of thumb, through developments by the USBR in the 1960s and 1970s to other researchers
in the 1980s. Building on the results of relatively recent research, Fell et al (2005) recommend a nine-
step methodology for filter design. Other sources of guidance can be found in ICOLD (1994 and 2014),
3
USACE (2004), and more recently in USBR (2011).

‘Standard’ or ‘general’ filter designs are not appropriate as gradings should be tailored to that of the
specific base soil grading. Consequently, suitable filter materials are rarely available on site and are
4
usually purchased from aggregate suppliers or processed from materials near to the site.

Laboratory filter tests are normally required to inform the selection of suitable filter gradings.
5
In addition to being compatible with the embankment materials being protected, filter materials should
be able to sustain deformation without cracking. Fine, poorly graded filters have poorer self-healing
characteristics than broadly graded, coarser filter materials. However, segregation can be more of a
problem in broadly graded filters. 6
The filter material should be hard and durable and not subject to segregation during transport,

7
placement or compaction. The use of narrowly graded materials helps to prevent the segregation
problem. The grading should be chosen to avoid excess water pressure build-up. A two-layer filter may
be required for zoned embankment dams where the core and shoulder zones are very different. Filter
materials should not be placed in very wet or freezing conditions as compaction will be impaired.

The leakage and seepage flow intercepted by the filter diaphragm needs to be safely transferred to
the downstream toe of the dam by means of a granular drain. This is typically used as bedding for the 8
remaining conduit length, and can alternatively form a full surround to the conduit, which is also known
as a filter collar. The drain should be surrounded by filter material to avoid choking with fines washed
from the surrounding fill.

A drainage monitoring point should be provided at the outlet to the conduit drainage system. Such a
9
measurement point is typically a V-notch weir within a manhole chamber. In such instances the weir

10
plate should preferably be arranged to allow reading without entering the chamber. Ocherous deposits
can often build up behind weir plates and distort readings, so arranging the system so this area can be
easily cleaned is also beneficial.

9.2.2 Construction process

11
9.2.2.1 Control of reservoir water level during construction works

12
Replacement of conduits should be undertaken by:

‰‰ isolating the existing conduit from reservoir and potential flood waters
‰‰ making the necessary excavation
‰‰ removing the existing conduit if required
‰‰ constructing the new conduit
‰‰ covering it with compacted backfill so that the original dam profile is reinstated and the required
A1
long-term freeboard is re-established.

Dams and reservoir conduits 135

Before starting the works, any required environmental and planning approval for the works should be
obtained, and measures and programming undertaken to avoid detrimental effects on protected species
and habitats or to prevent the spreading of invasive species (see Section 8.2).

The draw-down and control of water in the reservoir before and during the works can be a major
undertaking, using existing pipework, temporary pumps/siphons and normally a temporary cofferdam
designed to retain an appropriate flood rise in the reservoir basin during construction (for impounding
reservoirs). Control of water level during the replacement of a conduit is probably the largest single
construction risk associated with such work (see Section 8.2.1).

9.2.2.2 Removal of the existing conduit

The next step in replacing an existing conduit may be the removal of the existing one. This requires an
open trench excavation to expose the existing conduit.

The full excavation at the level of new conduit installation needs to be wide enough to allow compaction
on either side, with sloping trench sides to avoid arching of the fill above the conduit. This excavation
will be much wider at the crest than that required simply to remove the existing conduit, which could
be undertaken using temporary trench boxes to limit the trench width. So, the excavation could be
undertaken in two stages:

1 A narrow trench to enable removal of the existing conduit.

2 Construction of the new one, to be widened later for compaction of the new conduit backfill.

This method could help mitigate the risk of poor weather affecting exposed core materials.

Where the excavation passes through a puddle clay core, it is important that this layer is protected from
the elements when exposed, for example, by a tarpaulin and suitable drainage.

Excavated fill that is suitable for reuse should be stockpiled nearby in segregated piles, which are
protected from damage.

Excavation into old embankment dams can often expose unforeseen conditions, particularly redundant
drains or other structures. Any exposed features should be brought to the attention of the QCE. Any
affected drains should be re-established or flows brought safely to the downstream toe. Animal burrows
should be excavated as much as possible, although excessive burrows may instead be filled by grout,
foamed concrete or polymer resin. The exact remedial works should be agreed with the QCE.

Unforeseen geotechnical conditions may also be uncovered, particularly relating to wet areas due to
springs or reservoir leakage. Again the response should be agreed with the QCE. Removed material
that is unsuitable for reuse should be stockpiled separately and clearly marked until disposed of in an
appropriate location.

When exposed, the conduit should be removed and disposed of appropriately. Samples of old conduits
may be of interest to research institutions and universities for material investigation. Masonry units and
undamaged modern pipelines may be reused on other structures.

Asbestos cement pipes have previously been used in the UK water industry. They were not typically used
within embankment dams, although their presence has been confirmed at some dam sites. If they are
encountered, they require additional safety controls during pipeline removal to mitigate the release of
asbestos fibres into the atmosphere and prevent inhalation by demolition
Asbestos personnel and others. The removal of asbestos is regulated in the UK
Guidance can be found at: under the Control of Asbestos Regulations 2012. Further information
http://www.hse.gov.uk/asbestos can be found on the HSE website (see Asbestos box).

136 CIRIA, C743

9.2.2.3 Placement of new conduit and compaction
The new conduit should be constructed, placed within a cradle (where applicable) and tested for leaks
before compaction of the surrounding fill. Compaction should progress as soon as the cradle has set, and 1
the tests have been successfully completed.

The shoulder fill (or the general fill in homogeneous dams) should be placed in thin layers close to the
optimum water content and compacted in accordance with a suitable earthworks specification. The width
of the replacement trench should be sufficiently wide to allow free passage of the compaction plant on
2
both sides of the cradle. Where fill is placed against in situ soil, the side of the trench should be trimmed

3
back to form a bench of the same height as the layer being placed, and the entire layer compacted evenly
together to achieve a consistent density (see Figure 9.10).

The compaction of soils in the conduit replacement trench may be specified by ‘method’ or by
‘performance’. If a method specification is used, it is necessary to specify the source materials, moisture
content, layer thickness, compaction plant type and weight and number of passes. Layer thicknesses
would typically lie in the range 100 mm to 250 mm for this type of application. In a performance or ‘end 4
product’ specification, the source materials, particle size, Atterberg limits, moisture content and density
ratio are specified. The density ratio is the ratio of the dry density achieved against the maximum dry
density achieved in standardised laboratory compaction tests.

A ratio of 95 to 100 per cent (2.5 kg rammer) would be typical for compaction of fill around a culvert.
5
6
Both forms of specification should be preceded with sufficient laboratory or field testing to allow the
various criteria to be well defined before the replacement works start. The performance specification
requires that further testing is carried out throughout the compaction works to ensure the specified
end product is met. Compliance testing will normally be based on achievement of the specified density

7
ratio, but other criteria such as moisture content and grading can also be checked. The frequency of
compliance testing should be chosen to give a reasonably representative set of data to control the works
without interfering unduly with the smooth execution of the works.

Care is required when compacting over the crown of the conduit not to over-stress the conduit structure.
Vibrating compaction methods should rarely be used, for example, close to the crown of a pipe. A
different method of compaction may be specified for the zone within a set distance of the pipe crown.
However, the other compaction criteria (ie layer thickness, number of passes) needs to also be adjusted to
8
ensure that the soil is properly compacted.

Where the new conduit goes through the clay core of the dam, it is imperative that the replacement
clay core is compacted sufficiently. Generally the clay core should be raised in layers roughly in
9
parallel with the raising of the shoulder material. However, it may be helpful to raise the shoulder

10
material first and slightly over-fill, trim back to the exact profile of the clay core, then place the clay
core material up against the shoulder material. This should give the greatest opportunity to fully
compact the core material.

The clay core material should be selected to be compatible with the existing material. Often, the
excavated clay core material may be suitable for reuse. However, its moisture content should be
carefully checked to ensure that it has not been allowed to dry out or wet up during storage. The
compaction of the clay core should also be specified as previously indicated, although the compliance
11
criteria may be different.

While undertaking these works, other aspects of construction good practice are also critical to the
success of the operation. The management of surface water in the excavation is crucial to prevent the
12
construction materials from becoming unsuitable through being excessively wet. Conversely, if the

A1
existing soils are exposed for long periods in dry weather they may dry out at the surface and desiccation
cracks may form. Any desiccated material should be scraped off before fresh backfill is placed. It is
always prudent to avoid delays to the works to minimise the risk of such effects.

Dams and reservoir conduits 137

When the embankment reconstruction is complete, other construction works can be carried out to
complete the works, such as topsoiling, seeding and reinstatement of access routes.

9.3 REPLACEMENT OF PIPES WITHIN CULVERTS

In most large UK earth embankments constructed for water supply, pressurised pipework drawing water
from the reservoir is transferred through the dam via pipes within masonry or concrete culverts. There
is a huge variety of culvert arrangements at UK dams. These culverts are normally accessible for human-
entry from the downstream end and in many instances from the base of an intake tower at the upstream
end as well (see Figure 9.14).

The size of such culverts was generally based

on a combination of pipe footprint and human-
entry requirements, and in many cases the
latter assumed smaller stature of personnel
than today’s inspecting engineers, together
with lower standards of accessibility!

Culverts in embankment dams are often

classified as confined spaces. Work within them
should be avoided if at all possible, and where
it is required, adequate safety precautions
need to be undertaken that are generally more
onerous than completing the same work in an
unconfined space. Section 4.1.2 provides more
Figure 9.14 Restricted access in conduit (courtesy Canal &
River Trust) information on confined spaces.

The methodology for replacement of pipes within culverts should be determined with specialist advice to
ensure risks to the dam embankment are assessed and mitigated as appropriate.

9.3.1 Replacement planning

Pipes in culverts are generally supported on concrete or masonry cradles at regular intervals, or are
continuously supported through being cast into the culvert invert. Localised areas of damage can occur
due to excessive stress in the pipe wall. This can occur due to a wide range of reasons such as:

‰‰ Material weakness inherent in the pipe from manufacture, exacerbated by corrosion by internal flows.
‰‰ Poor workmanship during construction (eg uneven support).
‰‰ Movement due to pipe expansion or dam settlement, where no allowance for this was made in the
original design.
‰‰ Sudden movement due to earthquake or surge loading during rapid valve closure.
‰‰ External damage from concentrated drips or scour flows.
‰‰ External corrosion where maintenance painting has been insufficient (typically where hard to reach).

The replacement of a section of the pipework (or the entire pipe) may prove preferable to lining or
painting it. Alternatively, a section of pipe or an existing valve may need to be removed to install a new
valve into the pipeline.

Where other options are available, the designer of the replacement works should establish the following
to check the feasibility of replacement:

1 Access for removal of the damaged section and insertion of the new piece.
2 How the pipe can be sufficiently isolated from the reservoir.
3 How the function of the pipe can be provided elsewhere when the replacement works are taking place.

138 CIRIA, C743

4 What pipe material is available, which meets the requirements of the replacement, including

1
jointing with upstream and downstream sections.

Once these points have been confirmed, the following needs to be determined in further detail:

1 Points between which the length of pipe is to be removed.

2
3
Demolition process including disconnection from support, and removal from culvert.
Jointing and testing of the new pipe section. 2
These points are discussed in further detail as follows.

9.3.2 Replacement feasibility 3

9.3.2.1 Access
Access into reservoir culverts is often dependant on negotiating complex towers or chambers, while
the actual culvert is never bigger than the designer considered necessary. In some instances the pipe
4
diameter is larger than the tunnel radius, and so it is not possible to remove and replace a full section of

5
the pipe without first removing the pipe body between the section in question and the access point.

Where the access is constrained by a particular chamber, consideration may be given to its removal,
particularly if this removes a hazard to be otherwise faced by O&M staff in the future. Where the entire
structure cannot be removed, access points can be increased in size or sections of wall removed. Such
works have structural integrity implications and should not be undertaken without appropriate checking
first. Significant structural changes should also be accompanied by monitoring for deflection throughout
6
the works (see Chapter 6).

Embankment dam culverts are also often partially blocked by walls used to support pipes and possibly
the actual culvert. Walls within culverts may be original or subsequently added for stability reasons or
because the contents of the culvert have changed. Again, the structural implications of their removal
7
should be assessed.

9.3.2.2 Isolation from the reservoir 8

Section 8.2 provides techniques for isolation of the working area from the reservoir head.

For works to remove pipework in culverts, double isolation in retained upstream pipework is the most
common method. This is where two separate barriers exist between the reservoir and the working area,
9
formed by closed valves or bungs in the upstream section of pipe.

In some instances isolation should also be provided on other pipework. For example, where work is to be
undertaken on a supply pipe in a culvert in which a scour valve is present at the upstream end, the scour
pipe should also have adequate (normally double) isolation.
10
9.3.2.3 Temporary function replacement
During works to replace a section of pipe (see Figure 9.15), the function of that pipe needs to be
11
provided from elsewhere if it is necessary for the safe operation of the dam.

For draw-off pipes within water supply reservoirs the most economical solution to this problem is to
supply the applicable demand with water from other reservoirs. Where this is not possible an alternative
12
draw-off system needs to be arranged, for example by using temporary pumps or siphons. Care is
needed to avoid erosion of the dam in the event that the pumping main should burst.

For compensation pipework, compensation can normally be provided through a scour valve at the dam A1
or one just downstream on the supply network.

Dams and reservoir conduits 139

Works that prevent operation of the scour valve at an
impounding reservoir are most problematic as this
function could be required if an emergency occurs
during the works that requires a draw-down of the
reservoir. The engineer responsible for the works
needs to consider the likelihood of an emergency
draw-down being required, which depends on the
condition of the entire dam structure. Where the
reservoir level cannot be drawn-down significantly
before the works, the minimum mitigation measure
would be the planning and arrangement for
emergency pump installation at short notice. At higher
risk sites, provision for emergency draw-down will
need to be provided throughout the works, possibly in
conjunction with a preliminary draw-down.

9.3.2.4 Pipe material and jointing

The replacement pipe should be as durable as possible
without adversely affecting the loading on the existing
pipework or requiring a weak connection between
the two. Normally the pipe is replaced on a like-for- Figure 9.15 Wachs cutter in operation before pipe
replacement (courtesy Mott MacDonald Ltd)
like basis. However, some old pipe sizes are no longer
commonly available, and CI pipes are often replaced by DI alternatives.

140 CIRIA, C743

10 Techniques for abandonment
of conduits
1
2
This chapter details the process to be followed for the abandonment and removal of conduits, including
inspection, planning, specifications and filling of the abandoned conduit.

Where the function provided by a conduit is no longer required, has been replaced or is better provided
3
elsewhere, the optimum action to reduce long-term dam safety risks may be to remove or abandon

4
the conduit. This is particularly the case where the conduit has already deteriorated and the risk of
structural collapse is high, with serious potential implications for dam stability.

Equally, it may be decided to abandon a conduit where a dam is to be raised and the conduit does not
have the required strength to resist the greater embankment loads. Another purpose of filling a conduit
may be to remove a confined space from the site in the interests of operative personnel safety.
5
The most significant decision relating to the abandonment or removal of a conduit is whether and how
the function of the conduit is to be provided after abandonment or removal is achieved. This chapter
discusses the processes of abandonment and removal of conduits on the assumption that this important
question is already resolved. 6
10.1 ABANDONMENT
Filling the conduit with grout or foamed concrete is typically the optimum method of abandonment, as
7
the grout will fill the void within the conduit and support the remaining conduit structure, preventing

8
future collapse. Considerable care is required when selecting the material for filling such a conduit to
ensure that shrinkage is avoided and potential future leakage paths are eliminated. This method also
avoids the dam safety risks associated with an open trench through the dam core, which would normally
be required to remove the conduit fabric altogether. A recently developed alternative to grout is by using
expanded polymer resin to fill the conduit.

Historically, some conduits have been abandoned at dams in the UK through blocking a short length with
a plug. Such structures still form a hazard to the dam as they continue to form preferential leakage paths
9
within the surrounding fill and could yet deteriorate to the point of collapse. In addition it may be difficult

10
or impossible to inspect them or assess their condition. Where an inspecting engineer has particular
concerns about such a conduit, they may choose to complete its abandonment by locating and filling all
voids on either side of the plug. This removes the risk posed by deterioration of the pipe shell material.

11
The main costs and risks of abandoning a conduit are noted in Table 10.1.

Table 10.1 Conduit abandonment – summary

Cost or risk Project element Comments

12
Mobilisation and labour/plant ‰‰ generally low, although depends on access for fill material
Construction costs delivery.
cost
Typical material costs ‰‰ generally low, although depends on size of pipe.

Damage to existing structure ‰‰ unlikely.

A1
Construction ‰‰ poor workmanship can leave inaccessible voids within conduit
risk Other construction risks forming points of structural weakness and increasing the risk of
leakage path formation.

Dams and reservoir conduits 141

 Cost or risk Project element Comments

‰‰ complete removal
Capacity reduction
‰‰ capacity needs to be provided elsewhere.

‰‰ cost of future maintenance reduced, valves will no longer need

Operational cost operation
Surveillance/maintenance/ ‰‰ some surveillance still required as abandoned conduit length
running costs downstream of a plug or locked off valves still poses a reservoir
safety risk
‰‰ no surveillance needed if the pipe is filled over the whole length.

‰‰ abandoned human-entry conduits may no longer be accessible

Good installation for visual inspection of leakage or further structural deterioration
‰‰ good for health and safety, bad for deterioration identification.
Operational risk ‰‰ poor fill material can allow leakage paths to form, which may only
become apparent at a later date
Poor installation
‰‰ higher risk for long narrow conduits, conduits inadequately
prepared, filling from dam surface.

Note
See Section 8.1 for guidance on applicability and colour coding.

10.1.1 Abandonment process

Once it has been decided to abandon a conduit, the project should be progressed through the
following stages:

‰‰ inspection
‰‰ planning – access and isolation requirements
‰‰ specification of fill material
‰‰ specification of filter collar (if applicable)
‰‰ preparation of conduit
‰‰ filling process and recording
‰‰ post-construction monitoring.

Further information is given in Sections 10.1.2 to 10.1.9.

10.1.2 Inspection
The conduit should be inspected to confirm that the proposed fill material can fill all voids. Large
obstructions caused by debris, local collapse, sediment or sludge or joint offsets should be identified.
Such obstructions can be overcome by introducing the fill material from multiple locations, or adjusting
the mix design. However, some defects identified in the inspection, such as large voids, may imply a
significant risk associated with the surrounding fill, which may require further investigation or remedial
work before abandonment.

The inspection process is best carried out following isolation of the conduit from reservoir water.
However, in many conduits this is not possible without considerable investment in temporary works.
Where alternative options to abandonment are available, isolation of the conduit for inspection may not
prove cost effective. In such cases, inspection using divers or underwater ROVs inserted into the conduit
inlet may be sufficient.

10.1.3 Planning – access and isolation requirements

Abandonment of a conduit requires appropriate fill materials to be transported to site and then into the
conduit. The former will require adequate access roads, including considering the setting time where
ready-mixed materials are to be used.

142 CIRIA, C743

The conduit should also be isolated from the reservoir water where possible. This can be done by

1
construction of a cofferdam (see Section 8.2) or by closing valves and/or inserting an inflatable bung into
the upstream end of the conduit, which can resist the reservoir head.

Where a bung is used, the capacity of the conduit structure to resist the pressure from the bung should
also be checked. Where abandoning a pipe under reservoir head, double isolation is usually required for
safe working conditions.

Two types of material are commonly used for filling conduits – cementitious grouts/concrete and
2
polymer resins.

Where grout or concrete is used, it is typically inserted at a low pressure into the conduit at one location,
and sometimes at multiple locations depending on the conduit size and constraints. Typically the
3
material is inserted at the upstream or downstream end of the conduit, depending on ease of access and

4
slope of the conduit. Temporary bungs that can withstand the pressures imposed by the fill material are
required at either end. An allowance for the removal of air is required.

Filling conduits through holes drilled into the top of the conduit from the embankment crest or
shoulders is possible and has been completed in the USA. However, there are significant dam safety risks
that may preclude such activity at high consequence dams, namely:

‰‰ Improper drilling procedures increase the potential for hydraulic fracture. A contractor may
5
use drilling fluids such as water or bentonite to enhance removal of drill cuttings. Such fluids

6
should be avoided when drilling in the fine grained embankment zones, as the fluids can be
quickly pressurised by the pump to pressures that exceed the surrounding earth pressure, causing
hydraulic fracture of the embankment material.
‰‰ Poor drilling accuracy or drill bit migration can mean that not all parts of the conduit are equally

7
served by fill holes. It can also be difficult to confirm by this method that all parts of the conduit
have been adequately filled.
‰‰ Holes drilled into the embankment fill (particularly weak puddle clay) can act as leakage paths and
weaken the surrounding soil.

The technique also relies on knowledge of the conduit location – multiple holes that fail to connect into
the conduit will cause both high costs and greater damage to the dam. 8
As with cementitious materials, fill materials based on polymer resin can be injected into the conduit at
one or multiple points. As such materials can be pumped over several hundred metres – vehicular access
to either end of the conduit may not be required. 9
10.1.4 Specification of fill material
Traditionally foamed concrete or a cementitious grout is used to abandon pipes or caverns. Concrete
is more economical than grout for larger voids, as a standard large diameter concrete pump truck can
10
be used with supply from concrete mixer lorries. Grout injection is typically carried out using smaller

11
diameter pump lines and an on-site mixer. It also requires a longer time to fill the conduit.

The material requires careful specification to ensure that no voids or leakage paths remain after
application. To achieve this, it needs to have high density and reasonably low permeability, and be
applied suitably into the conduit. Voids may remain following a grouting operation for a range of
reasons, including:

‰‰ formation of air pockets – local obstructions or blockages caused by inconsistencies in the fill
12
material

A1
‰‰ plastic shrinkage – occurs during the initial setting of the material during hydration of the cement
‰‰ bleeding – occurs when solid particles within the grout or concrete settle leaving water at the surface
‰‰ drying shrinkage – occurs when the water in the grout mixture is removed by hydration,

Dams and reservoir conduits 143

 evaporation and absorption by the surrounding media. The damp enclosed conditions within a
reservoir conduit mitigate this action
‰‰ thermal expansion – occurs during setting of the material, although generally the low strength of
the fill materials prevents this from being a major problem.

Due to the elongated nature of conduits, the fill material needs to be workable. This is most easily
controlled through amending the water/cement ratio. However, this reduces the strength of the grout/
concrete, and makes it more susceptible to bleeding. Small amounts of admixtures such as plasticisers
can be introduced to the mix to balance the requirements of workability and durability. Trial mixes
should be created to verify the suitability of any admixtures.

Alternatively, conduits can be filled using expanded polymer resin. This is mixed on site and pumped
into the conduit. This has some advantages over grout as a fill material, as it does not require shuttering,
it expands in every direction to fill voids and it pushes out debris, liquid or air as it fills the conduit.
It is also removable if necessary. The material is also suitable for sites with difficult access, and can be
pumped over several hundred metres (beyond the limits of grout pumping).

However, the material needs to be used in conjunction with a watertight plug at a suitable point along
the conduit. It also has lower compressive strength (0.15 to 0.3 N/mm² at 28 days) compared to grout
(typically 3 to 12 N/mm² for void filling).

Polymer resin fill materials can allow the passage of some water entering the conduit to escape. Where
this is a particular concern, or where the flows are high, positive drainage can be provided by fixing a
porous material to the conduit walls and invert before installation of the polymer fill.

10.1.5 Specification of filter collar

The earthfill/conduit interface of an abandoned conduit will still form a preferential leakage path
through the embankment. This hazard can be effectively mitigated through the construction of a filter
diaphragm or collar to intercept such flows at the downstream end of the pipe. Such a collar might be
recommended when abandoning conduits in high consequence dams or conduits being abandoned due
to past leakage problems. However, in small dams, the QCE may not consider the collar to be necessary.
Advice on such collars is included in Section 9.2.1.

10.1.6 Other design considerations

Other works may be required as part of an abandonment, such as partial or full demolition of ancillary
structures (eg intake towers, gate houses), removal of gates or valves, or filling of shafts.

10.1.7 Preparation of conduit

Before filling, every effort should be made to clean out all loose materials from the conduit. Projections
that may impede the flow of the fill material should also be cut off or ground down. Bulkheads should
be installed at either end of the pipe, and a vent pipe (or pipes) should be installed at the crown of the
conduit if required. Grout pipes of a suitable diameter (depending on pumping distance) should also be
installed at the crown of the conduit.

The potential effect of locally increasing the foundation loading by filling large conduits with, for
example, mass concrete, should be considered in planning the works. The loading conditions on the
conduit walls during the filling operation should also be considered.

10.1.8 Filling process and recording

The proposed filling process should be detailed in a method statement to the satisfaction of the QCE
and applicable safety requirements. The method statement should detail the process of mobilisation,

144 CIRIA, C743

access, and set-up, fill material mixing, fill sequencing, procedures for handling waste, method of

1
communication, and emergency actions.

The material should be mixed as appropriate to obtain the required consistency. Large conduits filled
by concrete will often use ready-mixed concrete delivered to site. Quality assurance at the batching plant
should meet recognisable quality standards (such as BS EN ISO 9001:2008). The time taken to travel to
site needs to be considered when batching and should form part of the trial mix process.

Grout fill is likely to be mixed on site. Quality control of constituents is essential. Water for use in the
2
mix should be clean, and grout materials stored appropriately indoors before use. Unless the grout
is expected to pass through very small gaps, a normal paddle mixer is usually suitable. All of the
constituents of the grout (with the possible exception of admixtures) should be batched by weight. The
careful calculation and batching of all constituents is required to ensure that the production grout mix
3
is of the quality and consistency required by the design and proven in any trials. Suitable means of

4
testing and quality control should be present and used on site before grout is delivered. Simple errors in
batching are feasible and can have large consequences (see Case study A1.21).

Mixed grout should be sieved to extract any lumps and then pumped into the conduit as soon as
possible. If its use is delayed to the extent that its workability or admixture effectiveness is reduced, the
batch should be discarded.
5
The mixing and pumping equipment should be kept clean and flushed with water at intervals during
long grouting operations and after each use.

The grout or concrete should be pumped into the conduit. It is important to measure and monitor the
pressure of the grout and the volume inserted. Estimated volumes are difficult with complex conduit
6
shapes. Pressure should ideally be measured at the crown within the conduit. The allowable pressure

7
ceiling should be determined before work is completed and adhered to.

Where polymer resin is used as fill material, the filling process should be undertaken according to the
manufacturer’s instructions. The volume of material used should be recorded.

As with all grouting operations, success depends significantly on the experience of the operating staff and
their familiarity with and the suitability of the grouting equipment employed. In this regard employers
should take particular care to engage suitably qualified and experienced companies and individuals.
8
10.1.9 Post-construction monitoring
The performance of the filled conduit (and filter collar if installed) should be monitored. In particular,
9
drainage flows emanating from the embankment at the downstream end of the conduit should be
monitored for colour and quantity. Any increases in outflows from other conduits should also be noted.
Other signs of leakage paths such as lush vegetation or settlement should be watched. In conjunction
with monitoring crest settlement, checking settlement along the conduit alignment would be prudent if
10
concerns remain about the conduit.

10.2 REMOVAL 11
The removal of conduits has several disadvantages when compared to abandonment by filling, namely:

‰‰ isolation from reservoir water is required during removal

‰‰ access onto the dam by excavation and compaction plant is required
12
‰‰ the impermeable zone of the dam is disturbed by the works.

However, where a conduit is shallow a temporary draw-down can be achieved economically and the
dam impounds a low hazard reservoir, removal may be more appropriate than abandonment by filling.
A1
Dams and reservoir conduits 145
Removal is also more appropriate than abandonment for timber conduits, as the timber could rot and
leave voids in the embankment fill in future, even if the conduit has been filled.

The main costs and risks of removing a conduit are noted in Table 10.2.

Table 10.2 Conduit removal – summary

Cost or risk Project element Comments

‰‰ potentially very large water control and earthworks costs

Mobilisation and labour/plant
‰‰ likely to be most effective for shallow conduits with good plant
Construction cost costs access and where draw-down is easily achievable.

Typical material costs ‰‰ cost of additional/replacement materials are minor.

Damage to existing structure ‰‰ significant risk of damaging core.

Construction risk ‰‰ flood overtopping when trench excavated through dam can
Other construction risks
cause breach.
‰‰ complete removal reduces capacity to zero unless a
Capacity reduction
replacement is provided.
‰‰ cost of future maintenance reduced, valves will no longer need
Operational cost
Surveillance/maintenance/ operation, and conduit no longer requires internal inspection
running costs ‰‰ enhanced surveillance required soon after works, which disrupt
core, reduced in the long-term.
Good installation ‰‰ complete removal of conduit hazard.
Operational risk
Poor installation ‰‰ damage to core causing leakage and internal erosion.

Note
See Section 8.1 for guidance on applicability and colour coding.

10.2.1 Removal process

Removal of the conduit should be completed as set out in Section 9.2.2.

In particular, careful management of the excavation and subsequent placement and compaction of
replacement fill is essential. Care is needed to slope the trench sides adequately to avoid significant
arching, which can lead to the formation of voids.

Unlike most conduit replacements, additional fill will be required to fill the volume of the removed
conduit. Additional imported fill should match closely the characteristics of the existing fill. Note that
the opportunity may be taken to add drainage to the downstream shoulder or toe within the trench
footprint.

The removed conduit should be disposed of appropriately. Samples of old conduits may be of interest to
research institutions and universities for material investigation. Masonry units and undamaged modern
pipelines may be reused in other structures.

146 CIRIA, C743

11 Optimising the remedial works
1
This chapter provides guidance on the steps to be taken in the planning of the remedial works.
This includes summary flow charts to assist in the rapid identification of deterioration mechanisms
2
and solutions, and an overview of the cost and risks of various renovation, repair, replacement and

3
abandonment options discussed in this guide.

Identifying the optimum solution to a problem or concern at a reservoir conduit requires several steps to
be taken, in a staged process. A suitable process is set out in Figure 11.1.

Further information on each step is provided in the following sections.

4
Undertake routine maintenance/
Undertake periodical or ad hoc inspection
monitoring or periodical assessment of
of the reservoir

5
conduit structural or hydraulic capacity

Identify problem/need for action

Engage appropriate specialist advice 6

Undertake precautionary action if necessary (see Chapter 7)

Gather further data (see Chapters 4 to 6)

7
Identify deterioration mechanism and failure mode
8
Identify feasible solutions considering:

9
‰‰ quality of available information on reservoir construction and
conduit deterioration
‰‰ hazard posed by reservoir to public
‰‰ current and future needs relating to the conduit.

Rank feasible solutions considering:

‰‰ appropriateness of solution
10
‰‰ construction risk (including environmental risks, safety risks to
construction and operational staff, risk of damage to existing

11
structure, ‘standard’ reservoir risks, unforeseen conditions)
‰‰ construction cost (including temporary loss of reservoir
during drawdown and supply during conduit works)
‰‰ operational cost changes (positive and negative)
‰‰ operational risk (consider reservoir hazard level).

Identify optimum solution

12

Figure 11.1
Identify optimum procurement route

Process for optimum solution identification

A1
Dams and reservoir conduits 147
11.1 IDENTIFYING PROBLEMS
The identification of a problem at a conduit that requires action to be taken may occur during routine
maintenance or monitoring of the structure, or as a result of a periodic risk assessment. Reservoir owners
with thorough routine maintenance and monitoring regimes, and with staff who are trained to monitor
change and recognise hazards, are more likely to identify the need for action early. This can avoid costly
emergency actions, and allow for more time to undertake investigations, carry out preparatory works or
wait for a suitable occasion to carry out operational requirements and optimum weather conditions. It
can also help considerably to deliver the best solution. Further information on inspection procedures,
which can assist in the enhancement of surveillance, can be found in Chapter 4.

Although continuing deterioration should ideally be picked up during routine inspections and dealt
with as part of routine maintenance where possible, it is recognised that deterioration in UK dam
conduits may not always be identified in the timeliest manner for reasons such as difficult access. So,
the identification may be more likely during an inspection of the structure, such as under Section 10
of the Reservoirs Act 1975. This reflects the main value of these inspections, which are undertaken
by independent expert engineers who understand the potential risks posed by conduits, can review
their current condition in comparison to that during previous formal inspections, and may also have
experience in conduit remediation measures.

11.2 ENGAGING APPROPRIATE SPECIALIST ADVICE

Where a significant change has occurred to any part of a dam structure, or the reservoir owner wishes
to undertake improvement works to a dam conduit, an appropriately qualified and experienced
reservoir engineer should be employed. For a conduit through an embankment dam, a significant
change would include:

‰‰ movement or deformation of fill around the conduit or on the dam surface above the conduit
‰‰ movement or deformation of the conduit structure
‰‰ deterioration of parts of the structure that may indicate or cause overloading
‰‰ water flow in the fill around the conduit, which may be apparent through visible flow, enhanced
vegetation growth, or changing pore water pressures (where piezometers are installed)
‰‰ water flow into or through the conduit. New leaks or existing leaks becoming cloudy, and any
evidence that conduit discharge is carrying sediment.

This list is not exhaustive – if there is uncertainty seek advice.

As noted in Section 1.4.2, statutory reservoirs in the UK require

APREs
an appropriate engineer to inspect and supervise the reservoir. A list of APREs is given on the Gov.uk
This person is appointed to one of the reservoir engineer panels, website: http://tinyurl.com/pz4b3q7
typically the supervising engineer panel (though engineers
appointed to any of the other panels can also fulfil this role). The supervising engineer should know
the reservoir well and should be the first point of contact. They will then be able to advise on getting
further assistance, usually an ARPE. These are highly experienced dam engineers. An up-to-date list
of all ARPEs is available from Gov.uk (see ARPEs box).

If advice is sought regarding a non-statutory reservoir, then it should be directly from an ARPE.

Where new works may affect the safety of a statutory reservoir the dam owner should appoint a QCE
to oversee the design and construction of the works. The ARPE who is invited to give advice on the
problem can subsequently provide this service. However, an alternative ARPE can be engaged to do this
if desired. Some engineers will have more experience in dealing with particular types of conduits than
others. The QCE may also be able to help identify a suitable contractor for the works with a proven track
record in carrying out similar works.

148 CIRIA, C743

11.2.1 Emergency actions
Emergency actions will need to be taken if there is a significant risk that the dam may breach, resulting
in an uncontrolled release of water. Chapter 7 contains further information on emergency actions. An 1
ARPE, working with the supervising engineer, is best placed to provide advice on emergency actions.

11.2.2 Gathering further information 2

Where a new defect is reported, a desktop study for the collation and analysis of all available information
relating to the conduit is necessary. Particularly useful items are:

‰‰ construction drawings and records (original and any remedial works)

‰‰ previous inspection reports and supervision statements
3
‰‰ records of reservoir water levels and conduit operation
‰‰ records of leakage flows and settlement above or along the conduit
‰‰ testimony of operators and regular visitors to the dam 4
‰‰ records of geotechnical and other investigations of the site.

Where changes are occurring that are not causing danger of imminent collapse or serious leakage, it is
likely that further information can be gathered. Further site or structural investigations or examinations
can confirm the structural dimensions and the properties of the fabric of the conduit, and also the
5
geotechnical properties of the embankment fill and possibly the geological properties of the dam
foundation (see Chapter 5).

The most valuable additional data is often an enhanced time series of measured flows, levels and movements.
6
In many cases these can be cheap to obtain although they require regular visits to the reservoir by the

7
recording person. This can also help by ensuring regular visual monitoring (see Chapter 6).

11.3 IDENTIFICATION OF DETERIORATION

MECHANISM AND APPROPRIATE SOLUTIONS
Deterioration of conduits occurs due to physical loading or chemical reaction between the conduit
8
materials and air, dam fill, water or materials carried by water. Several deterioration mechanisms may

9
occur at the same time, and exacerbate the problems caused by each other. Common deterioration
mechanisms are discussed in Chapter 3.

Identification of the deterioration mechanism of a conduit is often difficult, due to changes hidden in the
embankment fill, uncertainty about leakage paths and multiple potential reasons for visible movements.
Ideally the engineer would be able to determine the types, locations and extent of both flow and
deterioration in the conduit. Difficult access and the fact that the most dangerous flow is hidden within 10
the embankment fill means that the determination of all of this data is rarely possible.

Although further testing and investigation may help to support a hypothesis about the causes of
deterioration, there is always a level of uncertainty regarding hidden parts of the structure. Sometimes it
is uneconomical to attempt to remove this uncertainty entirely through exhaustive further investigations
11
and monitoring. Indeed if a failure process is taking place time can be of the essence. A decision is

12
needed as to when the limit in value relating to gathering further data has been reached. This again is a
difficult decision, and one for which previous experience is invaluable.

The wide range of conduit loads, shapes and sizes means that sound judgement is necessary to
determine which failure mode is most likely and what extent of the conduit structure is most prone to
failure. So, most works to conduits will involve repair or strengthening of considerable lengths if not
the entire structure. A1
Dams and reservoir conduits 149
11.3.1 Identification of appropriate solutions
Once the most likely deterioration mechanisms and failure modes are known, a list of appropriate
solutions can be identified. At this point all feasible solutions that can improve long-term integrity of the
structure to an adequate extent should be listed. These can then be ranked considering costs and risks as
discussed in Section 11.4.

11.3.2 Summary flowcharts

The following two flowcharts assist in the rapid identification of deterioration mechanisms and
solutions. They relate to buried pressurised pipes (see Figure 11.2) and buried human-entry conduits
(see Figure 11.3). They cannot consider site-specific issues that are likely to affect decisions relating to
actual reservoir conduits and so should not be considered to be applicable to all situations. The need to
obtain suitable advice from experienced engineers is vital.

Note the flowcharts include preference for solutions that will reduce the long-term risk associated with a
conduit. The ranking of feasible solutions is discussed in detail in Section 11.4.

150 CIRIA, C743

 1
2
3
4
5
6
7
8
9
10
11
Figure 11.2 Decision flowchart for buried conduit normally containing water under pressure
12
A1
Dams and reservoir conduits 151
Figure 11.3 Decision flowchart for non-pressurised conduit (eg human-entry culvert, spillway pipe)

152 CIRIA, C743

11.4 RANKING FEASIBLE SOLUTIONS

11.4.1 Cost and risk summary

1
Table 11.1 provides an initial, first order summary of the cost and risk of the various renovation,
repair, replacement and abandonment options discussed in this guide. This table should only be used
to compare options within the same category, eg renovation/repair, replace or abandon, and decisions
should be made separately.
2
3
Table 11.1 Cost and risk – comparison

Renovation/repair Replace Abandon

same alignment
Cured-in-place

In situ sprayed

Replace with

Replace with
4
Sliplining

Replace on
Sliplining

Close fit

Remove
siphons
pumps

Block
Mobilisation and labour/plant costs

5
Construction cost
Typical material costs

Damage to existing structure

Construction risk
Other construction risks

Operational cost
Capacity reduction

Surveillance/maintenance/running costs 6
Good installation
Operational risk

7
Poor installation

The impact of each main cost or risk type has been colour coded as follows:

‰‰ red – this solution has a major disadvantage or cost/risk implication

‰‰ orange – medium disadvantage or cost/risk implication
‰‰ yellow – minor disadvantage or cost/risk implication
8
‰‰ green – advantageous solution with cost or risk benefit.

These have been based on the consideration of ‘typical’ UK sites, as follows:

‰‰ buried pipe – 300 mm to 450 mm ID pressurised CI or DI pipes, or 600 mm ID concrete/brick

9
spillway pipe, buried in a low, old embankment dam, either homogenous or with central core

10
‰‰ pipe in culvert – 300 mm to 900 mm ID pressurised CI or DI pipe in a masonry culvert in an old
embankment dam with clay core
‰‰ human-entry culvert – more than 2000 mm nominal diameter circular or horseshoe/modified
horseshoe section masonry or concrete culvert containing pipework, with space for human-entry

11
and access from the downstream end.

However, many conduits will deviate from these typical arrangements and all sites will have conduits
that, due to dimensions or materials, require bespoke solutions to problems.

Note the costs and risks considered in Table 11.1 relate only to the technical solution, and do not
consider the priorities of the reservoir owner, the hazard posed by the dam, or the quality of information 12
available to the project team. The implication of these on construction and operation costs and risks are
discussed in the following sections. Various weightings on each cost/risk may be applicable on a case-by-
case basis.
A1
Dams and reservoir conduits 153
11.4.1.1 Construction cost
The assessment of the construction cost does not take into account the capabilities of the reservoir
owner. Costs associated with the procurement of works can be significant, and the capabilities of the
reservoir owner’s maintenance team should also be considered. The cost of construction of a method,
which is within the skills base of the maintenance team, is likely to be significantly lower than the cost of
procuring an outside team with the necessary skills.

11.4.1.2 Construction risk

The construction risk will vary from dam to dam based on the quality of information available on the
reservoir construction. For many older embankment dams, details of the dam foundation, fill and core
may be unavailable and as such the conditions that will be encountered during construction may be
unknown. This increases the risks, especially those associated with options that require excavation into
the embankment.

The level of risk is also determined by the quality and extent of information available regarding the
condition of the existing pipe and of the embankment. The construction risk attributable to renovation
of pipes can be lowered by carrying out a high quality CCTV survey, or visual survey (for human-entry
culverts) of the pipe, which adequately identifies the level of water ingress into the conduit as well as
any deformation along its length. However, CCTV surveys may not adequately identify the structural
integrity of the conduit.

Where delays in, or failure of, construction could result in loss of some public service (such as water
supply), or lead to failure of a dam where flooding could result in the loss of life, greater emphasis should
be placed on consideration of the construction risks.

For remedial works at impounding reservoirs that require draw-down of the reservoir, construction
flood risk will often be a significant risk to mitigate.

Any repairs requiring the unusual loading of the dam crest for construction plant access or operation
should consider the possible effect of the loading on slope stability.

11.4.1.3 Operational cost (and benefit)

Conduit repairs may lead to changes in conduit dimensions and/or hydraulic roughness that may affect
reservoir operation. In some cases the changes may lead to functionality having to be supplemented by
additional facilities such as mobile pumps, which would increase operational costs.

The requirements for monitoring and surveillance vary with the type and condition of conduits
and evaluation of risk through statutory inspections. Repairs or replacement of conduits will affect
operational costs associated with monitoring and surveillance.

11.4.1.4 Operational risk

The level of tolerable operational risk relates to the current and future requirements of the conduit.
Where the conduit provides the means for emergency draw-down or acts as an important water supply,
only a low level of operational risk may be considered tolerable. Critical conduits may need preventative
maintenance in the short-term to allow improvement works to be planned for the longer term. This will
reduce the disruption that may occur from progressive deterioration and incidents.

154 CIRIA, C743

12 During and after works
1
This chapter provides details of the as-built record to be kept, ongoing monitoring and surveillance, and
the importance of post-incident reviews and lessons that can be learnt.
2
12.1 AS-BUILT RECORDS 3
On completion of any remedial works, full details of the works should be retained in the dam’s asset files
for future reference. This does not just extend to that information required in compliance with current
health and safety regulations, but includes all works no matter how minor. Such information would include:

‰‰ findings of the desktop study before works being completed

4
‰‰ historical drawings or data uncovered in the desktop study

5
‰‰ ground investigation records from recent investigations
‰‰ any design report including analyses
‰‰ construction records
‰‰ records of decisions made to amend the design during the construction works
‰‰ material performance records and Control of Substances Hazardous to Health (COSHH)
statements (HSE, 2002)
6
‰‰ photographs of works progressing

7
‰‰ as-built drawings
‰‰ details of monitoring installations or procedures
‰‰ records of all maintenance and repairs.

8
It is crucial that changes made on site due to unforeseen conditions are recorded within the as-built
drawings. Where changes are made because the hidden elements of the existing structure were not as
expected, this information should be recorded and retained for review when works to other parts of the
dam structure outside the scope of the current remedial works are being considered.

Unexpected or unusual ground or groundwater conditions should also be recorded on the as-built records.
9
12.2 MONITORING AND SURVEILLANCE
Where appropriate, it is important to have a monitoring and surveillance system in place to continuously or
periodically measure and record the performance of the repair or alteration. This is particularly the case
10
for when the performance of a repair or alteration to the structural elements of a conduit is uncertain, such

11
as if a new technique or material has been used, or a known one is used in new circumstances.

This may involve a periodic visual check carried out in the course of routine inspections or the long-
term installation of monitoring instrumentation. Procedures should be set up to capture and review this

12
information and make use of it for immediate and future needs (see Chapter 6).

12.3 LESSONS LEARNED

Lessons can be learned from all construction works, but are rarely shared widely enough to prevent costly errors
being repeated. This is particularly the case for emergency or responsive works, where urgency prevents both a
thorough review of previous works before implementation, and careful record keeping during the works.
A1
Dams and reservoir conduits 155
The phrase ‘lessons learned’ can often focus the minds of engineers on the purely technical lessons, such
as unforeseen conditions or problems during construction that were not accounted for in the design, or
the sheer technical challenges of construction in difficult and bespoke circumstances. However, many
lessons may also relate to the earlier decisions in the planning and procurement of the project.

The UK dam engineering community frequently shares lessons learned through the publication of
papers in the journals and conference proceedings of the British Dam Society (BDS) and the Institution
of Civil Engineers (ICE). Particularly interesting schemes are often presented at BDS evening meetings,
and now most are recorded for access to members on the BDS website. Published papers relevant to
British dams are available from Charles and Tedd (1996).

When an incident occurs at a dam, this should be reported to the Environment Agency, as is prescribed
in the Reservoirs Act 1975 following amendment by the Flood and Water Management Act 2010. This
allows incidents to be classified and recorded in the National Database to inform research projects. The
Environment Agency also provides annual reviews on post-incident reporting and investigation.

Some relevant questions that may be suitable for a post-project or post-incident review are set out as
follows. Note that these should be asked with a ‘no blame’ overtone.

For operation:

‰‰ Would more vegetation management/surveillance/operator training during the reservoir operation

(before the start of the project) have allowed the problem to be identified earlier, and would this
have led to savings in the project?
‰‰ Should the vegetation management/surveillance at this or other reservoirs be changed in future?
‰‰ Should operator training be adjusted?
‰‰ To what extent was instrumentation useful in identifying the problem?

At planning stage:

‰‰ Was an effective level of investigation undertaken to define the problem within a reasonable
programme? Was the investigation technique successful?
‰‰ Was sufficient time and action taken to consider all feasible solutions and possible future operating
changes?
‰‰ Was the procurement of the design and construction of the works carried out in the most effective
way? Could further overlap of activities have helped make the project cheaper without increasing
risk to an inappropriate level?
‰‰ Was there any opportunity to undertake other works at the site, and was this incorporated into the
works package effectively?
‰‰ Was the programme driven effectively to avoid winter working (where this is important)?

Performance:

‰‰ Were the design deliverables of sufficient detail to allow works to progress without delay caused by
seeking additional detail?
‰‰ Were technical and safety risks effectively communicated around the project team to allow
management? (Did somebody know of a risk and not tell others?)
‰‰ Was the reservoir isolation/draw-down plan effective?
‰‰ Were the access constraints understood by the entire project team?
‰‰ Were environmental and weather risks adequately managed during construction?
‰‰ Was safety managed proactively throughout the project delivery?
‰‰ What new monitoring and surveillance measures have been put in place following the project to
monitor performance?
‰‰ Are there any aspects of the project that should be shared with the UK reservoir community to
make improvements in reservoir risk management or research?

156 CIRIA, C743

References

ABDOUN, T, BENNETT, V, DESROSIERS, T, SIMM, J and BARENDSE, M (2013) “Asset management

and safety assessment of levees and earthen dams through comprehensive real-time field monitoring”
Geotechnical and Geological Engineering, vol 31, 3, Springer, UK, pp 833–843

ALMOG, E, KELHAM, P and KING, R (2011) Modes of dam failure and monitoring and measuring
techniques, SC080048/R1, Environment Agency, Bristol, UK (ISBN: 978-1-84911-230-7).
Go to: http://tinyurl.com/mm724ky (accessed 24/06/2015)

ANDREWS, M E (1998) “Large diameter sewer condition assessment using combined sonar and CCTV
equipment”. In: Proc APWA International Public Works Congress and Exhibition, NRCC/CPWA seminar series,
Innovations in urban infrastructure, Las Vegas Convention Center, Las Vegas, USA, 14–17 September 1998,
pp91–103. Go to: http://tinyurl.com/o2gef6y (accessed 24/06/2015)

ANDREWS, M E and DORNSTÄDTER, J (2000) “Investigations into seepage at Rotton Park Reservoir
using temperature distribution measurement”. In: P Tedd (ed) Dams 2000, British Dams Society,
Institution of Civil Engineers, London, UK (ISBN: 978-0-7277-2870-8) pp 388–401.
Go to: http://www.icevirtuallibrary.com/content/book/100249 (accessed 24/06/2015)

ASCE (2000) Guidelines for instrumentation and measurements for monitoring dam performance, American
Society of Civil Engineers, Reston VA, USA (ISBN: 978-0-78440-531-4)

ATKINS (2007) Reservoir safety research and development strategy, final report, Flood and Coastal Erosion
Risk Management Research and Development Programme, Department for the Environment, Food and
Rural Affairs, London, UK

ATHERTON, D L (2000) Electromagnetic method for non-destructive testing of prestressed concrete pipes for broken
prestressing wires, Patent US6127823 A, USA. Go to: www.google.com/patents/US6127823#backward-citations

ATHERTON, D L, TOAL, C J and SCHMIDT, T R (1989) “Investigations of the remote field eddy
current technique in large diameter pipeline” British Journal of Non-Destructive Testing, vol 31, 9, British
Institute of Non-Destructive Testing, Northampton, UK, pp 485–488

AURIAU, L, MÉRIAUX, P, LACOMBE, S, MARMU, S, MAURIN, J and BOULAY, A (2013) “The

airborne high-resolution LiDAR as an efficient tool for topographical survey and detection of surface
anomalies on good protection dike”. In: Proc second European conf on flood risk management: science,
policy and practice, closing the gap, 20–22 November 2012, Rotterdam, The Netherlands. F Klijn and T
Schweckendiek (eds) Comprehensive flood risk management: research for policy and practice, CRC Press, UK
(ISBN: 978-0-41562-144-1) pp 104–105

BASSETT, R H, KIMMANCE, J P and RASMUSSEN, C (1999) “An automated electrolevel deformation

monitoring system for tunnels” Proceedings of the ICE – Geotechnical Engineering, vol 137, 3, Institution of
Civil Engineers, London, UK, pp 117–125

BASSETT, R H (2011) A guide to field instrumentation in geotechnics: principles, installation and reading,
Taylor & Francis Group, Florida, USA (ISBN: 978-0-41567-537-6)

BATEMAN, J F L (1884) History and description of the Manchester waterworks, T J Day, London, UK (ISBN:
978-1-23130-351-1)

BECK, Y L, KHAN, A A, CUNAT, P, GUIDOUX, C, ARTIÈRES, O, MARS, J and FRY, J J (2010)

Thermal monitoring of embankment dams by fiber optics”. In: Proc eighth ICOLD European club
symposium, dam safety – sustainability in a changing environment (IECS 2010) Innsbruck, Austria, 22–23
September 2010, pp 444–448. Go to: http://tinyurl.com/km6aflg (accessed 24/06/2015)

Dams and reservoir conduits 157

BETH, M and OBRE, X (2005) “Ground movement monitoring at King’s Cross Station, London” Proceedings
of the ICE – Geotechnical Engineering, vol 158, 3, Institution of Civil Engineers, London, pp 125–133

BINNIE, G M (1981) Early Victorian water engineers, Thomas Telford, London, UK (ISBN: 978-0-72770-128-2)

BRE (2000) Corrosion of steel in concrete – investigation and assessment, part 2, BRE Digest 444, BRE Press,
Building Research Establishment, Bracknell, UK

BROAD, R (2000) “Construction of a cement-bentonite slurry trench cut-off at Pebley dam”. In: Proc of
the eleventh biennial British Dam Society conference, University of Bath, UK, 14–17 June. In: P Tedd (ed)
Dams 2000, Thomas Telford, London (ISBN: 0-72772-870-9) pp 417–424

BROWN, A J (2009) “Low level outlets 2: risk assessment” Dams & Reservoirs, vol 19, 1, ICE Publishing,
Institution of Civil Engineers, London, UK pp 11–14

BROWN, D H (2010) “Elton reservoir visit” Dams & Reservoirs, vol 20, 1, ICE Publishing, Institution of
Civil Engineers, London, UK, pp 5–7

BROWN, D H (2012) “Recent incidents at canal reservoirs”. In: Proc of the seventeenth biennial British Dam
Society conference, University of Leeds, UK, 12–15 September 2012. In: A Pepper (ed) Dams: engineering in a
social and environmental context, Institution of Civil Engineers, London (ISBN: 978-0-72775-799-9) pp 501–510

BROWN, A J and GOSDEN, J D (2000) “Safety issues at small reservoirs”. In: Proc of the eleventh biennial
British Dam Society conference, University of Bath, UK, 14–17 June. P Tedd (ed) Dams 2000, Thomas
Telford, London, UK (ISBN: 0-72772-870-9) pp 159–172

BROWN, A J and GOSDEN, J D (2004) Interim guide to quantitative risk assessment for UK reservoirs,
Thomas Telford, London, UK (ISBN: 978-0-7277-3267-5)

BROWN, D H and HOWLETT, P (2011) “Recent canal reservoir improvements in north-west of England”
Dams & Reservoirs, vol 21, 3, ICE Publishing, Institution of Civil Engineers, London, UK, pp 113–122

BRUGGEMANN, D A and FRANCIS, O J (2014) “Remedial grouting at Shon Sheffrey Dam, Wales”. In:
Proc 18th Biennial British Dam Society Conference: Maintaining the safety of our dams and reservoirs, Queens
University, Belfast, 3–6 September 2014

BURGESS, T (2004) Investigation of Fli-Map system for flood defence asset monitoring, R&D Technical Report
W5A-059/1/TR, Defra/Environment Agency Flood and Coastal Defence R&D Programme, Environment
Agency, Bristol, UK (ISBN: 1-85705-885-2). Go to: http://tinyurl.com/knj99bj (accessed 24/06/2015)

BUSSELL, M N (1997) Appraisal of existing iron and steel structures, Steel Construction Institute, Berkshire,
UK (ISBN: 1-85942-009-5)

CARINO, N J (2001) “The impact-echo method: an overview”. In: Proc 2001 Structures Congress &
Exposition, 21–23 May 2001, American Society of Civil Engineers, Reston VA, USA, pp 1–18

CBDG (2002) Guide to testing and monitoring the durability of concrete structures, Technical Guide 2,
Concrete Bridge Development Group, Surrey, UK

CHARLES, J A and TEDD, P (1996) Bibliography of British dams: a companion report to the register of British
dams, BR310, BRE Press, Building Research Establishment, Bracknell, UK

CHARLES, J A, TEDD, P, HUGHES, A K and LOVENBURY, H T (1996) Investigating embankment dams:

a guide to the identification and repair of defects, BR303, BRE Press, Building Research Establishment,
Bracknell, UK (ISBN: 1-86081-069-1)

CHARLES, J A, TEDD, P and WARREN, A L (2014) Lessons from incidents at dams – an engineering guide,
SP167, CIRIA, London, UK (ISBN: 978-86017-735-7). Go to: www.ciria.org

CONCRETE SOCIETY (1991) Patch repair of reinforced concrete subject to reinforcement corrosion, TR38,
The Concrete Society, Surrey, UK

158 CIRIA, C743

CONCRETE SOCIETY (2000) Diagnosis of deterioration in concrete structures. Identification of defects, evaluation
and development of remedial action, TR54, The Concrete Society, Surrey, UK (ISBN: 978-0-94669-181-4)

CONCRETE SOCIETY (2004) Design guidance for strengthening concrete structures using fibre composite
material, second edition, TR55, The Concrete Society, Surrey, UK

CONCRETE SOCIETY (2005) Mortars for masonry: Guidance on Specification, types, production and use,
Publication 159, The Concrete Society, Surrey, UK

CONCRETE SOCIETY (2014) Analysis of hardened concrete: a guide to test procedures and interpretation of
results, second edition, TR32, The Concrete Society, Surrey, UK (ISBN: 978-1-90448-282-6)

CONTRERAS, I A and HERNÁNDEZ, S H (2010) “Techniques for prevention and detection of leakage
in dams and reservoirs”. In: Proc 30th Annual USSD Conference, collaborative management of integrated
watersheds, Sacramento, California, 12–16 April 2010, United States Society on Dams, Denver, USA
(ISBN: 978-1-88457-551-8) pp 785–814

DE LAFONTAINE, Y and COSTAN, G (2000) The unfolding story of the Zebra mussel in the St. Lawrence
River, fact sheet, Environment Canada, St Lawrence Centre, Gatineau QC, Canada

DEFRA (2011) Invasive species to look out for, GB non-native species secretariat (NNSS) York, UK. Go to:
http://tinyurl.com/o4r8429 (accessed 24/06/2015)

DORNSTÄDTER, J (1997) “Detection of internal erosion in embankment dams”. In: Transactions of

the nineteenth international congress on large dams, Florence, Italy, 26–30 May 1997, vol 2, Q.73R.7.2,
International Commission on Large Dams (ICOLD) Paris, France, pp 87–101

DORNSTÄDTER, J (2013) “Leakage detection in dams – state of the art”. In: Proc int symp on dam
engineering in Southeast and Middle Europe – recent experience and future outlooks, Ljubliana, Slovenia, 16
October 2013, pp 77–86. Go to: Downloaded from http://slocold.si/symp20years/110_Z15_77-86_
Dornstaedter.pdf (accessed 24/06/2015)

DORNSTÄDTER, J, DUTTON, D, FABRITIUS, A and HEDINGER, P (2006) “Is internal erosion

detectable”. In: Hewlett H (ed) Improvements in reservoir construction, operation and maintenance proceedings
of 14th conference of the British Dam Society, University of Durham, 6–9 September 2006, Thomas Telford,
London, UK (ISBN: 978-0-72773-470-9) pp 129–143

DUFFY, M, HILL, C, WHITAKER, C, CHRZANOWSKI, A, LUTES, J and BASTIN, G (2001) “An

automated and integrated monitoring program for Diamond Valley Lake in California”. In: Proc of the
10th FIG int symp on deformation measurements, Orange, California, USA, 19–22 March 2001, pp 1–23.
Go to: http://tinyurl.com/qajt4mx (accessed 24/06/2015)

DUNNICLIFF, J (1988) Geotechnical instrumentation for monitoring field performance, John Wiley and Sons,
New York, USA (ISBN: 978-0-47100-546-9)

DUNNICLIFF, J (1993) Geotechnical instrumentation for monitoring field performance, John Wiley and Sons,
New York, USA (ISBN: 978-0-47100-546-9)

DUNNICLIFF, J (2012) “Types of geotechnical instrumentation and their usage”. In: Burland, J,
Chapman, T, Skinner, H Brown, M (eds) ICE Manual of geotechnical engineering, Chapter 95, ICE
Publishing, Institution of Civil Engineers, London, UK (ISBN: 978-0-72773-652-9)

DURAN, O, ALTHOEFER, K and SENEVIRATNE, L D (2002) “State of the art in sensor technologies
for sewer inspection” IEEE Sensors Journal, vol 2, 2, Institute of Electrical and Electronic Engineers, New
York, USA, pp 73–81

DUTTON, D (2001) “British Waterways’ dams – the last 25 years” Dams & Reservoirs, vol 11, 2, ICE
Publishing, Institution of Civil Engineers, London, UK, pp 19-27

EA (in press) Determining the optimum hydraulic capacity of low level outlets for reservoirs, SC130001,
Environment Agency, Bristol, UK. Go to: http://tinyurl.com/l8s4ejg (accessed 24/06/2015)

Dams and reservoir conduits 159

EA (2013) How to comply with your environmental permit, Document 43311 (version 6) Environment Agency,
Bristol, UK. Go to: http://tinyurl.com/kckjunm (accessed 24/06/2015)

EA/NIEA/SEPA (2007) Pollution Prevention guidelines (PPG) 5 Works and maintenance in or near water,
Environment Agency, Bristol, UK. Go to: http://tinyurl.com/pnoonuh (accessed 24/06/2015)

EA/NIEA/SEPA (2012) Pollution Prevention guidelines (PPG) 6 Working at construction and demolition sites,
second edition, Environment Agency, Bristol, UK. Go to: http://tinyurl.com/nand3q4 (accessed 24/06/2015)

EARP, D N W (1995) “Reservoir outlet works in the mid-1990s” Water and Environmental Management, vol
9, S1, Wiley Online, UK, pp 7–14

EDGELL, G (ed) (2005) “Testing of ceramics in construction” Testing in Construction Series, vol 2, Whittles
Publishing (ISBN: 978-1-870325-43-1)

ÉKES, C and NEDUCZA, B (2012) “Pipe condition assessments using pipe penetrating radar”. In: Proc
14th int conf on ground penetrating radar, 4–8 June 2012, IEEE, New York, USA (ISBN: 978-1-46732-662-9)
pp 840–843

ESR TECHNOLOGY (2005) Paints/polymeric coatings for corrosion protection, Water Industry Mechanical and
Electrical Specification (WIMES) 4.01, ESR Technology, Warrington, UK. Go to: http://tinyurl.com/kupjt9a
(accessed 24/06/2015)

FELL, R, MACGREGOR, P, STAPLEDON, D and BELL, G (2005) Geotechnical engineering of dams,

Taylor & Francis Group LLC, New York, USA (ISBN: 978-0-41536-440-9)

FEMA (2005) Technical manual: conduits through embankment dams. Best practices for design, construction,
problem identification and evaluation, inspection, maintenance, renovation, and repair, Federal Emergency
Management Authority, Washington, DC, USA. Go to: http://tinyurl.com/q2ywnh9 (accessed 24/06/2015)

FEMA (2007) Technical manual: plastic pipe used in embankment dams: best practices for design, construction,
problem identification and evaluation, inspection, maintenance, renovation, and repair, Federal Emergency
Management Authority, Washington, DC, USA. Go to: http://tinyurl.com/n3998ag (accessed 24/06/2015)

FERC (2012) “Dam safety performance monitoring program” Engineering guidelines for the evaluation of
hydropower projects, Chapter 14, Federal Energy Regulatory Commission, Washington DC, USA. Go to:
http://tinyurl.com/meega2x (accessed 24/06/2015)

FOSTER, M, FELL, R and SPANNAGLE, M (1998) “Risk assessment – estimating the probability of
failure of embankment dams by piping”. In: Proc of the 1998 ANCOLD/NZSOLD conference on dams,
University of New South Wales, Sydney, Australia, 29 August 29 to 4 September 1998

FUGRO (2013) Fugro assists with rigorous tunnel checks at Glendoe Hydro Electric Scheme, UK, Fugro Aperio
Limited, Cambridge, UK. Go to: www.fugro-aperio.com/news/13_02_10_glendoe.html (accessed 24/06/2015)

GARDINER, K D, HUGHES, A K and BROWN, A J (2004) “Lessons from an incident at Upper

Rivington reservoir – January 2002” Dams & Reservoirs, vol 14, 2, ICE Publishing, Institution of Civil
Engineers, London, UK, pp 17–20

GIBSON, S P (1992) “Gorpley reservoir tunnel grouting” Dams and Reservoirs, vol 2, 2, ICE Publishing,
Institution of Civil Engineers, London, UK, p 31–35

GORZELANCZYK, T, HOŁA, J and SCHABOWICZ, K (2012) “Non-destructive tests aimed at

determining the thickness of the concrete shell of a heat pipe carrying tunnel”. In: Proc 18th World
conference on non-destructive testing, Durban, South Africa, 16–20 April 2012, pp 1–8.
Go to: www.ndt.net/article/wcndt2012/papers/211_wcndtfinal00211.pdf (accessed 24/06/2015)

GRENERCZY, G and WEGMÜLLER, U (2011) “Persistent scatterer interferometry analysis of the

embankment failure of a red mud reservoir using ENVISAT ASAR data” Natural hazards, vol 59, 2,
Springer, UK, pp 1047–1053

160 CIRIA, C743

HAMMOND, M (1981) Bricks and brick-making, second revised edition, Shire Publications,
Buckinghamshire, UK (ISBN: 978-0-74780-067-5)

HAO, T, ROGERS, C D F, METJE, N, CHAPMAN, D N, MUGGLETON, J M, FOO, K Y, QWANG,

P, PENNOCK, S R, ATKINS, P R, SWINGLER, S G, PARKER, J, COSTELLO, S B, BURROW, M
P N, ANSPACK, J H, ARMITAGE, R J, COHN, A G, GODDARD, K, LEWIN, P L, ORLANDO, G,
REDFERN, M A, ROYAL, A C D and SAUL, A J (2012) “Condition assessment of the buried utility service
infrastructure” Tunnelling and Underground Space Technology, vol 28, March, Elsevier BV, UK, pp 331–344

HARRIS, R and DOBSON, C (2006) “Sewer pipe infiltration assessment: comparison of electro-scan,
joint pressure testing, and CCTV inspection”. In: Proc ASCE speciality conference. Pipelines 2006: service to
the owner, Chicago, Illinois, USA, 30 July to 2 August 2006, American Society of Civil Engineers, Reston
VA, USA, pp 1–10

HSE (2012) Control of Substances Hazardous to Health (COSHH) Health and Safety Executive, London, UK.
Go to: www.hse.gov.uk/nanotechnology/coshh.htm (accessed 24/06/2015)

HSE (2014) Safe work in confined spaces. Confined Spaces Regulations 1997. Approved code of practice,
regulations and guidance, Health and Safety Executive, London, UK (ISBN: 978-0-71766-622-5). Go to:
www.hse.gov.uk/pubns/priced/l101.pdf (accessed 24/06/2015)

HUDSON, M and HODGSON, J (2009) “Ground penetrating radar and sub-surface laser scanning – the
full 3D picture” Civil Engineering Surveyor, vol XX, Chartered Institution of Civil Engineering Surveyors,
Cheshire, UK, pp 22–23

ICOLD (1994) Embankment dams – granular filters and drains, Bulletin 95, International Congress on
Large Dams, Paris

ICOLD (2014) Internal erosion of existing dams, levees and dikes, and their foundations, Bulletin 164,
International Congress on Large Dams, Paris

ICE (1996) Floods and reservoir safety, third edition, Institution of Civil Engineers, London (ISBN: 978-0-
72772-503-5)

JOHANSSON, S and SJÖDAHL, P (2009) A guide to temperature measurements for seepage investigation and
monitoring of embankment dams, T062700 0214, Centre for Energy Advancement through Technological
Innovation, Montreal, Canada. Go to: www.ceati.com/publication-details?pid=0214 (accessed 24/06/2015)

JOHANSSON, S (2004) “Detection of internal erosion in embankment dams–possible methods in theory and
practice”. In: Proc int sem on stability and breaching of embankment dams, Oslo, Norway, 21–22 October 2004

KASTNER, R, JAMES, R and KJEKSTAD, O (eds) (2003) Avoiding damage caused by soil-structure
interaction: lessons learnt from case histories, ICE Publishing, Institution of Civil Engineers, London, UK
(ISBN: 978-0-72773-124-1)

KENNARD, M R, OWENS, L and READER, R A (1996) Engineering guide to the safety of concrete and
masonry dam structures in the UK, R148, CIRIA, London (ISBN: 978-0-86017-432-5). Go to: www.ciria.org

KOERNER, R M, MCCABE, W M and LORD, A E (1978) “Acoustic emission monitoring of soil stability”
Journal of the Geotechnical Engineering Division, vol 104, 5, American Society of Civil Engineers, Reston
VA, USA, London, pp 571–582

KOFOED, V O, GARDINER, K D and GEORGE, A A (2008) “Locating the leakage route at Torside
Reservoir using the Willowstick AquaTrack system”. In: Proc of the fifteenth biennial British Dam Society
conference, University of Warwick, 10–13 September 2008, British National Committee on Large Dams,
Institution of Civil Engineers, London, pp 1–12. Go to: http://tinyurl.com/qc5rze8 (accessed 24/06/2015)

KOFOED, V, JESSOP, M and WALLACE, M (2014) “Investigating seepage at the Bartley dam”, UK,
Hydropower and Dams, vol 4, Aqua-Media International Ltd, Wallington, Surrey, UK, pp 2–4. Go to:
http://tinyurl.com/nu84t97 (accessed 24/06/2015)

Dams and reservoir conduits 161

KONTOGIANNI, V and STIROS, S (2003) “Tunnel monitoring during the excavation phase: 3D kinematic
analysis based on geodetic data”. In: Proc 11th FIG symp on deformation measurements, Nomicos Conference Centre,
Santorini Island, Greece, 25–38 May 2003, S C Stiros and S Pytharouli (eds) publication 2, Department of Civil
Engineering, Patras University, Patras. Go to: http://tinyurl.com/oou969o (accessed 24/06/2015)

LILLIE, K, REED, C and RODGERS, M (2006) Workshop on condition assessment inspection devices for water
transmission mains, American Water Works Association Research, Denver, USA (ISBN: 978-1-84339-916-2)

LIU, Z and KLEINER, Y (2013) “State of the art review of inspection technologies for condition
assessment of water pipes” Measurement, vol 46, 1, Elsevier BV, UK, pp 1–15

LIU, Z, KLEINER, Y, RAJANI, B, WANG, L and CONDIT, W (2012) Condition Assessment technologies
for water transmission and distribution systems, EPA/600/R-12/017, US Environmental Protection Agency,
Washington DC, USA, Go to: http://nepis.epa.gov/Adobe/PDF/P100E3Y5.pdf (accessed 24/06/2015)

LONG, G, MAWDESLEY, M, SMITH, M and TAHA, A (2012) Performance based inspection of flood defence
infrastructure: Integrating visual inspection and quantitative survey measurements, Research Report SWP4.2,
Flood Risk Management Research Consortium, Nottingham University, UK. Go to: http://tinyurl.com/
npumbn8 (accessed 24/06/2015)

LONG, G, MAWDESLEY, M, SMITH, M and TAHA, A (2013) Quantitative assessment methods for the
monitoring and inspection of flood defences: new techniques and recent developments, C717, CIRIA, London, UK
(ISBN: 978-0-86017-720-3). Go to: www.ciria.org

MAKAR, J M and RAJANI, B. (2000) “Gray cast-iron water pipe metallurgy” Journal of Materials in Civil
Engineer, vol 12, 3, American Society of Civil Engineers, Reston VA, USA, pp 245–253

MALHOTRA, V M and CARINO, N J (eds) (2004) Handbook of non-destructive testing of concrete, second
edition, CRC Press LLC. Boca Raton, Florida, USA (ISBN: 978-0-84931-485-8)

MARSHALL, J and FISK, P (2010) “Prestressed concrete cylinder pipe decay” Pipelines 2010, American
Society of Civil Engineers, Reston VA, USA, pp 548–557

MARTELL, K, FEENEY, C and TUCCILLO, M E (2011) Field demonstration of condition assessment

technologies for wastewater collection systems, Report no EPA/600/R-11/078, US Environmental Protection
Agency, Washington DC, USA. Go to: nepis.epa.gov/Adobe/PDF/P100C8E0.pdf (accessed 24/06/2015)

MCKIBBINS, L, ELMER, R and ROBERTS, K (2009) Tunnels: inspection, assessment and maintenance,
C671, CIRIA, London, UK (ISBN: 978-086017-671-8). Go to: www.ciria.org

MOFFAT, A I B (1975) “Problems of embankment dams and remedial measures”. In: Proc BNCOLD
Symposium, Discussion: technical session 5, D5/7, 24–26 September 1975. A I B Moffat (ed) Inspection,
operation and improvement of existing dams, Department of Civil Engineering, University of Newcastle

NICHOLSON, D, TSE, C-M and PENNY, C (1999) The Observational Method in ground engineering:
principles and applications, R185, CIRIA, London, UK (ISBN: 978-0-86017-497-4). Go to: www.ciria.org

NCE (2012) No 32 Heathrow Express collapse, November 2012, New Civil Engineer, London (online). Go to:
http://tinyurl.com/pn73v5x (accessed 24/06/2015)

PARKINSON, G and ÉKES, C (2008) “Ground penetrating radar evaluation of concrete tunnel linings”.
In: Proc 12th int conf on ground penetrating radar, Birmingham, UK, 16–19 June 2008. Go to: http://
tinyurl.com/mrg658q (accessed 24/06/2015)

PECK, R B (1969) “Advantages and limitations of the observational method in applied soil mechanics”
Geotechnique, vol 19, 2, ICE Publishing, Institution of Civil Engineers, London, UK, pp 171–178

PECK, R B (1972) “Observation and instrumentation: some elementary considerations” Highway Focus,
vol 4, 2, Transport Research Board, Washington DC, USA, pp 1–5

162 CIRIA, C743

PENMAN, A D M and CHARLES, J A (1975) Deformation and stresses in rockfill over a rigid culvert, CP
100/75, BRE Press, Building Research Establishment, Bracknell, UK

PETHER, R and ROWLAND, A (2009) Scoping study on the need for additional research and/or guidance on
reservoir conduits, Report SC080049/R1, Environment Agency, Bristol, UK (ISBN: 978-1-84911-148-5). Go
to: http://tinyurl.com/oup6dqz (accessed 24/06/2015)

RAILTRACK (1996) Guide tunnel management, Network Rail, UK (internal guidance document, not in
the public domain)

RAJANI, B (ed) (2000) Investigation of grey cast iron water mains to develop a methodology for estimating service
life, American Water Works Association, Denver, USA (ISBN: 978-1-58321-063-5)

RAUCOULES, D, COLESANTI, C and CARNEC, C (2007) “Use of SAR interferometry for detecting
and assessing ground subsidence” Comptes Rendus – Geoscience, vol 339, 5, Elsevier BV, UK, pp 289–302

RAWLINGS, C G, HELLAWELL, E E and KILLKENNY, W M (2000) Grouting for ground engineering,

C514, CIRIA, London, UK (ISBN 978-0-86017-514-8). Go to: www.ciria.org

READER, R A, KENNARD, M F and HAY, J (1997) Valves, pipework and associated equipment in dams –
guide to condition assessment, R170, CIRIA, London, UK (ISBN: 978-0-86017-469-1). Go to: www.ciria.org

RIGBY, P J and GARDINER, K D (2012) “Remedial grouting at Upper Rivington Reservoir”. In: Proc of
the seventeenth biennial British Dam Society conference, University of Leeds, UK, 12–15 September 2012. A
Pepper (ed) Dams: engineering in a social and environmental context, Institution of Civil Engineers, London,
UK (ISBN: 978-0-72775-799-9) pp 489–500

SANSALONE, M and CARINO, N J (1989) “Detecting delaminations in concrete slabs with and without
overlays using the impact-echo method” Materials Journal, vol 86, 2, American Concrete Institute, MI,
USA, pp 175–184

SANSALONE, M J and STREETT, W B (1998) “The impact-echo method” Journal of Nondestructive

Testing and Ultrasonics, vol 3, 2, NDT.net, Germany. Go to: www.ndt.net/article/0298/streett/streett.htm
(accessed 24/06/2015)

SJÖDAHL, P and JOHANSSON, S (2012) “Experiences from internal erosion experiences from internal
erosion detection and seepage monitoring based on temperature measurements on Swedish embankment
dams”. In: Proc sixth int conf on scour and erosion (ICSE6) Paris, France, 27–31 August 2012, pp 198–205. Go
to: http://scour-and-erosion.baw.de/icse6-cd/data/articles/000287.pdf (accessed 24/06/2015)

SMITH, M R (ed) (1999) Stone: building stone, rock fill and armour stone in construction, Geological Society
Engineering Geology Special Publication No. 16, Geological Society Publishing House, Bath, UK (ISBN:
1-86239-029-0)

TEMPLE, B P, and KENNEDY, A (1989) The engineering properties of old bricks, Technical Memorandum
TM CES 128. British Rail Research, Derby, UK

THOMPSON, J and BATTELLE, L W (2009) Condition assessment of ferrous water transmission and
distribution systems. State of technology review report, EPA/600/R-09/055, US Environmental Protection
Agency, USA. Go to: http://nepis.epa.gov/Adobe/PDF/P1004Q38.pdf (accessed 24/06/2015)

TOMÁS, R, CANO, M, GARCÍA-BARBA, J, VICENTE, F, HERRERA, G, LOPEZ-SANCHEZ, J M

and MALLORQUÍ, J J (2013) “Monitoring an earthfill dam using differential SAR interferometry: La
Pedrera dam, Alicante, Spain” Engineering Geology, vol 157, 8, Elsevier BV, UK, pp 21–32

TUCCILLO, M E, JOLLEY, J, MARTEL, K and BOYD, G (2010) Condition assessment technology of

wastewater collection systems, EPA/600/R-10/101, US Environmental Protection Agency, Washington DC, USA

USSD (2013) Routine instrumented and visual monitoring of dams based on potential failure modes analysis,
USSD Committee on Monitoring of Dams and Their Foundations, United States Society on Dams,
Denver CO, UA. Go to: www.ussdams.org/pfma.PDF (accessed 24/06/2015)

Dams and reservoir conduits 163

USACE (1995) Evaluation and repair of concrete structures, EM 1110-2-2002, US Army Corps of Engineers,
Washington DC, USA. Go to: http://tinyurl.com/n4nfu69 (accessed 24/06/2015)

USACE (2004) Earth and rock-fill dams: general design and construction considerations, US Army Corps of
Engineers, Washington DC, USA (ISBN: 978-1-41021-712-7)

USBR (2011) Reclamation. Managing water in the west. Design Standards No 13 Embankment Dams, Chapter 5
Protective filters, phase 4 (final) US Department of Interior Bureau of Reclamation, Washington DC, USA.
Go to: www.usbr.gov/pmts/tech_services/engineering/design/DS13-5.pdf (accessed 24/06/2015)

VOEGE, M, FRAUENFELDER, R and LARSEN, Y (2012) “Displacement monitoring at Svartevatn dam

with interferometric SAR”. In: Geoscience and Remote Sensing Symposium (IGARSS) 2012 IEEE International,
22 – 27 July 2012, Institute of Electrical and Electronic Engineers, New York, USA, pp 3895–3898

WALKER, J P and TIETAVAINEN, M T (2014) “Sliplining bottom draw-offs”. In: Proc of the eighteenth
biennial British Dam Society conference, University of Belfast, UK, 3–6 September. A Pepper (ed)
Maintaining the safety of our dams and reservoirs, ICE Publishing, Institution of Civil Engineers, London,
UK (ISBN: 978-0-72776-034-0) pp 11–20

WICKHAM, D B (1992) “Collapse of an earth embankment dam” Dams & Reservoirs, vol 2, 3, British
Dam Society, Institution of Civil Engineers, London, UK, pp 18–19

WILLOWSTICK (2014) White paper on Willowstick technology used for mapping and modelling groundwater
systems, Willowstick Technologies LLC, Utah, USA. Go to: http://tinyurl.com/qcpb9vs (accessed 24/06/2015)

WINDSOR, D M (2012) “Pebley Reservoir (Derbyshire) emergency drawdown”. In: Proc of the seventeenth
biennial British Dam Society conference, University of Leeds, UK, 12–15 September. A Pepper (ed) Dams:
engineering in a social and environmental context, Institution of Civil Engineers, London (ISBN: 978-0-
72775-799-9) pp 11–20

WRc (2001) Sewerage rehabilitation manual, fourth revised edition, WRc Publications, UK (ISBN 978-1-
89892-041-0)

WRIGHT, P (2011) “Assessment and monitoring of London Underground tunnels. Some views from a clients
perspective”. In: Proc MONICO Workshop, Athens, Greece, 18 March 2011. Go to: http://tinyurl.com/mlpxjz5
(accessed 24/06/2015)

Statutes
Acts
Flood and Water Management Act 2010 (c.29)
Natural Environment and Rural Communities Act 2006 (c.16)
The Animal Welfare Act 2006 (c.45)
The Conservation (Natural Habitats &c) Regulations (Amendment) 2007
The Countryside and Rights of Way Act 2000 (c.37)
The Protection of Badgers Act 1992 (c.51)
The Wild Mammals (Protection) Act 1996 (c.3)
The Wildlife & Countryside Act 1981 (c.69)
Reservoirs Act 1975 (c.23)
Reservoirs (Scotland) Act 2011 (asp 9)
Water Industry Act 1991 (c.56)
Water Act 2003 (c.37)

Codes
ATV-M 127-2:2000 Static calculations for the rehabilitation of sewers with lining and assembly procedures,
Hennef (German)

164 CIRIA, C743

Regulations
Control of Asbestos Regulations 2012 (SI No 632)
The Conservation (Natural Habitats &c) Regulations (1994) (as amended)
The Diving at Work Regulations 1997 (No.2776)
Construction (Design and Management) (CDM) Regulations 2015
The Water Environment (Controlled Actitivies) (Scotland) Regulations 2011 (No. 209)

British Standards
BS 78-2:1965 Specification for cast iron spigot and socket pipes (vertically case (and spigot and socket fittings.
fittings (withdrawn)
BS 5628-1:2005 Code of practice for the use of masonry. Structural use of unreinforced masonry
BS 5911-1:2002 Concrete pipes and ancillary concrete products. Specification for unreinforced and reinforced concrete
pipes (including jacking pipes) and fittings with flexible joints (complementary to BS EN 1916:2002) (+A2:2010)
BS 9295:2010 Guide to the structural design of buried pipelines
BS EN 295-1:2013 Vitrified clay pipe systems for drains and sewers. Requirements for pipes, fittings and joints
BS EN 295-2:2013 Vitrified clay pipe systems for drains and sewers. Evaluation of conformity and sampling
BS EN 295-7:2013 Vitrified clay pipe systems for drains and sewers. Requirements for pipes and joints for pipe jacking
BS EN 545:2010 Ductile iron pipes, fittings, accessories and their joints for water pipelines. Requirements and test
methods
BS EN 771-5:2011 Specification for masonry units: Manufactured stone masonry unit
BS EN 771-6:2011 Specification for masonry units: Natural stone masonry units
BS EN 1295-1:1997 Structural design of buried pipelines under various conditions of loading. General
requirements (incorporating corrigenda May 2006, July 2008, February 2010 and March 2010)
BS EN 1504-3:2005 Products and systems for the protection and repair of concrete structures. Definitions,
requirements, quality control and evaluation of conformity. Structural and non-structural repair
BS EN 1504-4:2004 Products and systems for the protection and repair of concrete structures. Definitions,
requirements, quality control and evaluation of conformity. Structural bonding
BS EN 1504-5:2013 Products and systems for the protection and repair of concrete structures. Definitions,
requirements, quality control and evaluation of conformity. Concrete injection
BS EN 1504-9:2008 Products and systems for the protection and repair of concrete structures. Definitions,
requirements, quality control and evaluation of conformity. General principles for use of products and systems
BS EN 1504-10:2003 Products and systems for the protection and repair of concrete structures. Definitions.
Requirements. Quality control and evaluation of conformity. Site application of products and systems and quality
control of the works
BS EN 1610:1998 Construction and testing of drains and sewers
BS EN 1916:2002 Concrete pipes and fittings, unreinforced, steel fibre and reinforced (incorporating corrigenda
December 2003, December 2006 and April 2008)
BS EN 1996-1-1:2005 Eurocode 6: Design of masonry structures. General rules for reinforced and unreinforced
masonry structures (+A1:2012) (incorporating corrigenda February 2006 and July 2009)
BS EN 12504-1:2009 Testing concrete in structures. Cored specimens. Taking, examining and testing in compression
ISO 1438:2008 Hydrometry – open channel flow measurement using thin-plate weirs
BS EN ISO 9001:2008 Quality management systems. Requirements (incorporating corrigendum July 2009)
PD 6697:2010 Recommendations for the design of masonry structures to BS EN 1996-1-1 and BS EN 1996-2

USA Standards
ASTM F2550-06 Standard practice for locating leaks in sewer pipes using electro-scan--the variation of electric
current flow through the pipe wall

Dams and reservoir conduits 165

Further reading

AROSIO, D, MUNDA, S, ZANZI, L, LONGONI, L and PAPINI, M (2012) “GPR investigations to assess
the state of damage of a concrete water tunnel” Journal of Environmental and Engineering Geophysics, vol
17, 3, Environmental and Engineering Geophysical Society, Denver, CO, USA, pp 159–169

ACKERS, J C (2000) “Early siphon spillways”. In: Proc of the eleventh biennial British Dam Society conference,
University of Bath, UK, 14–17 June. P Tedd (ed) Dams 2000, Thomas Telford, London, UK (ISBN:
0-72772-870-9) pp 54–63

BINNIE, G M (1987) Early dam builders in Britain, Thomas Telford, London, UK (ISBN: 978-0-72774-701-3)

DEMPSEY, P and MANOOK, B A (1986) Assessing the condition of cast iron pipes, Source Document No. 9,
WRc, Swindon, UK (ISBN: 0-90215-634-9)

GREENEMEIER, L (2009) “Divine idea: plugging dams and tracking underground water, using an earth
MRI” Scientific American, 8 May 2009, Scientific American, Inc, USA. Go to: http://tinyurl.com/pvvvrf
(accessed 24/06/2015)

GUDGEON, D L and WALBANCKE, H J (1985) “Design and performance of 19th century earth dams”.
In: Proc 11th int conf on soil mechanics and foundation engineering, San Francisco, USA, 12–16 August 1985,
CRC Press, UK (ISBN: 978-9-06191-566-9)

HUGHES, R A N and KELLY, P (2002) “Remedial works at Brent Reservoir to address leaking sluice gates”.
In: Proc of the twelfth biennial British Dam Society conference, Trinity College, Dublin, 4–8 September 2002. P
Tedd (ed) Reservoirs in a changing world, Thomas Telford, London (ISBN: 0-72773-139-4) pp 224–235

ROLLINS, P (2011) “Diagnosing dam seepage” International Water Power and Dam Construction, vol 63, 2,
Global Trade Media, Sidcup, Kent, UK, pp 22–25

SANSALONE, M and CARINO, N J (1986) Impact-echo: a method for flaw detection in concrete using
transient stress waves, NBSIR 86-3452, National Institute of Standards and Technology, Colorado, USA

SANSALONE, M J and CARINO, N J (eds) (1991) Handbook of NDT of concrete, CRC Press, Boston, UK
(ISBN: 978-0-84931-485-8)

SANSALONE, M J and STREETT, W B (1997) Impact-echo: non-destructive evaluation of concrete and

masonry, Bullbrier Press, Jersey Shore, PA, USA (ISBN: 978-0-96126-106-1)

TEDD, P, DUTTON, D P M and HOLTON, I R (1998) “Investigating internal erosion at Brent Dam”. In:
Proc of the tenth biennial British Dam Society conference, University of Wales, Bangor, UK, 9–12 September. P
Tedd (ed) The prospect for reservoirs in the 21st century, Thomas Telford, London (ISBN: 978-0-72772-
704-6) pp 70–78

TEITSMA, A and MAUPIN, J (2006) Reduced mandated inspection by remote field eddy current inspection of
unpiggable pipelines, technical final report, DE-FC26-04NT42266, Gas Technology Institute, Illinois, USA.
Go to: http://tinyurl.com/o37oscc (accessed 24/06/2015)

ZARGHAMEE, M S, OJDROVIC, R P and NARDINI, P D (2012) Best practices manual for prestressed
concrete pipe condition assessment: what works? What doesn’t? What’s next? Water Research Foundation, US
Environmental Protection Agency, Denver, Colorado, USA.
Go to: www.waterrf.org/publicreportlibrary/4233.pdf (accessed 24/06/2015)

166 CIRIA, C743

Statutes
British Standards
BS 5628-2:2005 Code of practice for the use of masonry. Structural use of reinforced and prestressed masonry
BS EN 295-4:2013 Vitrified clay pipe systems for drains and sewers. Requirements for adaptors, connectors and
flexible couplings
BS EN 295-5:2013 Vitrified clay pipe systems for drains and sewers. Requirements for perforated pipes and fittings
BS EN 295-6:2013 Vitrified clay pipes systems for drain and sewers. Requirements for components of manholes
and inspection chambers

Dams and reservoir conduits 167

168 CIRIA, C743
A1 Case studies
1
The following case studies provide further details on guidance given in earlier chapters of the guide.
They have been categorised according to the topic(s) discussed (see Table A1.1).
2
3
Table A1.1 List of case studies by topic(s)

Monitoring and testing

Conduit abandonment
CIPP – water cured

Emergency works
External grouting

Upstream control
Segmental lining

Internal grouting

Discontinuance
localised repair

Other risk item

4

Shaft sinking
Ref. Reservoir

Pipejacking
Re-pointing
Willowstick

New siphon
Swageling

CIPP – UV
Sliplining

Augering
Guniting

Timber
Sonar

5
1 Birkenburn

2 Brent

3 Brunclough

6
4 Cofton

5 Daventry

6 Elton

7 Harthill

8
Knight and
Bessborough 7
9 Lebgei and Dubh

10 Napton

11

12
Pebley

Redbrook
8
13 Rudyard

14

15
Shon Sheffrey

Tardebigge
9
16 Torside

10
Un-named new
17
reservoir
18 Upper Rivington

19 Warmwithens

20

21
Yeoman Hey

Redmires
11
12
A1
Dams and reservoir conduits 169
Case study A1.1 Birkenburn reservoir

Background
This case study describes the method whereby
a 510 mm diameter CI outlet pipe through an
earth embankment dam, with downstream
valves, was successfully grouted against the
full hydrostatic head of the reservoir (see
Figure A1.1).

The operational discharge and emergency

draw-down functions of the outlet pipe were
replaced by two new siphon pipes.

Birkenburn reservoir was built around 1820

to augment the water supply to the Forth and
Clyde canal. The earth dam is 14.5 m high and
Figure A1.1 Reservoir outlet pipe downstream valve, fitted
impounds 780 600 m3 of water. The catchment with end plate and three glands, ready for
area is 4.4 km2. The reservoir is located in a grouting (courtesy A Ross)
remote part of the Campsie Fells to the north of Kilsyth and is 431.3 m above sea level. The original
draw-off system comprised a CI pipe at the base of the dam, with three downstream valves.

A statutory inspection was carried out on 18 May 2006. One of the reservoir inspecting engineer’s
recommendations for safety reasons was: “control the risk posed by the unprotected outlet pipe by lining the pipe
or by other appropriate measures at the dam.”

The inspecting engineer recommended that the works be completed by 31 March 2010.

The reservoir owner’s initial plan was to line the outlet pipe with a 450 mm (OD) HDPE pipe and grout
the annulus. With this in mind, some initial improvements were made to the access track to the reservoir
in the summer of 2006.

In summer 2009, the reservoir water was pumped out and a rockfill cofferdam constructed around the
outlet pipe at the upstream toe of dam. The plan was to locate the pipe under the two metres of peat and
silt, then to break into the pipe to install the HDPE pipe insert. However, the works had to be abandoned
in autumn 2009, as it was not possible to remove the many tonnes of silt over the pipe.

Conduit type/material/use
The conduit function as an outlet pipe under the dam and is constructed from circular socket and spigot
CI pipes, with an internal diameter of 510 mm, with three valves on the downstream side. The facility is
provided for operational and emergency drawdown. The conduit is about 150 m long, extending out into
the reservoir be well beyond the upstream toe of the dam.

Justification for works

At the time of the statutory inspection in 2006, the CI outlet pipe was 186 years old. The position of the
valves at the downstream end of the pipe meant that the conduit was under hydrostatic pressure at all
times. While the pipe had not failed, the condition of the pipe was unknown. Potential failure modes
would include internal erosion and operation failure loss of ability to draw-down the reservoir.

A diver survey of the upstream end of the outlet pipe in September 2007 revealed that there was up to
two metres of peat and silt in the bed of the reservoir. A full draw-down of the reservoir was attempted in

170 CIRIA, C743

October 2007 so that the inside of the pipe could be inspected, but this was halted in the latter stages. A

1
subsequent diver survey revealed that a substantial volume of peat and silt had mobilised and buried the
upstream end of the outlet pipe.

Remedial works
A workshop to consider engineering options was held in March 2010. Subsequently, the reservoir owner
opted to construct two permanent siphons at the dam and to permanently seal the existing CI outlet 2
pipe by filling with cementitious grout. The siphon pipes (365 mm ID and 455 mm ID) were sized for
operational use as well as for emergency draw-down.

These works were successfully carried out in summer 2011. 3

The flushing, internal inspection and grouting of the CI outlet pipe was achieved by the following methods.

Pipe flushing:

‰‰ A large steel flange plate, with a smaller flanged pipe welded to the plate, was fabricated and bolted
4
to the exposed flange of the lowest downstream valve.

5
‰‰ A temporary valve was connected to the smaller flanged pipe, and a ‘Bauer’ connector fitted to the
temporary valve.
‰‰ A pump intake was then attached to the Bauer connector, and the pump discharge flexible hose
run up and over the dam, back into the reservoir, and away from the overflow weir.
‰‰ All existing and temporary valves were opened and the water and peat pumped back into the
reservoir. Turbidity levels were recorded using a turbidity tube, and the flushing of the pipe was
undertaken until an acceptable reading was recorded.
6
‰‰ All valves were then closed and the pump and temporary pipework disconnected.

Pipe internal inspection: 7

‰‰ A short length of steel flanged pipework, equal in diameter to the lowest downstream valve, was
fabricated and bolted to this valve.
‰‰ A steel end plate of the same diameter, was also fabricated with a double sealed gland.
‰‰ An underwater pipe inspection vehicle (PIV), with camera and sonar, was placed into the short
8
length of pipework with the power and communications cable passed through the double sealed
gland.
‰‰ The end plate was bolted in place and all three of the outlet pipe valves opened.
‰‰ The PIV was driven up the outlet pipe, visually recording the condition of the inner face, as well as
9
providing sonar data of the internal condition of the CI pipe.

10
‰‰ Sonar data was later used to produce pipe cross-sections to look for any cracks or material defects in
the CI. No such defects were identified, and planning for grouting the outlet pipe was then initiated.

Grout pumping trial:

11
‰‰ A successful grout pumping trial of 2.5 m3 of the micro concrete mix, using a 270 mm diameter by
50 m long PVC coated fabric sleeve, was conducted at a low level site compound.

Pipe grouting:

‰‰ A steel end plate with three glands was fabricated and fitted to the open end of the lowest
downstream valve.
‰‰ Three mild steel grout injection tubes, of varying lengths, were welded on site. Each tube was fitted
12
with a quarter turn valve and a pumping pipe connector.
‰‰ The injection tubes were inserted into the loose glands with the quarter turn valves of the injection
tubes closed. The glands were then tightened enough to seal the glands against the injection tubes,
but loose enough to allow insertion of the injection tubes into the reservoir outlet pipe.
A1
Dams and reservoir conduits 171
‰‰ All three of the outlet pipe valves were
then opened.
‰‰ The grout tubes were pushed up into the
outlet pipe.
‰‰ The grout pump was connected in
sequence to the three grout tubes,
starting with the shortest and progressing
to the longest the volumes of grout take
were carefully monitored at each stage
of the process to gauge the filling of the
outlet pipe.
‰‰ The outlet pipe was successfully grouted
over the required minimum length of 97
m under the embankment dam. The total Figure A1.2 Site team preparing to start grouting of reservoir
outlet pipe (courtesy A Ross)
grout take was 22 m3.
‰‰ When the grout had set, a 500 mm long double flanged steel pipe was fitted to the end of the gland
plate, and then capped with a steel end plate to provide an additional seal.

The 2006 statutory inspection report had another recommendation for safety reasons: “Install provisions
for enhanced monitoring of seepage and settlement on the line of the outlet pipe”.

Although the outlet pipe had been permanently sealed, it was recognised that a seepage path along the
outside of the pipe could still be an ongoing risk to the long-term integrity of the embankment dam.

The reservoir owner installed deep toe drains either side of the outlet pipe to a depth below the bottom of
the outlet pipe. This picked up any reservoir water that could travel along the outside of the outlet pipe.

A series of chambers were constructed to direct the toe drainage flow through a measuring flume and
to a stilling tube. The water level in the tube is monitored in real time by supervisory control and data
acquisition (SCADA) equipment. The toe drainage flow is constantly monitored for any signs of change
that would indicate leakage.

Lessons learnt
‰‰ Access difficulties up a steep hillside and across an exposed upland moor.
‰‰ Severity and unpredictability of weather and site conditions working at such a high altitude.
‰‰ Difficulty in dealing with large volumes of peat and silt in the bed of a reservoir.
‰‰ Environmental constraints requiring removal of fish at the latter stages of reservoir dewatering and
strict control on the turbidity levels of water discharged to downstream watercourses.
‰‰ Difficulty in locating and exposing an intake that was buried under two metres of peat and silt.
‰‰ Need for a viable and cost-effective alternative option, if the first plan did not succeed.
‰‰ Need for a specialist contractor who can deliver a novel pipe grouting method under the full
hydrostatic head of the reservoir.

172 CIRIA, C743

Case study A1.2 Brent reservoir
1
Background
Brent dam is a nine metre high embankment with a puddle clay core. 2
The draw-off system comprises an upstream brick culvert 900 mm high by 980 mm wide and 15.5 m

3
long, twin 225 mm diameter pipes about 5 m long through the core. There are valves at the head of
a brick built human-entry sized downstream culvert 3 m high, 2 m wide and 15.5 m long. Normally a
trickle feed is maintained to the canal throughout the year by partial opening of one of the valves.

There are two outlets, a low level system used for releasing water to the River Brent downstream and a
canal feeder draw-off at a slightly higher level. A diagrammatic cross-section is shown in Figure A1.3.
4
5
6
Figure A1.3 Cross-section at draw-off (courtesy Canal & River Trust)

The low level system, which is controlled by two large sluices at the upstream end, comprises twin
parallel CI boxes 1.78 m high by 0.895 m wide, through the base of the masonry spillway weir, which
is located at the centre of the dam. Automatic sluices at the upstream ends of these conduits operate in
7
stages to maintain the operating level of the reservoir 1.03 m below the weir in normal conditions.

The leakage at the head of the draw-off ‘tunnel’ is easily accessed via a permanent walkway suspended
above the submerged invert. There is a forced ventilation system.
8
Measuring devices at the site included:

‰‰ measuring boxes/spouts at the leakage points

‰‰ crest levelling stations
9
‰‰ measuring points across certain cracks in the concrete crest slabs
‰‰ disused piezometers.

Figure A1.4 shows the seepage at the

10
head of the tunnel and the collection

11
trough in 1987.

The inspection report of 2009

recommended for safety reasonsthat

12
the monitoring of the seepage should
be improved and that turbidity be
measured for safety reasons. Crest level
monitoring was also to be carried out
more frequently in an attempt to identify
movement above the line of the culvert.
The more intensive monitoring was to be Figure A1.4 Seepage at the head of the tunnel, and collection trough A1
carried out for three years and presented (courtesy Canal & River Trust)

Dams and reservoir conduits 173

to a QCE. The engineer considered that the rate of settlement was marginally higher than might be
inspected and that there was potential for internal erosion, but that the situation was stable and it could
be controlled by continued monitoring.

While certifying the safety measure complete, the engineer advised that the seepage should be
investigated further. A groundwater temperature investigation was commissioned, which was carried
out in February 2014. Probes were driven downstream from the core to a depth of 10 m on either side of
the conduit. A single probe was driven to a depth of 8 m above the pipes to avoid damaging them. The
temperature of the reservoir water was 5.7°C (see Figure A1.5). The temperature of the leakage water
into the culvert were 12°C. This gave an immediate indication that the reservoir water could not be
following the short leakage path alonside the pipes.

Also, the probing identified a significant

temperature anomaly, indicating seepage
through the core of the dam, centred 20 m
from the draw-off, at a depth of 5 m to 7 m
below crest level and extending over a length
of 30 m. This coincided with an area where
seepage at the toe had been seen in 1994. It is
plausible that this leakage was finding its way
into the culvert, the head of which is a similar
distance from the temperature anomaly as the
toe. The leakage water is thought to be warmer
than the reservoir water because of the length
of time taken for the water to pass through
the embankment. A spring source is unlikely
Figure A1.5 Installing temperature probes (courtesy Canal &
because the reservoir is founded on 40 m of
River Trust)
London Clay. Weekly temperature monitoring
of the leakage water and the reservoir water has been started to study seasonal variations in order to
better understand the phenomenon.

Crest levels are recorded annually and cracks in the concrete crest slabs are measured annually using
Vernier callipers.

Further reading includes Gudgeon and Walbanke (1985), Binnie (1987), Tedd et al (1998), Ackers (2000),
Brown and Gosden (2000, and Hughes and Kelly (2002).

Conduit type/material/use
The conduit is made up of brick lined culverts housing CI valves and pipework.

Reasons for intervention

There has been longstanding turbid leakage through the brickwork at the upstream end of the human-
entry section of the draw-off culvert. This was first discussed in the statutory inspection report in 1946,
when it was described as being a “slight percolation of old standing and of no consequence”. The problem was
not discussed again by the inspecting engineers until 1983.

Quantitative leakage monitoring on a quarterly basis was instructed in the inspection report of 1983. Chemical
analysis carried out at the same time indicated that the seepage and reservoir waters were dissimilar.

The hydraulic gradients are high in this area and internal erosion caused by leaking water tracking on
the outside of the pipes was giving cause for concern that voids could be forming in the core.

The dam is designed to be overtopped and has a concrete slab on the crest. Settlement associated with
internal erosion alongside the conduit would be hidden by the slab spanning voids.

174 CIRIA, C743

By 1996, the seepage had become a cause for concern to the supervising engineer, who invited another

1
ARPE to give an opinion. The ARPE believed that some internal erosion was taking place and suggested
investigation by GPR to seek for voids, that flows be measured weekly and that the sediment should be
collected and analysed chemically.

The 1999 inspection report confimed that the seepage was of concern and recommended it be further
investigated for safety reasons including a CCTV of the CI pipes. The reservoir was drained in 2002, the
upstream culvert was inspected manually and minor repairs were carried out. The CCTV inspection of 2
the CI pipes was carried out. The CI pipes were encrusted, but not cracked or open jointed. However,
they were lined with HDPE liners while the reservoir was drained. The time available was restricted due
to pressure from leisure users and SSSI interests.

Erosion of material through the brickwork, which forms the draw-off culvert, raised concerns that voids
3
were being formed in the core. Investigations to determine the extent and cause of the internal erosion

4
were carried out.

GPR was used to search for such voids. A number of cores through the slab were needed for calibration,
however nothing definitive was found. GPR was also used in the downstream tunnel locating, but not
identifying, features behind the walls.

Two boreholes were sunk from the crest to points near to the top of rear of the bulkhead wall at the head 5
of the downstream culvert. Permeability and piezometric pressure measurements were carried out.

6
Samples of eroded material were analysed and erosion rates estimated at 0.5 l/year.

The thickness of the bulkhead was measured by drilling. It was 1.6 m thick and backed in firm grey clay
with no trace of the orangey material collected from the seepages. No evidence of any erosion of the

7
brickwork was identified.

A report was delivered to the undertaker in 1997 and the study was presented in Tedd et al (1998) at the
BDS biennial conference that year.

The reservoir is only of modest operational value to the CRT, supplying water to the Grand Union Canal.
However, it serves as a valuable amenity of regional importance, a key flood alleviation feature in the
downstream watercourse and is a SSSI. Discontinuance has been considered but has not been taken forward.
8
9
The turbidity of the seepages has been monitored since 2005, but the results are scattered and no trends
have been identified to date.

Remedial works
The conduits were lined with GRP (see
Figure A1.6) in 2002 for precautionary
10
reasons because their condition was

11
unknown and regular flow, turbidity and
settlement was established. This was mostly
successful, but locally the lining bulged (see
Figure A1.7) and failed to adhere to the
CI, requiring it to be cut away and patched
(Hughes and Kelly, 2002).

The downstream draw-off tunnel,

12
which has forced ventilation, is entered
weekly for surveillance purposes. The
flow from the two measuring boxes is
measured using jugs and measuring Figure A1.6 GRP liners (courtesy Canal & River Trust)
A1
Dams and reservoir conduits 175
cylinders. Once a month, the turbidity of these
leaks is also measured. The data is reported
to the supervising engineer, who was advised
by the last inspecting engineer to increase
surveillance to three times a week if there were
to be any significant change in the regime. The
engineer would also seek advice from an ARPE
if this were to happen.

Lessons learnt
‰‰ Leakage is often not where you expect it
to be.
‰‰ Results of each phase of investigation
should be scrutinised and the need
Figure A1.7 Bulging GRP liner (courtesy Canal & River Trust)
for further investigations established
before reaching conclusions based on
preconceived ideas and incomplete information.

176 CIRIA, C743

Case study A1.3 Brunclough reservoir
1
Background
Brunclough reservoir is retained by a 10.5 m high × 230 m long embankment dam, believed to
incorporate an impervious clay core. The dam has a CI pipe under the dam with downstream control
2
(see Figure A1.8). The risk associated with this arrangement was highlighted at a statutory inspection.

3
4
5
Figure A1.8 Cross-section through draw-off works (courtesy Canal & River Trust)

There is no instrumentation present at the dam. 6

In a report of the 2007 inspection, the inspecting engineer included the following discussion:

The old outlet pipe has no upstream control and is, therefore, pressurised through the core of the dam.

Various suggestions have been made including the following:

7
(a) fit upstream control to the pipe

(b)

(c)
grout up the pipe

fit an internal structural liner

8
(d) fit an internal structural liner and also an upstream control valve

The pipe is an important safety feature and grouting up is, therefore, not favoured. Whilst
temporary pumps could be deployed in an emergency it is not good practice to rely too heavily on
9
such temporary equipment.

10
The installation of an upstream valve is superficially attractive but suffers from a number of practical
difficulties. To avoid the need for a valve tower and bridge it would need to be either a hydraulically
operated valve or, alternatively, a gate valve with an inclined operating rod. Neither of these
arrangements is ideal and vandalism might be a problem. Furthermore the installation of an upstream
valve would complicate operation and subject the pipe to operating conditions not hitherto experienced.

The undertaker had many old dams without upstream control and carried out a portfolio study 11
to identify the extent of the challenge and consider the possible solutions. Considerations of cost,
maintenance liability and the risks of changing the operating conditions of the pipes by no longer
maintaining an internal pressure, possibly leading to new challenges, led the undertaker to prefer the
lining option at this and other similar dams. The facility to add an upstream valve at a later date is
considered good practice.
12
A1
Further information is given in Brown and Howlett (2011).

Dams and reservoir conduits 177

Conduit type/material/use
A 229 mm diameter CI pipe buried under the
dam with a 150 mm diameter downstream
valve in a very short downstream culvert or
buried valve house constructed of sandstone
masonry. The valve is used to supply a feed to
the canal and the reservoir is drawn down in
the summer (see Figure A1.9).

Reasons for intervention

The need for this intervention was identified
by statutory inspection carried out in 2007.
Figure A1.9 Valve pre-works (courtesy Halcrow)
The recommendations given in the report, for
safety reasons, included a requirement to carry
out a CCTV inspection of the outlet pipe and to install a structural liner.

Should a leak develop in the very old CI pipe at a joint or through deterioration of the material of the
pipe, water under reservoir pressure would enter the embankment fill, leading to rapid erosion of the fill
and probable failure of the dam.

Remedial works
The pipe was refurbished using a CIP liner, which was introduced, inflated and cured using water
pressure and temperature.

The undertaker had experience of using successfully CIP liners on other non-reservoir culverts. The CIP
liner was at the time rarely used on reservoirs and
the QCE needed to be convinced of the suitability
of the system. Other lining systems would have
reduced the pipe diameter by too great an
amount, restricting outflows.

The pipe was lined using the CIP system with hot
water curing. The reservoir had to be completely
drained to give access to both ends of the pipe.
There were no fish in the reservoir and silt
discharge could be controlled using straw bales.
There was good vehicle access for the heavy
equipment needed for this technique.

The job was successful and a similar solution

could be adopted in the same circumstances in
the future.

The installation of the CIP liner at various stages

of the installation is shown in the Figures A1.10
and A1.11 together with the new valve installed in
Figure A1.12.

The final cost was about £250 000.

Figure A1.10 CIP liner emerging from upstream end under

water pressure (courtesy Canal & River Trust)

178 CIRIA, C743

Lessons learnt
‰‰ There are alternative ways of addressing the risk of pressurised pipes under existing dams.
‰‰ Pipe lining was a proportionate solution in this case.
1
2
3
4
Figure A1.11 Liner cured and trimmed, downstream end
(courtesy Canal & River Trust)
Figure A1.12 New valve (courtesy Canal & River Trust) 5
6
7
8
9
10
11
12
A1
Dams and reservoir conduits 179
Case study A1.4 Cofton reservoir

Background
Cofton reservoir constructed in 1815 is retained by a 11.2 m high × 160 m long homogeneous
embankment dam. There is two-level draw-off arrangement with CI pipes described as follows (see
Figure A1.13).

Following an incident where sandy embankment fill was discharged from a little used outlet, both outlets
were lined with a CIP liner inflated by air and cured using UV light.

Figure A1.13 Cross-section through draw-off works (courtesy Canal & River Trust)

The defect was identified by routine surveillance in 2008, and is discussed by Brown (2012).

There is a single V-notch monitoring f low from a spring buried under the dam and piped to the
downstream toe.

During a routine valve test in June

2008, a defect was identified in the
upper valve operating mechanism.
It was decided that the water level
would be reduced using the lower
valve to allow the upper valve
rodding to be repaired without the
need for divers. There was no great
urgency in carrying out the repair.
The on-site emergency draw-
down plan was revised to include
additional pumping in the unlikely
event of an emergency. When it
was operationally appropriate, the
lower valve was opened on 1 August Figure A1.14 Leakage of defective joint three metres from the valve in the
lower draw-off pipe (courtesy Canal & River Trust)
2008. On the following Monday,
routine surveillance identified
that discoloured water and sand was being discharged from the lower pipe. The supervising engineer
instructed that the valve be closed immediately to prevent any further erosion. A CCTV survey was
carried out later in the week. Two defective joints close to the upstream valve were identified. The joints
appeared to have opened up and had debris lodged in them. The debris was thought to be deflecting the
flow of water from the pipe into the embankment fill behind (see Figure A1.14).

The upper draw-off was also subject to a CCTV survey and a known leak from the downstream end of
the pipe was traced to slight leakage from joints in the body of the dam (see Figure A1.15).

180 CIRIA, C743

 1
2
3
4
Figure A1.15 Joint seepage in draw-off pipe (courtesy Canal & River Trust)

Conduit type/material/use

5
The upper draw-off is a 300 mm diameter
CI pipe under the dam and 4.3 m below
the top water level (TWL) with a short brick
downstream culvert. There is a single upstream

6
‘teapot lid’ valve operated by inclined rodding
and a submerged crank (see Figure A1.16). The
lower draw-off is similar, but 150 mm diameter
and 8.3 m below TWL.

Normally the lower draw-off is not used and

water is passed to reservoirs lower down the 7
valley using the upper draw-off valve.
Figure A1.16 Teapot lid valve (courtesy Canal & River Trust)

Reasons for intervention 8

The investigations suggested that there was internal erosion of a sandy embankment fill taking place due
to defects in the CI draw-off pipe. On this basis it was decided that measures should be taken to reduce
the potential for loss of embankment material into the pipe.
9
Remedial works

10
The reservoir is required operationally, so the problem needed to be addressed. Had it been impractical
to line the small diameter lower pipe, it would have been grouted. In small diameter pipes such as these,
there is no lining alternative other than CIP, which does not severely reduce pipe diameter and flow.

The matter was discussed internally by the undertaker’s reservoir, engineering and operational teams.
Advice was sought from an ARPE. The use of such liners by the undertaker was already well-established
for repairing culverts under canals, so the advice needed from the specialist subcontractors was site- 11
specific only. The work was carried out under a framework contract and the partnering contractor
contributed to the buildability of the chosen solution.

Fitting of additional guard valves was considered, but was dismissed as being too complex and costly. 12
The reservoir was completely emptied by means of siphons and pumps. A piled cofferdam had been

A1
driven around the lower valve in 2005 and left in situ. It served to hold back the mud in the reservoir
basin. Both the lower and upper draw-off pipes were lined with CIP sleeves. These were pulled into the
pipes with a winch, inflated with air pressure and cured using UV light. This system needed no use of

Dams and reservoir conduits 181

cranes, unlike the hot water cured system and
left no contaminated water to dispose of. The
opportunity was taken to replace the upper
valve, together with its operating mechanism
and grille.

There has been ongoing continual seepage

along the annulus between the liner and the
CI pipe on both draw-offs, since the lining
was carried out (see Figure A1.17). This is
unfortunate, but does not threaten reservoir
safety because the flow is constrained by the
size of the annulus. The specification of the
liner size, the cleaning of the pipe and the lower
Figure A1.17 Seepage along the annulus between the liner and
pressures used to inflate a UV cured liner
the CIP pipe (courtesy Canal & River Trust)
compared with the water cured system were the
reasons why this occurred.

A similar solution would be adopted in the future in the same circumstances, but with refinements in the
specification and execution to address the leakage along the annulus.

Lessons learnt
‰‰ Other work was carried out at the same time and the lining costs are not readily separated.
‰‰ The specification of liners of this type and the process of their installation are critical to obtaining
the best results.

182 CIRIA, C743

Case study A1.5 Daventry reservoir
1
Background
The inspecting engineer was concerned about clay being extruded into a brick built downstream access
valve culvert. Monitoring was attempted, but the matter was resolved by repointing (see Figure A1.18).
2
3
4
5
Figure A1.18 Cross-section through draw-off works (courtesy Canal & River Trust)

Nothing has been published elsewhere on this intervention.

6
The defect was identified during a statutory inspection in 2005 and the inspecting engineer expressed
concern over the loss of material into the bottom outlet, particularly as it had been going on for
some time. The inspecting engineer recommended that the phenomenon should be investigated and
quantified as far as practicable and proposed the following steps: 7
‰‰ The walls, soffit and invert should be washed and brushed down to remove the extruded clay.

8
‰‰ The culvert should be inspected at quarterly intervals to check for the reappearance of clay.
‰‰ Once this has occurred then an effort should be made to collect and weigh the deposits so as to
determine the rate of movement.
‰‰ A collection trough should be placed beneath the two dripping cracks so as to collect not only the

9
water but also the clay and ocherous material (see Figure A1.19).
‰‰ The rate of water ingress shall be recorded and the solid material shall be retained, weighed and
tested to establish its composition.
‰‰ Records of leakage and settlement transport shall be maintained for a period of at least five years
or until the end of December 2011, whichever is the sooner.
‰‰ The monitoring results should be
reviewed regularly by the supervising
10
engineer and submitted at the end
of the monitoring period to an
ARPE, who shall then give further
directions with regard to either further
11
monitoring or remediation.

The inspecting engineer advised that it would

be acceptable for the undertaker to bring
12
forward the remedial works to avoid the costs
associated with a long monitoring period, then
the measures outlined in the section below on
the remedial works would be satisfactory.
Figure A1.19 Clay intrusions (courtesy Canal & River Trust)
A1
Dams and reservoir conduits 183
Visual monitoring continues. For instance on 10 June 2010, the supervising engineer recorded:

“Valve tunnel – walls wet just upstream of and in vicinity of valves. Further in is dry before
becoming wet at upstream end. Bulk-head wall dry. Portal dry. No clay exudations noted.
Clay materials collected mainly from the shroud position at the downstream pipe.
Both collecting bowls full of water with virtually no clay sediment.”

Crest levelling pins were installed in 2009 and are levelled annually. No significant settlement has been
noted above the valve culvert.

Conduit type/material/use
A human-entry oval brick built culvert 1.2 m wide, 1.5 m high and 20 m long passes under the
downstream shoulder ending at a brick bulkhead. Two CI pipes pass through the bulkhead. One is an
operational 200 mm diameter pipe, which passes along the invert of the culvert to two in-line valves near
the portal. The second pipe is 150 mm in diameter is blanked off and has been encased in concrete. It is
not known if the pipes continue under the upstream shoulder or if there is an upstream culvert section.

This is the only outlet from the reservoir and, particularly in the summer months, a feed to the canal is
maintained by making releases through this system.

Reason for intervention

The inspecting engineer was concerned that if allowed to progress, the extrusion of clay would lead to
loss of embankment material resulting in internal erosion and settlement of the dam. If left unchecked,
internal erosion could lead to failure of the embankment. Settlement reduces the freeboard and could
render the embankment susceptible to damage from concentrated overtopping.

Remedial works
The inspecting engineer advised that works that would be satisfactory included:

‰‰ Crack repairs to the brickwork

and pointing of the mortar joints,
or formation of a new structural
lining with sprayed mortar would
be potential options, but would
both be difficult given the limited
space inside the culvert.
‰‰ Completely fill the inner section of
the culvert with foamed concrete.
This would be relatively simple to
carry out and would also enhance
safety by surrounding the aging
exposed section of main upstream
from the control valves.
Figure A1.20 Attempts to monitor clay intrusions (courtesy Canal &
River Trust)
The culvert was cleaned and attempts
were made to collect and weigh the intrusions. However, this proved difficult and inconclusive as some
material was lost into the invert and washed away (see Figure A1.20).

A more permanent and less onerous solution was preferred by the undertaker.

As noted, repair works considered included filling the culvert with concrete. However, this was not
pursued because the pipe would no longer be able to be inspected. Also, there was the possibility of
seepage being diverted to the track behind the culvert walls, emerging diffusely rather than being

184 CIRIA, C743

able to be identified and monitored easily

1
from within the culvert. Guniting of canal
structures was common in the 1960s, but
is now known to trap water within the old
brickwork leading to hidden deterioration.
The culvert was too small for a sprayed
concrete lining acting independently rather
than compositely with the existing.
2
The culvert was repointed using lime mortars
by the undertaker’s framework contractor
(see Figure A1.21). The use of lime mortars
is normal in the conservation environment of
3
the waterways and the contractors and direct

4
labour are well versed in the materials and Figure A1.21 Repointed culvert (courtesy Canal & River Trust)
techniques.

Repointing open joints using lime mortars was the adopted technique. This was reasonably successful

5
and would be used again in similar circumstances. In worse cases, pressure pointing and grouting
through the repaired barrel could be employed.

Lessons learnt
‰‰ The costs are not easily identified, but were modest.
‰‰ Monitoring of the rate at which fine clay particles erode through painting or brickwork is difficult
6
and the results can be inaccurate and inconclusive.

7
8
9
10
11
12
A1
Dams and reservoir conduits 185
Case study A1.6 Elton reservoir

Background
Elton reservoir was constructed in 1808 and incorporates an 8.8 m high × 600 m long embankment dam
with a clay core. The draw-off works comprise a siphon and low level outlet constructed from CI pipes
(see Figure A1.22).

A displaced and open joint in a CI pipe was found during a CCTV inspection. This was repaired using a
localised CIP liner and voids outside the pipe were grouted.

Figure A1.22 Cross-section through draw-off works (courtesy Canal & River Trust)

This intervention is mentioned by Brown and Howlett (2011).

There are crest levelling pins and V-notches, but they were not relevant in this case.

Conduit type/material/use
The low level outlet is a 229 mm diameter CI pipe with upstream and downstream valves. The
downstream end of the pipe is in a brick culvert about 2 m high and 1.5 m wide. The nearby siphon
outlet is a 300 mm diameter CI pipe also with upstream and downstream valves. There is a priming ‘T’
in the crest of the dam.

This reservoir is not currently required for supplying water to an operational canal and all the valves are
usually kept closed.

Reasons for intervention

The recommendations from the 2008 statutory inspection report included a requirement that the
upstream valves on the syphon and low level outlets, which had defective draw-bars, be repaired. The
reservoir needed to be drained for this purpose and this was done in 2009 to 2010. During the works,
following heavy rain, silt and debris were washed into the scour pipe causing a partial blockage. A
CCTV survey led to the discovery of an open and displaced joint located about three metres upstream
of the bulkhead wall. It might have been caused by mining subsidence although a pillar of coal had been
purchased to protect this part of the dam.

Access to the downstream end of the pipe was via the downstream culvert. There was a double bend in
the pipe downstream from the bulkhead to take the pipe down to the culvert invert. A temporary hole
was cut in the pipe to facilitate access for the CCTV and patch lining to avoid the need for dismantling
the joints.

186 CIRIA, C743

Remedial works
The pipe was needed for emergency draw-down, so could not be grouted.
1
The defect was in the core area and it was agreed with the QCE overseeing the other work that the
joint be repaired with a local liner and any voids in the embankment grouted through the bulkhead.
The means of lining was advised by the specialist subcontractor. The ‘Epros’ system uses a glass fibre
laminate, impregnated with Epros resin. 2
A one metre long Epros patch was placed over the defect, inflated using an air bag and allowed to cure

3
chemically. Two further patches were later placed in the same way, overlapping the first to give a double
thickness over the joint, 1 m long and 0.5 m wide single thicknesses at each end.

The Epros patches were supplied by Future Environmental Services and tested and certified by Bodycote
Testing Ltd.

A polymer modified grout, GW1, was used to fill any voids and was introduced into the area around the 4
patch by drilling through the bulkhead wall and injecting through lances at a pressure of 1 bar.

5
The pipe was again surveyed by CCTV to check the result, which was satisfactory. A similar solution
would be adopted in the future in the same circumstances.

Lessons learnt
‰‰ Using Epros patches provides an effective and economical method of repairing a cracked and
displaced joint in a small diameter pipe.
6
7
Further information is given in Brown (2009), Brown (2010) and Brown and Howlett (2011).

8
9
10
11
12
A1
Dams and reservoir conduits 187
Case study A1.7 Harthill reservoir

Background
Harthill reservoir constructed in 1796 is retained by a 7.5 m high × 400 m long embankment dam (see
Figure A1.23). The upstream culverts were inspected by a ROV. There was zero visibility, but the sonar
readings showed that there were no significant defects.

Figure A1.23 Cross-section through draw-off works (courtesy Canal & River Trust)

Nothing has been published elsewhere on this case study.

The investigations described below took place in 2007.

Conduit type/material/use
The upper draw-off has its invert 2.77 m below the TWL. The upstream culvert is of brick construction
and is 0.76 m diameter. There is a 225 mm diameter CI pipe through the core area with a valve at the
downstream end. The human-entry sized downstream culvert is 2.45 m high × 1.14 m wide.

The lower draw-off is 7.09 m below TWL. The upstream culvert is of brick construction and is 0.76 m
diameter. There is a 225 mm diameter CI pipe through the core area with a valve at the downstream
end. The human-entry sized downstream culvert is 2.17 m high × 1.24 m wide.

The draw-offs are used to supply reservoir water to the Chesterfield Canal as and when required.

Reason for intervention

The inspecting engineer identified that the
upstream culverts had not been seen for 70
years and that their condition was unknown.

In the 2004 statutory inspection report, the

inspecting engineer recommended, for safety
reasons, that the condition of the brickwork
in the upstream and downstream culverts
should be ascertained and that the culverts be
strengthened if necessary.

The upper draw-off could be checked in the

dry, but not the lower one, which would have
involved draining the reservoir and tackling the
problems of fish rescue and silt management. Figure A1.24 ROV (courtesy Canal & River Trust)

188 CIRIA, C743

In August 2007 a CCTV survey of the lower draw-off using an ROV (see Figure A1.24). was attempted.

1
However, there was no visibility in the depths of the reservoir. The ROV was nearly the same diameter as
the culvert, passage was unobstructed and this indicated that there were no drastic defects, but no other
useful information was obtained.

A subsequent sonar survey by ROV (see Figures A1.25 and A1.26) indicated that the brickwork was free
from significant defects.
2
Remedial works
In this case the remedial works were restricted to an investigation only. The complete dewatering for
visual inspection was not practicable. 3
Lessons learnt
‰‰ Underwater CCTV surveys require good visibility if anything more than a close up of individual 4
bricks is required. Sonar is a technique that can provide useful information where visibility is limited.

5
6
7
Figure A1.25 Use of ROV (courtesy Canal & River Trust) 8
9
10
11
12
Figure A1.26 ROV control position (courtesy Canal & River Trust)

A1
Dams and reservoir conduits 189
Case study A1.8 Knight and Bessborough
reservoirs

Background
Two existing 1.37 m riveted steel pipes were replaced through the existing 11 m high continuous earthfill
embankment dam at the operational statutory Knight and Bessborough reservoirs (see Figure A1.27).

Leakage was observed in the embankment near the steps to the valve/outlet tower in 2007. Following the
leakage, the longer 1.37 m riveted steel water main, which is connected to a 1.22 m main running to the
centre of reservoir, was taken out of service. Continued leakage of the longer 1.37 m riveted steel main could
undermine the slope of the reservoir,
which overlooks a road and commercial
properties. A solution to the leakage
problem had to be found in order to
protect the reservoir embankment.

The preferred option was to

construct two new DN 1300 mm
steel replacement pipes placed
above ground on the slope of the
embankment and to grout the
existing pipes. Concrete leakage
collection channels were designed to
prevent catastrophic erosion of the
embankment material in the event of a
Figure A1.27 Two existing 1.37 m riveted steel inlet pipes from the
leak from either of the new pipes. Walton treatment works into Knight and Bessborough
reservoirs (courtesy Mott MacDonald Ltd)
The reservoirs are owned and
operated by Thames Water within the Thames Valley abstraction system. The reservoirs are retained by
an 11 m high continuous earthfill embankment with a clay core keyed into the underlying London Clay.
The existing twin 1.37 m riveted steel inlet mains are about 100 years old and carried raw water from the
Walton water treatment works into the reservoirs. The existing inlet pipes pass through the embankment
and cross the puddle clay core. Leakage was first observed in the embankment in 2007.

Conduit type/material/use
The conduits are twin 1.37 m diameter circular pipes constructed from riveted steel. The conduits
function as reservoir inlet culverts near crest level and function as water inlets from Walton water
treatment works.

Reason for intervention

Leakage was first observed in the embankment in 2007 near the steps to the valve tower and the longer
of the two pipes was decommissioned. The leakage was detected from one of the couplings on the longer
of the two existing pipes, although due to the age of the pipework it is considered likely that other
smaller leakages were also present.

A drop test carried out on the pumps in Walton water treatment works confirmed the pipe was leaking.
Thames Water undertook their own assessments and subsequently decided to include these works in the
capital improvement works to the networks in south London.

190 CIRIA, C743

Remedial works
The designer, client and delivery partner worked closely together to identify the following options:

1 Expose and repair leak on the inlet. It was considered possible that the leaking may have been
1
located at one of the couplings thought to be close to the downstream embankment face, which

2
could be accessed by means of excavation and repaired. The option was rejected due to concerns
over the quality of the remaining pipe length and the quality of the second pipe, which could be
subject to another failure in the near future.
2 Non-structutal linings such as epoxy or polyurethene spray or cement mortar lining, rejected due

3
to concerns of the structural integrity of the existing pipes.
3 Sliplining with PE, steel, rubber or GRP liner pipe, rejected due to requirement for a deep sheet
piled excavation at the top and foot of the embankment to allow access to the existing pipe.
4 New below ground pipe and grout existing pipes, rejected due to significant excavation requirements.
5 New above ground pipe and grout exisiting pipes, adopted as the preferred solution.

The main risks considered during the design were:

4
‰‰ condition of the existing pipe was unknown
‰‰ close to public road at the toe of the embankment
‰‰ environmental constraints relating to the SSSI designation of the reservoirs
5
‰‰ the requirement to maintain flat vehicle access along the dam crest

6
‰‰ apparent very loose condition of the embankment fill
‰‰ space requirements for construction compound, material storage and pipe access.

The proposed design involved the construction of twin DN 1300 mm welded steel pipes constructed up

7
to the reservoir embankment slope supported by cradles at regular intervals (see Figure A1.28). Due to
the loose condition of the embankment fill, the cradles were constructed on a system of Grundomat piles
driven through the embankment
fill into the underlying river terrace
deposits with cast in situ pile caps.
A leakage collection channel was
constructed under each pipe to
8
prevent erosion of the embankment
in the event of a pipe burst. A
connection was made with the
existing pipe at the reservoir end
to avoid extensive works within the
9
reservoir itself. Vehicle access along
the dam crest level was provided by
recessing the pipes into the dam crest
and providing sealing details around
10
the pipes that connected into, but

11
minimised disruption of, the existing
Figure A1.28 The constructed scheme (courtesy MG JV) clay core of the embankment. The
existing pipes were filled with grout.

Lessons learnt
‰‰ The decommissioning of existing pipes and construction of new pipes above ground can provide 12
a safe and efficient solution rather than sliplining or non-structural lining of defective pipe within
the embankment.
‰‰ If a reservoir is to be kept in operation, the construction of new culverts through an existing
embankment can be undertaken by using surface pipes, cutting through the embankment only A1
above the maintained TWL.

Dams and reservoir conduits 191

‰‰ The risk of future leakage from large diameter pipes over an existing earthfill embankment can
be mitigated by a concrete leakage collection channel beneath the pipes to prevent erosion of the
embankment. A critical element for the design of such channels is the likely maximum time over
which such leakage could remain undetected, and operation considerations become important.
‰‰ Proprietary pile systems are available, which can provide economic and effective foundations being
installed safely while working on a grass embankment slope of 1 in 3.

192 CIRIA, C743

Case study A1.9 Loch Lebgei and Dubh
Loch reservoirs 1

Background 2
Discontinuance works were carried out at Loch

3
Lebgei and Dubh Loch Reservoirs in 2013 (see
Figure A1.29).

An initial examination and identification of the

4
poor condition timber culverts at both these
disused reservoirs, and a subsequent inspection
by an independent reservoir inspecting
engineer, helped the reservoir owner recognise
the risks that a blocked or collapsed culvert
could pose. The problem was solved by
breaching the dams at both reservoirs. 5
The timber culverts were used to discharge

6
Note the stone pitching scour protection along base of channel
run-off from the natural lochs at both sites
while trapezoidal openings were excavated Figure A1.29 Completed trapezoidal channel cut through peat
dam at Dubh Loch reservoir (courtesy MacKenzie
down to original ground level through the low Construction)
level embankment (peat) dams. Run-off was

7
then re-directed through the new openings and the timber culverts blocked at the upstream end.

These two remote reservoirs in the Knapdale Forest of Argyll had not been used for many decades. The
upstream sluices on the outlet culverts were thought to have been removed in the 1950s, leaving the
reservoirs empty, but with natural lochs remaining at both locations.

There is no vehicle access to either reservoir, the nearest access point being Loch na Faoilinn reservoir, 8
about two kilometres away.

9
In 2007 the earth embankment dams and the timber box culverts were found to be in a very poor condition.

Both reservoirs were then formally registered with Argyll & Bute Council, and classified as ‘abandoned’,
under the Reservoirs Act 1975.

Conduit type/material/use 10
The conduits were provided as low level outlets for operational drawdown and were constructed from

11
timber. They were rectangular in cross section with the following dimensions:

‰‰ Loch Lebgei – 300 mm high by 400 mm wide internally by 14 m long.

‰‰ Dubh Loch – 280 mm high by 380 mm wide internally by 15 m long.

Reasons for intervention

Examination of the outlet culverts at both these disused reservoirs revealed that the timbers were in a
12
very poor condition and were in danger of collapse. Also, the low level peat dams at both reservoirs were
in a very poor condition.

The risk of a culvert collapse or blockage of either culvert by a fallen tree from the surrounding forests
A1
Dams and reservoir conduits 193
was highlighted, which would cause either
reservoir to fill with water to the overflow weir
level. This would lead to the risk of failure
of the poor condition peat dams and an
uncontrolled discharge of water and peat down
a steep hillside into Gleann Loch reservoir.

A reservoir inspecting engineer carried out a

statutory inspection of both reservoirs under
Section 10 of the Reservoirs Act 1975 on 1 April
2009. The engineers’ report contained the
recommendation to “maintain the flow continuity
of the bottom outlet”.

Note stone pitching scour protection along base of channel

Remedial works Figure A1.30 Completed trapezoidal channel cut through peat
dam at Loch Lebgei reservoir (courtesy MacKenzie
In 2010, the reservoir owner identified three Construction)
possible solutions:

1 Reinstate both reservoirs to full operational status.

2 Insert an HDPE pipe liner into both timber culverts and retain the reservoirs as ‘abandoned’ assets.
3 Permanently breach the dams and ‘discontinue’ both reservoirs.

Point 1 was discounted on cost. Also, the water that could be retained at these reservoirs is within the
catchment of a lower reservoir owned by the same operator.

The risks associated with pushing a pipe into a fragile timber culvert, together with the need for an upstream
intake grille and ongoing inspection and maintenance, meant that point 2 was discounted in favour of point 3.

Also, as both dams were only low level earth embankments, permanently breaching both sites and
handing them back to the landowner with no future inspection or maintenance liabilities was seen as the
most cost-effective and favourable long-term solution.

The discontinuance works were carried out in August 2013.

A QCE was appointed to supervise and certify the works, in accordance with Section 14 of the Reservoirs
Act 1975.

A trapezoidal channel, with a 2.5 m wide base (same width as the overflow weir) and side slopes of 2H:1V,
was excavated through the peat dams at both reservoirs (see Figures A1.30 and A1.31).

The stone pitching from the upstream face

of the dams was reused as protection on the
upstream area, the channel base, and the lower
slopes of the channel sides.

The location of the breach channels was

planned such that the existing timber culverts
were retained to allow undisturbed discharge
from the natural lochs while the breached
channels were excavated. The flow was then
diverted through the new openings in the
dams and the upstream ends of the timber
culverts were permanently blocked with
Figure A1.31 Final stage of work to cut trapezoidal channel excavated material (peat).
through peat dam at Dubh Loch reservoir
(courtesy A Ross)

194 CIRIA, C743

Lessons learnt
‰‰ The works were postponed from 2012 to 2013 due to the requirement of the SEPA for a licence
under the Water Environment (Controlled Activities) (Scotland) Regulations 2011. 1
‰‰ There was some difficulty in getting plant and operatives to these remote upland reservoirs, over
very soft peaty ground and steep terrain and two kilometres from the nearest access track. The
contractor hired a tracked all-terrain vehicle to transport personnel to and from the site daily, and
used a tracked excavator and tracked dump truck for moving the peat. 2
3
4
5
6
7
8
9
10
11
12
A1
Dams and reservoir conduits 195
Case study A1.10 Napton reservoir

Background
Napton reservoir was constructed in 1814 and is retained by a 6.1 m high × 850 m long homogeneous
embankment dam.

Following deterioration of the downstream culvert, it was strengthened with gunite. Upstream control
was established at the same time by construction of a wet well at the upstream end (see Figure A1.32).

Figure A1.32 Cross-section through draw-off works (courtesy Canal & River Trust)

Nothing has been published elsewhere on this intervention, which took place in 1985. Note that no
instrumentation was present at the time.

Conduit type/material/use
The outlet comprised an upstream curved brick retaining wall a short length of 180 mm diameter CI pipe
with a downstream valve at the head of a human-entry sized brick culvert under the downstream shoulder.

Reasons for intervention

In 1977, the downstream culvert brickwork was identified as bulging. By 1982, it was showing signs of
movement. The inspecting engineer advised that it should be shored without delay and repaired, but
that the work did not need to be overseen by a QCE.

The downstream culvert is of human-entry size and visual inspection and assessment was carried out.

Remedial works
In 1967, the wall of the culvert were cracked and leaked when the reservoir was near TWL. In 1985, the
reservoir was drained and a reinforced concrete valve tower (see Figure A1.33) was formed against the
upstream retaining wall. A 300 mm control penstock was provided at the head of the pipe and a 600 mm
diameter guard valve was installed on the upstream side of the chamber to allow it to be drained so that
the penstock could be serviced (see Figure A1.34). The downstream culvert was gunited and a concrete
channel was formed in the invert for the water to flow in. The original valve was retained, but is left
open (see Figure A1.35).

Lessons learnt
‰‰ It could be agreed that the arrangement has stood the test of time.

196 CIRIA, C743

 1
2
3
4
Figure A1.33 Valve tower (courtesy Canal & River Trust)

5
6
7
8
Figure A1.34 Guard and service valves in valve tower (courtesy Canal & River Trust)
9
10
11
12

Figure A1.35 Original valve in gunited culvert (courtesy Canal & River Trust)
A1
Dams and reservoir conduits 197
Case study A1.11 Pebley reservoir

Background
Pebley reservoir was constructed in 1776 and is retained by a 9.5 m high × 152 m long embankment dam
with a bentonite core.

A timber downstream culvert was identified as being in poor condition and erosion of the embankment
fill was taking place. It was grouted in 2012 and replaced with new pipe installed by auger boring. The
rest of the draw-off system was repaired and updated at the same time (see Figure A1.36).

Figure A1.36 Cross-section through draw-off works (courtesy Canal & River Trust)

The reservoir is used for supplying water to the Chesterfield Canal. Water is occasionally transferred via
the conduit to Harthill reservoir, which is in cascade downstream.

The need to carry out work is mentioned in Windsor (2012), which also includes a description of the
reservoir. For more information see Broad (2000).

The CCTV survey was undertaken in May 2009. The results were not conclusive, but the culvert
downstream of the Neptune Shaft appeared to have stone sidewalls with timber planking to the soffit.
There was evidence of voids behind the soffit joints. The full length of the culvert could not be inspected
because of steps and debris in the invert.

A void in the downstream shoulder just below the surface above the line of the culvert was identified
during routine surveillance and investigated by probing early in 2010. The supervising engineer
instructed that the draw-off system should not be used.

A statutory inspection was carried out in April 2010. The condition of the culvert was known by that
stage and repair or replacement was required as a recommendation for safety reasons. The inspecting
engineer was retained as the qualified civil engineer supervising the works.

There are levelling pins along the crest of the dam, but these did not pick up any sign of movement
above the line of the culvert. There are also V-notches, but they were not relevant to the conduit
problems. A few piezometers survive from an earlier investigation.

Once the void had been discovered, the draw-off was no longer used as advised by the supervising
engineer. A temporary emergency plan was prepared, substituting the use of the draw-off by pumps.

This plan was exercised in autumn 2011 (Windsor, 2012), serving the dual purpose of testing the plan
and reducing water levels for further investigation and works.

198 CIRIA, C743

Conduit type/material/use
Before reconstruction:

‰‰ Upstream circular brick culvert 900 m diameter leading to a wet well, the ‘Neptune Shaft’, at the
1
upstream side of the crest. Vertical timber plug valve in the shaft (see Figure A1.37). Downstream

2
a 0.3 m by 0.46 m timber box culvert, possibly with masonry side walls under the crest and
downstream shoulder.
‰‰ The original arrangement for lifting the plug involved an improvised lever system. This
arrangement was replaced with a purpose-made mechanical device.

3
4
5
6
7
8
Figure A1.37 Plug valve in the Neptune shaft in 1998 and 2012 (courtesy Canal & River Trust)

Reason for intervention

9
The deterioration of an aged timber box culvert under the downstream shoulder had led to voids being
formed in the embankment fill when the draw-off was in use by erosion through open joints.

The statutory inspection in 2005 had not raised any concerns about the draw-off.
10
11
The condition of the system was not known. For asset management reasons and taking the advantage of very
low water levels a CCTV survey was subsequently carried out. It was possible to dewater the entire system
without too much difficulty because the outlet is not at the bottom of the valley and an area of dead water
remains, safeguarding fish life and avoiding any risk of discharging silt into downstream water courses.

Remedial works 12
The reservoir is used to feed the canal and for fishing. Discontinuance was not an option. A reliable

A1
draw-off system was required.

The reservoir was emptied by pumping in 2011. A further CCTV inspection was undertaken to attempt
to see a greater length of the downstream culvert.

Dams and reservoir conduits 199

Lining of the downstream culvert was rejected because the diameter of the pipe, which could have been
practically installed, would have reduced the flow rate to too great an extent, prejudicing emergency
draw-down. To seal a liner into the base of the Neptune Shaft would also have been very difficult.

An on-line replacement, such as pipe bursting was not adopted because the geometry of the system did not
provide a suitable line for the heavy cables required to haul the pipe buster and liner through. There would
also have been a risk of voids being left within the dam among the displaced debris from the old culvert.

Excavation was not favoured because of the difficulties of supporting the sides of the trench and the rear
of the Neptune Shaft and in restoring a watertight dam.

An off-line replacement of the downstream culvert using trenchless techniques was adopted. Advice was
provided by the undertaker’s framework consultant, framework contractor and specialist subcontractors.

Human-entry techniques such as open face pipe jacking are no longer favoured for safety reasons. The
generally cohesive made ground of the embankment did not necessitate the use of closed face mini-
tunnelling machines with or without earth pressure balance. Access pits are also required at both ends of
the drive so that the machine can be recovered.

Augur boring from a thrust pit near the downstream toe to the rear of the wall of the Neptune Shaft was
the adopted technique (see Figure A1.38). If the pipe is pushed through the ground without the auger
protruding, there is little risk of creating voids or overbreak. A 336 mm ID steel pipe with a 10 mm wall
thickness was installed in this manner. The Neptune Shaft masonry was broken through to effect the
connection and the embankment fill in this
area was grouted to fill any voids. The steel
pipe was lined with a 6 mm thick CIP liner,
cured using UV light (see Section 8.3.4). The
liner was a full structural and watertight liner.

The thrust pit, which was offset from the line

of the original culvert, was converted into a
manhole. This was connected to the channel
below the old headwall by installing pipework
in shallow open excavation and building a new
headwall.

The upstream culvert, which had been in fair

condition, was lined with a 600 mm diameter Figure A1.38 Augering machine in thrust pit (courtesy Canal &
twin walled pipe and the annulus was grouted. River Trust)

The contractor suggested this approach as a more cost-effective, quicker, better and easier alternative to
brickwork repair and repointing. A new grille was installed.

The Neptune Shaft was repaired as required and a new penstock valve of modern design was fitted in
lieu of the original plug valve. The old valve was retained and will be conserved and put on display with
interpretation in due course.

The old culvert was filled with grout as were the voids in the downstream fill of the embankment.

The work on the outlet was part of a more extensive scheme of tube-à-manchette grouting to seal leaks
under and around the spillway, and the costs are not readily separable.

Lessons learnt
‰‰ It can be possible to replace a conduit satisfactorily using trenchless techniques.

For more information see Broad (2000), Dutton (2001) and Windsor (2012).

200 CIRIA, C743

Case study A1.12 Redbrook reservoir
1
Background
Redbrook reservoir was constructed in 1815 and is retained by a 18.6 m high × 278 m long embankment
dam reputedly with a clay core (see Figure A1.39).
2
3
The existing outlet pipe was lined with a die-drawn structural swage liner to reduce the risks associated
with downstream control of a very old pipe buried under the dam. During installation the liner stuck,
but it was easily cut out and replaced.

4
5

Figure A1.39 Cross-section through draw-off works (courtesy Canal & River Trust)
6
The works described took place between 2010 and 2011.

This intervention was discussed by Brown and Howlett (2011). 7

There is no relevant instrumentation present at the dam.

Conduit type/material/use
8
Before the works started there was an upstream masonry culvert 1.5 m wide × 1.8 m high.

The CI outlet pipe had a diameter of 300 mm and passed through the core of the dam with no upstream
control. Downstream of the core it was contained in a culvert 900 mm wide by 2500 mm high. This
9
culvert was originally of stone masonry, but was lined with brick in 1920.

At the downstream end of the culvert there was a valve house with two valves in series.
10
Reasons for intervention
In 2008, following a statutory inspection, the inspecting engineer advised that “the outlet pipe passes
through the core of the dam with no upstream control. This is considered to pose a risk to the integrity of the dam.”
11
The engineer recommended for safety reasons that the undertaker inspect the CI outlet pipe and then
fit it with an internal structural lining.

In order to access the upstream end of the culvert to carry out the work, the reservoir level had to be reduced
greatly. A steel piled cofferdam was required to hold back the silt and to provide a dry working area.
12
Remedial works
Lining the pipe had been specified for safety reasons.
A1
Dams and reservoir conduits 201
The solution, agreed with the QCE was developed by the undertaker with advice from its framework
consultant, framework contractor and specialist subcontractors. A similar draw-off system had been lined
earlier at nearby Swellands reservoir. The intended CIP liner could not be used because of the change
in cross-section between the pipes and the upstream culvert. A slip-liner had been used at Swellands.
This approach was not applicable at Redbrook because the reduction in pipe diameter would have
compromised emergency draw-down capacity. A 280 mm diameter rolldown/swage liner was specified.

Swage lining involves the temporarily reduction in diameter of the liner by pulling it through dies. The
liner will then slowly return to its original size. Before this occurs, it can be drawn through the existing
pipe using a winch. Once in place, the liner expands until it is in close contact with the existing pipe. The
first attempt to line the pipe at Redbrook failed when the liner became stuck before it was fully installed.
This is not uncommon in the gas industry, where swage lining is often employed and a solution to the
problem has been developed. A slot can be cut along the length of the liner using an internal cutter. The
liner can then be pulled out. This was done and a second attempt was successful.

The liner was concreted into the upstream culvert. Care was needed to ensure that the pipe did not
float during this operation. An upstand was provided at the upstream end with a flange to facilitate the
addition of an upstream valve in the future if required.

The cost of this and other works at the reservoir was about £600 000.

A new DI pipe was installed strapped to concrete plinths in the downstream culvert. Two new valves were
fitted at the downstream end.

Lessons learnt
‰‰ This was an appropriate solution and gives confidence in the suitability of the swage lining process.
‰‰ A slight ochreous seep between the old pipe and the liner has developed since refilling. This is of
no consequence because the flow rate is minimal and the seepage path along and of minimal cross-
sectional area.

202 CIRIA, C743

Case study A1.13 Rudyard reservoir
1
Background
The reservoir, constructed by John Rennie in 1797–1798, has a puddle clay core, a cut-off trench
and many other features, which became standard in later dam construction. It has required little
2
intervention over the years to upgrade it to modern standards.

By the 1980s the upstream valve

was ineffective and upstream
3
closure of the pipes through the

4
core was called for by the inspecting
engineer (see Figure A1.40). The
system was renewed in 1992 with
a similar layout, but using modern
materials.

This intervention took place in

1992.
5
6
No other published work is
available.

There is a V-notch at the dam

monitoring very longstanding
Figure A1.40 Original upstream valve operating mechanism (courtesy Canal
& River Trust)
seepage, but this is not relevant
here.
7
Conduit type/material/use
As-built, there was a sloping upstream timber paddle operated from the surface by inclined rods at
8
the head of a 1.65 m diameter sandstone masonry culvert under the upstream shoulder. Twin 300 mm
diameter CI pipes passed through the core with valve at their downstream ends in a dry well located
at the rear of the crest of the dam. There is a masonry culvert under the downstream shoulder. This is
usually nearly full of water retained by a compensation weir downstream. 9
The outlet is used as required for supplying water to the Trent and Mersey Canal.

Reason for intervention 10

The 1988 statutory inspection report noted that the drawbars for the upstream valve were too frail to

11
be operated and leakage was suspected past the upstream valve into the upstream masonry culvert.
The inspecting engineer stated that it was essential that a means of upstream closure of these pipes
be provided lest a pipe collapse or a joint defect lead to rapid erosion of the core. The engineer
recommended for safety reasons that the draw-off be sealed or if they could be shown to be safe, they be

12
provided with a reliable means of upstream closure.

There is vehicle access to the dam and a spiral staircase down the dry well.

Remedial works
A draw-off was required operationally and for lowering the level in the early stages of an emergency
A1
Dams and reservoir conduits 203
before pumps could be mobilised. It
was not recorded if alternatives to
repairing the original system were
considered at the time nor from
whom advice was sought other than
from the QCE.

In order to provide an upstream

closure, it was intended to jack a
900 mm diameter concrete pipe
from the dry well (see Figure A1.41)
to replace the upstream culvert,
to grout in twin 300 mm DI pipes
through the core area and replace
the upstream valve. This would
have been a replication of the
original layout in modern materials.
It was found that the upstream Figure A1.41 Dry well (courtesy Canal & River Trust)
culvert was in excellent condition,
so it was retained and lined with 825 mm GRP pipes, with appropriate type of grout being used on the
annulus and CI pipes.

The upstream valve was replaced by a new DI unit with new inclined rodding and a vent pipe in a new
duct under the pitching.

The work was not entirely satisfactory and water was still able to bypass the upstream valve. In 1994,
the upstream valve was adjusted and appropriate waterproofing was applied to all the joints in the
GRP liner.

Had this work been done today, a CIP liner would have provided a more satisfactory solution to the
upstream culvert, but would have been difficult to introduce due to the change in diameter. Had pipe
jacking still been considered as an option to replace the conduit through the core, the whole length of
concrete/masonry culvert could have been lined with a single liner, before installing the pipes through
the core area. These would have been bigger to better manage emergency draw-down.

Lining the pipes that pass through the core with CIP liners due to pipe jacking minimised the loss in
diameter, but would have failed to address the problems with the upstream culvert. Today, it might be
possible to line both the core pipes and the upstream culvert with a CIP system, but discussions with
specialists would be needed to explore this possiblility.

The operating rods for the upstream valve had proved not fit for purpose and being buried in ducts
under the pitching were difficult to modify, so in 2013 to 2014 a replacement hydraulically operated
upstream valve was fitted by divers.

Lessons learnt
‰‰ Fully filling the annulus behind a liner is difficult. It is not possible to determine how effective
grouting has been. In this case, ingress into the lined culvert can be seen and measured, but in
other cases the defect may remain hidden.

For more information see Binnie (1987) and Earp (1995).

204 CIRIA, C743

Case study A1.14 Shon Sheffrey reservoir
1
Background
Shon Sheffrey is an impounding reservoir in Wales with a Pennine-type, puddle clay core embankment
dam. The embankment is 14.72 m high and 250 m long completed in 1895.
2
3
There is a culvert through the dam that is brick lined with two types of brickwork, not engineering type,
with ash or lime mortar. The tunnel is circular about 1.7 m in diameter with a rough benched concrete
floor. The culvert has an end plug and a 13.5 cm diameter CI supply pipe running through it. The pipe
may have lead-run joints, and does not appear to be secured.

Three reservoir keeper visits are carried out a week. Five V-notches are present in the tunnel, four of
which are bolted to the culvert wall. 4
Clay ingress was apparent into the culvert with multiple water leaks (see Figure 3.1), which was detected
by human-entry inspection.

Targeted tube-à-manchette cementitious grouting completed June 2013 from the surface. The reservoir
5
was 700 mm below TWL.

Water and clay ingress into the tunnel was ongoing in recent years and accelerated in 2010, with the
discharge of jets of water. The deterioration was initially apparent at TWL then subsequently at lower
6
reservoir levels.

Two new piezometers were installed in June 2013, which were reading constant levels. Five V-notches
were installed on the side walls of the tunnel. A Willowstick leakage survey was carried out in 2010. The
survey identified two leakage paths – a high level leak into the tunnel joints and a deeper leak, possibly
7
through the clay core.

Conduit type/material/use
8
Brickwork tunnel with ash or lime mortar, carrying a 135 mm diameter CI raw water supply main using
potable water.
9
Reason for intervention
There was concern regarding internal erosion along the leakage paths, which if left untreated, could
affect the safety of the dam. 10
Remedial works
Grouting was deemed to be the appropriate solution. 11
Targeted tube-à-manchette grouting from crest level. Some grout was detected in the culvert during the

12
grouting operations (see Bruggemann and Francis, 2014).

Lessons learnt

A1
‰‰ The site work started before spring and the ambient temperature dropped below 4oC on a number
of occasions and grouting was suspended. Grouting work should be programmed to avoid periods
when low ambient temperatures are likely to occur.

Dams and reservoir conduits 205

Case study A1.15 Tardebigge reservoir

Background
Tardebigge reservoir was constructed in 1815 and is retained by an 18.6 m high × 460 m long
homogeneous clay embankment dam (see Figure A1.42).

In 1984, the brick culvert under the dam collapsed under the downstream shoulder. A shaft was sunk
to investigate the defect and the culvert was lined in both directions. The original upstream valve was
replaced by a new control valve in the shaft with upstream diver operated guard valve at the upstream
end of the conduit (see Figure A1.43).

Figure A1.42 Cross-section through pre-1987 draw-off works (courtesy Canal & River Trust)

Figure A1.43 Cross-section through present day draw-off works (courtesy Canal & River Trust)

The incident took place in 1984.

Nothing has been published elsewhere on this incident.

At the time there was no instrumentation present at the dam.

Conduit type/material/use
Before the incident in 1984, the draw-off comprised a 600 mm diameter brick culvert under the dam
with a 250 mm diameter teapot lid valve at the upstream end. Vent pipes were provided buried under
the upstream face of the dam to release any air trapped in the culvert.

The outlet was/is in regular use for supplying water to the canal as required for lockage purposes.

206 CIRIA, C743

Reason for intervention
In 1984 the culvert collapsed 15 m downstream from dam crest. There was a 2 m deep hole 2.5 m in
diameter at the surface. A 2.5 m diameter shaft (shaft 7) of segmental concrete construction was sunk at
1
the site of the collapse and the culvert connected in. A CCTV survey of the 52 m long culvert upstream
from the shaft was carried out, identifying that it was in poor condition in places.

A statutory inspection in 1985 instructed that the culvert be repaired or replaced within the next three years. 2
Vehicle access to the dam is available. It was necessary to put a temporary bridge across a canal lock

3
for heavy plant, and temporary working platforms were put in place to access the upstream end of the
conduit (see Figure A1.44).

Remedial works
It is not recorded what alternatives were considered at the time or from whom advice was taken. 4
In 1987 the culvert was lined
with a 300 mm diameter liner
with integral screw joints. It was
introduced from the new shaft, 5
had a seal at the upstream end
to butt up against the rear of the
smaller diameter valve block,
carried its own grout lines and
was grouted in place with grout.
6
The culvert downstream from
the new shaft was lined with a
450 mm diameter PE pipe with
the annulus grouted.
7
8
The teapot lid valve was removed
and replaced by a diver operated Figure A1.44 Access to the upstream end of the conduit during construction
(courtesy Canal & River Trust)
upstream valve, intended for
maintenance purposes only, fitted to the original masonry valve block. In the new shaft MH 7, a benched
channel was created and a new gate valve for flow control was fitted to the downstream end of the pipe.
This valve was operable from the surface.
9
Today, consideration would have been given to retaining the original upstream valve as a guard valve
operable from the surface.

The provision of a larger diameter liner would also be investigated. This reservoir needs auxiliary
pumping to meet the undertaker’s emergency draw-down criteria.
10
Lessons learnt
‰‰ The valve in Shaft 7 was underspecified and had to be replaced in 2003–2004 by a new needle
11
valve with a knife type guard valve immediately upstream.

12
‰‰ The 1987 diver operated guard valve proved ineffective because the old vent pipes had corroded
and bypassed the valve. When this was resolved, the upstream end of the pipe still proved
impossible to seal, so the reservoir had to be drained and a small cofferdam driven. It proved that
the 1987 work was not of a high standard and that the pipe flange had been damaged.

A1
‰‰ Valves need to be of good quality and correctly specified. Cheap valves are a false economy.

Dams and reservoir conduits 207

Case study A1.16 Torside reservoir

Background
The draw-off pipes buried beneath the embankment pulled apart soon after construction of the
embankment. Despite several repairs they were abandoned and a tunnel constructed through the
abutment instead.

Torside is second in a cascade of five reservoirs constructed in the Longdendale valley for the City of
Manchester by J F Bateman. Construction of the dam began in 1849 and following several setbacks and
incidents was eventually completed in 1865. Two CI outlet pipes were sunk a considerable depth into
solid ground below the base of the embankment and embedded in concrete except at the centre where
they were supported by puddle.

For further information see Charles et al (2014), Bateman (1884) and Binnie (1981).

Conduit type/material/use
The conduits were constructed from CI pipes.

Reason for intervention

In 1851 cracks were observed in the outlet pipes and the damaged pipes were strengthened internally
with cylinders assembled inside them. On 17 November 1854, when the reservoir was within three
metres of being full, fractures on the outlet pipes occurred and a considerable amount of water escaped
near the foot of the outside slope of the embankment. The south pipe, which appeared the worst
affected, was shut down and the north outlet used to draw-down the reservoir level.

Leakage from the pipes, if allowed to continue, would have caused failure of the embankment by
internal erosion.

On examination of the south range it was found that the discharge pipes had ‘drawn asunder’ (Bateman,
1884). Similar, though less severe, damage was found in the north range and there was some localised
settlement of the embankment surface. About 4.5 m of pipe were crushed into an elliptical form. Later in
the year the embankment ‘stretched upon its base’ and an extension of 1.5 m was measured in the south
outlet pipe.

In 1855, further movement occurred, which led Bateman to abandon the two outlet pipes under the
dam. The probable cause of movement was the existence of a bed of hard clay 1.5 m to 8 m thick in the
valley, which allowed movement of pipes with the base of the embankment as it consolidated.

Remedial works
Repairs were made to the pipes by installing smaller pipes inside the originals to span the gaps, but
eventually Bateman abandoned them. A new 2.5 m diameter discharge tunnel containing 1140 mm and
610 mm diameter pipes was constructed in the right abutment rock. Bateman was concerned that the
puddle clay core of the dam may have been compromised so a new puddle trench was sunk into the clay
near the foot of the upstream slope and the slope was lined with a clay blanket.

The pipes were not repaired but abandoned.

Lessons learnt
‰‰ Conduits within the foundation can move with the embankment as it consolidates, fracture and
pull apart.

208 CIRIA, C743

Case study A1.17 Un-named new (raw
water storage) reservoir 1

Background 2
A new water treatment works has

3
recently been constructed to increase
potable water supply. Abstracted
river water is stored in a new storage
reservoir before being treated and

4
transferred to the supply network.
The new reservoir was formed with
an oval earth-filled embankment with
a cut/fill balance. The outlet pipe was
designed as a 670 mm PE pipe within
an outer 750 mm diameter segmental
concrete pipe with a concrete haunch 5
and with compacted clay backfill above
(see Figure A1.45). The haunching
was initially constructed using no-fines
concrete, which was an unsuitable
material. The error was spotted and
6
Figure A1.45 A 750 mm diameter outer concrete outlet pipe, laid on dry-
remedial measures were instigated mix concrete haunching (courtesy Mott MacDonald Ltd)
during construction.

The reservoir is about 300 m by 200 m and roughly oval. It stores a total of around 300 000 m3 of water.
7
It is constructed on the top of a hill, so it is non-impounding and the inlet is pumped. The oval dam is

8
formed in stiff clay by a combination of excavation to around 6 m depth and the construction of new
embankments from 5 m to 7 m high. Upstream (inner) slopes are 1:4 and downstream (outer) slopes
are 1:3. The dam is of homogeneous construction, but with a re-worked clay ‘liner’ over the natural
(cut) sections. The downstream slope is formed of wind-blown sand, which had to be removed from the
surface before construction began.

Construction of the reservoir fill started in 2011 and was fully completed in 2013. The reservoir was 9
filled completely in the same year.

Conduit type/material/use 10
The outer conduit is a 750 mm diameter segmental concrete pipe with spigot/socket connections, set
in a concrete haunch. Inside the concrete pipe is a 670 mm OD pipe held centrally within the bore.
Concrete haunching was to be class C20/25 mass concrete. To allow for movement in the outer pipe,
joints were formed in the pipe bedding using bentonite-enhanced sand with permeability no greater
than 3 × 10 -10 m/s.
11
12
Backfill above the haunch was specified as selected clay fill, having a permeability less than 10 -8 m/s.
Concrete stanks were to be provided about every 10 m to inhibit flow along the outside of the concrete
pipe. At the downstream end is a sand filter, 18 m long, with a sand filter collar.

Reason for intervention

In July 2012, a site inspection revealed that the pipe bedding as laid was in fact a semi-dry C20/25
A1
Dams and reservoir conduits 209
concrete mix. This was not acceptable as the pipe bedding was intended to be less permeable than the
surrounding clay. In addition there should be a satisfactory bond between the pipe and the bedding
concrete. It was considered that the sand filter on the downstream end of the pipe would have prevented
any catastrophic washout of any material, but the reservoir leakage rates would have been likely to
exceed acceptable levels.

The contract drawings indicated that the pipe surround should be a C20/25 mass concrete. The pipes
were laid by a specialist pipeline subcontractor who was engaged for all the other pipeline works, for
which a switch to dry-mix concrete haunching was acceptable. However, this subcontractor did not have
particular experience of reservoir pipelines.

The fault was observed by the site supervision team and was immediately brought to the attention of
the main contractor and the designer, and a corrective action plan was made. No other investigations or
assessments were required specific to this issue.

Remedial works
At the time the issue of the incorrect bedding material was identified, most of the outlet pipe had been
laid, but little backfill had been placed over the pipe. Removal of the entire length of the laid pipe would
have had adverse programme implications. An assessment was made that concluded:

‰‰ the bentonite enhanced sand ‘stanks’ in the haunching had the effect of preventing continuous flow
paths existing along the haunching
‰‰ the downstream longitudinal sand filter
‰‰ the downstream filter collar at the end of this longitudinal filter.

It was highly unlikely that the dry-mix concrete haunching would lead to any loss of serviceability of the
reservoir, let alone a more serious impairent of the structure. So, the full length of the dry-mix concrete
haunching was not removed.

Further measures were introduced

to ensure that an appropriate level
of protection against leakage was
maintained. These measures comprised
the removal of sections of the dry-mix
concrete haunching at certain critical
locations, including the upstream
end of the pipe, and its replacement
with mass concrete, together with the
provision of additional anti-seepage
concrete collars at intervals along the
pipe (see Figure A1.46).

These works were carried out during

construction and the amount of
leakage noted in the downstream filter Figure A1.46 Concrete anti-seepage collars along the conduit length
observation chamber has been negligible. (courtesy Mott MacDonald Ltd)

Lessons learnt
‰‰ A subcontractor with specialist expertise in one area may have a ‘blind spot’ in another area.
‰‰ Adequate supervision should always be provided for safety-critical works, as highlighted here.
‰‰ Design solutions, which include multiple barriers, can produce the most robust solutions as they
may allow for one component to fail without jeopardising the functionality of the whole structure.

210 CIRIA, C743

Case study A1.18 Upper Rivington reservoir
1
Background
Leakage into the brick culvert beneath the Yarrow embankment of the Upper Rivington reservoir
caused the reservoir to undergo an emergency draw-down. Remedial works were required before it
2
could be brought back into supply (Gardiner et al, 2004, Charles et al, 2014).

Subsequent grouting is described by Rigby and Gardiner (2012).

3
The Yarrow embankment of the Upper Rivington reservoir is an earthfill embankment, 12 m high and
270 m long and was built in 1852 to 1857. The incident was recorded on 9 January 2002.

Discoloured water was observed emerging from the culvert by a member of the owner’s staff. 4
The leak was plugged with rags and a wooden board, which was strutted in position off the opposite

5
wall of the culvert. This stemmed most of the flow and limited the loss of material, but had the effect of
pushing the leakage further down the culvert into other weepholes and out of the low wall at the toe of
the embankment. These flows were still coloured, but were dispersed and had less erosive power.

6
An ARPE was called out by the owner to give advice on emergency action.

The reservoir was overflowing by about 50 mm at the time of the incident and rain was forecast.
The overflows of two reservoirs upstream of Upper Rivington were raised with sandbags to provide
additional flood storage. It was estimated that this would provide two to three days storage protection in
the event of the forecast moderately heavy rainfall and much more if the weather remained dry.

To draw-down the reservoir the scours were opened into the culvert, but at a restricted flow so that
7
the temporary plug was not washed away. Pumps were brought to site the following morning and the

8
temporary seal was replaced with a more robust plug and an ‘Acrow’ prop. The scours were then opened
fully, more pumps were brought to site and the reservoir was drawn-down.

The ARPE advised that the embankment was to be inspected three times a day, the culvert was to be

9
checked for structural defects once a day and the draw-down should continue with auxiliary pumping to
a level seven metres below TWL. This level was reached on 18 January. The leakage into the culvert and
from the toe of the dam ceased when the reservoir level was about 5.5 m below TWL.

At the time of the incident only seepage flows into the natural rock spillway and the culvert were
measured. A seepage flow from the roof of the culvert near to the weephole was regularly measured, but
did not change during or before the incident. 10
Conduit type/material/use
A 1.8 m diameter brick culvert was constructed at the foundation level beneath the dam. The culvert
receives scour and compensation flows through a brick bulkhead at its upstream end. The culvert
11
normally conveys the compensation flow to the River Yarrow.

Reasons for intervention 12

The flow into the culvert through a half brick sized weephole was laden with sediment, which was

A1
deposited on the invert. The water flow was at high pressure impinging upon the far wall of the culvert.
It was estimated that about 1 m3 of fill had been removed from behind the culvert wall. If this had been
allowed to continue the culvert would probably have collapsed due to lack of horizontal support. This
may have led to the collapse of the embankment above and failure of the dam by internal erosion.

Dams and reservoir conduits 211

The crest of the dam is about 10 m above the soffit of the culvert.

The culvert could be entered relatively easily from the downstream portal and the base of the spillway cascade.

Remedial works
It was concluded that the source of the leak had to be a fault in the puddle clay core in the embankment.
The solution identified was to grout the core locally.

After a limited borehole investigation and installation of piezometers, a grouting programme was
designed with advice from the ARPE. This programme sought to grout the upstream and downstream
shoulders and the foundation and then the core. The grouting was undertaken during summer 2002
and the reservoir was slowly refilled during the autumn with TWL being reached on 23 February 2003.
During the grouting operation grout was found to have sealed the weephole from which the leak had
emerged indicating that a fault within the clay core was the cause of the problem.

The performance of the embankment was closely watched during this refilling and daily seepage flow
and piezometer readings were taken until November 2003. The reservoir has been inspected, and
readings taken, by the owners staff every 48 hours to date. The incident led directly to the owners
increasing the surveillance on all their embankment dams from three times a week to every 48 hours
and this remains the standard.

Monitoring revealed that leakage into the culvert and the spillway increased when the reservoir level was
higher than one metre below TWL and the reservoir was held below this level until further grouting was
carried out in 2012. Piezometers showed that there was leakage through the rock beneath the dam and it
was considered, following a risk assessment, that the grout curtain should be extended from the culvert
area to the spillway and the rock/clay core interface should be grouted over the full length of the dam.

Lessons learnt
‰‰ The need for regular and frequent surveillance in order to detect a serious defect at a dam while
there is still time to take effective action.
‰‰ The need for an effective on-site emergency plan, which details the actions that should be taken in
the event of serious progressive internal erosion being detected, including means of diverting inflows.
‰‰ The need to have contractors on call that can carry out works at very short notice, overnight
if necessary.
‰‰ The value of a functioning bottom outlet, which allows the reservoir to be lowered rapidly in the
event of a problem.
‰‰ Incidents can develop very rapidly.

212 CIRIA, C743

Case study A1.19 Warmwithens reservoir
1
Background
Warmwithens reservoir near Oswaldtwistle, Lancashire was built to supply a local textile mill near
Accrington. The 10 m high embankment, which impounded the uppermost reservoir in a chain of
2
three reservoirs, was described as “consisting of clay filling on the upstream side of the centre and there is

3
some evidence there having been some form of clay core” (Charles et al, 2014). The draw-off consisted of a
250 mm diameter outlet CI pipe laid beneath the embankment, which was controlled by a valve at
the downstream end. Between 1964 and 1966, a 1.5 m diameter tunnel was constructed through the
embankment to contain new

4
outlet pipes. The original outlet
pipe was plugged at the upstream
end and filled with cement/
bentonite grout from the lower
end (Moffat, 1975).

Serious leakage occurred near 5

to a recently installed tunnel
leading to erosion. This resulted
in catastrophic failure of the
embankment dam in November
1975 (see Figure A1.47).
6
7
This account has been taken
from Charles et al (2014), and is
also discussed in Moffat (1975) Figure A1.47 Breach of the embankment along the line of the recently
and Wickham (1992). constructed draw-off tunnel (from Charles et al, 2014)

Conduit type/material/use 8
The tunnel was formed from 1.5 m diameter precast concrete segments.

Details of failure 9
The first indication of escaped water was detected at 07:30 on 24 November 1970, although the water

10
level recorder indicates it started the evening before (Wickham, 1992). The outflow reached a maximum
within two hours. The dam was breached to foundation level by 13:30. The breach, which occurred over
the line of the new draw-off works, was 20 m wide at crest level and extended down to the tunnel. Large
sections of the concrete tunnel segments were washed out and deposited downstream. Internal erosion
appears to have taken place along the line of the tunnel. This led to a cavity being formed and later
subsidence and eventual overtopping of the crest.

Water impounded by the dam was discharged into the two lower reservoirs. The embankment of Cocker
11
Cobbs was overtopped, but did not fail and the water passed over the spillways of the lowest reservoir.

12
Had a cascade failure of the two lower dams taken place, the resulting flood could have caused serious
damage in Oswaldtwistle.

The incident occurred four years after construction of the new draw-off works.

On inspection there was little evidence of grout adhering to the tunnel segments.
A1
Dams and reservoir conduits 213
Remedial works
The dam was never rebuilt following the incident.

Lessons learnt
‰‰ As far as is known, the embankment had performed satisfactorily for 100 years before the incident.
‰‰ Construction of a new draw-off tunnel through the embankment was likely to have provided
leakage paths along the tunnel.
‰‰ Tunnelling always reduces the stresses near to the tunnel to zero, which may have led to hydraulic
fracture between the lining and ground.
‰‰ Annulus grouting will not necessarily re-establish sufficient earth pressure to prevent hydraulic fracture.
‰‰ This internal erosion incident shows the speed with which this type of failure can develop, and
emphasises the need for careful monitoring and interpretation of seepage flows.
‰‰ Remedial works to embankment dams need to be carefully designed and executed with full
understanding of how the work could change the behaviour of the dam in terms of stress changes
and stability.

214 CIRIA, C743

Case study A1.20 Yeoman Hey reservoir
1
Background
Yoeman Hey reservoir was constructed around 1880 and is retained by a 19.8 m high × 250 m long
embankment dam with a clay core. Considerable leakage into the brick culvert beneath the earth
2
embankment dam started almost as soon as the reservoir was completed and continued until 2008 when

3
the cracks were successfully grouted.

The incident has not been documented elsewhere.

There was no incident as such. Cracks occurred in the culvert at the point where it crosses the concrete
cut-off trench soon after the construction of the dam. A drawing dated 1885 shows the old and new
cracking pattern. This pattern did not change significantly in 130 years and it was assumed that the 4
leakage rate remained about the same. The leakage flow continuously deposited a layer of ochre on the
invert of the culvert, which required cleaning
out from time to time. The leakage was mainly
from the walls and soffit of the culvert, but
there were also small waterspouts from the
5
invert (see Figure A1.48). The leak was obvious

6
to anyone entering the culvert.

The reservoir is retained by an earthfill

embankment about 19.8 m high and 250 m

7
long across the Greenfield Brook with a puddle
clay core and concrete filled cut-off trench.

A new reservoir was constructed in 1967,

which impounds against the downstream
toe of Yeoman Hey. A valve house was built
at the downstream portal of the culvert to
prevent its inundation.
Figure A1.48 Reservoir leakage through the cracked brickwork
culvert (courtesy United Utilities)
8
9
The leakage flow into the culvert was sufficient to need continuous pumping to prevent its inundation.
The flow was monitored by measuring the run time on the pumps.

A survey was carried out in 2005 to identify the source and route of the leakage. This revealed that the

10
flow was coming from the reservoir passing through the upstream shoulder of the dam and flowing
vertically into the culvert via the cracks upstream of the clay core.

Conduit type/material/use
The culvert is constructed of brick and is generally founded on natural ground. It is ‘S’ shaped in plan.
The culvert is surrounded in concrete where it crosses the concrete filled cut-off trench.
11
There are two 61 cm diameter CI pipes in the culvert. One is the scour pipe, which discharges into
the downstream reservoir, and the other is the supply pipe. The pipes are strapped to the invert of the
culvert and to the walls at the bends. 12
No flows are discharged into the culvert.

A1
Dams and reservoir conduits 215
Reason for intervention
In a report in 2006 the last inspecting engineer recommended for safety reasons that the leakage into
the culvert be stopped and any voids found on the outside of the culvert be grouted. This was because an
earlier analysis of the leakage water had revealed that it was carrying clay particles.

The loss of material behind the wall of the culvert could lead to loss of support and collapse of the
culvert. This could then have led to a breach in the clay core and failure of the dam by internal erosion.

The leakage into the culvert was taken into account by the risk assessment that was carried out on
Yeoman Hey dam by the owner’s staff using the New South Wales method (Foster et al, 1998).

The culvert can be inspected at the upstream end from the valve tower in the reservoir, via multiple
ladders, and by stairs and a ladder from the downstream valve house.

The culvert is considered to be a confined space by the owners and gas detection equipment is required
to be carried by those inspecting it.

There is a thick layer of ochre on the invert making it slippery underfoot (see Figure A1.49).

Remedial works
The dam owner’s engineering section
recommended that specialist advice be sought
on the best way to stop the leak. Based on this
advice it was decided that the leak could be
stopped using a hydrophilic grout injected into
the leakage flow.

A hydrophilic epoxy grout material was

injected into the leakage flow without lowering
the reservoir. This stopped the leak almost
entirely and was deemed to have filled any
voids behind the brickwork.
Figure A1.49 Ochre deposited on invert of culvert (courtesy
The culvert leakage was about 120 litres per United Utilities)
minute before the works were carried out and
reduced to less than 20 litres per minute afterwards. This includes the ingress from the downstream
reservoir into the culvert, which has not been resolved.

The method was successful and the owner has used this technique on other leaks, although not in conduits.

The leakage continues to be monitored by the owner’s staff and by the supervising engineer.

Lessons learnt
‰‰ Cracking may occur in a brick culvert if it can settle either side of a hard support such as a rock
outcrop or a concrete cut-off trench.
‰‰ High leakage flows can continue for a considerable time depositing ochreous material on the invert
of the culvert (in this case for over 100 years) and failure of the culvert will not necessarily occur.
However, because the quality and quantity of leakage flow has not changed for a considerable time,
this does not mean that it will not/should not be the subject of a recommendation for safety reasons.
‰‰ Modern grouts can seal a severe leak in a brick, masonry or concrete effectively, from the
downstream side, without lowering the water level in the reservoir.

216 CIRIA, C743

Case study A1.21 Redmires reservoir
1
Background
Construction of the earthfill clay core dam at
Redmires was completed in 1854. The dam is
2
15 m high and 686 m in length.

An inspection report identified problems

with two existing CI draw-off pipes that pass
3
through this Victorian dam. In accordance

4
with the requirements of the inspecting
engineer, the two CI pipes were to be lined
with a structural liner. The technique adopted
was sliplining using PE pipes. Problems were
encountered in grouting the annulus between
the PE liner and the CI pipes. This led to a
delay in the completion of the works, although 5
ultimately the problems were rectified and the Figure A1.50 Upstream end of scour pipes in access shaft
lining process was successful. (courtesy Mott MacDonald Ltd)

This incident is discussed by Walker and Tietavainen (2014). 6

Defects in the draw-off pipes were identified and discussed in the inspecting engineer’s report of July

7
2006, which recommended that “the two pipes [draw of pipes from the (new upstream) shaft to the scour channel]
should be lined with a structural liner”. A contract to undertake these works for safety reasons was let in
early 2009. An initial CCTV survey showed there was an existing CIPP lining inside the CI scour pipes
(see Figures A1.50). This dated from the 1970s.

Pressure jetting of the CI pipes allowed further detailed surveys to be undertaken, which revealed there
was about 300 mm of settlement over a 10 m to 15 m length of both scour pipes. This was roughly where
they pass through the clay core of the dam, as well as a partial blockage on the left hand pipe.
8
9
The main remedial works were also undertaken in 2009, with the grouting done in June. The reservoir
was drawn-down to allow the works to take place.

No instrumentation was employed in the assessment of the initial problem.

Conduit type/material/use 10
The twin 350 mm diameter scour pipes were made of CI. These pipes were exposed in a new access

11
shaft constructed in the upstream dam face to house new upstream valves. They exit the toe of the
downstream slope through a masonry headwall into a scour channel.

Reasons for intervention

The failure (or defect) in the original scour pipes was caused by long-term settlement within the dam
core. The 300 mm of settlement, suggested by the surveys, had occurred over a period of 150 years since
12
construction. It is not known to what extent any leakage had been occurring as a result of this settlement,
but it was clearly a potential, if not an actual concern. In addition, the left hand pipe was partially
blocked by some lightweight concrete that had become lodged in the pipe.

Detailed inspections and assessment of the two scour pipes was undertaken as an initial phase of the
A1
Dams and reservoir conduits 217
remedial works in 2009, although earlier inspections had identified the problems in the first place. The
detailed survey was undertaken by CCTV.

Remedial works
Before the works were undertaken, different methods of lining the scour pipes were considered. These
included:

1 Rolldown/swagelining. This was ruled out as the presence of the historic CIPP lining meant there
was no guarantee of a tight fit.
2 Removal of CIPP, then rolldown. This was ruled out as it could not be guaranteed that the lining
could be fully removed without damaging the CI pipe.
3 Sliplining the pipes using a 280 mm OD PE pipe and grouting the annulus. This was selected as
the preferred option as it was considered the lowest risk engineering solution, even though there
was a reduction in scour capacity.

It was agreed to line each CI pipe with a

black 280 mm OD HPPE pipe and grout
up the annulus (see Figure A1.51).

The HPPE liner pipe was welded in 12

m lengths using a butt fusion machine.
A nose cone was then attached and it
was pulled into the host pipe. Both PE
pipes were pulled into the host pipes
without problem. The pipes were filled
with water and pressure tested.

Link seals were installed at both ends

of each pipe, and a 50 mm diameter
grouting hose was connected into
the outer (CI) pipe at each end, one
for grout supply and one for venting. Figure A1.51 Remedial works under way – note the black HPPE pipe
inside the original CI pipe (courtesy Mott MacDonald Ltd)
Grouting started by injecting from the
downstream end a very dilute ‘pre-grout’ wash through the annulus, which was completed successfully.

However, problems arose during the injection of the main grout mix. From the time taken to mix the
grout, together with the grout mix consistency meant that the grout was injected in two batches. The
second batch required a high pressure to force the grout in. On pipe A, the high grout pressure caused
a failure in the inner pipe and a large bulge appeared on the inside of the pipe. On pipe B the upstream
link seal was blown out.

An investigation into the causes of the difficulties during these remedial works highlighted the following
contributing factors:

‰‰ The high grout pressures may have been caused by a blockage somewhere along the pipe annulus.
‰‰ The error in batching the grout mix may have led to this, or at least exacerbated it, so that high
pressures built up in the annulus.
‰‰ There was no pressure gauge on the grout pump to give early warning of the pressure build-up.
‰‰ Under these conditions the inner HPPE pipe was damaged.

Pipe A was not usable, and had to be repaired. As the bulge in the pipe was close to the downstream end,
it was feasible to dig the affected area out in an open excavation supported by sheet piles. This was done
and the damaged section of HDPE pipe (plus the CI pipe to that point) was dug out and replaced.

218 CIRIA, C743

Lessons learnt
‰‰ Ensure the maximum external pressure rating of the sliplined pipe is sufficient for the maximum
head of the reservoir and grouting pressures. 1
‰‰ Fill the pipe with water to reduce floatation and the risk of buckling.

2
‰‰ Use spacers to control floatation.
‰‰ A grout mix with slow set time, good flowability and low solids content is preferred.
‰‰ Reduce the grout pumping length, where possible, with multiple injection points.
‰‰ Pump grout in a continuous operation to prevent blockages.
‰‰ Use a pressure gauge to monitor grout delivery pressure, and to prevent structural damage (see
Figures A1.52 and A1.3) caused by high pressures). 3
4
5
6
7
Figure A1.52 Damage to structural lining pipe (note bulge in top right of pipe)
caused by excessive grout pressures (courtesy Mott MacDonald Ltd)
8
9
10
11
Figure A1.53 Link seal blown out of upper end of pipe by high grout pressures
(courtesy Mott MacDonald Ltd)
12
A1
Dams and reservoir conduits 219
 Core and Associate members
AECOM Ministry of Justice
Arup Group Ltd Morgan Sindall (Infrastructure) Plc
Atkins Consultants Limited Mott MacDonald Group Ltd
Balfour Beatty Civil Engineering Ltd Mouchel
BAM Nuttall Ltd MWH
Black & Veatch Ltd Network Rail
Buro Happold Engineers Limited Northumbrian Water Limited
BWB Consulting Ltd Rail Safety and Standards Board
Cardiff University Royal HaskoningDHV
Environment Agency RSK Group Ltd
Galliford Try plc RWE Npower plc
Gatwick Airport Ltd Scottish Water
Geotechnical Consulting Group Sellafield Ltd
Golder Associates (Europe) Ltd Sir Robert McAlpine Ltd
Halcrow Group Limited SKM Enviros Consulting Ltd
Health & Safety Executive SLR Consulting Ltd
Heathrow Airport Holdings Ltd Temple Group Ltd
High Speed Two (HS2) Thames Water Utilities Ltd
Highways Agency Tube Lines
HR Wallingford Ltd United Utilities Plc
Imperial College London University College London
Institution of Civil Engineers University of Reading
Lafarge Tarmac University of Sheffield
Laing O’Rourke University of Southampton
London Underground Ltd WYG Group (Nottingham Office)
Loughborough University
Maccaferri Ltd April 2015
 C743
Reservoir owners and safety engineers have long known or suspected that conduit problems
have led directly to a significant number of dam failures and incidents. The mechanisms and
modes of failure are well understood, but what is lacking is a single guidance document
providing comprehensive information on inspection, monitoring, investigation, maintenance
and repair of conduits in their various forms specific to UK reservoirs and dams.

Dams and reservoir conduits. Inspection, monitoring, investigation, maintenance and repair
A conduit, in the context of this guide, is considered to be a means to convey water from a
reservoir through, around or under an embankment dam in a controlled manner, or to
provide access.
This guide covers a wide range of aspects to conduits such as forms and materials used
in their construction, failure modes, methods for inspection and assessment (ie surveys,
integrity testing, and leakage investigations), and the instrumentation that can be used
for such testing.
There is also guidance on identifying the optimum solution with respect to the renovation,
Dams and reservoir conduits
repair, replacement or abandonment of conduits.
Steps to take in the event of a significant failure (ie emergency actions) are given and the Inspection, monitoring, investigation,
importance of accurate record keeping for current and future management of conduits
are included.
maintenance and repair
This stand-alone guide is aimed at reservoir owners and engineers concerned with
reservoir safety. In addition, reference to other documents from the UK and those
highlighting practice overseas has been provided where appropriate.

CIRIA

C743 9 780860 177487