Chapter 2 Hazards to Engineering Lifelines in Christchurch

Please Note: The contact details given in this 1997 publication are no longer correct. Current contact information (2010) is: New Zealand Centre for Advanced Engineering Private Bag 4800, Christchurch 8140, New Zealand

Phone: +64 3 364 2478 Fax: +63 3 364 2069 E-mail: [email protected] Web: www.caenz.com Risks Realities

A Multi-disciplinary Approach to the Vulnerability of Lifelines to Natural Hazards

Report of the Christchurch Engineering Lifelines Group

Centre for Advanced Engineering University of Canterbury Christchurch New Zealand ISBN 0-908993-12-9

First printing, November 1997

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, transmitted, or otherwise disseminated, in any form or by any means, except for the purposes of research or private study, criticism or review, without the prior permission of the Centre for Advanced Engineering.

Project Manager John Lamb, Christchurch Engineering Lifelines Group.

Editorial Services, Graphics and Book Design Charles Hendtlass and Úna O'Grady, Centre for Advanced Engineering.

Cover Photograph by Michael Provost Photography. View looking west from above Banks Peninsula showing: Lyttelton Port; the Port Hills; the Estuary; the Sewage Treatment Works; the Central Business District of Christchurch; Christchurch Interna- tional Airport; the Southern Alps.

Cover Design and Printing Print City, Christchurch.

Disclaimer It will be noted that the authorship of this document has been attributed to the many individuals and organisations involved in its production. While all sections have been subject to review and final editing, the opinions expressed remain those of the authors responsible and do not necessarily reflect the views of the Centre for Advanced Engineering. This report by its nature is general in its application and subjective in its recommendations and is intended as an initial guide only. While every effort has been made to ensure the accuracy of the report, no liability whatsoever can be accepted for any error or misprint. Any persons concerned with any of the issues raised in this report should consult their professional advisers before undertaking any action, and no liability whatsoever shall attach to any party associated with the report for any reliance that any person may purport to place in the report. Centre for Advanced Engineering

Establishment Function The Centre for Advanced Engineering was founded in The Centre is managed by a Board of Directors compris- May 1987 to mark the centenary of the School of Engi- ing representatives from industry, the engineering pro- neering at the University of Canterbury. It was estab- fession and the University of Canterbury. Chairman of lished by means of an appeal fund launched in conjunc- the Board is Mr Brian Wood of Christchurch. The Board tion with the centennial celebrations. To date approxi- selects the title for each project undertaken by the Centre mately $2 million has been raised, contributed by 150 and approves the level of funding. A Steering Committee corporate donors and 750 individual donors. The earn- is then appointed, initially to carry out detailed planning ings from this capital sum assist in funding the activities for the project and then to provide overall direction. The of the Centre. A 10th Anniversary Appeal has been Steering Committee appoints Task Group Leaders and a launched in 1997 to provide additional financial support Project Manager. to the Centre in its ongoing projects. Detailed work on the project is carried out on a voluntary basis by the members appointed to each Task Group. The Centre arranges to bring to New Zealand, at the Objective appropriate time, one or more Visiting Fellows to work The objective of the Centre is to enhance engineering with members of the Task Groups, bringing to the project knowledge within New Zealand in identified areas judged the latest relevant information from overseas. to be of national importance and to engage in technology transfer of the latest research information available from The Centre also undertakes a variety of smaller projects overseas. The Centre is not concerned with basic engi- and produces publications on engineering subjects of neering research but with the application of research current concern, and arranges lectures and seminars on findings to engineering problems and practice. appropriate topics as the occasion arises. The objective is achieved for each major project under- Contact: taken by bringing together a selected group of practising Centre for Advanced Engineering and research engineers and experts in the particular field University of Canterbury from both New Zealand and overseas to: Private Bag 4800 Christchurch • consolidate existing knowledge; New Zealand • study advanced techniques; Street Address: 39 Creyke Road • develop approaches to particular problems in engi- Christchurch 8004 neering and technology; Telephone: +64-3-364-2478 • promote excellence in engineering practice; and Fax: +64-3-364-2069 • disseminate findings through documentation and e-mail: [email protected] public seminars. http: www.cae.canterbury.ac.nz A unique forum for co-operation among industry, the Executive Director: John P Blakeley engineering profession and university research engi- Projects Director: John L Lumsden neers is thus provided.

NOTE: Although the Christchurch Engineering Lifelines project has been based upon earlier work in 1990/91 on the Wellington Case Study of Lifelines in Earthquakes, the Christchurch project has been undertaken by a separate Christchurch Engineering Lifelines Group, not under the general direction of the Centre for Advanced Engineering (CAE). John Lumsden, Projects Director of CAE, has been on the Steering Committee of the Christchurch Group and CAE has been responsible for the publication of this book. B Contents

Preface ...... v

Project Team Participants ...... vii

Acknowledgements ...... ix

Executive Summary ...... xi Photographs ...... following page xiii

Introduction ...... xxxi

1 Risk Assessment, Methodology, Vulnerability, Impact and Importance ...... 1

2 Hazards to Engineering Lifelines in Christchurch 2.1 Seismic Hazards ...... 17 2.2 Earthquake Hazards ...... 18 2.3 Waimakariri Flood Hazard ...... 29 2.4 Local Flooding Hazard ...... 30 2.5 Tsunami Hazard ...... 31 2.6 Extreme Wind Storm Scenario ...... 34 2.7 Snowstorm Scenario ...... 39 2.8 Slope Hazard and Damage to Services on Hills ...... 45

3 Seismic Liquefaction and Lifelines 3.1 Introduction ...... 47 3.2 Fundamentals ...... 50 3.3 Level Ground Liquefaction ...... 51 3.4 Christchurch Lifelines Study ...... 57 3.5 Sustained Shear Stress ...... 58 3.6 Counter Measures ...... 60 3.7 Conclusion ...... 61 3.8 References ...... 62 3.9 Summary of Liquefaction Effects...... 66

4 Civil Services 4.1 Introduction ...... 71 4.2 Land Drainage ...... 72 4.3 Sewer System ...... 79 4.4 Water Supply ...... 84 4.5 Petroleum Products ...... 88

5 Electrical and Communications 5.1 Telecom New Zealand Limited ...... 91 5.2 New Zealand Fire Service Communications ...... 110 5.3 New Zealand Police Communications ...... 112 5.4 Trans Power (NZ) Limited Electrical System ...... 115 5.5 Southpower Distribution Network...... 122 5.6 Broadcasting System ...... 142

6 Transport 6.1 Christchurch Transport System ...... 147 6.2 Strategic Role of Transport System in Emergency ...... 147 6.3 Description of the Transport System ...... 149 6.4 Vulnerability — Earthquake ...... 156

i 6.5 Mitigation Measures — Earthquake ...... 164 6.6 Vulnerability — Major Windstorm ...... 184 6.7 Vulnerability — Major Snowstorm ...... 185 6.8 Vulnerability — Waimakariri River Flooding ...... 187 6.9 Vulnerability — Local Flood Hazard ...... 188 6.10 Vulnerability — Tsunami ...... 190 6.11 Vulnerability — Slope Hazard ...... 191 6.12 Mitigation Measures — Status of Measures...... 194

7 Emergency Buildings 7.1 Buildings Investigated ...... 197 7.2 Building Services ...... 198

8 New Zealand Fire Service 8.1 Scenario ...... 207 8.2 NZFS Initial Reaction Plan...... 208 8.3 Hazardous Substances Scenario Problems ...... 208

9 New Zealand/Los Angeles Workshop 9.1 Lessons from the Northridge and Loma Prieta Earthquakes ...... 219 9.2 Geotechnical Aspects ...... 232 9.3 Transportation ...... 237 9.4 Water and Wastewater Utilities ...... 241

10 Interdependence ...... 247

11 Summary of Benefits and Work Undertaken or Proposed 11.1 Christchurch City Council Water Services ...... 251 11.2 Christchurch City Council ...... 254 11.3 Trans Power New Zealand Limited ...... 261 11.4 Southpower ...... 261 11.5 Telecom ...... 262 11.6 Christchurch International Airport Limited ...... 262 11.7 Lyttelton Port Company ...... 263 11.8 Tranz Rail ...... 263 11.9 Transit New Zealand ...... 263 11.10 Petroleum Products ...... 263 11.11 Conclusion ...... 263

12 Ferry Road/Ferrymead Liquefaction Investigations 12.1 Ferrymead Bridge ...... 265 12.2 Ferry Road ...... 271

13 Continuing Work 13.1 Introduction ...... 275 13.2 Hazards ...... 275 13.3 Use of GIS ...... 277 13.4 Risk Balancing ...... 278 13.5 Use of TCLEE Guide for Review of Vulnerabilities ...... 278 13.6 Underground Services Damage Research ...... 280 13.7 Critical Areas and Critical Buildings...... 280 13.8 Interdependencies ...... 280 13.9 Combined Exercise/Response Plans ...... 281 13.10 Mutual Aid Agreements ...... 281 13.11 Internet ...... 281 13.12 Risk Management Seminar ...... 281 13.13 Review Project with Social Scientist ...... 282

ii 13.14 Task Group Meetings ...... 282 13.15 Promulgation ...... 282 13.16 National Forum ...... 282 13.17 Summary ...... 282

Maps Reference List for Colour Maps ...... 283 1: Location of Emergency Services and Major Contractors Yards with Seismic Hazard ...... 284 2: Location of Emergency Services and Major Contractors Yards with Waimakariri River Floodplain ...... 285 3: Location of Emergency Services and Major Contractors Yards with Local Rivers Floodplain ...... 286 4: Tsunami Hazard ...... 287 5: Slope Hazard Zones for Hill Areas...... 288 6: Transport System Network with Seismic Hazard...... 289 7: Bridge Vulnerability ...... 290 8: Slope Hazard — Transport System Locality Plan ...... 291 9: Seismic Hazard and Telecom Cables ...... 292 10: Local Flooding Hazard and Telecom Cables ...... 293 11: Tsunami Hazard and Telecom Cables ...... 294 12: Waimakariri Flooding Hazard and Telecom Cables ...... 295 13: Seismic Hazard and Fire Services ...... 296 14: Seismic Hazard and Police Radio Stations ...... 297 15: Seismic Hazard and Electricity Network...... 298 16: Electricity Network...... 299 17: Seismic Hazard and Broadcasting Services...... 300 18: Seismic Hazard and Stormwater System ...... 301 19: Foulwater Sewer and Pressure Mains with Seismic Hazard ...... 302 20: Seismic Hazard and Water Services (Major Pipelines in Metropolitan Area) ...... 303 21: Seismic Hazard and Petroleum Products ...... 304

Bibliography ...... 305

Project Team Participants — Contact Details ...... 309

iii iv Preface

The Christchurch Engineering Lifelines Project has been widely accepted and Christchurch is fortunate to have had the wide involvement of engineers and managers from local authorities and utilities which serve metropolitan Christ- church. There was tremendous enthusiasm in response to the approach from the Regional Controller of Civil Defence, Mr John Lamb, himself a Civil Engineer, to take part in the work (which is still ongoing). The aim of the project is to make Christchurch better able to withstand the natural hazards which have the potential to affect so many people.

The Lifelines work in New Zealand was initiated by the Centre for Advanced Engineering which, in 1991, published the Project Report on pilot work undertaken in the Wellington area. We are fortunate that the Centre for Advanced Engineering is also based in Christchurch and is readily available for ongoing consultation on Lifelines work; our appreciation is expressed to the Centre for the publication of this book.

Although Christchurch is seriously threatened by a major earthquake originating from the main alpine fault, or from other faults in the area closer to Christchurch, the public perception is of many threats to Christchurch from other hazards. The most notable in the eyes of the public is the threat of flooding from the Waimakariri River on whose floodplain Christchurch is situated. For this reason an assessment of the vulnerability to natural hazards other than earthquakes was undertaken. The project highlighted our inadequate detailed knowledge on many of the hazards, notwithstanding the considerable work which has been done and is ongoing, by the Regional Council in identifying these hazards.

Lifeline engineering is ongoing work and I am very impressed with the considerable amount of mitigation work undertaken as a result of the initial project (the public workshop was held in October 1994). I am confident that the investigations and ongoing action in Christchurch means that it can serve as a model for many similar investigations, not only in New Zealand, but in other places in the world.

I commend this book to readers in the hope that by studying the findings, they may be impressed by the enthusiasm and commitment of all of those who are taking part in this valuable work in Christchurch and be inspired to undertake similar projects in their own area.

Richard Johnson

Chairman, Canterbury Regional Council

(on behalf of all of the organisations involved in “Engineering Lifelines in Christchurch”)

v vi Project Team Participants

STEERING COMMITTEE Mr John Lamb Project Manager Mr Kevin O’Kane Ministry of Civil Defence Mr Brian Hasell Ashburton District Council Mr Allan Watson Christchurch City Council Mr Stephen Franklin Telecom New Zealand Limited Mr Mark Gordon Christchurch City Council Mr John Lumsden University of Canterbury Mr Richard Keys Otago Regional Council Mr Ian McCahon Steering Committee members (1994) Soils & Foundations Limited Back row (l-r): Alan Watson, Stephen Franklin, Richard Keys, Mark Gordon, Gordon Hood Mr Jim Williamson Christchurch City Council Front row (l-r) Jim Williamson, Brian Hasell, John Lamb, John Lumsden

Hazards Task Group Dr Bob Kirk Electrical & Communications University of Canterbury Mr Ian McCahon Networks Task Group Mr Mark Yetton Soils & Foundations Limited Mr Stephen Franklin Geotech Consulting Limited Mr David Bell Telecom New Zealand Limited University of Canterbury Mr Ray Basher Civil Services Task Group Mr Tony Boyle Transpower New Zealand Limited Canterbury Regional Council Mr Allan Watson Mr Peter Brash Christchurch City Council Dr Ian Owens Mr Roger Smithies University of Canterbury Mr Neil Bennett Telecom NZ Limited SERCO Mr Derek Todd Mr Russ Botting Tonkin & Taylor Mr Neville Stewart Telecom New Zealand Limited Christchurch City Council Mr John Weeber Mr Bud Chapman Canterbury Regional Council Mr Mike Berry Communications Centres Christchurch City Council Mr John Coleman Hazards Task Group — Other Mr Ken Couling NZ Police Engineering Workshop Christchurch City Council Advisors Mr Noel Maginnity Dr John Berrill Mr Allan Dowie Radio Network Shell NZ Limited University of Canterbury Mr John MacKenzie Dr Neil Cherry Mr Dave May Montgomery Watson Limited Christchurch City Council Lincoln University Mr Grant Roberts Mr Ken Couling Mr Alan Marshall Lincrad Aerials Telecom New Zealand Limited Christchurch City Council Mr John O’Donnell Dr Derek Goring Mr Bob Watts Southpower National Institute of Water & Atmos- Christchurch City Council pheric Research A list of the Project Teams current contact details is shown at the end of the book (page 309)

vii Transportation Task Group Mr Mike McGlinchey Building Services Task Group Lyttelton Port Company Limited Mr Mark Gordon Mr Iain Drewett Christchurch City Council Mr Neil McLennan Consulting Engineer Lyttelton Port Company Limited Mr David Bates Mr Grant Wilkinson Transit NZ Mr John Reynolds Holmes Consulting Group Limited Opus International Consultants Limited Mr Tony Barnett Mr Graeme Wilson Mr John Robb City Streets Unit Mr Russell Herbert Tranz Rail Limited Fire Services Task Group Mr Ken McAnergney Christchurch International Airport Mr Barry G J Shields Limited NZ Fire Services

A list of the Project Teams current contact details is shown at the end of the book (page 309)

viii Acknowledgements

The work of investigating the vulnerability of engineering lifelines in Christchurch to natural hazards has involved many people and organisations, and appreciation is recorded as follows:

• The members of the various Task Groups and particularly the Chairmen who were involved in many investigations and meetings, much of which was in addition to their normal work and in their own time.

• The more than 100 participants in the public workshop in October 1994 and, in particular, to Dr David Hopkins, the Deputy Chairman of the Steering Committee of the Lifelines in Earthquakes, Wellington Case Study, who was the Invited Reviewer.

• Mr Ron Eguchi, Vice President of EQE International, a US-based consultancy that specialises in earthquake engineering, who was the keynote speaker at the Workshop and who acted as a Visiting Fellow for the Project.

• Mr David Brunsdon, the Project Manager of the Wellington Earthquake Lifelines Group, and Mr John Norton who was the initial Project Manager for the Wellington Case Study. These people, together with other members of the Task Groups of the Wellington Case Study, were very willing to provide information and comment on the Christchurch work, which, as far as possible, was based on the Wellington study.

• Mr John Lumsden, the Projects Director at the Centre for Advanced Engineering in Christchurch, who was very supportive of the Christchurch Study as a Steering Committee member and who has had a continuing interest in the Christchurch group. He, together with Charles Hendtlass and Una O’Grady, has been responsible for the publication of this book.

• Mr Brian Hasell (the initial Steering Committee Chairman) of the Canterbury Regional Council who made available the Council's resources for controlling the finances of the Project, meeting rooms, word processing and general administration, and the Councillors and Staff of the Regional Council who have been very supportive of the ongoing work.

The following major sponsors for their financial support of the Project: • Earthquake Commission; • Ministry of Civil Defence; • Christchurch City Council; • Telecom NZ Limited; • Southpower Limited; • Trans Power New Zealand Limited; • Transit New Zealand; • Christchurch International Airport Limited; • Canterbury Regional Council; • Banks Peninsula District Council; • Shell New Zealand Limited; • Centre for Advanced Engineering, University of Canterbury; and

• AMI Insurance.

AMI Insurance sponsored the cost of the colour photographs and colour maps in this book.

ix x Executive Summary Executive Summary

The results of an investigation into the vulnerability of effects being likely as a result of extreme rainfall or the infrastructure serving metropolitan Christchurch earthquake. (including Lyttelton) is the main content of this book. The work was undertaken by the Christchurch Engi- A severe windstorm and severe snowstorm were inves- neering Lifelines Group whose objectives are: tigated although it was not possible to produce a hazard map for these. The worst windstorm is likely to be a • To identify the vulnerability of engineering lifeline nor-wester with gusts on the open plains of up to 110 services to damage from earthquakes, flooding, knots with this being increased over hills. tsunami and meteorological hazards. The maximum snow depth on the plains of metropoli- • To identify practical engineering strategies for re- tan Christchurch is expected to be up to 30 cms with ducing the risk or impact of such damage and for approximately a metre on average on the Port Hills. providing for reinstatement following such events. The various coloured maps showing the hazard and the • To communicate the issues to people involved in utility services investigated are on pages 283 to 304, at the management of these services and to raise the the end of the book. awareness of the public to their importance. Four main task groups and two smaller ones were An engineering lifelines investigation is about reduc- established to undertake the detailed work of assessing exposure to risk from natural hazards, so risk ment of the infrastructure of Christchurch. The main management is addressed. The techniques have much task groups were: Hazards, Civil Services, Electrical to offer those responsible for reacting to emergencies and Telecommunications and Transport. The smaller and dealing with the aftermath. task groups were Buildings and Fire.

The engineering lifelines project considered a range of The utility networks are described and for the purpose natural hazards in the Christchurch area and the table of risk analysis, each network was broken up into below shows the hazards and the annual exceedance components, which were then assessed for their vul- assessed for the scenarios used: nerability to each hazard scenario. This was done by overlaying electronically the networks over the haz- Hazard % Probability of ards map which was then examined to determine exceedance importance, vulnerability and impact of damage. In 1 year In 50 years Local flooding 0.1 5 In this study, officers with a day-to-day involvement Waimak flooding 0.2 10 with the particular lifeline assessed both the vulner- Seismic 0.67 28 ability and importance of each element, and its depend- Wind 0.67 28 ence on other services and recommended mitigation Tsunami 0.67 28 Snow 2.0 64 measures for the particular hazard. Slope hazard 0.67 28 The vulnerability charts which were produced from the Christchurch being partly located on a saturated, sand- risk analysis were then used for identification of the and silt-rich prograding coastline, is potentially at risk need for, and priority of, mitigation measures. Any from widespread liquefaction. The areas likely to be measure which would reduce importance (e.g. by re- affected are shown on the earthquake hazard map and dundancy), reduce vulnerability (e.g. by strengthen- there is a detailed explanation of liquefaction, the ing) and/or reduce impact (e.g. by alternatives or potential for it and methods of testing. contingency planning), would help to mitigate the disaster. In many cases, the mitigation measures iden- The Christchurch area is subject to two major sources tified were very inexpensive, or could be easily inte- of flooding, one from the Waimakariri River and the grated with ongoing maintenance and replacement other from the local rivers. Maps showing the areas programmes. likely to be flooded were produced. In addition, however, a tsunami involving a total water variation at the The analysis described above was carried out network open coast of 10 metres inclusive of tide (i.e. 5 metres by network and then the consequences of interdepend- above and below msl) could potentially affect Christ- ence assessed both in operation (if A fails, B fails), and church. The likely effects are described in detail. The in response (need to fix A to get to B), as the interde- slope hazard on the Port Hills is reported, including pendence could be a critical factor.

xi Risks and Realities

The work has served to open up the subject of the Northridge investigation and this book also reports on performance of Christchurch’s lifelines during severe the workshop from a Christchurch perspective. Al- natural events and how the managers responsible for though some members of the group also went to Kobe those services are now ensuring that the work of in August 1995 to learn from the lessons of the Great assessing and prioritising the many mitigating recom- Hanshin Earthquake, the results of this investigation mendations is being pursued and appropriate measures are not reported in this book, as it was adequately implemented. reported in the 1995 report of the Wellington Earth- quake Lifelines Group, to which members of the Christ- The results of the analysis and recommended mitiga- church group contributed. tion measures were considered in two workshops during 1994 (March and October ), the October workshop Some measure of the success of the work so far is the being organised over three days with an attendance of extent to which budget provision has been made for the over 100. The purpose of this workshop was to allow various mitigation measures which have been identi- peer review of the project. fied. These are set out in Chapter 11, “Summary of Benefits and Work Undertaken or Proposed”, and a That workshop was addressed by Mr Ron Eguchi, as perusal of this list will show that the engineering the keynote speaker. He is a Vice-President of EQE lifelines investigation undertaken during 1993 and International, a US-based consultancy that specialises 1994 was not just an academic exercise. in earthquake engineering. Mr Eguchi, speaking about the Wellington and Christchurch Projects, said “I think One of the areas in Christchurch most likely to be badly a lot of the ideas and concepts that are really being affected by earthquakes, tsunami or flooding is the developed by these two projects are very forward- Ferry Road/Ferrymead Bridge area and the investiga- thinking and in fact in many respects they are much tions undertaken in this area as a “critical area” are more forward-thinking than a lot of programmes we reported. have in the US.” Twenty-one of the various hazard maps produced have Members of the Wellington and Christchurch Lifelines been reduced in size from the working A0 size. These Groups had visited Los Angeles in August 1994 (fol- are included at the end of this publication. lowing the Northridge Earthquake) and during a New Zealand/Los Angeles workshop, members of the team A summary of the work undertaken is depicted in the had met Mr Eguchi. The 1994 Wellington Lifelines 16 aerial photographs that follow, and which form part Group report contained a large section dealing with the of the Executive Summary.

xii Executive Summary

Photographs The following sixteen photographs taken by Michael Provost Photography show some of the principal aspects of the Christchurch investigation.

Some of the major components of the Christchurch infrastructure are identified (such as the airport, port, tunnels, central city, sewage treatment works, etc.) together with the vulnerabilities assessed. Some mitigation measures are mentioned as is some proposed and ongoing work.

Photo 1: Christchurch International Airport

Photo 2: The Port of Lyttelton

Photo 3: Lyttelton rail and road tunnels

Photo 4: Christchurch central city

Photo 5: Central city, showing the Avon River and the police station

Photo 6: Central city, looking southeast

Photo 7: The northeast central city

Photo 8: The Ferrymead Bridge

Photo 9: Clifton Hill

Photo 10: South Brighton Spit and Clifton Hill

Photo 11: View to the north-west over the sewage treatment ponds

Photo 12: Lower reaches of the Avon River

Photo 13: Southpower yard and Madras Street

Photo 14: The Trans Power Islington substation

Photo 15: Sugarloaf television transmission tower

Photo 16: August 1992 snow storm

xiii Risks and Realities

Photo 1: Christchurch International Airport

This view, taken looking to the west over the technology park in the foreground, shows Memorial Avenue leading to Christchurch International Airport.

Important communication linkages for air traffic control run between the Air Traffic Control building (foreground) and the airport. Steps have been taken to provide diversity in the link. The airport itself will be of major importance as the primary transport link to the outside world immediately following an earthquake. Fortunately the Airport Authority employed consultants to assess vulnerability and mitigation measures and much mitigation work has been done.

Although the majority of the engineering services to the airport are dependent on the city-wide services, the water supply stands alone and has an emergency power plant.

The airport is on the flood plain of the Waimakariri river although the protection provided by the extensive stopbank system should considerably lessen the likelihood of flooding from this source. The Canterbury Regional Council proposes further work constructing a major stopbank to return any floodwater to the river and this work was considered adequate so that no special work for the airport (or Christchurch itself) has been recommended as a mitigation measure for flooding from this source.

However it was discovered that the apron of the terminal buildings slopes towards them and an exceptional localised heavy rainfall could cause problems to the emergency services located in the basement area.

xiv Executive Summary

Photo 2: The Port of Lyttelton

This view, looking to the west, shows part of the port at Lyttelton with the large building in the foreground being the Port Com- pany building. A roundabout can be seen which is in front of the road tunnel portal and just to the right of the roundabout can be seen three levels — rail and two roads.

The Lyttelton Port road and rail portals can be seen and the rail portal is under the lower of the two roads. The major concern is the failure of the bridge walls lining the inside of the rail tunnel portal and the possible collapse of the road structures above. Questions still revolve around who “owns this problem”. On the road tunnel, damage to the ventilation structure is likely in an earthquake, and there is potential for slips on to the roundabout.

The main port authority building is not likely to suffer damage, but jetties will probably be pulled away from land. However, the Port has a response plan. The Cashin Quay, which is out of the photograph to the right, is likely to suffer slip circle failure into the harbour but there are no mitigation measures planned.

Just out of the photograph to the left is the bulk fuel installation serving the Christchurch area and opinions differ on the propensity towards liquefaction or seaward slumping of the Lyttelton Tank Farm.

Lyttelton has acknowledged the need for extensive work on its water and sewage pipelines and a severe earthquake will undoubtedly cause considerable damage to the reticulation.

xv Risks and Realities

Photo 3: Lyttelton rail and road tunnels

This view, which is looking southeast, shows the Lyttelton rail tunnel portal on the left centre and the portal of the road tunnel at the upper right.

In the earthquake event, the Tunnel Road skirting the Port Hills is likely to be blocked by slips, with the ventilation tower at the Tunnel portal and its associated equipment being vulnerable and requiring more assessment. With the bridges on Tunnel Road (Port Hills and Horotane Valley underpasses) being vulnerable, the only other access to the Tunnel would be via Ferrymead Bridge and Bridle Path Road, both of which are also vulnerable. The tunnels themselves are unlikely to suffer major internal damage. Access to the rail tunnel is likely to be affected by embankment, bridges and cuttings failures.

Access to the Port will be most important for the recovery stage from an emergency and even in the response stage many supplies could come to Christchurch by sea if the access to the north of the City is disrupted (which is most likely in a major earthquake).

Water supply to Lyttelton and a Telecom fibre cable also use the tunnels.

It is quite common for damage as a result of an earthquake to be magnified in a zone at the foot of hills (possibly from the less compacted soils that have washed off the hills and reflection effects) and a similar effect is expected in Christchurch.

xvi Executive Summary

Photo 4: Christchurch central city

This view of the central city area looking southeast shows Colombo Street in the centre of the photograph and the Square with the Cathedral visible almost in the centre.

Traffic control in the central city is largely controlled by traffic signals which are connected to computers in the City Council offices. There are one-way signal linked streets, the overall impact of loss of the system would be severe in traffic management terms. Also, within the central city the full or partial collapse of vulnerable buildings would block access to much of Manchester Street, and also parts of Lichfield Street and Tuam Street near centre, top of picture.

Although the bridges shown over the Avon River (the Avon river is shown flowing right to left in the foreground) are unlikely to sustain much damage in a large earthquake, consolidation of the bridge approaches could cause damage to services with consequent disruption to the central city area.

The Southpower building (mid left) was identified as likely to sustain considerable damage in an earthquake in the area occupied by the control room and this has now been moved to a safer location and the old part of the building demolished.

Basement flooding arising from a localised extreme rainfall event could cause difficulties in an emergency as it is suspected that several of the central city buildings have electricity installations and other standby plant located in basements. As a result of the lifelines investigation, one large building now has sandbags stored on site.

xvii Risks and Realities

Photo 5: Central city, showing the Avon River and the police station

This view looking south shows the Avon river with the Police building sign identifiable.

The police were involved in the investigation and several modifications were made to their communication system as a result.

The road bridges over the Avon River are robust and require no further mitigation measures but other services using them may be vulnerable if the approach fills consolidate.

Within the central city, the Avon river is relatively deeply entrenched and fortunately there is little likelihood of flooding from the river. However it has to be remembered that in February 1868, following a severe rainstorm in the “back country”, i.e. the foothills of the Southern Alps which are part of the catchment of the Waimakariri river, floodwaters flowed through the central city when the Waimakariri overflowed and joined the Avon. That was before the stopbanks had been established along the Waimakariri and the operations of the modern day catchment authorities.

The people of Christchurch still perceive that the greatest threat to Christchurch is the Waimakariri River. Although in the view of the Engineering Lifelines Project, an earthquake presents a larger threat, flooding is still possible. If a flood occurs, its effects are much more likely to be greater in the lower reaches, and in the small floods from rainfall affecting the catchment of the Avon River or other local rivers.

xviii Executive Summary

Photo 6: Central city, looking southeast

In this view looking to the southeast, Colombo Street runs diagonally across the picture with the railway line just visible across the top third. The major vulnerabilities to the transport system are the overbridges on Moorhouse Avenue and Colombo Street over the railway. Both will suffer major damage during an earthquake. Major expenditure on both bridges would be required and further detailed investigation of these structures is programmed. Moorhouse Avenue is an important primary route providing access around the city core.

Because of likely building debris on the streets after an earthquake, access may be difficult to the Christchurch City Council Offices, which will have a vital role in response to an emergency. The CCC building is midway between Colombo Street and Manchester Street, nearly halfway up the picture. The building itself should be relatively undamaged and a programme is under way to seismically restrain essential elements such as fittings and computers.

Fire fighting in high rise buildings may be affected by power failure if the booster pumps supplying pressure to the upper storey sprinkler systems have no auxiliary power supply. Also, damage to the city reticulation itself could leave viable sprinkler systems unable to operate, unless the building has its own water storage. Backup power or water storage are not compulsory and vulnerability therefore varies from building to building.

xix Risks and Realities

Photo 7: The northeast central city

This view, looking southeast, shows the Avon River with the Fire Service central fire station nearby and the central business district in the background (the Southpower building is at the top centre just to the right of the Manchester Street parking building).

The central fire station is a reminder that widespread damage coupled with power failure may limit the availability of water for fire fighting. River water could be a viable alternative, but the practicalities need more investigation. Equipment suitable for this duty must be available and means established of protecting above-ground long, high-pressure hoses at road crossings, etc. The four bridge crossings also bring to mind the adopted mitigation measure of providing water supply valving at each end of bridge crossings. However it is not expected that the bridges themselves will be extensively damaged.

It is expected that most damage to underground services will be from differential settlement, particularly caused by backfill of excavated areas. In the case of bridges this is on the approaches, but the effect may be caused simply by backfilled trenches from one service crossed by another service at a different level. In buildings, the likely damage area for the services is where pipes enter foundation walls.

Heavy winds can cause considerable disruption, as described in the scenario on page 34 and the heavily treed banks of the Avon River would be of concern.

xx Executive Summary

Photo 8: The Ferrymead Bridge

The Ferrymead Bridge is in the centre of the photograph.

The bridge is likely to suffer extensive damage in even a moderate earthquake. The damage, including services breakdowns, will cause severe traffic disruption during restoration. Investigations are in progress to strengthen the bridge and/or relocate services off the bridge. Damage to the approaches and the liquefaction risk are currently being assessed in detail, but it would appear that the foundation conditions are some of the worst in Christchurch. The hazard assessment reveals that all of the land between the Heathcote River and the Estuary is likely to be subject to liquefaction in a severe earthquake. All of the services crossing the bridge also pass through this land.

One long-term possibility could be a bridge connecting Rocking Horse Road (in photograph 10) across the mouth of the estuary, possibly just for services. The mitigation measures are currently the subject of a team investigation by members of all of the service authorities affected. This investigation was commenced in 1996.

The Tunnel Road and tunnel portal can also be seen in the background . Both are important transport routes as referred to in Photograph 3.

xxi Risks and Realities

Photo 9: Clifton Hill

This view looking southwest shows Clifton Hill in the foreground with a cellular VHF site for Mount Pleasant and Gebbies Pass

Earthquakes would be likely to severely affect road access to residential areas on Clifton Hill. Foundations of structural crib walls could fail, resulting in loss of the road. Mitigation measures will need to be investigated. The Main Road along the foreshore would also be subject to inundation from a Tsunami with the resultant likelihood of damage to all of the engineering services along the road.

All development east of the Ferrymead bridge is supplied with water from wells located west of the bridge, leaving this area vulnerable. Duplication of the McCormacks Bay Reservoir has substantially boosted storage in this area and investigation work is proceeding to strengthen the Ferrymead crossing, one result of which will be the reduction in water supply vulnerability for this area.

The extent of seismic landslip will depend on the season and the water content of the surface layers. An earthquake during the winter is likely to cause much more damage than one in the summer. An engineering geologist accompanied a minibus load of representatives from the service authorities on the main roads of the Port Hills, as the vulnerability of the various services was assessed in relation to landslip and other seismically induced hazards.

xxii Executive Summary

Photo 10: South Brighton Spit and Clifton Hill

This view, looking towards the southeast, shows South Brighton Spit in the foreground with Clifton Hill in the background.

The Main Road to Sumner around the base of the cliffs and estuary foreshore would be inundated under a tsunami. It is anticipated that tsunami effects will cause inundation of properties on the Spit and also of low-Iying land in McCormacks Bay, Redcliffs and Sumner.

Minor flooding occurs at present on the roadways from extreme tidal events, so that any rise in water level will cause noticeable effects.

All of the services to the Spit area are dependent on Rocking Horse Road and access through South Brighton. The Hazard Assessment reveals that there is possibility of liquefaction so that in any severe earthquake at least one service is likely to be affected.

One of the possibilities to provide an alternative for services (and access) east of the Ferrymead bridge is a new bridge from the end of the Spit across the mouth of the Estuary, but no site investigation has been done . If the proposal is the subject of a detailed report, then that report will have to deal with access for all services across the lower reaches of the Avon river and through South Brighton. Unlike most of the rest of the city, there are few alternative routes available.

xxiii Risks and Realities

Photo 11: View to the north-west over the sewage treatment ponds

This view, looking to the north-west, shows Dyers Road in the foreground where the road travels across the sewage treatment ponds with the sewage plant in the background.

At the top of the picture is part of the eastern suburbs of Christchurch which the project identified as being likely to be subject to liquefaction in the event of an earthquake.

Seismic damage to the embankments of the sewage treatment ponds in the form of slumping or liquefaction is a threat, the extent of which has not yet been established. Although a cone penetration test at the buildings of the sewage treatment plant revealed that liquefaction should not be a problem at the location tested, it is likely that the embankments could be severely damaged. Repair of internal embankments will be able to proceed independently of tidal effects, so attention needs to focus on the vulnerability of the external embankments fronting the Estuary.

Only minor damage to the sewage treatment plant oxidation ponds is anticipated from tsunami effects, due to the satisfactory elevation of the retaining embankments.

Dyers Road, a “ring road” which is an important arterial route, passes through the middle of these pond areas. However, it does not warrant the planning of any mitigation measures at this stage against possible subsidence risk.

xxiv Executive Summary

Photo 12: Lower reaches of the Avon River

This view, looking to the southeast, shows the lower reaches of the Avon River with the sea just visible at the top of the picture, Pages Road bridge in the background and the Wainoni Road bridge in the foreground.

Both bridges may suffer considerable damage. Mitigation measures being planned include a new expressway bridge in the immediate foreground, and the likely replacement of the Pages Road bridge within the next 10 years. The Pages Road bridge carries very significant Telecom cables (fibre optic and 2000 pr. cable) and a large feeder watermain. High tide levels in the Lower Avon are above surrounding residential land which is protected by stopbanking. There is concern that stopbank damage through slumping or liquefaction will allow tidal water to flood into residential areas.

A high tide and some flooding is of concern at any time, but the possible problems following an earthquake which may cause slumping of the stopbanks and damage the pipework allowing stormwater to discharge into the river will be of much greater concern. A solution to this problem is still being sought and extensive testing is still required to determine how many, if any, banks may be affected. Nearly all of the area has been identified in the hazard assessment as being likely to be subject to liquefaction.

This area is also subject to possible flooding arising from the effects of extremely low barometric pressure coinciding with snow and rainfall and in 1992 these effects occurred together.

Tsunami effects are also significant in this locality, including localised flooding close to the coast and damage to or overtopping of stopbanks upstream caused by the passage of tidal ‘bores’.

xxv Risks and Realities

Photo 13: Southpower yard and Madras Street

This view, looking southeast, shows the Packe Street Southpower yard with Madras Street in the foreground.

As a result of the project, Southpower have relocated emergency spares to a much more secure area for storage which was identified in a detailed assessment of the buildings.

Most service authorities do not now carry a large amount of spares (fittings, pipes etc.) and the modern storekeeping with “just in time” stores control will not assist in prompt repair in the event of extensive damage following an earthquake. To minimise the effects of this, a project has been initiated to arrange for “mutual aid” in advance. In California such arrangements are formalised in contracts and many authorities now use similar materials and methods which will make mutual aid much easier.

Madras Street is just one access road to the north — not “primary” but parallel to the Cranford Street primary route. Christchurch is fortunate in having a basic “grid pattern” of streets which provides many alternatives in the event of one street being closed as a result of an emergency. Likewise, other services generally have many alternatives and not services all located on one route down a valley, which so often is the case in other cities. However, outside the area of the investigation, the city is dependent on resources such as replacement spares being delivered from the main trunk railway line or State Highways and, other than the airport and seaport, these are really the only two routes into the city.

xxvi Executive Summary

Photo 14: The Trans Power Islington substation

This view of the Trans Power Islington substation is looking east. This is on the south-western edge of the city and is the major substation for Christchurch.

Trans Power have had a very responsible attitude to the minimisation of risk and already had undertaken a seismic investigation of their installations. However, a walk over survey was done of the transmission lines which lead from Islington to the east of the city to ascertain the stability of the foundations. Trans Power New Zealand Limited has a policy in place with the objective of being able to maintain power supplies during a MM IX intensity earthquake and to restore power supplies to earthquake damaged areas within three days.

Wind, snow, electrical storm and flooding are other hazards that may impact on the continuous supply of electricity. The Islington substation is on the edge of a possible flood flow from the Waimakariri and it was determined that no insurmountable problems should arise from this source.

Notwithstanding the considerable work already done, Trans Power will continue to address the issues raised and ensure adequate contingency plans are in place.

Subsequent to the Wellington Lifelines Study, a new national system of equipment spares and inventory is in place and this will greatly assist the reinstatement operation.

It has to be borne in mind that the limitations on the extent of the lifelines project caused the placement of artificial boundaries on the considerations, and Christchurch is dependent on many services outside the metropolitan area. Eventually, Christchurch (and the rest of New Zealand) will be much more able to withstand the effects of natural disasters when lifeline studies are carried out nationwide.

xxvii Risks and Realities

Photo 15: Sugarloaf television transmission tower

This view looking over the Port Hills to the north-east shows the Sugarloaf Tower, which is the major transmission tower for television services in Christchurch.

A review, in conjunction with BCL, of the Sugarloaf transmission site was undertaken as a result of the project. Very minor modifications were made to the fixings in the buildings, some of which had not been seismically restrained and the communication to the public necessary following a disaster should be much more secure. Similar checks were made to all masts on the Port Hills, with most being found to be in secure locations, although access to them may be difficult following an earthquake.

One of the specific mitigation measures identified for the broadcasting system was a review of the overall robustness of each network and the establishment of a plan to manage with a reduced system (i.e. a disaster response plan), but there have been many changes in staff since the project and this may still need to be done.

One of the considerable advantages in having such a large group of people involved in the various Task Groups is the establishment of the informal, first name contacts between the various organisations which would be useful in an emergency. However, with the passing of time personnel change and the contacts are lost. The continuing nature of the lifelines work will assist to some degree, but it will be important to devise some method of keeping the contacts current.

xxviii Executive Summary

Photo 16: August 1992 snow storm

This photograph was taken during the aftermath of the August 1992 snow storm and illustrates the havoc that can be caused to transport.

The depth of a snowfall will vary over the city and the worst effects are likely to be on the hills. Snow could cut off road access throughout the whole city, possibly bringing down overhead power lines and making access for public transport, emergency vehicles and the public very difficult. The city’s business sector would virtually come to a standstill because of the lack of transport. However, disruption is unlikely to last more than a few days. The main mitigation measures are to prepare a response plan and to place underground the services on key transport routes.

Snow causes problems with the difficulty for melt water to flow and get access to sumps. Snow can block sumps and sometimes open waterways. Overhead lines carrying services can be brought down by snow but one of the worst clean up problems may well be the foliage that has been brought down. This was certainly the case in the 1992 snowfall.

The loss of electricity to essential services caused by snow will be felt very quickly by community homes for the elderly and infirm as very few, if any, have emergency power available. Although the project has attempted to be concerned with effects on engineering services only, it has to be borne in mind that these services are for people, not just the services themselves.

Civil Defence can therefore usefully learn from the project and it is hoped that in the future more work will be able to be done on investigating the best way to make the supply of services to essential locations more secure. The interdependence of the various services also has yet to be further investigated and mitigation measures put in place.

xxix Risks and Realities

The Ambulance in the Valley ( ANON )

Twas a dangerous cliff, as they freely confessed, Though to walk near its crest was so pleasant: But over its terrible edge there had slipped A duke, and full many a peasant. The people said something would have to be done, But their projects did not at all tally. Some said ”Put a fence ‘round the edge of the cliff,” Some, “An ambulance down in the valley”.

The lament of the crowd was profound and was loud, As their tears overflowed with their pity: But the cry for the ambulance carried the day As it spread through the neighbouring city. A collection was made, to accumulate aid, And the dwellers in highway and alley Gave dollars and cents - not to furnish a fence - But an ambulance down in the valley.

“For the cliff is all right if you’re careful.” they said: “And, if folks ever slip and are dropping, It isn’t the slipping that hurts them so much As the shock down below - when they’re stopping.” So for years (we have heard), as these mishaps occurred Quick forth would the rescuers sally, To pick up the victims who fell from the cliff. With the ambulance down in the valley.

Said one, in a plea, “It’s a marvel to me That you’d give so much attention To repairing results than to curing the cause; You had much better aim at prevention. For the mischief, of course, should be stopped at its source: Come, neighbours and friends, let us rally. It is far better sense to rely on a fence Than an ambulance down in the valley.”

xxx Executive Summary Introduction

This book has arisen out of the Christchurch Engineer- and cities in New Zealand. It involves a detailed, ing Lifelines project and contains the report of the structured investigation of the services, the hazards to workshop held in October 1994. It is believed to be the which they are vulnerable, the mitigation measures to first book of its kind to address the effects of a range of reduce the effect of hazardous events and considers the natural hazards on emergency services using a multi- dependence the various lifelines have upon each other. disciplinary approach. It is one of the tools of risk management now made possible because of modern technology and is really The initial work on the project was undertaken during only part of responsible asset management. 1993/94 and this followed on from the successful major project of the Centre for Advanced Engineering Notwithstanding all the investigation and mitigation (Lifelines in Earthquakes - Wellington Case Study). work carried out, it is inevitable that in a severe event This Wellington-based project was a prototype for some damage to the infrastructure will occur. The work New Zealand and developed methodologies which undertaken and proposed will, however, make Christ- were used, as far as was possible, in the Christchurch church better able to withstand the effects of natural work. However, unlike Wellington, where only the hazards. effects of earthquakes were studied, the objectives of the Christchurch Engineering Lifelines study were: In this book the investigations focus on the Christ- church Metropolitan area, including Lyttelton and the • to identify the vulnerability of engineering lifeline Port, with a few extensions where damaged engineer- services to damage from earthquakes, flooding, ing installations close to the area would affect the tsunami and meteorological hazards. operation of the lifelines within Christchurch, e.g. a repeater station in North Canterbury and the bridges on • to identify practical engineering strategies for re- nearby rivers. ducing the risk or impact of such damage and for providing for reinstatement following such events. Any investigation of the effects of hazards involves consideration of risk management so the book incorpo- • to communicate the issues to people involved in the rates a brief overview of the issues involved in consid- management of these services and to raise the ering risk. However, although in the Christchurch awareness of the public to their importance. work, high probability/high consequence risks, high This book, then, also addresses many of the essential probability/low consequence risks, low probability/ elements of a lifelines investigation and will be of use high consequence risks and low probability/low conse- to other areas of New Zealand and in other countries. quence risks were considered, it has to be borne in mind Similar investigations have now commenced in Auck- that the physical work identified as desirable need not land and Otago regions, with many other areas consid- be very costly. The book contains reference to much ering being involved. simple work that has been undertaken at relatively low cost. Lifelines are not only those resources that enable the over-wrought to make contact in times of emotional The book does not deal with response planning in detail need, but also is the generic term used to describe all of although this is at present being undertaken in Christ- the engineering services that enable people to live with church as a result of the lifelines project. Those mem- the standard of living we have come to expect in New bers of the team who were privileged to visit Northridge Zealand. In other countries the term “infrastructural and Kobe after earthquakes in those areas have no assets” is being used. doubt that response planning, at least in general terms, is a very necessary for engineering lifelines as well as A lifelines investigation is not an Emergency Services emergency services. or Civil Defence exercise (although Emergency Serv- ices and Civil Defence will receive considerable ben- Although much of the text of the book is based on the efit from one), but is essentially an investigation under- situation pertaining in 1994, it has been updated to taken by the providers of the various services forming 1996 insofar as has been possible. the infrastructure that permits settlement of the towns

xxxi Risks and Realities

xxxii Risk Assessment, Methodology, Vulnerability, Impact and Importance • 1

Chapter 1 Risk Assessment, Methodology, Vulnerability, Impact and Importance

This chapter attempts to set the work on engineering much to offer those responsible for reacting to emer- lifelines in the context of risk management, and ini- gencies and dealing with the aftermath. tially presents excerpts (in italics) from material published by two acknowledged experts in risk management — Patrick Helm and Janet Gough. What follows Government Perspective is a description of risk management considerations discussed at the time of the project workshop in Octo- Government’s interest in risk management in respect ber 1994, the methodology and the project structure. of dealing with emergencies and natural disasters The various steps of the analysis are then described, began about ten years ago. With the transfer of respon- which should prove useful for any group embarking on sibility that occurred through the reforms in local their own project. The way in which mitigation meas- authorities in the 1980s, central government put in ure implementation was, and continues to be, under- place policies designed to encourage more effective taken is then discussed. safety and loss prevention strategies. These were enun- ciated, for example, in a set of principles developed for In the June 1996 volume of Tephra, the magazine the 1987 Recovery Plan for Natural Disasters which published by the Ministry of Civil Defence, a compre- placed considerable emphasis on risk management hensive article dealing with “Integrated Risk Manage- and mitigation. ment for Natural and Technological Disasters” was written by Patrick Helm of the Department of Prime The essential idea was that central government would Minister and Cabinet in Wellington. The article intro- accept shared responsibility for the restoration of duces the concepts and principles of risk analysis and damage from natural disasters only if the local author- outlines a practical risk management strategy. Al- ity concerned had done its part to minimise, mitigate, though the term ‘lifelines’ is not used, its equivalent and manage the risk to its assets. The expectation was term is ‘infrastructure’. that local authorities would not simply provide insurance cover, but would seek to protect life and property Patrick Helm’s article deals with disasters and their by managing all the risks they faced. That is, they were consequences, not only lifelines. The relevant portions expected to take all reasonable steps to reduce the of the article are set out below and very adequately possibility of adverse events occurring (or follow-on introduces the theme of this book. The issues raised and secondary events), to put in place protection and the practical applications suggested are very similar to damage limitation measures that would reduce the that developed in the Christchurch Engineering Life- consequences, to examine the efficacy of response lines Project. mechanisms, and generally to improve the way that emergencies and disasters were managed. Introduction By obliging local authorities and other asset owners to accept a share of the responsibility for restoring dam- Risk assessment has been little used in New Zealand for aged infrastructure, central government’s intentions the management of emergencies and natural disasters. were to transfer some of the risk, to limit its potential At best, it has been regarded as a passive activity to financial exposure, and to shift the focus to loss preven- help insurance companies set premiums for coverage tion and better overall risk management. As the “own- of infrastructural damage. ers” of the assets, local authorities are best placed to identify local hazards and to implement strategies for Yet the techniques now becoming available for analys- ameliorating the consequences of any disasters. The ing and quantifying risk can prevent or minimise outcome in the long run should be better protection of disasters, can improve safety, and can markedly re- public assets, a safer environment for employees and duce societal disruption following disasters. They have society generally, less frequent interruption of essen- 2 • Risks and Realities

tial community services, and a reduction in the finan- different risk practitioners. This paper is most closely cial impact of losses. aligned with the definitions used in the Australian/ New Zealand Standard Risk Management, AS/NZS With time, however, some local authorities have started 4360:1995. to acknowledge the wider benefits of pro-active risk strategies to eliminate or reduce potential losses. This is an encouraging trend. But it has also revealed the Definitions complexity of, and paucity of knowledge about, risk Risk Analysis assessment involving disaster potential in New Zea- land. A systematic use of available information to determine how often specified events may occur and the magnitude of their consequences. Risk Assessment Risk Assessment In New Zealand, those responsible for dealing with emergency and disaster situations have been slow to The process used to determine risk management pri- adopt risk control techniques. This is partly ex- orities by evaluating and comparing level of risk against plained by the public scepticism that has developed predetermined standards, target risk levels or other over recent years in response to “expert” assur- criteria. ances about risks over which individuals have no control, or about which they hold different values Risk Management (e.g. nuclear safety, mad cow disease, etc.). But it is The systematic application of management policies, also because the concept of risk can be difficult to procedures and practices to the tasks of identifying, grasp, dealing as it does with chance and uncer- analysing, assessing, treating and monitoring risk. tainty somewhere in the indeterminate future. At first sight, risk analysis seems to lack the rigour of The essential idea is that it involves a formal, quanti- some other disciplines and even as a process of tative evaluation of potential injury or loss over a applied science it appears to have methodological specified period of time, or the prospect of future mal- shortcomings. performance of a safety or security system.

Notwithstanding these perceptions, risk analysis can be a powerful aid in decision-making involving public safety or in dealing with potential emergen- Approaches to Assessment cies and disasters. It forms an overlay on the emer- Risk techniques do not eliminate uncertainty, but help gency/disaster management process (i.e. the four put it in context. Provided there is some appreciation phases of Mitigation, Preparation, Response and of the degree of uncertainty innate of the factors in a Recovery) which can help evaluate the contribution defined situation, risk assessment can make an impor- of each phase to overall safety management. Risk tant contribution to reducing potential adverse effects. methodologies are useful not only for well-under- In practice, there are usually no direct linear tech- stood situations where good empirical evidence and niques for assessing risk. Experts tend to “gravitate” statistics are available (bridge design, fire suppres- towards a conclusion through cyclical processes that sion, river control, etc.), but also for situations in may involve several independent approaches. For ex- which there may be inadequate direct experience ample: (e.g. a large volcano, epidemics, or environmental issues such as stratospheric ozone depletion). • With natural hazards such as flooding, there may well be considerable local experience on which to Its particular strength for analysing situations of base extrapolation to more serious risks. uncertainty stems from the fact that it offers a structured, systematic and consistent approach that • In some situations observed records and expert forces the analyst into understanding the total risk views will be highly valuable. picture. Provided that hazards are identified with • In other situations where events are rarer (e.g. care and consistency, that causal models are ana- tsunami) it may be instructive to draw on experi- lysed logically, and that data is subject to strict ence and practices elsewhere. quality control, the results of risk analysis will make a practical contribution to public safety and loss In general, using a variety of approaches will yield a prevention. Risk assessment may be defined and more robust assessment and help avoid problems of undertaken in many ways. Terminology varies among systemic bias. It will also increase the likelihood of Risk Assessment, Methodology, Vulnerability, Impact and Importance • 3

exposing rogue conditions or interrelationships. For accidents occur that kill 200 to 300 people at a time. effective risk management, these different techniques Society would probably not accept say, one jumbo jet should be utilised in ways that lead to a quantitative crash per week, but seems to tolerate one per year. outcome where possible. The better the quantitative Clearly, when dealing with natural disasters in par- base, the better will be decision making, resource ticular, it is not practicable to seek to achieve zero risk; allocation, mitigation success and, ultimately, public the investment to make life totally safe would be beyond safety or loss prevention. any government. This is more feasible when dealing with technological hazards but, even here, economics, risk acceptability, and the need for the technology have Measuring Risk to be balanced carefully. Risk management experts recommend that, whenever feasible, those responsible Quantitative risk assessment combines three key ideas: try to eliminate high severity risks that might occur in a typical lifetime. • the chance of something going wrong; For any specific risk, they recommend trying to reduce • the consequences if it does; and dangers to just below a level commensurate with • the context within which the situation is set. reasonable cost — the “As Low as Reasonably Prac- ticable” (ALARP) principle. (Too conservative an ap- For any given set of circumstances, the level of “Risk” proach to safety can easily end up costing more than may be calculated as the product of the “Probability” the benefits.) A great deal of work has been done to try of an event or adverse outcome (chance/likelihood/ to devise standards of tolerability, but as yet they have frequency, expressed as occurrences per unit time) and not received universal acceptance. Desirably, risks a measure of the “Consequences” of the event (dam- associated with all hazards, natural and technologi- age/detriment/severity, expressed numerically as a cal, should be reduced to the point where they are low specific value measure such as lives lost or financial compared with levels that are widely accepted by the damage per event). In symbolic terms, we can write the community without concern. equation R = P x C. Figure 1.1 brings together several different concepts It has to be stressed that this simple product is not to do with the tolerability of risk. Derived from risk sufficient in itself to fully describe the real risk, but for guidelines developed in the United Kingdom, it depicts a given situation in which the terms may be specified risk thresholds in terms of local acceptability of deaths with reasonable accuracy, it provides an adequate from industrial and other accidents. Plotted as basis for comparing risks or making resource deci- exceedance or disasters involving a given number (N) sions. of fatalities. The “Local Tolerability Line” defines a region which is characterised by both high frequencies Tolerable Risk and severe consequences — the “Intolerable” region. The region between this line and the “Local Scrutiny Decisions on the appropriate level of investment for dealing with natural and technological hazards de- 10-1 pend critically on judgements about the acceptability Local Tolerability Line of risk. Strictly speaking, no level of risk is “accept- INTOLERABLE able” but, as a point of principle, risk can be consid- 10-2 Local Scrutiny Line ered tolerable when there are commensurate benefits. Possibly Unjustifiable Risk Safety does not require all risk to be eliminated: rather, 10-3 that there be an appropriate balance among costs, risks and benefits. As in any analysis of this type, there 10-4 Negligibility Line will inevitably have to be value judgements made of ALARP one kind or another: what level of risk will individuals REGION 10-5 or society tolerate? What proportion of ratepayers’ funds should be invested in mitigation? What criteria -6 should be applied? 10 Frequency of N or more fatalities, F NEGLIGIBLE In just the way that individuals tend to set themselves 10-7 personal thresholds of tolerable risk, so too communi- 1 10 100 1,000 10,000 ties have informal but real perceptions of societal risk Number of Fatalities, N thresholds. Travellers the world over continue to fly in jetplanes notwithstanding the fact that, occasionally, Figure 1.1: Indicative frequency-fatalities curve 4 • Risks and Realities

Line” is a region of possibly unjustifiable risk. Be- can be modified in many ways: by good safety engi- tween this latter line and the “Negligibility Line” is a neering, redundancy in design, self-regulation, fail- region which is judged to be tolerable but for which all safe mechanisms, defence-in-depth, and proper train- reasonably practicable steps should be taken to reduce ing, operations and maintenance, for example. In both the hazard further. This is the ALARP region (As Low situations we should also put in place protective or As Reasonably Practicable). All combinations of fre- damage limitation measures, and other provisions quency and number of fatalities which fall below the (preparation and response etc.), to reduce the unpleas- “Negligibility Line” are considered to be negligible. ant effects should the event happen. The difficulty, of course, is in deciding on the balance of effort that Above a certain level, a risk is regarded as intolerable should be applied to prevention (or reduction of like- and cannot be justified in any ordinary circumstances lihood) versus mitigation (e.g. containment, diversion, (see Figure 1.2). Below such levels, an activity is protection) or to any other aspect of damage limitation allowed to take place provided that the associated risks in the response and recovery phases. Low probability, have been made as low as reasonably practicable high consequence events present the greatest diffi- (ALARP). In pursuing ALARP, account can be taken of culty. Some highly unlikely circumstances such as cost. It is, in principle, possible to apply formal cost- meteorite impact are most efficaciously dealt with benefit techniques to assist in making judgements of simply by repairing the damage — the “do nothing” this kind. [Source: The Tolerability of Risk from Nu- option. clear Power Stations, Health and Safety Executive, London, 1992.] Implementation Issues Management Strategies In practice, the local authority risk manager has the difficult task of controlling the risks from all public The risk equation (R = P x C) points us to the two basic hazards in ways that not only maintain community strategies for managing risk: we can try to lower the safety, but minimise physical damage, and reduce likelihood of an event happening; or we can try to social and economic disruption. More than most, that reduce the consequences by putting in place suitable person must understand the big picture and avoid provisions for ameliorating the worst effects. excessive reliance on any one sector such as the In practice we should do both. With natural hazards it monitoring of developments, response or insurance. is often difficult to do anything about modifying the Their task is to devise a management strategy that likelihood of the primary event occurring, but we ensures every facet of the risk management process is should be aware of possible problems from closely understood, is operating well, and is in appropriate coupled systems and should aim to reduce the chances balance. They need to anticipate potential weak points of further adverse events being triggered (i.e. second- since failure at a single point in the logic or practice of ary events such as flooding leading to dam collapse). a safety process may jeopardise the success of the By contrast, the likelihood of technological accidents whole. For example, where a known hazard exists (such as the

Risk cannot be possibility of a river control system failing catastrophi- Unacceptable justified except in cally) the risk manager has to understand the impor- region extraordinary circumstances tance of each and every factor in the risk equation, and the relative contribution each makes to saving lives Tolerable only if risk reduction is and minimising damage. The following are some of the RISK impracticable or if its cost is grossly more significant considerations: disproportionate to the The ALARP or improvement gained tolerability region • Know what can go seriously wrong, when, and with [Risk is undertaken only what forewarning. if a benefit is desired] Tolerable if cost of reduction would exceed the improvement gained • Examine ways to reduce the chance of it happening. Broad acceptable region Necessary to maintain [no need for detailed working assurance that risk to demonstrate ALARP] remains at this level • Put in place measures to contain the worst effects, divert them away from communities, or otherwise Negligible Risk reduce the severity of the impact.

• Anticipate what might go wrong with the mitigation Figure 1.2: Risk tolerability levels measures and any new risks they may introduce. Risk Assessment, Methodology, Vulnerability, Impact and Importance • 5

• Test whether the warning time is sufficient to evacu- represents the processes that a local authority might ate people (and decide whether the warning system go through in doing a comprehensive risk study. For and communications can be relied upon to work the sake of clarity, the sequence has been broken into properly). five broad steps: analysis, estimation, evaluation, control and verification. Other approaches are possible • Be familiar with all aspects of the response mecha- depending on the nature of the risk, but the process of nisms (who, how, when, how reliable, contact num- quantitative risk analysis would normally involve most bers, availability, alternates etc.). of the steps detailed.

• Consider specific public education or training that In an ideal world, a local authority or other risk might improve the community’s reaction to the manager would undertake a series of risk analyses warning. covering all the risks falling within their jurisdiction in • Anticipate what might go wrong in the response order to rank them in importance and to apportion phase, and what new risks it might introduce. funding for mitigation. In practice, experience and pragmatism would mean that only the most significant • Decide how the injured will be dealt with to mini- risks would be subject to comprehensive analysis of mise further loss of life. this type. The degree of analytical effort applied would tend to reflect the scale of risk, potential benefit, and • Develop plans for recovery (people, materials, familiarity with the hazard (e.g. expert judgement and finance, insurance, etc.). experience might be given greater weight than other techniques in determining risks). But, regardless of the • Look for opportunities to test and evaluate the entire system (especially with small events), and degree of formal evaluation, there are unquestionable make adjustments. benefits in knowing as much as possible about local hazards and in improving the understanding of risk and its control. Experience overseas has shown that Resource Allocation risk management is most effective when those respon- An important corollary is that there is no point in sible for the risk fully accept ownership of the assess- having any one step functioning excessively well. Each ment and control processes. In particular: step should yield an outcome of equivalent quality in terms of its contribution to the overall safety process. • Local authorities need to fully understand their Anything else represents a less than optimum use of local risks themselves. The practice of employing resources. That is not to suggest that effort and re- external contractors does not encourage those with sources should be applied uniformly to each step, but the ultimate responsibility to gain the qualitative rather that each should receive support commensurate insights necessary for informed management. And, with its importance or potential for improving the in-house knowledge is essential to understand the outcome. Moreover, in complex situations involving implications for safety under unusual circumstances many variables and uncertainties the precision with or when elements of the hazard change with time. which the overall risk can be determined will be • Assessment should be specific to a category of risk dominated by the most uncertain or crudely known (e.g. flooding of a defined section of a lake or river) factor. or to a particular facility or hazard (e.g. hydro-dam or volcano). Results from similar situations elsewhere may be relevant, but considerable care is Practical Application needed in adapting from other experiences.

It is not the intention of this paper to detail all of the • Risk management can not be done purely on the steps used in developing risk strategies. That will basis of codes of practice or regulations. Risk depend on the risk being analysed. The process, how- managers need to take responsibility for the total ever, typically involves three broad phases: risk situation, not merely demonstrate compliance with general safety regulations. • assessing the risks quantitatively;

• applying reasonable and effective safety and loss control measures; and Conclusion • evaluating the effects of the overall risk manage- Disasters can be difficult problems to address. They ment programme. are inevitably complex and characterised by high levels of uncertainty. Typically they involve low prob- These are set out in greater detail in Table 1.1, which abilities and high consequences, making them difficult 6 • Risks and Realities

ANALYSE Understand the context of the hazard, existing controls, and safety objectives. Treat all aspects of the hazard and its management as an integrated system. Identify sources of all hazards, vulnerabilities, threats and potential losses associated with the event, activity or system. Clarify potential problems, trigger mechanisms, and conditions of exposure. Develop models, and establish relationships between cause and effect. Analyse consequences of all possible outcomes, especially safety aspects. Consider threats to life, property and environment separately. Consider records, empirical evidence, experience elsewhere, and expert opinion. ESTIMATE Quantify all factors objectively, and determine uncertainty. Carry out sensitivity analysis for the dynamic situation. Consider physical limits and worst credible bounds (e.g. using statistical inference and scientific postulation). Define character and magnitude (size or severity) of consequences. Estimate chance (likelihood or frequency) of event or condition. Calculate component risks and overall risk using Probabilistic Risk Assessment. EVALUATE Determine significance of estimated risks (in absolute and comparative terms). Consider acceptability (in terms of both individual and societal tolerance). Study economic impact and funding options for response and recovery. Examine costs and benefits of control for most serious risks. Assess cost risk benefit balance. Decide to accept, reduce, or transfer risk. CONTROL Minimise, Mitigate and Manage risks, i.e.: develop ameliorative measures that: - lessen likelihood of event and/or consequent system failure; - lower magnitude of consequences; - provide resources for response and recovery. Consider all possibilities for risk reduction: - design for safety, using prevention, protection, and damage limitation; - reduce uncertainty, monitor constantly, maintain and upgrade systems; - set standards and apply quality control at all steps; - develop defence-in-depth (layered response) to counteract small failures; - reduce likelihood of human error or perversity (training and safety culture). VERIFY Test effectiveness of risk reduction strategies. Obtain independent safety audit and inspection. Establish incident reporting methods that include responses etc. Establish feedback mechanisms to learn from experience, then re-prioritise. Develop compliance programmes covering management, training and procedures. Assess cumulative system risk across all stages (including any additional risk introduced by intervention, i.e. mitigation, preparation, response and recovery). Establish quality assurance mechanisms to have all parts optimally balanced. Continually monitor, review and improve systems.

Table 1.1: Risk management strategy for natural and technological hazards Risk Assessment, Methodology, Vulnerability, Impact and Importance • 7

to analyse. The most serious events are (fortunately) tainty, the magnitude of technological and environ- the least likely to happen, but this also means there is mental risk is also increasing. Man is now able to often a lack of reliable data. Data that does exist may create his own catastrophic events, without the aid of be held anecdotally or in a variety of forms by different God. agencies. And, not least, because of their potential to affect people adversely, they can end up being the Some of the types of risks we encounter as a society subject of conflict or controversy. include environmental, psychological, physical, future oriented and political risk. It is noteworthy too that as The analytical techniques now becoming available for well as having varying perceptions as to the degree of interpreting situations of this type have much to offer risk involved in a particular activity, individuals and those responsible for public safety and loss prevention. groups have different perceptions of the types of risks Risk management strategies of themselves cannot guar- involved. antee better performance because of both the role that chance and uncertainty play, and the vagaries associ- People concerned with the study of risk include phi- ated with human intervention. But the methodologies losophers, sociologists, business managers, econo- used for assessing risk can contribute to understand- mists, engineers and politicians. Of these, philoso- ing where the most serious components lie. They can phers, sociologists and economists are concerned pri- point to the more promising control options, assist marily with the characteristics of risk and choice under policy development, and inform the allocation of re- uncertainty, engineers are concerned with quantifying sources. Risk analysis is therefore going to be used risk, while managers desire to manage and reduce risk. increasingly in the management of emergencies and Politicians and decision makers rely on information natural disasters in New Zealand. The message from obtained from the other groups to make decisions government is clear. Those responsible for public which will in part reflect the risks involved and in part safety and for managing infrastructural assets in New be the result of other contributing factors. Zealand are obliged to assume full responsibility for managing the risks, i.e. to identify the hazards, to Page 3 assess the risks, and to take whatever precautions are Risk-related problems tend to enter the public domain required. Support from central government is condi- when the magnitude of the potential outcome is so tional upon proper risk management having been great that it might have possible severe or even cata- demonstrated. These are not matters that can be regu- strophic consequences for large land or sea areas or lated in Wellington. It is up to the local authorities or big population groups. These population groups may asset owners to take responsibility for balancing the be identified by geographic, demographic or other costs, risks and benefits in the best interests of the social boundaries. communities they serve.

Page 6 Relevant Quotations The decision-making process begins with the decision that a problem exists, and in any decision-making The following quotations are from Risk and Uncer- process there will be a number of value elements about tainty by Janet Gough (Information Paper No.10, Cen- which value judgements must be made. These judge- tre for Resource Management, July 1988). ments are necessary because there is seldom a commonly accepted ‘correct’ approach. Rowe (1977) di- Page 1 vides these value judgements into three groups: Risk is an important part of our everyday existence. We (1) technical value judgements; continually expose ourselves or are exposed to risk over which we may have little or no control. Our (2) social value judgements; and perception of the risks we encounter varies according to factors such as whether our exposure is voluntary or (3) managerial value judgements. involuntary, how much control we feel we have over the Technical value judgements are important, since very risk, and whether or not we feel that the risk is ‘fair’. often the ‘experts’ who make technical risk estimates Increased knowledge, as well as technological and do not recognise that these estimates are in fact value institutional changes are giving us greater control judgements because of the uncertainties and assump- over our environment and at the same time allowing us tions involved in their calculation. Social and manage- to modify it at a much faster rate than previously. The rial value judgements are more readily accepted as number of risks involved is increasing and as greater such because they are generally more open to scrutiny knowledge does not necessarily reduce total uncer- by external observers. 8 • Risks and Realities

Page 15 • historical levels of risk which continue to be an Statistical estimates of risk are used when there are acceptable one; and sound statistical data available for the particular event being studied. In this way, we can make reasonable • risk which is deemed worth the benefits by the risk estimates of the risk to the child from a mother smoking taker. during pregnancy (using long run average frequency He refers to these as the threshold condition, the status estimates). The difficulties arise when, for example: quo, the regulatory condition, the de facto condition (1) the historical data being used is not sufficiently and the voluntary balance condition. These are exam- specific for the purpose for which it is being used; ples of risks which we argue should be termed accepted risks. We further contend that there is no such thing as (2) the data does not cover a sufficiently long period; acceptable risk, but only accepted risk, and that this is what is commonly meant by the term acceptable risk. (3) the estimate obtained is applied to a different population to that from which it was derived; or

(4) the probability of occurrence is very low. Risk management considerations and the Page 19 community Fischhoff (in Covello et al., 1981) lists six reasons why disagreements occur between the public and experts: The following discussion presents some issues of relevance to risk management beyond those specifically (1) the distinction between ‘actual’ and ‘perceived’ considered in the project work. To some degree, these risk is misconceived; are implicit in what has been done, however there are gaps in the work and these are recorded to qualify the (2) lay people and experts are talking different lan- results to date and to flag areas for attention in any guages; future effort. (3) lay people and experts are solving different problems; Risk acceptance criteria In compiling mitigation measures, the project identi- (4) debates over substance may disguise battles over fied many things of modest cost that have already been form and vice versa; adopted. Other sensible measures can be phased in (5) lay people and experts disagree about what is through ongoing maintenance and replacement pro- feasible; and grammes, often with little or no extra cost. Where the requirement is costly, however, and may have some (6) lay people and experts see the facts differently. urgency, there is a need for a rational approach to decision-making. Page 31 The following discussion describes how decisions Fischhoff (1978) defines acceptable risk in the follow- could be made regarding proposed mitigation meas- ing way: “The acceptable level is the level which is ures in terms of the effects on the community of the ‘good enough’, where ‘good enough’ means that you particular hazard scenarios considered. Largely, the think the advantages of increased safety are not worth decision whether or not to invest in mitigation rests on the costs of reducing risk by restricting or otherwise the level of risk that is acceptable to those affected. The altering the activity.” more risk that may be accepted, the less need there is Rowe (Goodman and Rowe, 1979) describes a set of for mitigation. conditions which he suggests support the existence of The perceived magnitude of risk consequences, and acceptable risk. They are: the probability of occurrence in relevant forward time • risk which is perceived to be so small as to be frames, will determine acceptability or non acceptabil- deemed negligible; ity to those affected.

• risk which is uncontrollable or unavoidable with- In discussing the concept of risk, it is suggested that out major disruption in lifestyle; people are concerned about effects that may realistically occur within their “cognisant lifespan”, say, 50 • acceptable risk levels established by a credible years. To the extent that people’s views can be ex- organisation with responsibility for health and tended to society as a whole we should consider impact safety; through loss of lifelines on that basis. So an imperative Risk Assessment, Methodology, Vulnerability, Impact and Importance • 9

for action is the likelihood or probability of the out- being exceeded in a 50 year period. This is equivalent come of concern occurring, or, more properly, being to a one in ten year occurrence event. exceeded within a 50 year period.

We therefore have the two-dimensional problem of Reasonably probable selecting hazards that occur often enough, and which Probability of exceedance greater than 50% and less are severe enough in terms of the consequences of than 99.5% in a 50 year period. lifelines failure that we should do something about them. Remote Figure 1.3 illustrates this problem and proposes the Probability of exceedance greater than 5% and less existence of decision spaces which may assist in differ- than 50% in a 50 year period. entiating between acceptable risk and risk which may require action. Extremely remote Obviously, if a frequently recurring event (lifelines Probability of exceedance less than 5% in a 50 year failure) causes catastrophic effects in the community, period. This is equivalent to a 1 in 1000 year event. the risk of failure is unacceptable and action is indicated. Conversely, if extremely rare occurrences result The problem with these definitions is that they place in only minor impact, the risk of failure will be accept- the majority of our hazards in only two classes. They ably low and no action would be required. In the range do, however, reflect an intuitive understanding of the between these cases there will be a transition from terms, “frequent” and “extremely remote” in particu- “yes, we must act” through ambivalence to “no, we lar. don’t need to act.” The return period of exceedance and the probability of The assignment of hazards to decision spaces is de- exceedance for identified hazard scenarios are given in pendent upon the meaning of “frequent”, “remote”, Table 1.2. The relationship between return period of “major”, “catastrophic”, etc. Some definitions of terms exceedance, and probability of exceedance for engi- have been drafted and these are given below. These neering lifelines is discussed below. definitions, and the perception of the hazards issues, are then used to assign them to decision spaces. Effects Parameters: The effects of lifelines failures on the community may Probability parameters definitions be expressed in terms of inconvenience, economic cost, threats to life and health, social impacts, and Frequent threats to the natural environment. A significant factor A 99.5% probability of the particular hazard scenario will be recovery time.

Increasing probability

Manage risk Frequent

Accept risk Reasonably Probable Conditional

Remote

Extremely Remote

Minor Major Critical Catastrophic

Increasing effect on community due to lifeline failure

Figure 1.3: Risk acceptance criteria 10 • Risks and Realities

Hazard Return Period of % Probability of exceedance exceedance (years) (indicative) In 1 year In 50 years Local flooding 1000 0.1 5 Waimak flooding 500 0.2 10 Seismic 150 0.67 28 Wind 150 0.67 28 Tsunami 150 0.67 28 Snow 50? 2.0 64

Table 1.2: Probability and return period of exceedence for identified hazard scenarios

Minor effects Catastrophic effects Temporary loss of one or more utility services, recov- Total loss of most lifeline services would occur requir- ered generally (may be residual local loss) in less than ing major external support effort and emergency regu- one week with essential services restored in one to two lation for many months. There would be significant days. Loss of life would be unlikely, however there loss of life (one hundred to low thousands). There may be freak accident scenarios. The well-being of would be major social impacts both locally and prob- people on life support or equivalent systems may be ably New Zealand wide. Serious health issues would placed at temporary risk. Economic impact would be arise that could not be properly addressed. There would generally low but could be selectively significant for be a devastating impact on the local economy with very some businesses. Social impacts are largely of very difficult consequences for the national economy. temporary inconvenience. On that basis the hazard scenarios could be plotted as shown in Figure 1.4. Major effects Short-term loss of more than one service (could be The outcome of this provisional evaluation is that only through interdependency), say, two to three weeks, seismic hazard is clearly out of the acceptable risk which would require importation of resources to re- category, but it is still not clearly of the “definite action tain essential services. Possible loss of life, but low required” category in terms of fundamental commu- numbers (single figures?). Some at-risk people may nity needs. That is not to say that action is not relevant die through loss of life support, etc. Social impact to optimise the performance of utilities, or that co- would largely be one of sustained inconvenience with operation between utilities owners in prioritising miti- associated stress and conflict. There would be a sig- gation and recovery action is not of social benefit. nificant impact on the local economy, but this would We may be looking at fine tuning rather than a need for be generally recoverable with local resources. There a fundamental shift in policy in lifelines management. may be some business failure. Who defines acceptable risk Critical effects What is an acceptable level of risk depends upon who Medium-term (three to six months) loss of more than is affected by the hazard. Each community of interest one service, which would require substantial assist- will have different needs. Those concerned will in- ance and temporary provision for meeting population clude utility owners, utility operators, utility customers needs (for example, evacuation, water delivery, tem- and the community as a whole. The wider community porary field hospital facilities, etc.). There would be needs may be met to some extent by regulations (e.g. moderate loss of life, say less than one hundred. Public the Building Act), however, in this case, regulations do health would be a serious management issue, people not fully cover the issues concerned. It is suggested that with marginal antecedent health likely to die. There the lifelines participants can represent the interests of would be major social and economic impact on the owners and operators reasonably well. The needs of local community that would require large-scale relief customers and the wider community (is there a differ- from outside the region. This would provide difficult ence?) is something else again. but manageable resourcing issues for central government. In applying the above method to assessing acceptable risk, the assignment of issues to decision spaces rests Risk Assessment, Methodology, Vulnerability, Impact and Importance • 11

Increasing probability

Manage risk Frequent

SNOW/WIND Accept risk

Reasonably SEISMIC Probable Conditional LANDSLIP TSUNAMI WAIMAK FLOOD Remote LOCAL FLOOD

Extremely Remote

Minor Major Critical Catastrophic

Increasing effect on community due to lifeline failure

Figure 1.4 : Acceptability of lifelines hazards risk heavily on the definition of terms and the individual tial of all possible event scenarios for the period under perceptions of anyone brave enough to use it. The consideration. “right” decision would rely upon it being in accord with wider community perceptions. Failure of our In theory, mitigation can be optimised by designing to system management will be signalled by adverse pub- a particular exceedance probability such that the damage potential saved divided by (the cost of mitigation lic reaction. Of course the wider community would plus the residual damage potential) is a maximum. have limited understanding of the effects of loss of (The cost of mitigation depends on probability as well.) lifelines until the hazard actually occurs. In the mean- This approach is routinely used for flood mitigation time, the best we can do is use our collective experi- decision making in New Zealand. Also, by considering ence, and the experience of others, and propose a the full range of event scenarios, knowledge of the course of action given overt assumptions. This is what effects of super design events will be of value to is being done in this project. The community at large preparedness and response planning, if not mitigation. should be consulted in some way to check the validity of these assumptions, which would in turn require public education to ensure consultation is informed. This is a job for the utilities owners over time, with appropriate technical input. However, it could also be Maximum credible damage considered in any future lifelines efforts.

Hazard scenarios Residual damage potential The project used discrete hazard scenarios selected on Damage saved by “100 year the basis of the knowledge and intuition of the hazards standard” mitigation task group, as moderated by other participants in the project. There are some inherent difficulties in this Potential damage rate (e.g. $) approach; in particular, the single scenario may leave some surprises for us that a broader view may have 0 0.01 0.02 0.03 avoided, and the single scenario may not be the opti- Probability of event scenario exceedence in x (e.g. 1) years mum in terms of return on investment in mitigation.

By studying a range of scenarios the above difficulties Figure 1.5: Hazard damage/probability can be overcome to some extent. relationships A plot of potential damage rate against probability of But an approach of this type would have required exceedance for a range of event scenarios typically immensely more work (on a component by component takes the form shown in Figure 1.5. The area under the basis) which could have been unsustainable in terms of graph equates to the total unmitigated damage poten- utilities support for the project. Moreover, an exhaus- 12 • Risks and Realities

tive approach of this type would not be appropriate It is acknowledged that the methodology closely fol- given the high levels of uncertainty involved in our lows that adopted for the Wellington Earthquake Life- knowledge of the hazards, and the network responses lines Project (Centre for Advanced Engineering, 1991), to them. to which we are greatly indebted.

It is therefore considered that the approach taken — the selection of a realistic scenario and mitigating its Risk management effects — is a reasonable first step. One thing that In its pursuit of more resilient networks, any lifelines would be of value for future study would be to put the project is essentially an exercise in risk management. work in the perspective of possible super design events. Because we have no uniquely defined problems in The historical experience in flood mitigation is that front of us (an actual disaster) we have to deal in design standards are regularly exceeded in nature and possible disaster outcomes which we hope will bear it is now common practice to at least consider the some resemblance to events when they do occur in the effects of maximum credible events, rare though they future. Experience elsewhere, and our knowledge of may be. the Christchurch networks and likely hazards, was combined to forecast, with no little uncertainty, what can go wrong and from that what may be done to Project Methodology prevent or mitigate it. Risk is often expressed numerically as the product of Introduction quantified outcomes (e.g. $ damage) and probability Earthquakes and other natural hazards can damage the of occurrence. Outcomes can then be compared and utility networks that provide essential support to hu- the preferred one chosen. man life and well-being. Severe social consequences can arise through network failures. Network owners As in most risk situations, the definition of natural need to allow for appropriate levels of hazard mitiga- hazards risks to Christchurch lifelines is fraught with tion in their asset management strategies. They also uncertainty. Numerical methods of risk assessment need to consider how they will respond to disasters therefore have their limitations. They are, however, a when they occur and it is useful if network owners useful aid to decision making, as long as uncertainty collaborate in this. Resources can be shared in the is taken into account. definition of hazards, and in working through their The project used a broad-brush approach to quantify- implications. But, most importantly, in most cases ing risk. This provided useful comparative informa- there is a high level of interdependency between life- tion, but it did not promote absolute figures for dam- line services. Each lifeline generally needs the others age which could be related to cost of mitigation. in some way. Collaboration in mitigation is therefore essential if the best results are to be obtained. Utilities operators need to do that for themselves. Risk management may be considered to include the fol- A key function of the project was to provide hard lowing steps: information to facilitate hazard mitigation measures which meet the needs of each network and the lifelines • definition of the system at risk, in this case a utility system as a whole. network, and its components;

It recognised that natural hazards form only part of the • description of the hazards faced by the system, and overall concerns relating to network vulnerability. its vulnerability to those hazards, in this case Other hazards include human error in operation, defec- natural hazards; tive plant, fire, chemical spills and sabotage. Budget • description of the damage which may be sustained decision makers will also consider factors such as by the system and its wider impacts; operational costs, service demands, wear and tear, etc. Often factors other than natural hazards will prevail in • identification of appropriate mitigation measures; the work and purchase programme, but even so it will and often be a simple matter and relatively inexpensive to incorporate mitigation measures to avoid the worst • preparation and implementation of an action plan. effects of hazardous events. While the lifelines project broadly followed this proc- This section outlines the methodology adopted for the ess, there are some areas of acknowledged weakness project and places it in a context of risk management. which have been accepted for expediency, and prac- Limitations of the approach are discussed and sugges- ticality. These are discussed in “Risk management tions are made regarding any possible future work. considerations and the community” (page 8). Risk Assessment, Methodology, Vulnerability, Impact and Importance • 13

Project structure Risk analysis Project participants were structured into task groups Each network was broken up into components, which firstly to identify the hazards and then representing were then assessed for their vulnerability to each network types: hazard scenario. This was done by overlaying electronically the networks over the hazards map which • hazards; was then examined to determine importance, vulner- • civil services; ability and impact of damage.

• electrical and communication; A scale of 1 to 3 was used to define component vulnerability: 3, 2, 1 corresponding to high, moderate, • transport; or low probability or distribution of failure. (“Probabil- ity” relates to components that are discrete elements, • buildings; and “distribution” relates to components that are grouped • fire. elements, (e.g. networks)). ‘0’ indicates no susceptibility to damage. The hazards group defined the natural hazards to the The consequences of failure of each network compo- networks, such as seismic, flooding, tsunami, landslip, etc., and provided ongoing advice as the networks task nent were then assessed, taking into account its impor- groups worked through the detail of their assessments. tance to the network concerned, and the wider impact of its being damaged. The networks task groups were responsible for de- Importance was ranked 1 to 5, with 5 being most scribing the networks considered to be critical to delivery of lifeline services, and for working through the important. A component’s importance would be ranked risk management process to the action phase. highly if it were essential to the function of the network as a whole. It would be ranked low if it could be The project organisation is shown in Figure 1.6. bypassed.

Impact related to the degree of disruption caused or the Hazard scenario selection effort required to reinstate it. Impact was consequently A disaster scenario was described for each natural assessed for three timeframes: hazard under consideration. The approach taken was to select a scenario which, on the one hand could be • immediately after the hazard event; considered to pose a significant hazard, but on the other • the period following the hazard event; and had some realistic probability of exceedance in forward planning timeframes. • the time for return to normality.

Wherever possible, the hazard was represented in a “Immediately after the hazard event” covers the period hazards map, otherwise the hazard was described as a for recovering a minimum emergency service by scenario, e.g. wind and snow. backup, bypass or temporary repair.

Some limitations of this approach are discussed in The “period following the event” is the period to “Risk management considerations and the commu- restore a full service, albeit perhaps on a temporary nity” above (page 8). basis.

Steering Committee

Project Manager

Hazards Civil Electrical Transport Buildings Fire Services Telecommunications

Figure 1.6: Project organisation chart 14 • Risks and Realities

The “return to normal” period relates to full restoration was undertaken. The rankings are based upon the task of reliable services. groups perception of community needs.

Impact factors reflect the community needs for rein- The vulnerability, importance and impact rankings statement of services in the timeframes concerned, but were assembled on a vulnerability chart (see Figure it should be noted that no survey of community needs 1.7).

Utility: Regional/Local Network:

Vulnerability to Hazard Impact of Damage

Component Element Ground Shake Comment Importance 1 - 5 Liquefaction Landslide Ground Settlement Zone Boundary During Earthquake Immediately After Period Following Return to Normality

Vulnerability Chart: Define components and elements of network at Regional and District level For each component: Assess importance 1 to 5 — 5 most important Assess vulnerability on hazard map 1 to 3 — 3 most vulnerable Assess impact of damage 1 to 3 — 3 most impact

Figure 1.7: Sample of vulnerability chart Risk Assessment, Methodology, Vulnerability, Impact and Importance • 15

Mitigation measures factor. It is no good for the wider community if indi- The vulnerability charts were then used for identifica- vidual utilities go their own way in ignorance of the tion of the need for mitigation measures. Any measure impacts of other utilities on them, and vice versa. that would reduce importance (e.g. by redundancy), Interdependencies are discussed further in Chapter 10. reduce vulnerability (e.g. by strengthening), or reduce impact (by alternatives or contingency planning), would mitigate the impact of the disaster. The rankings pro- Risk in Relation to Mitigation vide a basis for prioritising action where budgets may be a limiting factor. In many cases, however, mitiga- The ultimate measure of the success of the whole tion measures identified were very inexpensive, or project was the extent to which budget provision was could be easily integrated with ongoing maintenance made for the various mitigation measures that have and replacement programmes. been identified. To assist in this it was necessary to attempt to translate the technical assessments under- An holistic approach embodying all opportunities for taken into understandable information so that the deci- mitigation will include strategies at the asset planning, sion-makers (which includes the public) can them- engineering and operational levels. selves have some conception of the risk involved and the justification for any expenditure. Utility service provision can be planned to encompass: It is important that as a result of publicity regarding the • hazard awareness programmes; project, people do not become unnecessarily con- • planning for redundancy; cerned. It has to be remembered that the work was essentially an engineering lifelines project, the result • locating to avoid hazards; of which will mean improved security of the engineering lifelines after some catastrophic event. It does not • review of alternatives for critical components; and attempt to address Civil Defence concerns. (Neverthe- less the whole project will be of inestimable value to • systems monitoring and response planning. Civil Defence in the event of an emergency.) Engineering design and detailing can include: The public sometimes have such a perception of a risk • hazard-wise layout and conceptual design; that, however unlikely the event is to occur, they are prepared to pay for the mitigation of the hazard produc- • hazard-wise design standards; ing that risk. For example, the people of Christchurch have always been concerned about flooding from the • hazard-wise code levels; Waimakariri and although it would appear the likeli- • hazard-wise procedures and practices; hood of Christchurch flooding from the Waimakariri is very much less than the risk from the other hazards • hazard-wise detailing; and identified, the public of Christchurch still want the city protected from the Waimakariri River flooding. • special measures for critical components. Reference to some of the quotations on pages 7 and 8 Each utility’s operations management can incorporate: draws attention to the different way in which risks are • preparedness and response planning; perceived by different people, and concern could be expressed that those involved in the lifelines project • appropriate spares holdings; are mainly engineers or managers. Although these people are decision-makers in their own right, most • equipment inventories; ultimately report to Councils or Boards of Directors • rerouting options; whose members are most often non-engineers. It was thus necessary to attempt to express the likelihood of • damage assessment; and the various events and the justification for the mitigation measures in understandable terms for the decision- • planning staff training and exercises. makers, which include the Councils, Boards of Direc- tors, and ultimately, of course, the public. Interdependence In endeavouring to tease out proposed mitigation meas- The methodology described above was carried out ures in one of the Task Groups, one engineer made his network by network. The interdependence of net- own value judgement to the effect that “the (decision- works, both in operation (if A fails, B fails), and in making body) would never give us funds for those sorts response (need to fix A to get to B) can be a critical 16 • Risks and Realities

of measures. It is a waste of time identifying commercial light and the impact of being unable to them. We are much better to recommend some- operate a business for some time may warrant expendi- thing that may be financially viable”. In the ture that could be said to be in the form of an insurance. context of the lifelines analysis this should not However, in some cases there will not be a commercial be the approach. Ultimately, someone has to loss but a very great inconvenience and, in fact, even take the various recommended mitigation meas- danger to the public. It should be possible to distinguish ures from the report and recommend their adop- between these — it is not simply an engineer’s decision tion in an environment that at budget time was as to what work will be done. almost inevitably one of pruning desirable works. The person making those recommenda- In effect, the project drew to the attention of the tions, who is part of the decision-making proc- participants in the project the risks that existed. The ess, has a very real responsibility to understand extent of investigation and funds allocated is left to the what is involved and be best able to assist the individual authorities. Problems may arise when these decision-makers. authorities take decisions that affect the interdependence of other lifelines. Interdependence issues must be The real test as to whether or not a community taken into account when budget allowance is made. It is prepared to accept the risk occurs at budget is also important to regularly review progress on the time, when the authorities decide whether or mitigation measures and this alone would justify a not to allocate money for mitigation measures. continuing lifelines group. It is therefore most important that the decision- makers have the risk assessment expressed in Those involved in Civil Defence are aware of the understandable terms. general “motherhood” status of Civil Defence in that everybody agrees that it is necessary but very few do Budget needs by public authorities are assessed anything actively about it. An engineering lifelines often with little flexibility because of the fixed investigation must not have that same status. charges that have to be met and the maintenance requirements, so the “spare money” is very The investigation associated with this project was small and there is a lot of competition for the technically very interesting but would have failed in its discretionary budget. objectives if the work on the project was not translated into budget provision for mitigation measures to be Some decisions may be made simply in a hard undertaken (see Chapter 11). Hazards to Engineering Lifelines • 17

Chapter 2 Hazards to Engineering Lifelines in Christchurch

2.1 Hazards Considered explanatory notes and background information for each hazard are presented in the following sections. The Engineering Lifelines Project has considered a number of natural hazards in the Christchurch Area. These are: earthquake; flooding; tsunami; severe wind Earthquake hazard storm; severe snow storm; and slope hazard. The earthquake scenario is a generalised one applicable to a range of earthquake magnitudes with epicen- Only one scenario for each hazard has been considered tres at various distances from the city. It is based on for this study, mainly to keep the overall project to a recent work reviewing the seismic hazard in Christ- manageable size. Further work on Christchurch Life- church, and postulates the effects of a 150-year return lines could include considering a number of scenarios period earthquake. This return period was originally for each hazard to allow a more deliberated risk assess- selected as it corresponds to the return period incorpo- ment and mitigation programme to be made. However, rated into the structural loading codes, and is large whether or not the increased certainty of the risks enough to cause very significant damage, but is not so involved justifies more detailed work in task groups is remote a probability as to be able to be discounted as questionable. More work on the seismic hazard risk is irrelevant. definitely worthwhile.

Initially the hazards group intended to consider hazard Flooding hazard scenarios with the same return periods, but the prob- Two flooding hazards were considered — a breakout abilities of the various hazard scenarios as produced of the Waimakariri River, and local flooding in the vary. The annual exceedance can be summarised as: immediate study area catchments. The Waimakariri River event has been sourced directly from work Hazard % Probability of carried out for the Canterbury Regional Council exceedance Waimakariri Floodplain Management Plan. It assumes In 1 year In 50 years a 500-year return period flood, as lower return period Local flooding 0.1 5 floods would not affect the city, and at this level of flow Waimak flooding 0.2 10 there is only a 50% chance of breakout. Seismic 0.67 28 Wind 0.67 28 The local flooding hazard has been taken from Tsunami 0.67 28 floodplain management studies by the Christchurch Snow 2.0 64 City Council, which take a 1 in 500-year return period Slope hazard 0.67 28 flood in each of the Styx, Avon and Heathcote Rivers, coinciding with extreme water levels in the Estuary The reason for these differences is entirely pragmatic and Brooklands Lagoon, to give an overall return — the Waimakariri Flood Hazard is non-existent for period of about 1,000 years. less than a 500 year return period flood, and the local flood and slope hazard have been based on studies already carried out. To revise these studies for the 150- Tsunami hazard year return period events would have required a level The Tsunami event was assessed specifically for this of effort that was not considered warranted given the project and assumes a large remote seismic source. budgetary and programme constraints for the lifelines The return period cannot be defined accurately, but can study. be assumed to be in excess of 150 years.

The hazards have been presented in map form with 1 The coloured maps illustrating the various hazards are on pages explanatory notes, wherever possible1. The wind and 283 - 304 and are presented with an overlay of various features (e.g. emergency services and contractors yards) rather than showing the snow hazards cannot be realistically mapped, and no hazard only. The working maps used by the Task Groups were maps have been prepared for these storm events. The available at much larger scales. 18 • Risks and Realities

Both the extreme wind storm, and severe snow storm Alpine Fault. The magnitude, M, of an earthquake is scenarios have been based on historical events, but measured on the Richter Scale and relates to the energy enhanced to an estimated 150-year return period. released by the earthquake at the epicentre, or source of the earthquake. The effect of the earthquake decreases Slope hazards with increasing distance from the epicentre. Hence any one of the above earthquakes could produce a similar Slope hazard affects a relatively small proportion of level of shaking in Christchurch. the study area, and significant problems with lifelines on the hills is most likely to occur in conjunction with earthquake or local flooding. This hazard has been Intensity based on a study for planning and building consent The effects of an earthquake are measured on the purposes carried out for the City Council, and is for a Modified Mercalli (MM) scale of felt intensity. This is triggering event of a 100-year return period rainstorm, a descriptive scale which reflects the intensity of shak- or 100- to 150-year return period earthquake. ing according to damage and felt effects. While the magnitude of an earthquake event is a single measur- Fire able value, the intensity will vary according to distance from the earthquake source and the ground and struc- Fire is a likely secondary effect of a major earthquake. tural conditions of a particular location. A scenario for the post-earthquake fire situation has been assessed by the Fire Service, and is considered separately to the earthquake in Section 7.2 (“Building Local geology Services”) section. The shaking intensity at a site is affected by the ground conditions, with sites of deep soils generally showing Volcanic Activity amplified motion compared with sites on rock. Christ- church is situated over geologically recent deposits of Volcanic activity is not regarded as being a hazard in alluvial gravels laid down by the Waimakariri River, Canterbury, and has not been considered in the study. and fine marine sediments deposited on the coastal margin of the floodplain and in estuaries and lagoons. The sediments are about 700 m deep, lying on 200 m to 2.2 Earthquake Hazards 300 m of volcanic rock overlying a greywacke basement at about 1 km depth. To the south of the city the Magnitude sediments become shallower against the weathered These notes should be read in conjunction with seismic volcanic cone of Banks Peninsula. The Port Hills are hazard maps 1, 6, 9, 13, 14, 15, 17, 18, 19, 20 and 21 mantled with loess soils over the basalt rock. (page 284 et seq). For further information on the seismicity of the region, Christchurch is situated near the edge of a tectonically and the geology of the region and the city, refer to active region, with active faults to the north and north- Brown and Weeber (1992) or Elder et al. (1991). west in Canterbury, and the Alpine Fault to the west. These faults mark the boundary between the Indian Earthquake scenario event Plate to the north and west and the Pacific Plate to the The earthquake event adopted for this study is a major south and east, which are moving obliquely to each earthquake producing shaking intensities with a return other. Earthquakes occur as movement between the period of about 150 years. Shaking intensities of VIII plates is accommodated along the faults. The nearest to IX on the Modified Mercalli Scale would be ex- known active faults to Christchurch are the onshore pected over most of the Christchurch area. Such shak- Ashley Fault 30 km to the north and the offshore ing would most likely be caused by a moderately-large Pegasus Bay Fault about 30 km to the north east. to large earthquake in the Canterbury foothills or North Known active faults in Canterbury are shown in Figure Canterbury. A very large earthquake on the Alpine 2.1. Fault would also be likely to produce these shaking The maximum magnitude of earthquakes on the faults intensities or greater. It is also conceivable that an has been assessed from geological evidence, and gen- earthquake centred close to Christchurch under the erally increases with distance away from Christchurch. Canterbury Plains or Christchurch itself could produce The maximum magnitude of earthquakes on the faults intensities of MM VIII. closest to Christchurch is postulated at M6.6, those on The effects of any of those three possible earthquakes faults in the foothills region at about M7.5, and earth- will be similar, although the longer duration shaking quakes on more distant faults at M8 or M8.5 on the probable with the larger, more remote earthquakes Hazards to Engineering Lifelines • 19

171 172 173174 200 174 200

T L U F F A E L R F E L F U T GLASGOWE A K A Y IR W L C A A W

E 42

T I

42 200 H IT F 200 W G CLARENCEFID F N

? F E R TE WA A ELLIOT F ? Crustal seismicity 1942-1964 HOPE F

Crustal seismicity 1964-68 E JOLLIES E (magnitudes known accurately) 1888 AMURI PASS F L ? M = 7 - 7.3 A ND CLARENCE F HU O F AP ? KAK 150 ALPINE FAULT HOPE F 1901 CHEVIOT F M = 6.5 - 7.0

BALMORAL A

F A F LLY W ? E I K A K

) 1922 E? UT F MOTUNAU SP TE F DI HI M = 6.7 (IN Y U W UNA t E T 43

M MO

R G OMIHI 43 ESK F

BRUCE FZ HARPER F FZ E N SS O LE Y OR SS LEY F F N D PA ASH Y A N T A C E S B S R R US U T TE AS S L R EG A A O P G R P E T P N E BENMORE C

RAILWAY F

+ 1869 NEW BRIGHTON M = 5.7

LAKE HERON RAKAIA 100 FZ GORGE

50 IRISHMANSCK FOX'S 44 PEAK 1:1,000,000 44 FZ 150 100 10 0 10 20 30 40 50 60 70 80 90 100 (km)

OSTLER FZ 200

170 171

Figure 2.1: Active faults within 200 km of Christchurch such as the Alpine Fault would produce more damage Appropriateness of the earthquake and more extensive liquefaction. There are few geo- scenario for the Christchurch lifeline logical indications of surface fault traces either in the study Canterbury Plains or Banks Peninsula, and the likeli- The Christchurch Lifeline Study has adopted a 150- hood of any surface rupture in Christchurch is very year return period earthquake of shaking intensities remote. VIII-IX over most of the Christchurch area. This Because the most damaging earthquake shaking in reflects primarily the conclusions of Elder et al. (1991). Christchurch will be from earthquakes at some dis- This study first modelled the shaking intensities in the tance, the pattern of seismic hazard will be similar for bedrock under the city and then allowed for expected earthquakes with a range of return periods. The same amplification effects, particularly in the top 20 m of areas susceptible to liquefaction at 150-year return young, soft sediment. periods are also susceptible at 50- or 500-year return period events, but clearly the proportion of these areas At the Lifelines workshop in October 1994, David that will liquefy and the extent of resultant damage will Dowrick and, to a lesser extent, Warwick Smith, who be very different. This is different to cities such as are scientists with the Institute of Geological and Wellington, where movement on an active fault within Nuclear Sciences (IGNS), disagreed with this intensity the city will produce a different distribution of shaking scenario for the 150 year time period adopted. and hazard than would be produced from a remote “regional” earthquake. Limited discussion followed which created uncertainty as to the real extent of seismic hazard in Christ- Only one scenario has therefore been adopted for the church. The principal issues raised are discussed study. below. 20 • Risks and Realities

Recorded seismicity record - problems Overshadowing the whole question of predicting the with predicting the frequency of future recurrence interval of earthquakes in the central South events Island is the Alpine Fault and the associated problem of Warwick Smith pointed out a mismatch within the the “seismic gap” (Adams, 1980). Once again, on this Hope Fault Seismic region of the Elder et al. model fault, there appears to be a significant gap in the between the occurrence of recorded small-to-medium expected small- and medium-seismicity normally ex- earthquakes since 1964 (when records began), and the pected adopting the Gutenberg-Richter model. corresponding predicted return period of large earth- The limited geological work to date suggests a large quakes. earthquake (of approximately Magnitude 8) occurs on If the small events are used in isolation to predict the the central Alpine Fault at around 500 year intervals frequency of the larger ones a much lower occurrence with the last event approximately 550 years ago. If of large events is predicted than that established from correct, this raises the likelihood of an Alpine Fault geological studies. Cowan (1989) obtained reliable earthquake in the next 150 years. Apart from the direct recurrence data for large earthquakes of around magni- damage caused, this may in turn affect the activity of tude 7 to 7.3 on this fault at approximately 130 year the nearby active faults in the Canterbury foothills and intervals. The recorded seismicity since 1964 predicts mountain areas in the years following a large Alpine a much longer recurrence interval. Fault event.

This same mismatch of record seismicity with reliable Dowrick views faults in these areas as being the most capable of generating large earthquakes in Christ- geological evidence of larger events has been observed church. Unfortunately, with the exception of areas of where data is available in studies in California (Cop- the Porters Pass fault, these faults have not been studied persmith and Schwartz, 1986). This has led to sugges- in much detail. After discussion with University of tions that the traditional Gutenburg-Richter log rela- Canterbury geologists, Elder et al. adopted recurrence tionship between the frequency of small and large intervals of 500 to 4000 years. In contrast, we under- events may not be the most appropriate. stand IGNS geologists prefer 2000 to 5000 years. In Geological studies effectively sample the activity of either case the recurrence intervals are long and the large earthquakes over hundreds of years. Where this historical record is not long enough to adequately information is available this is preferable to the use of define their activity. a short interval of 30 years of recorded small events to predict the next 150 years. Differences in the earthquake scenarios This is particularly so given the good evidence sug- For the reasons outlined above, the prediction of future gesting that the last 30 years have not been typical, but seismicity should be done with great caution and the instead may represent a relatively quiet period in the numerous unknowns involved demand a conservative seismic history of the area. Figure 2.1 shows the record approach. of historical earthquakes for the Canterbury area. Look- However, at present the difference in the various earth- ing at the frequency of these events it is obvious that the quake scenarios are not as great as might be expected. period of 60 years from 1869 to 1929 was much more Return periods for given intensities from the two Smith active than the period since. and Berryman models and the Elder model are shown There is no way to know with confidence what number in Table 2.1. of small earthquakes were occurring prior to 1929, but If the most recent Smith model is adopted, the 150 year it is reasonable to assume their activity was also greater. scenario would be MMVII to MMVIII whereas the Unfortunately the 30 years from 1964 definitely looks Elder et al. model predicts MMVIII to MMIX. quiet by comparison. In effect, Smith and Berryman would suggest the This phenomena has been earlier recognised by adopted scenario should have a 250-year plus return Robinson (1979) who pointed out that the overall rate period, not the 150 years used. But, most importantly, of energy release for shallow earthquakes in New there is no suggestion that shaking of this intensity will Zealand from 1950-1977 is only about 12% of the not occur in the future. average rate for the longer period of 1840-1977. A pragmatic alternative approach in selecting the ap- In the 150 years of European settlement for which we propriate scenario is to say that we should expect in the have records of large earthquake events, we have not next 150 years an event at least as great, if not a little experienced earthquakes on the active faults most greater, than that which the city has actually experi- critical for the Christchurch area. enced in its first 150 years. Hazards to Engineering Lifelines • 21

Smith & the earthquake shaking by modifying the ground accel- MM Berryman Elder et al Smith Dowrick eration, velocity and displacement at any frequency. In (1983) (1992) (1997) many areas of the city, the earthquake vibrations will VI 14 years 13 years 21 years 18 years be amplified. As a result, the overall average hazard for VII 48 years 35 years 70 years 85 years the city increases when compared to areas on bedrock VIII 160 years 120 years 250 years 650 years (for example most of Banks Peninsula), by approxi- IX 600 years 1200 years 980 years — mately 0 to 2 intensity units, or by 0 to 1 units when compared to areas on ‘average ground’ (comprising Table 2.1: Earthquake return periods for shallow sediments). different models Typical acceleration response spectra for Christchurch In 1869, a relatively small magnitude earthquake, are shown in Figure 2.2. This figure includes the estimated to have been around magnitude 6, had an response spectra at bedrock, which will be similar to epicentre close to the city on an offshore active fault of the response spectra on much of the Port Hills, and the which there is nothing known. The resulting shaking elastic response spectra from NZS4203:1984 for com- intensities have been estimated by Dibble Ansell and parison. Of particular note is the amplification at Berrill (1980) to have been MMVII to MMVIII. On longer periods. this basis alone, an earthquake of at least MMVIII seems a very reasonable scenario on which to plan. (a) Response Acceleration

1.6

Conclusion 5.0 percent damping Return period = 150 yr 1.4 Ground profile... The scale of lifelines projects requires much simplifi- A cation and the inevitable grappling with unknowns. It C 1.2 E is very hard to predict how a given lifeline service is going to behave in any earthquake, let alone predict 1.0 which services, or sections of services, will fail in an MMVIII event as opposed to a MMIX. 0.8

Work is planned over the next three years on the Alpine 0.6

NZS 4203:1992 Spectral Acceleration (g) Deep soil site Fault, and as part of this work the seismicity model and z = 0.8 0.4 μ = 1.0 predictions for the future shaking in the City will be revisited. This may lead to some further revision of 0.2 statistical estimates, but it will make no significant Bedrock spectra 0 difference to the need to plan for the effects of a 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 damaging earthquake on lifelines in the future. Period (sec)

The differences between the predictions of future shaking in Christchurch City are not important in the (b) Response Acceleration lifelines context. 1.6

5.0 percent damping Return period = 150 yr Current lifeline mitigation and planning strategies for 1.4 Ground profile... B earthquakes are simply not sensitive enough to make D subtle differences in the order of one intensity unit 1.2 F meaningful. 1.0

Earthquake effects 0.8 The main effects in Christchurch likely to be generated 0.6

by a major earthquake are discussed in the following NZS 4203:1992 Spectral Acceleration (g) Deep soil site z = 0.8 sections. 0.4 μ = 1.0

0.2 Earthquake shaking Bedrock spectra The ground shaking in Christchurch during an earth- 0 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 quake will be markedly affected by the deep, relatively Period (sec) soft sediment underlying the city and by variations in Figure 2.2: Site specific response grouped the soil profile within the sediments, particularly in the according to site stiffness (a) soft and (b) stiff top 30 m. This creates major changes in the nature of 22 • Risks and Realities

The variation in shaking across Christchurch is illus- • foundation failures as the liquefied soil loses its trated in Figures 2.3 and 2.4, which show the spectral shear strength and its ability to support foundation acceleration at 0.7 and 1.5 sec period respectively, at loads. points on a 0.5 km grid over the city. It can be seen that amplification effects are pronounced to the north of the Damage from liquefaction-induced lateral movement central city and in scattered south western areas. is usually much more extensive and serious than from any settlement and the magnitude of the movement is Figures 2.3 and 2.4 are illustrative only, and the soil much greater. Hence areas along river banks are profiles used at each grid node are frequently averaged particularly susceptible. In Christchurch this includes from sparse borehole data. The soil variation within most of the lower Avon and Heathcote Rivers. The the Christchurch area is shown by two representative Waimakariri upstream of the bridges has gravel banks soil type maps, Figure 2.5 (0 m to 2 m depth) and Figure and liquefaction is not expected to occur. 2.6 (2 m to 5 m depth). Figure 2.6 appears to show more uniform soil conditions with depth, but this is a reflec- The potential for liquefaction is so significant in Christ- tion of the smaller amount of borelog data available at church that John Berrill gave a specific lecture on the deeper depths and variability is not considered to subject at the Workshop which is summarised in change markedly with depth. If it is important to know Chapter 3. the response spectra for a particular structure, a specific study for that site must be carried out. Rockfill and slips on steeper hillsides For Christchurch, damage by landslide on a significant Liquefaction of loose sand and silt soils level is likely if the earthquake occurs in the two to four Christchurch, located near a saturated, sand- and silt- month period of mid-winter to early spring when soil rich coastline, is potentially at risk from widespread moisture levels are high enough to reduce the apparent liquefaction. This phenomenon occurs when the ten- cohesion of the loessial soils on the Port Hills. Local dency for loose granular materials to compact during sites may retain high moisture contents over much of earthquake shaking results in a pore water pressure the year, and therefore be susceptible over longer increase, and reduction or total loss in soil strength. periods. During drier conditions, damage is likely to be The pore water pressure increase occurs over a number confined to shallow soil falls from steep batters, rockfall of shaking cycles, and the extent of liquefaction is from bluffs and cliffs and rockfall from higher up the greater for earthquakes of longer duration. hillsides. The most vulnerable areas are generally at the foot of the steep slopes adjoining the valleys and Liquefaction-induced soil deformation can occur as: flood plain.

• Flow failure, where ground on even very gentle (Refer to Section 2.8 “Slope Hazard” for further infor- slopes moves laterally. In Christchurch this may mation). occur wherever lateral support to the soil is low, such as along river banks or the edges of the Hazard Zones estuary. The areal extent of the inferred shaking effects are • Ejection of sand onto the ground surface. shown on the seismic hazards map (see page 284 et seq), as three ground shaking zones and two liquefac- • Post liquefaction consolidation, with consequent tion zones. The boundaries to these zones are approxi- ground settlement. mate only, being based on sometimes very limited information (see section on “Map Limitations” on page • Large ground oscillations. 26. For any site of particular importance a specific site Liquefaction is known to reoccur numerous times at study is needed. the same site — liquefaction once does not secure a site against future liquefaction. Damage from liquefaction Ground shaking hazard zones is commonly seen as: Zone 1 • flotation of buried structures (e.g. manholes and large pipelines); Zone 1 is the least hazardous zone and is underlain by bedrock at shallow depths (generally less than 5 m) and • lateral spreading of ground on gentle slopes; shows very low to low amplification of seismic waves. Shaking intensity MM VII - VIII, and peak spectral • settlement of large areas due to consolidation and acceleration of 0.45 g at 0.2 sec period. liquefied soil being ejected through surface cracks; and Typical damage is loose brickwork dislodged, some

(continued on page 26) Hazards to Engineering Lifelines • 23

Peak Spectral Acceleration for Upper Bound Bedrock Spectrum as/g (T) at T = 0.7s 0.5 0.6 0.7 0.8 0.9

1.0

1.1

1.2

1.3

1.4

1.5

Transition Zone Not Modelled

Figure 2.3: Peak spectral accelerations at 0.7 second period

PeakPeak SpectralSpectral AccelerationAcceleration forfor UpperUpper BoundBound BedrockBedrock SpectrumSpectrum aass/g/g (T)(T) atat TT == 0.7s0.7s 0.50.5 0.60.6 0.70.7 0.80.8 0.90.9

1.01.0

1.11.1

1.21.2

1.31.3

1.41.4

1.51.5

TransitionTransition ZoneZone NotNot ModelledModelled

Figure 2.4: Peak spectral accelerations at 1.5 second period

24 • Risks and Realities

P M

Sumner

X X

F P

P X

F X P

F M

X F

M X

X F

M M

F X F M

F P

M M

X C

F F

M X

P 012

M F

M M M

M P

C M

M F

M M

C P

M M

M F

M M

F F

F M

C F

P C

X M

M P

X M

M X

M M

M F

X X

X F

F X

M C

F C

C F

X X

M F C

C Predominantly GRAVEL - C Predominantly GRAVEL gravelly sand to gravel M Predominantly SAND - silty sand to sand. May be interbedded with some silt or rare gravel F Predominantly fine grained soils. CLAYEY SILTS to SILTS soils. CLAYEY SANDY SILTS. May be SILTS. SANDY interbedded with some sand or organics P PEAT or highly organic fine PEAT P grained soils X FILL

C F

Figure 2.5: Representative soil type map of Christchurch, 0 - 2 m depth

Hazards to Engineering Lifelines • 25

Sumner

F F

P M

M P X

M F

012

F C

C M P

F C

M F

C P

P F M

C F

C C

Figure 2.6: Representative soil type map of Christchurch, 2 - 5 m depth 26 • Risks and Realities

chimneys brought down, stone walls cracked, some Liquefaction zones brick veneers damaged, small slips and ground cracking on steep ground. There is likely to be marked Zone A — high susceptibility variation within the zone due to topographic shielding Underlain predominantly by sands between 2 m and and focusing effects. 10 m depth. Generally water table within 1 m to 1.5 m of the ground surface. Insitu densities unknown. Zone 2 Twenty percent to 30% of zone A could liquefy. Zone 2 is underlain with sediments less than 50 m deep. Lateral spread of some river banks is likely. Ground It represents a transition zone between the low ampli- cracking, ejection of sand and water, and settlement of fication areas of zone 1 and the high amplification the ground surface will occur in the areas most affected areas of zone 3. Shaking intensity MM VIII. Peak by liquefaction. The pattern of liquefaction within spectral accelerations are likely to be intermediate zone A is unknown, but will reflect depositional envi- between those of zone 1 and zone 3. (There has been ronments, with previous estuarine areas as the most no study of any sites in this zone). affected. Damage could be intermediate between zones 1 and 3. However, the 1994 Northridge earthquake produced Zone B — moderate susceptibility some concentrations of damage with severe shaking on Underlain predominantly by silts and sandy silts be- thinning wedges of alluvium at the edge of basins. tween 2 m and 5 m or deeper depth. Water table between 1 m and 2 m of the ground surface (10% to Zone 2A has the same shaking hazard as zone 2 but is also underlain by soils highly susceptible to liquefac- 15% of zone B could liquefy). tion, whereas zone 2B has the same shaking hazard as Liquefaction effects will generally be less pronounced zone 2 but is also underlain by soils moderately suscep- than for zone A. Lateral spread of river banks is likely tible to liquefaction. to be limited in extent. The comments re pattern of liquefaction within zone A also apply to zone B. Zone 3 Zone 3 is underlain by more than 50 m and up to 800 Map Limitations m depth of sediments. High to very high amplification of earthquake ground motion, relative to bedrock. Mapping accuracy Shaking intensity MM VIII - IX, peak spectral accel- The hazard map zone boundaries have been deter- eration 0.6 - 0.7 g at 0.2 - 1.0 sec period. mined from geological maps of the area. These in turn are derived largely from borelog information from Typical damage includes: chimneys, monuments, tow- about 15,000 sites in the Christchurch area. Of these, ers and elevated towers may be twisted or brought down. the majority are shallow (less than 3 m deep), and are Panel walls can be thrown out of frame structures and often concentrated in particular areas (e.g. the central some brick veneers damaged. Frame houses not secured city). The boundaries between soil types is therefore to the foundations may move. Cracks may appear in wet frequently ill-defined, particularly with increasing ground, some ground deformation possible. depth. In places the boundaries can be moved two or Within zone 3 there will be areas of pronounced three hundred metres to either side, and still fit the amplification of earthquake ground motion, with shak- available data. ing intensity IX, and peak spectral acceleration of at The accuracy of the hazard zone boundaries is even least 1 - 1.2 g at 0.2 - 0.4 sec period. less than for the soil type boundaries as these relate to Typical damage includes: masonry buildings built the complex 3-dimensional changes in soil types be- before 1935 likely to be seriously damaged with some neath the city. For any critical structure, a specific destroyed. Frame structures racked and distorted, site study is needed to determine the actual degree brick veneers fall, frame houses not secured to the of hazard. foundations shift off, damage to foundations xcommon. Cracking of the ground conspicuous, damage to roads Ground shaking zones and paths. The area of zone 1 — bedrock — is clearly defined by Zones 3A and 3B have the same shaking hazard as zone the toe of the Port Hills. While areas on the hills are 3, but are underlain by soils that are respectively overlain with loess deposits up to 20 m thick in some highly, and moderately susceptible to liquefaction (see gullies and valleys, when compared to the flat areas the below). shaking will be essentially that of bedrock. Amplifica- Hazards to Engineering Lifelines • 27

tion of ground motion due to topographic focusing largely conjecture as to the proportion of areas under- effects has not been considered. However these usu- lain with susceptible soil types that are also sufficiently ally localised effects can be pronounced. loose to liquefy. However, three-quarters of the 16 sites investigated by the University of Canterbury as Zone 2 is shown as a transition zone, somewhat arbi- part of this project showed some liquefaction hazard, trarily taken as the area from the toe of the hills to where typically within the 2 m to 3 m of soil immediately the sediment reaches 50 m depth. The 50 m depth below the water table. contour on bedrock is approximate only, and is based on only 18 well-logs between Halswell and Redcliffs As the zones are based only on documented soil type, (Brown and Weeber 1992). No study has been at- the actual distribution of liquefaction effects that are tempted to determine typical soil profiles, or amplifi- likely to occur in an earthquake is largely conjecture cation effects in this zone. with our present level of knowledge. The suggested 20% to 30% of the area in zone A, and the 10% to 15% Zone 3 essentially covers the whole of the flat area of area in zone B, is simply a subjective assessment. Christchurch. Amplification effects appear to be most sensitive to the soil profile within 30 m of the surface. Soil profiles are known to change quite abruptly within Damage Assessment this depth range (at least in the central city area where Possible damage for the earthquake scenario is sum- the most data is available). Hence two nearby, or even marised in a damage assessment chart for different adjacent sites can have quite different response spec- structures in the different hazard zones (Table 2.2). tra. Little information on damage ratios for structures other than buildings has been found, and the proportion of damage in the chart is necessarily somewhat subjec- Liquefaction tive. It is based on a similar chart used in the Welling- The liquefaction zones are determined by soil type: ton Lifelines Project, and information from the zone A (shown as zone 2A and 3A on the map to Edgecumbe Earthquake. distinguish the respective shaking zones) is the area dominantly underlain by sand between 2 m and 5 m References depth or deeper, plus areas around known liquefiable sites; zone B (shown as zones 2B and 3B on the map) Adams (1980). “Paleoseismicity of the Alpine Fault represents areas of predominantly silt to sandy silt Seismic Gap”, Geology (8), pp 72-76. between 2 m and 5 m or deeper. The zones have been Brown and Weeber (1992). Geology of the Christ- terminated to the west where the water table exceeds church Urban Area, Institute of Geological and 2 m in depth. The increased effective stress and Nuclear Sciences Ltd. confining effect of the soil above the water table will generally prevent any significant liquefaction in the Berrill, Davis, McCahon (1992). “Christchurch seis- predominantly silt soils further to the west. mic hazard pilot study”, Bulletin of the New Zealand National Society for Earthquake Engi- The zones are based solely on the soil type, and do not neering, Vol 26. show: Cowan (1989) An Evaluation of the Late Quaternary • areas of other soils (i.e. gravels, deep peat layers) Displacements and Seismic Hazard associated within the zones that are not liquefiable; with the Hope and Kakapo Faults, unpublished MSc thesis, University of Canterbury. • some areas, such as the higher dune area in Linwood, Bromley and Brighton, which have water tables at Dibble, Ansell and Berrill (1980). Report on a Study of some depth, and are unlikely to liquefy; and Seismic Risk for BP NZ Ltd Sites at Woolston and Lyttelton. • density variations. Elder, McCahon and Yetton (1991). The Earthquake The relative density of the soils is the other prerequisite Hazard in Christchurch, Research Report to for liquefiable soils. The insitu density of the soils will EQC, Soils and Foundations vary markedly because of the range of environments they were deposited in. In high energy environments, Robinson (1979) “Variation of energy release, rate of such as steep river beds or a beach exposed to the open occurrence and b-value of earthquakes in the sea, deposited sand is likely to be dense. In low energy main seismic region, New Zealand”, Physics of environments such as swamps, or estuary, the sands are the Earth & Planetary Interior, 18, 209-220. likely to be loose. There is very little information on insitu densities over most of the city and it remains Schwartz and Coppersmith (1986). “Seismic hazards: 28 • Risks and Realities

Liquefiable Sites Liquefiable Sites Zone 1 Zone 2Zone 2 Zone 3 Zone 3 (Zones 2A/2B) (Zones 3A/3B)

Shaking MM VII - VIII MM VIII MM VIII - IX 0.45g 0.6 - 0.7g Intensities Areas of MMIX 1.2g

Structures Well designed Well designed Most damaged and Well designed structures Severe damage structures minor structures minor unserviceable, serviceable. Many non likely unless specific damage. damage. unless specific seismically designed foundation design. Non ductile Non ductile foundation design structures damaged structures some structures damaged. and unserviceable. damage. In areas of higher shaking some well designed structures damaged and unserviceable.

Fixings designed for 20% failure likely 30% failure likely 40% failure likely 50% failure likely 70% failure likely seismic loads

Equipment not fixed 40% failure likely 70% failure likely 90% failure likely 90% failure likely 100% failure likely or fittings not designed for seismic loads

In Ground Pipework Welded steel OK Should be OK Damage possible at Damage possible at Likely damage at Polyethylene entry to structures entry to structures entry to structures and at junctions. and at junctions. and at junctions.

Moderately ductile No mains damage. Mains damage Mains damage likely. Mains damage likely. Mains damage likely. pipes concrete with 5% entries and possible. 10% 40% entries and 50% entries and 70% entries and rubber joints junctions fail. entries and junctions fail. junctions fail. junctions fail. Verical Steel and cast iron junctions fail. displacements with rubber joints possible.

Low strength or Mains damage Mains damage likely. Mains damage Mains damage Mains damage low ductility pipes possible. 10% 40% entries and occurs. 70% of probable. 60% occurs. 80% Earthenware with entries and junctions fail. entries and junctions entries and junctions entries and junctions rubber joints junctions fail. fail. fail. fail. Asbestos cement Cast iron with lead joints

Non-ductile pipes Mains damage Mains damage Mains damage Mains damage Mains damage Ceramic cement possible. 20% occurs occurs packed joints entries and Brick junctions fail.

Note: Pipe failure in zone 1 may lead to slope instability

Transport

Roading Rockfall and small Some damage to Damage to kerbs. Some damage to Extensive damage to slips on steep kerbs. Some Distortion and kerbs. Distortion kerbs. Sumps batters. distortion and cracking of seal. and cracking of seal. damaged. Extensive cracking of seal. distortion and cracking of seal.

Railway Rockfall and small Some distortion of Distortion of rail Some distortion of Distortion of rail slips on steep rail lines lines, some rail lines. lines. Spreading of batters, failure of spreading of embankments. steep fills. embankments. Bridges

Structure As for structures above

Abutments Minor slumping Some slumpingSlumping likely, Some slumping Slumping likely possible failures possible if failures if not piled not piled.

Table 2.2: Damage assessment chart Hazards to Engineering Lifelines • 29

new trends in analysis using geologic data”, In overwhelmed by gravel and water generally decreases Active Geophysics, National Academy Press. further down the fan. However, the existence of old channels will direct the course of future flooding if Smith and Berryman (1983). “Revised estimates of water enters them, so that distance down a fan is not earthquake hazard in New Zealand”. Bulletin always a protection against flooding or aggradation. NZNSEE, Vol 16, No.4, pp 259-272. The particular topography of a fan will determine Smith (1992). Estimates of Earthquake Hazard in New which areas of the fan are at greatest risk. Zealand, EQC, Wellington. This succinctly summarises the Waimakariri situation.

2.3 Waimakariri Flood Hazard Historical flooding of the Waimakariri River At the time of European settlement, flooding from the Background Waimakariri was a major problem. Between 1860 and In pre-European times the Waimakariri River flows 1870, the Provincial Government constructed various could discharge to the sea anywhere from Lake works on the south side of the river aimed at cutting off Ellesmere to the Ashley River. In flowing over this the southerly overflows. area the river would drop out material (silt, stones etc.) it had been carrying. It was this mechanism that led to In 1869, a river district was formed that directed the the classic flood fan formation of the Canterbury construction of a number of groynes to block off plains. overflows to the basins of the Styx, Avon, Heathcote and Halswell Rivers. Inevitably, those dwellers on the Streamland 73, a DSIR publication, had this to say on north side of the river pointed out the inequity of this the formation of fans: situation and eventually, after much political furore, A fan is a more-or-less symmetrical cone of shingle the Waimakariri River Trust was set up in 1928 with often formed where mountain streams issue from steep powers and responsibilities relating to both sides of the side-valleys to flow over more gently sloping surfaces. river. When the slope of a stream decreases and its channel The Trust built a flood protection system to regulate the becomes wider, the streams ability to carry shingle lower 40 kms of the river by an extensive stopbank decreases and the shingle is deposited. This build up of system, together with diversions and other works to deposit of shingle is called aggradation. Aggradation provide a shorter, straighter course to the sea. occurs fastest near the head of the fan (at the apex of the cone). In a fan left in its natural state, the depositing of This system was upgraded extensively between 1960 shingle in the main channel causes the stream every so and 1990. Apart from one serious breakout of often to switch from one course to another. The stream floodwaters in 1957 both schemes performed very either creates a new channel or reoccupies an old well. They failed, however, to prevent accumulation of channel where it has flowed in the past. Water then gravel in the lower reaches, a problem now largely flows down an entirely different part of the fan, where addressed by commercial extraction. it may not have flowed for decades or even centuries. This sudden switching occurs when the stream is in Floodplain Management Plan flood and is termed an avulsion. The above schemes were built to contain floods up to Aggradation on a fan is not a steady continuous proc- a specified size (the so-called “100 year flood”). Dur- ess. Most occurs during rare major storms when mas- ing the late 1980s and early 1990s, Canterbury Re- sive amounts of shingle are deposited in a short time, gional Council and the community again addressed the mostly at the fan head. Then follow often long periods question of the Waimakariri flood hazard. As a further when the stream cuts into these deposits at the fan head response to the hazard, the Canterbury Regional Coun- and redistributes the material over the lower fan. cil prepared a draft Waimakariri Floodplain Manage- ment Plan. Because of the geometry of fans, any point below the fan head is lower than the stream not far up the fan and A floodplain management plan is a document describ- so is at risk from flooding. This may be brief inundation ing how an area of flood-prone land should be used and of areas adjacent to the channel, while the stream is in managed to reduce its susceptibility to flooding for a flood. But the more serious risk is the formation of a prescribed period of time. The aim of the plan was to new main channel, with flooding and the depositing of reduce the impact of flooding on the community by shingle on a different part of the fan. The risk of being minimising flood damage, which has the potential to 30 • Risks and Realities

disrupt normal activity and cause damage to life, limb, storm runoff may coincide with extremely high water property, the environment as well as tangible and levels downstream in the Avon/Heathcote Estuary and intangible assets. Although the Canterbury Regional Brooklands Lagoon. Council has withdrawn the plan as a legal document, it was of considerable use in the Lifelines Project. Flood maps The object was to quantify the hazard so that decisions The following text provides an explanatory back- could sensibly be taken regarding possible responses ground to the 1:35,000 scale “Styx, Avon and Heathcote which are effective both in terms of cost and location. Rivers Floodplains” (Map 3) prepared for the Lifelines Study.

Flood maps The blue cross hatched areas indicated on the map for The many and varied investigations carried out over a the Styx, Avon and Heathcote Rivers indicates the area period of three years resulted in the preparation of expected to be inundated at the peak of a 0.2% Annual flood maps, along with accompanying tables express- Exceedence Probability (AEP) flood event on each of ing the flood hazard in terms of depths, velocities, the rivers coinciding with extreme high water levels durations, rates of rise and probabilities of occurrence. downstream in the Estuary and Lagoon. (A 0.2% AEP The above relates the background to Map 2 (p 285). event is the same as a 1 in 500 year return period event). This extreme event was chosen for floodplain manage- In effect, the study area would not be affected by river ment planning purposes and a reduced probability floods having a return period of less than 500 years. overall of 0.1 % AEP (1 in 1,000 year return period) Breakouts from lower return period floods would not was adopted because it is unlikely that extreme storm reach the city due to a combination of attenuation and runoff and extreme estuary water levels would occur soakage to the west. simultaenously.

Map 2 is a composite map of four independent breakouts. Mapping of areas inundated along the margins of the At the 500-year return period level, it is estimated there Styx and Heathcote Rivers has been based on compre- would be only one break out. That is at Halkett, behind hensive flood studies involving the development of McLeans Island, through the Crossbank or to the north complex hydrological models of the catchments and at Baynons. There is a 50% chance that the existing hydraulic models of the river channels. Water depth, protection system would cope adequately with such a duration of flooding, velocity and frequency can be flood. The percentage chance of failure from the four provided at various locations along the river channels potential failure zones are indicated on the arrows and floodplains. A similar flood study has only just emanating from the potential failure points. Risk can begun on the Avon River and its catchments. The be obtained by the product of the chance of the flood inundated area indicated must be treated as preliminary occurring (1:500) and the percentage chance of failure. only and the other parameters are not available.

Average depths and velocities are shown on the map In each case the extreme event indicated on the map for various locations. The design hydrographs are of represents a river flood resulting from prolonged rain- the order of 24 hours through the city. fall within the catchment coinciding with an extreme high tide level at the river mouth. This situation repli- cates what occurred during the August 1992 snow 2.4 Local Flooding Hazard storm, but on a more extreme scale. Upstream catchment development in line with projects made for the Background new City Plan over the next 20 years have been The Christchurch urban area is drained by three rivers assumed. Existing channel conditions have been used — the Avon, Heathcote and Styx Rivers — whose in the models. catchments are entirely within the administrative bound- The purpose of the flood studies is to facilitate floodplain ary of Christchurch City. management planning jointly with the Canterbury Urban development has occurred on the rivers’ Regional Council. If the pattern of development within floodplains since the beginning of European settle- the catchments is substantially different to that as- ment. It is, therefore, not surprising that flood damage sumed or if channel conditions or storage characteris- resulting from storm runoff from within each of the tics are significantly altered in the implementation of three catchments has occurred on a number of occa- floodplain management measures, then the inundated sions. The most serious flood hazard occurs on the areas may change significantly and new maps will lower tidal reaches of each of the three rivers when need to be drawn. Hazards to Engineering Lifelines • 31

Styx River Flooding Floodwater depths of between 1 m to 1.5 m can be Discharge of the Styx River to Brooklands Lagoon is expected at various other locations on Eastern Terrace. controlled by a stopbank and tide gate located down- Elsewhere the flooding is shallower. stream of Kainga Road to prevent inundation of A number of bridges between Sparks Road and Opawa Brooklands by extreme high tides. Road are expected to be closed for a period of five to In the extreme flood event modelled, flood runoff from thirteen hours due to the depth of floodwater across the the catchment builds up behind the tide gates closed by approaches. Shallow saltwater flooding is expected in a higher downstream tide level. Approximately 360 the Woolston industrial area. Historically, blockage of hectares of rural land generally to the west of Lower the Sparks Road culvert by a barbecue table and partial Styx Road is inundated to a shallow depth and 182 blockage of a bridge waterway by a motorcar jammed residential properties in Brooklands are also inundated under the bridge deck have occurred. Unpredictable with floodwater over the floorboards of 48 houses. The events such as these are likely to occur and accentuate maximum depth of house flooding is 400 mm. The flooding problems and damage during an extreme depth of flooding on private properties and Lower Styx flood. Road is generally less than 1 metre.

Avon River Flooding 2.5 Tsunami Hazard Residential areas along the margins of the Lower Avon River are protected from flooding up to a level of Tsunami event approximately RL 10.9 m by semi-continuous Based on historical records, the most likely generating stopbanking from Porritt Park to Bridge Street. During source for a significant tsunami affecting Christchurch the August 1992 storm the Avon River overtopped the is from a large seismic event centred on coastal South stopbanks at a number of relatively low points along America (such as those that occurred in 1868 and their length. It has been assumed that widespread 1960). The tsunami event adopted for this study is from overtopping of the stopbank system could occur during this source. a more extreme event. The area indicated on Map 2 is land below a level of approximately RL 11.0 m. Under Magnitude present conditions inundation of this entire area could The magnitude of the tsunami adopted involved a total not occur simultaneously unless widespread failure of water level variation at the open coast of 10 m inclusive the stopbanks occurred during the extreme storm. of tide (i.e. 5 m above and below MSL). It was assumed It is likely that the detailed flood study underway will that these maximum water levels could enter Lyttelton indicate that a significantly smaller area is at risk of harbour but the level would be reduced to 3 m at the inundation during an extreme event. entrance to the Avon-Heathcote estuary and the Waimakariri River due to dissipation of energy in the limited water depths on the ebb tide deltas. Heathcote River Flooding The recently constructed Woolston Tidal Barrage is At the Sumner Esplanade from Scarborough to Cave closed under normal river flow conditions. In the event Rock, the tsunami height is assumed to be reduced to of a flood the radial gates will open to pass floodwaters 4 m due to shoaling in the shallow water at this location. through the Woolston Cut to the Estuary. In the For a worst case scenario, the initial tsunami wave extreme flood modelled, 192 existing house floors coincided with a high spring tide when water levels in between Lincoln Road and Radley Road are expected the estuary and Brooklands lagoon would be at 1 m to be inundated. Depth of floodwater exceeds 1 m in above MSL, hence maximum water levels at both of two cases and 500 mm in 44 cases. these locations would be 4 m above MSL.

The maximum main channel velocity is 1.4 m/s be- A tsunami wave period of three hours was assumed, tween Barrington and Colombo Streets. At local con- with minimum time from peak to trough of one hour for strictions such as bridge approaches maximum veloci- the first wave. Due to a falling tide, the second wave ties could be somewhat higher than this, but the mean would be 1 m lower than the first, the third wave velocity of flood flows over roads and private lots coinciding with low tide being 2 m lower than the first. adjacent to the river will be much lower than 1.4 m/s. Sea conditions at the time of a tsunami were assumed The maximum depth of inundation on the floodplain is to be normal with swell heights of 1 m, and a swell 1.8 m on Eastern Terrace near Birdwood Avenue. wave period of 8 seconds. 32 • Risks and Realities

Return period the south and west of Brooklands Lagoon which is It is difficult to determine the return period for the below the 2 m contour. This covers a total land area of adopted event as the majority of the historical informa- approximately 965 hectares, including the Spencer tion is anecdotal, and does not provide an accurate Park camping ground and sewage pond, Brooklands assessment of magnitude. Data from de Lange and Village, the Lower Styx Road, and the area west of the Healy (1986) shows that Canterbury has received nine Styx River to Chaneys Plantation. Water depths over tsunami in 142 years (1840 to 1982), and National Civil these land areas will be up to 0.75 m deep. Defence data shows that there has not been any tsunami reported since this time. Since the adopted tsunami is Bottle Lake Forest to South Brighton Spit larger than the biggest recorded event, a minimum At the Bottle Lake Forest and Waimairi Landfill return period of 150 years can be assumed. The actual areas the isolated nature of blowouts and small vol- return period may in fact be much longer. It would be umes of water overtopping (24,500 m3), should restrict of considerable benefit to more accurately establish the inundation to narrow track areas behind the foredune return period of such a tsunami in order to determine and be largely insignificant. With the majority of water what level of planning response is appropriate to deal entry points being along Bottle Lake Forest, the with its potential effects. Waimairi Landfill should not be affected by inundation. Tsunami effects At Waimairi Beach (Rothesay Rd to Pacific St), From the adopted tsunami magnitude, wave breaker overtopping will occur at six locations involving a and runup parameters were calculated for the Christ- water volume of 65,300 m3. The majority of this church beaches. It was found that at maximum tsunami volume would enter via the blowout located at the water levels, runup heights will exceed the 8 m RL outfall pipe north of Larnach St. Along this section this contour with large-scale overtopping occurring at sea- coast inundation should be limited to a narrow (200 m walls and dune blowouts resulting in widespread inun- to 250 m wide) area of approximately 28 hectares dation. Theoretical water flows into the Avon-Heathcote behind the dunes which lies between the 3 m and 4 m estuary and the Waimakariri River mouth were also contour and is bounded on the landward side by a ridge calculated along with inundation areas around the of higher older dunes. Inundation depths are calculated margins of the estuary and Brooklands Lagoon. to be approximately 0.2 m. A small amount of water The main effects possible in the Christchurch area are may also travel down Pacific St, but the inundation summarised below. areas and depths should be minimal. At North Brighton, overtopping will occur at three Waimakariri River and Brooklands Lagoon major locations (the surf club wall and dune lowering For an initial inflow duration of 30 minutes, approxi- at Effingham and Cygnet Streets), involving a water mately 8.91 x 106 m3 of water will flow in through the volume of 366,600 m3. The resulting inundation will Waimakariri River mouth with maximum velocities in cover approximately 135 hectares with general inun- the range of 7.5 m/s. It is estimated (Boyle pers comm.) dation depths being in the order of 0.3 m. Since the that 80% of this volume will travel up the main river most significant overtopping will occur at the sea wall channel and 20% will flow into Brooklands Lagoon. around the North Brighton Surf club, the major flow The southern stopbank on the lower 1.8 km of the path will be west down Bowhill Road and south west Waimakariri River will be overtopped resulting in across the Rawhiti Domain. approximately 3.42 x 106 m 3 of water spilling onto land to the south of the stopbanks. Some from the remaining Significant overtopping will occur along the New river flow will travel up the Kaiapoi River and some Brighton foreshore, notably at the sea walls, involving 6 3 into the Pines Beach area. a total water volume of 1.202 x 10 m . This will flow overland to the Avon River spreading out to inundate Along Brooklands Spit, dune overtopping will occur at the total area from Lonsdale Street to Rodney Street nine locations, including a 200 m long section at the (63 hectares). Taking into account the effect of build- site of the old Waimakariri River mouth, resulting in an ings, inundation depth will be 1.2 m and flow velocities additional 911,150 m3 of water entering the lagoon. A approximately 1 m/s hence it will take 10 minutes for further 14,000 m3 will be added from overtopping at the water to travel the 600 m to the river. seven locations around Spencer Park. Along the South Brighton coast (Rodney St to Cas- The combined volume of inundation water into the pian St) there are nine locations which will be Brooklands Lagoon area from all sources will be in the overtopped involving a total water volume of 141,900 order of 6.13 x 106 m3 which will affect all the land to m3. Due to the large number of entry points, it is Hazards to Engineering Lifelines • 33

assumed that the water flow will spread out to inundate peak due to the river system not having time to fully the total area of 193 hectares between the coast and the drain between successive tsunami peaks. Avon River. General inundation depths will be in the order of 0.45 m, however at Jellicoe Rd local depths Water velocities leaving the estuary have been calcu- could be up to 0.7 m due to higher discharges at this lated to be up to 7m/s, which is approximately seven location. Flow velocities will be approximately 0.13 times faster that normal tidal flows out of the estuary. m/s, hence will take 75 minutes for the water to travel This very rapid outflow could result in scour of sea the 600 m to the river. walls along the Moncks Bay and Redcliffs foreshore leading to possible wall failure with associated col- For South Brighton Spit, there are six locations be- lapse of the road network and services. Scouring could tween Caspian Street and Tern Street which will be also occur at the bridge abutments at Ferrymead and overtopped and all of the dune system south of Tern Bridge streets. The estuary will be “dry” (below low Street. The water volume involved will be approxi- spring tide level) for approximately 45 minutes be- mately 416,500 m3 spreading out to inundate the total tween the first and second tsunami waves. 81 hectares of the spit. Inundation depths will be in the order of 0.55 m, however around Heron Street local Considerably more detailed investigations and model- depths could be up to 1.0 m due to higher discharges at ling is required on tsunami effects in estuary and river this location. Flow velocities will be approximately systems, but whether this is warrented or not depends 0.33 m/s, hence will take 12 minutes for the water to on a more accurate assessment of the return period of travel the 250 m to the estuary. the tsunami scenario.

Avon-Heathcote Estuary Clifton and Sumner For an initial inflow duration of 30 minutes, approxi- The Sumner beach in front of Clifton Surf Club and mately 13.836 x 106 m3 of water will flow through the Cave Rock is very prone to tsunami inundation having estuary mouth with maximum velocities in the range of little dune protection and maximum beach heights only 8 m/s. With the addition of water entering the estuary 3 m RL. For this 200 m of beach, overtopping volumes 3 overland from the South Brighton Spit, the total estu- may be as high as 459,000 m all of which would flow ary volume added by tsunami is approximately 14.084 southeast into Sumner. The Sumner sea wall at a height x 106 m3. of 5 m RL affords some protection to the low lying land behind, but will still be overtopped at maximum tsu- Within the initial inflow duration of 30 minutes, the nami levels. The total water volume entering the Sumner tsunami will raise water levels right round the margins area will be approximately 476,000 m3, which will of the estuary. The following inundation could occur: cover the 70 hectares of land below the 3 m contour to an average depth of 0.7 m. • Moncks Bay to McCormacks Bay: 86.8 hectares, with maximum depths up to 1.8 m at Moncks Bay The road from Shag Rock to Clifton Surf club will and 1.2 m over the causeway at McCormacks Bay. also suffer from inundation, receiving 22,500 m3 of water across its 400 m length in the 15 minutes of • Ferrymead to Bromley: 210 hectares, at depths of maximum tsunami elevation. 0.3 m.

• Brighton spit and South Brighton: 190 hectares, at Lyttelton depths of 1.0 m and maximum depths of up to 2.0 Due to the water depths in Lyttelton Harbour, it is m at Tern Street. (Note that these areas are also assumed that the tsunami will be able to maintain its affected by beach overtopping). open coast height without breaking as it travels up the inlet. Hence, at the port, the maximum tsunami height There may also be some minor inundation at Bexley of 5 m RL will be retained and all of the port area below and Woolston. this contour will be affected by inundation. This equates A worst case scenario for tsunami bore travel distances to 80 hectares, which effectively covers the total port and inundation up the Heathcote and Avon River area with water depths between 1 m and 2 m for up to channels prepared by Barnett Consultants (1994) sug- 30 minutes. Port facilities affected include wharves, gests that some flooding may occur as far upstream as railway, marinas, graving dock and the bulk coal Fitzgerald Avenue on the Avon River and as far up- facilities. Although there will be water through the tank stream as Opawa Road on the Heathcote River. The farm, the embankments around the individual tanks are report also suggests that peak water levels in the rivers approximately 0.2 m to 0.3 m above the predicted may not occur until up to 10 hours after the tsunami water level, hence should be safe from inundation. 34 • Risks and Realities

There is also likely to be inundation of many low-lying Christchurch Star. Various Newspaper Articles On areas adjacent to pocket beaches within the harbour, The May 1960 Tsunami. and damage to jetties and wharves. One particularly badly affected location will be Teddington, where de Lange, W P and T R Healy (1986). “New Zealand despite a reduction in tsunami height due to shoaling in Tsunamis 1840-1982”. N.Z. J. Geol & Geophvsics :29, shallow water, inundation is likely to cover the road to p 15-134. Diamond Harbour and enter buildings. Heath, R.A (1976). “Response of several New Zealand There should not be any inundation of Lyttelton town- harbours to the 1960 Chilean tsunami”, Bulletin Royal ship. Maximum water level will be at least 1.2 m below Society of N.Z.: 15 p71-82. the top of the brick wall along the side of Norwich Lyttelton Port Company. Tidal Levels For May 1960 Quay. and August 1960 Tsunamis.

Hazard zones The areal effects of the predicted inundation from the 2.6 Extreme Wind Storm first tsunami wave are shown on the tsunami hazard Scenario Map 4. The zones are categorised depending on the source of inundation. There has been no attempt to categorise areas on depth of inundation. Introduction Severe wind storms in Canterbury are usually associated with northwesterly winds that have been acceler- Map limitations ated by the approach of a cold front from the south. Accuracy When a third factor, such as the development of a low pressure system within the frontal airmass, is also It should be emphasised that due to a lack of data and present, extremely severe winds are experienced, such information on a tsunami of this size, many assump- as occurred during the wind storm on 1 August 1975. tions have had to be made in calculating the above effects on beaches and structures, as well as the areas A second source of extreme winds is the passage of a and depths of inundation. The results of this study tropical cyclone such as Giselle (The Wahine Storm) in should be considered as being only indicative of the April 1968. When ocean temperatures around New order of magnitude of effects from a tsunami of this Zealand are much higher than average and a tropical size. There is scope for considerable review, and re- cyclone enters the New Zealand region, it may not die finement of the methods used and results obtained in but can be maintained and even grow because of local this study. energy and topographic factors. Giselle’s centre pressure dropped from 970 mb east of Cape Reinga to 965 Second wave effects mb over Napier. It then travelled down the east coast about 150 km offshore, bringing storm force SE winds The hazard maps only show areas predicted to be to Cook Strait, Kaikoura and Canterbury. All the time affected by inundation in the initial tsunami wave. it was expected that the cyclone would turn eastwards However, it is likely that the second wave may cause and move away from New Zealand. Instead, just SE of more flooding, although being 1 m lower in amplitude, Banks Peninsula, it made a right angle turn and moved due to increased blowout size, damage to sea walls and directly towards the Peninsula. This brought very larger river mouth areas as a result of the first wave. damaging SE winds to Christchurch and mid-Canter- The extent of inundation from this second wave has not bury, along with heavy rains and surface flooding. been calculated or mapped. A 1-in-150 year scenario for this situation would References involve 80 knot to 85 knot (41 m/s to 44 m/s) wind gusts, sustained over several hours, accompanied by a Barnett Consultants (1994). Christchurch Tsunami 1-in-50 to 1-in-100 year local flooding event for the Study Draft Report. Prepared For Christchurch City Avon and Heathcote catchments, accentuated by the Council. strong onshore SE winds and surges. Cerc (1984). Shore Protection Manual, US Army The most extreme wind events on their own, with a Corps Of Engineers. return period of about 150 years, are likely to be Christchurch Press. Various Newspaper Articles On produced by pre-frontal northwesterly winds with nar- The May 1960 Tsunami. rowing gaps between the isobars as the cold front moves northeastwards across the South Island towards Hazards to Engineering Lifelines • 35

a near stationary anticyclone over or to the east of the ing inland. Wednesday widespread nor’westers North Island. In addition there is the localised develop- with a late southerly change with showers. ment of a low pressure system in the frontal airmass Thursday, showers clearing and remaining cool. along the east coast of the South Island, as in the 1st Friday fine with increasing temperatures. August 1975 case. On Tuesday at 7 am the MSNZ forecast is: Hence, the best information on which to base a scenario of extreme winds over Canterbury is the 1st August Fine with afternoon nor’easters, nor’westers 1975 storm whose winds caused widespread damage inland reaching the coast at times. Tomorrow a with a maximum gust of 93 knots at Christchurch late southwesterly change with rain. Some thun- Airport. derstorms are likely with the arrival of the cold front.

The 150 year return period wind storm On Wednesday 7 am MSNZ forecast is: The 150 year return period wind storm in Canterbury Nor’westerly winds, gale force in the moun- is expected to be produced by the combination of at tains and in inland places, strong and gusty least three high wind factors occurring simultaneously: elsewhere. Overnight thunderstorms with cold • a strong pressure gradient in advance of a rapidly southwesterly winds and showers tomorrow. moving cold front from the south to south-south- Clearing later in the day but remaining cold. west; The actual winds during Wednesday were strong • a strong, near stationary anticyclone over or to the nor’easterlies in North Canterbury and Christchurch east of the North Island; and with strong and gusty winds in Christchurch and over most of the plains from Amuri to Timaru from late • the formation of a depression on the advancing cold afternoon. The anticyclone over the North Island slowed front which further accelerates the pre-frontal wind, to a stop and the cold front moved onto the southwest probably in association with the influence of Banks of the South Island, with the wind flowing parallel to Peninsula and the lee trough. the mountain ranges. The pressure gradient over the South Island began to increase with the northward The wind will be extremely gusty, as the extreme wind progression of the front, which intensified. sector is in the unstable nor’west air. Gusts will mainly be in the form of horizontal rolls of air moving at a At 6 pm Wednesday MSNZ issued a special weather mean speed of 70 knots (36 m/s), with the forward roll warning of gale to strong gale force winds along the taking the gust speed up to 103 knots to 110 knots (53 east coast of the South Island and strong gale force m/s to 56.6 m/s), with lulls of around 50 knots (25.7 m/ winds in the mountains. The AA issues a warning that s). These values apply to the open plains. In localised cars with caravans and camper vans are not allowed on places, speedup over hills or under the troughs of long the Lewis Pass, Arthurs Pass and Burkes Pass Roads wavelength lee waves will accelerate the wind by 15% overnight. It is carried on local radio at 6.30 pm and to 30%. each hour thereafter, on Teletext from 7 pm onwards, and is on CTV in the 7 pm News programme. TV1 and The extreme wind band will progress northwards in TV3 carry gale warnings for Canterbury on their early advance of the cold front, reaching a peak north of and late evening weather slots. Christchurch and reducing in intensity significantly as it moves north of Kaikoura. Overnight many people notice the warm evening and mild late night conditions. Early in the evening the The scenario wind speed drops, except in a band immediately to the It is not possible to produce a hazard map in the same lee of the mountains where the primary lee wave way as for the earthquake or flood hazards and there- produces a stiff nor’west wind with moderate gusti- fore a scenario approach has been adopted. ness. With the hot overnight temperatures the people of Canterbury generally find it hard to sleep. Dunedin is battered by very strong winds at about 3 am. Several The buildup trees are brought down taking power out over 25% of At noon on Monday a long range forecast for Canter- the city. bury is issued by Met. Service New Zealand (MSNZ):

Dry and settled weather with mild temperatures Thursday today. Tuesday, northeasterlies, partly cloudy The duty weather forecaster in MSNZ upgrades the with mild temperatures, nor’westers develop- gale warning to a storm force wind, issuing a new 36 • Risks and Realities

special warning at 4 am to the Civil Defence HQ, the hesitated until after 5 am when winds at Timaru had media, Canterbury Regional Council and District and risen to 35 knots, gusting to 50 knots. The mayor told City Councils, but no one is on duty except the radio him to call again if things got worse because the MSNZ station announcers who include the storm warning in often issued gale warnings which didn’t “turn out too weather bulletins from 5 am onwards, and the Duty bad”. Then he went back to bed and tried to sleep — but Officer at Civil Defence HQ, Wellington. the warmth, the noise of the wind and the thought of possible damage kept him awake. The Timaru District Civil Defence Officer was called by Wellington Civil Defence HQ at 4.15 am, making The Regional Civil Defence Officer was called at his his way to the District Council by 4.35 am. He imme- home by Civil Defence HQ at 4.17 am, and informed diately called the Duty Forecaster at MSNZ to confirm of the storm warnings and the power outage and the approach of the band of storm force winds. As he possible damage to Dunedin. He arrived at the CRC at drove to the Council he had to fight stronger wind gusts 4.35 am and immediately checked the fax and read the than he could remember for some time and thought that storm warning. He called Dunedin at 4.45 am and they were in for a good blow. Having talked to the Timaru at 4.55 am, confirming the power outage in forecaster, he thought he had better call the Mayor but parts of Dunedin and the rising winds in Timaru.

1020

1015

1010

1005

1000

995

990 L

985

980

0600 NZST

Figure 2.3: Synoptic chart for 0600 NZST 1 August 1975

Hazards to Engineering Lifelines • 37

170˚

160˚

H 1020

1010

1200 hrs NZST July 31 1975

1000

170˚ W

990

980

180˚

970

1010

1020

170˚ E

970

160˚ 170˚ 180˚

150˚

160˚

140˚

980

990

130˚

1000

150˚

1010

1020

140˚ 130˚

Figure 2.4: Synoptic chart for 1200 NZST 1 August 1975 38 • Risks and Realities

At 6.17 am, 6.25 am, 6.34 am and 6.37 am, accidents 16

360 360 occur on Highway 1 between Timaru and Ashburton as 90 270 180 180 360 90 15 trucks or cars are blown off the road. In one case a car collided with a truck and the driver of the car is killed

14 while the truck driver and two of the other drivers are badly injured.

N N 90 0 Just before 7 am the storm force winds hit Ashburton. 10 E W 80 KNOTS S S 70 W E 60 50 40 Strong nor’westers have been blowing most of the 12 night, with some old shelter trees being felled between 4 am and 7 am when the wind over a period of 30 11 minutes from 6.35 to 7.05 rose steadily to average 63

10 knots, gusting to 82 knots. Damage was occurring, 360 360 particularly in the vicinity of Highway 1 and in another 90 270 180 180 360 90 9 band about half way between Highway 1 and the

mountains because of the long wavelength lee waves CHRISTCHURCH

8 which had formed.

Around the band along Highway 1, dozens of grain 7

N N 0 90 10 silos and hay barns were completely demolished and E W 80 S S N E 70 hay was being blown with damaging force. Sheets of 6 aluminium roofing and corrugated iron from roofs and

5 sheds were being blown about. A power board lines- man was killed by flying corrugated iron as he at-

4 tempted to restore power.

360 360

90 270 1 AUGUST 180 180 270 360 90 3 Effects in Christchurch Between 7 am and 7.30 am, the people of Christchurch 2 began their day with very strong, gusty nor’westerly

1 winds and unseasonably warm temperatures. Many Wind direction Wind speed people were irritable because of their difficulty in Figure 2.5: Anemometer charts, Christchurch sleeping overnight. Almost all became aware of the Airport. Speed scale in knots (1 m/s = 2 knots) damage that had been caused in Dunedin and Timaru as they listened to the 7 am or 7.30 am radio news bulletins, which also carried storm force wind warn- The effects south of Christchurch ings for Christchurch. Most people did not appreciate At 6 am, the band of high winds immediately ahead of the gravity of the situation. Having lived through many the cold front reach Timaru, with mean speeds of 50 gale force nor’westers in their adult life without any knots, gusting to 75 knots, bringing down many trees harm or damage, most early starting workers (7 am to and taking out all of the power to the city, except the 7.30 am) were on their way to, or were at, work. hospital, radio station, and Timaru District Council, which go onto emergency generators, and taking down Between 7 am and 7.30 am the strength of the band of several key overhead telephone lines. The Telecom wind had increased even more as a low developed on microwave dish for the toll network is damaged and the cold front. This low deepened as it lifted over Banks blown out of line. Peninsula. The advancing destructive wind band, demolished most of the pine plantations on either side of The mayor of Timaru is talking to the Civil Defence Highway 1 from the Rakaia northwards. The strong Officer when the phone goes dead. He quickly goes nor’westers were also associated with extremely high through to his car in the garage and uses his car phone rainfalls in the Alps, with the mountain-fed rivers to re-establish contact. A fallen tree over his drive rising very rapidly. prevents him from driving to the Council. At 6.15 am the Timaru CDO calls the regional CDO on radio to After 7.30 am large trees began to topple in many parts report the wind strength, power and phone outages and of Christchurch, damaging dozens of houses and cars, house fires. Three houses are on fire where trees have and blocking or obstructing many roads, as the mean taken power lines down and the burning tree has wind speeds rose steadily from 40 knots, gusting to 65 ignited the houses. knots at 7.30 am, to 55 knots, gusting to 78 knots at 7.45 Hazards to Engineering Lifelines • 39

am. Many houses lost tiles or iron off their roofs during Kaiapoi received similar damage to Christchurch, but this time, with a great deal of other wind blown debris. with fewer casualties because of the broadcasts of At 7.38 am the first radio report said to definitely stay problems in Christchurch. Rangiora, being further home and stay inside. inland, received minor damage. Coastal properties and settlements, and inland places such as Hawarden and Dozens of people have been injured trying to secure Culverden experienced damaging winds. Kaikoura flapping roof sheets, three people had been killed by lost several vessels and three roofs, but no one was flying debris, 10 people were killed by trees falling on injured. Two motorists were killed by being blown off their cars, with 35 people being seriously injured. the road in the Hundalees. Off the coast, a crewman Ambulances couldn’t reach a serious accident at 7.45 was badly injured on a fishing boat, which couldn’t am because all approach roads were blocked with trees attempt to come into the harbour until after the storm and live power lines, and so four people died. Eight had passed. people were killed by electrocution as live wires fell on vehicles, and 12 people were burnt to death from house Extensive power loss due to fallen trees was experi- fires. enced throughout North Canterbury. Wind throw in the Eyrewell, Ashley, Balmoral and Hamner forests was By 8 am the wind speed over Christchurch has risen to widespread. 70 knots, gusting to 103 knots. Several aero-club planes were badly damaged. A ship was blown away from its mooring in Lyttelton Harbour and collided 2.7 Snowstorm Scenario with a tanker at the oil wharf, rupturing a 20,000 litre oil tank which began leaking into the harbour. Dozens Introduction of moored yachts and pleasure craft were blown off their moorings and driven onto rocks and beaches Because snowstorms are relatively rare events at low around Lyttelton and Akaroa harbours. Windows were elevations in New Zealand, snowfall magnitudes are blown in on several commercial buildings throughout not routinely recorded at our meteorological stations. the city. The power lines for the new tram were brought The only records available relate to the number of days on which snowfall occurred. In the Christchurch area down. Several homes along the Port Hills spurs lost this averages between two and three per year for the last their roofs. 100 years. At 8.20 am the wind speed had dropped a little and a Even at higher elevations the number and continuity of Radio news reporter spoke to the Forecast Manager at stations at which snow depth observations are recorded the Airport weather office who confirmed that the peak is limited (Fitzharris et al. 1992). There are, however, of the storm was passed but that the wind were still some case studies of conditions leading to heavy snow- gusting to over 80 knots. At 8.30 am, the radio station fall on the foothills and plains of the South Island and reported that the peak of the storm had passed. Many useful information can be obtained from these. With people ventured outside at hearing this news and sev- some exceptions (e.g. Harris 1981), relatively little has eral were injured and two people were killed by flying been recorded about the impacts of heavy snowstorms, corrugated iron. By 9 am the winds had turned to the though there may be information on file with local southwest, and dropped to 20 knots. The temperature authorities, power boards etc., which detail some of had dropped by 10˚C. The morning sun shone though these effects. the partly cloudy sky as people carefully moved outside to survey the extensive city-wide damage. Hardly By far the two largest snowstorms to affect Christ- a property was spared. Hundreds of trees in Hagley church in the last 100 years have been in July 1945 and Park were split, large branches broken off or were August 1992. While the former was slightly larger than completely uprooted. the latter, there is less information available regarding the 1945 event so it seems advisable to use the more The city was without power because of the overhead recent one as a basis for a scenario. The major argu- lines being blown down and the line to Islington was ment against this approach is that while there was little also down. difference in the amounts measured at the Botanic Gardens (28 cm versus 25 cm) the maximum reported Effects north of Christchurch elsewhere in the city and on the plains was 18" (46 cm) The Benmore/Haywards line was brought down in in 1945 and only 30 cm in 1992. North Canterbury at about 8:30 am, knocking out the As for the wind section, it is not possible to produce a supply to the North Island through the Cook Strait hazard map. cable. 40 • Risks and Realities

Characteristics of low elevation Snowstorm impacts snowstorms in the South Island The impacts most commonly reported as a result of low elevation storms are stock losses. However, damage to Meteorological conditions producing heavy buildings, other structures and power lines, and disrup- snowfall tion to traffic and communications are clearly much Neale and Thompson (1977) studied conditions bring- more important for urban areas. In addition, secondary ing heavy snowfall to low elevations in the South effects like snowmelt flooding need to be considered. Island by investigating ten storms for the years 1966 to 1975. Unfortunately few of these storms brought snow Snow loads and building codes. to coastal areas in Canterbury. However, the general Snow loads on structures are determined from the features they discussed are indicative of synoptic con- product of snow depth and density. Snow densities can ditions which produce snowfall at low elevations. be classified as shown in Table 2 .3. Because of our Following studies in North America, Neale and warm moist maritime environment, New Zealand snow Thompson noted that two types of storms could be densities are generally high. For instance, a density of recognised: 100 kgm-3 is commonly used as an estimate for average new snowfall in North America and Europe. In the • Warm advection storms often associated with a Craigieburn Range, new snow densities average about quasi-stationary or warm front (usually accompa- 130 kgm-3in windless conditions while stiff slabs with nied by north-west flow ahead of a trough in the densities up to 400 kgm-3 may be deposited when high New Zealand case). winds occur.

• Vorticity advection storms which are characterised However, most snow to low levels in Canterbury by the presence of a low pressure system to consid- occurs in relatively windless conditions. Nevertheless, erable elevation. At the surface, these storms fre- it is likely to have densities of over 200 kgm-3 soon after quently have well developed and sometimes com- deposition. Chinn (1981) and Hughes (1974) reported plex depressions located near or to the east of densities of about 300 kgm-3 in new snow at low central New Zealand. The North American studies elevations in Canterbury. As time passes and espe- defined the relative contribution of these advection cially if warm temperatures and/or rain occur, density processes using a dense network of radiosonde will increase rapidly. stations.

While this is not possible in New Zealand, Neale Type Density (kgm-3) and Thompson were able to classify storms accord- Wild snow 10 — 30 ing to the dominant process using a combination of New, immediately after falling in calm 50 — 70 upper air data and distinctive cloud patterns indica- Damp new snow 100 — 200 tive of vorticity advection on satellite images. Settled snow 200 — 300 Wind packed snow — soft slab 100 — 290 Arising out of these synoptic conditions, Neale and Wind packed snow — hard slab 290 — 450 Thompson identified four main factors contributing to Firn 400 — 800 snowfall to low elevations: Glacier ice 917

• ascending motion; Table 2.3: Snow-ice types and density • low temperatures in the lower troposphere (below (Fitzharris et al. 1992) approximately 5 km); The load is given by the equation: • abundant atmospheric moisture; and dρg L = (1) • persistence of these conditions at a given location 1000 for a significant period (at least several hours) so where L is snow load (kPa) that snow can accumulate. d is snow depth (m) Because of the required combination of all of the four -3 factors, heavy snowfall to low elevations is difficult to ρ is snow density (kgm ) forecast; that is, the absence of one factor will contrib- g is gravitational acceleration ute to a change of conditions and heavy snow will not (9.8ms-2) eventuate. Hazards to Engineering Lifelines • 41

For example, if snow depth and density are 0.5 m and 150 kgm3, the load is 0.74 kPa. 30 1945 The building code for snow (NZSS 4203) divides New 25 Zealand into five zones. For the plains area of Christ- 1992 church (zone 4) application of an elevation factor gives a load of less than 0.5 kPa (Tyndall, pers. comm.). 20 1918

Other impacts 15 1895 Several other impacts other than direct vertical loading 1901 arise from snowfall or snow in combination with other 10 1896

factors. Annual maximum snow depth (cm) ? 5 (a) Surface conditions In cities where snow is relatively rare, disruption to 0 traffic is common because suitable street clearing 5 10 15 20 30 40 5060 100 200 machinery is not available in large quantities to clear Return period (years) roads. Figure 2.6: Return period for Christchurch (b) Snow in combination with wind snowstorms Fortunately, windiness is not often associated with heavy snow to low elevations in Canterbury (Harris The August 1992 Christchurch 1981). When it does occur, snow or rime of suitable snowstorm cohesiveness will remain on structures or wires and greatly amplify the force of wind. Rime is ice deposited Synoptic conditions when supercooled water impacts and adheres to solid The August 1992 snowstorm (August 26-29), can be objects and is fortunately uncommon at low elevations. classified as a vorticity advection type storm because of the characteristic deep and complex low pressure that developed over central New Zealand (Kingsland (c) Snowmelt flooding 1992) as shown in Figure 2.7. Because snow to low elevations falls at temperatures near 0°C, rapid melting is possible. This may often be In 24 hours on the 26 August, the low pressure system associated with rain following snowfall and/or rain initially located to the west of the North Island, deep- preceding snowfall which will accentuate flood peaks ened rapidly from 1003 hPa to 979 hPa with a corre- from the resulting melt. sponding increase in activity. Equally important was the supply of cold air in the southerly flow that occurred once the depression was positioned east of Snowstorms in Christchurch central New Zealand from 28 August onwards. The Seasonal occurrence effects of this on temperature are clearly seen from midnight through to noon on Friday 28 August (Figure While the standard meteorological station is not help- 2.8). ful with regard to snow amounts, it does show that in general snow on the plains in the Christchurch area is restricted to the months of May to October and is most Precipitation and snow distribution common in June, July and August. Historical reports For the period 26 - 28 August, 100 mm to 140 mm of also indicate that the most serious storms have oc- precipitation was recorded on the plains in the Christ- curred in the June-August period. church area (Figure 2.9(a)). At the Botanic Gardens, a 72 hour precipitation of 120 mm has a return period of Return Period and Magnitude about five years (Griffiths and Pearson 1993) indicating that such a storm is not particularly unusual. How- From newspaper reports, particularly of the 1945 storm ever, the 24 hour total of 56.5 mm to 9 am on the 27th, (which contained comparative data for earlier storms), was the highest ever in August at the airport showing an approximate idea of return periods for storms can be that high precipitation is not common in the winter obtained as shown in Figure 2.6. This suggests that the months. On the Port Hills, precipitation totals of over 1992 storm has a return period of about 50 to 100 years. 200 mm were measured. 42 • Risks and Realities

Midday 24 August 1992 Midday 25 August 1992

Midday 26 August 1992 Midday 27 August 1992

Midday 28 August 1992 Midday 29 August 1992

Figure 2.7: Surface synoptic conditions August 24 - 29 1992

4 deg Noon — The snowfall distribution (Fig 2.9(b)) shows two main warmer air 11.30pm — cold arrives, snow differences when compared with total precipitation. air arrives over turns to rain. Christchurch. Firstly, the maximum amounts on the plains were 3 deg Heavy snow starts significantly further east into the eastern suburbs. This may represent a local circulation effect though more 2 deg probably reflects the coincidence of heavier precipitation with the period of colder temperatures in that area.

1 deg It seems unlikely that a corridor of consistently heavier snowfall is indicated. Secondly, the effect of elevation on temperature and the consequent effect on precipita- 0 deg 6am 9am noon 3pm 6pm 9pm midnight 3am 6am 9am noon 3pm 6pm tion type is also clear. The influence of elevation on Thursday Friday total precipitation amounts is well known and in this instance results in a ratio of 2:1 between the Summit Figure 2.8: Temperature variations during the Road and plains area total precipitation amounts. For well mixed air, a lapse of 0.65 C for each 100 m August 1992 snowstorm (Kingsland 1993) ° increase in elevation is expected so snow is increas- Hazards to Engineering Lifelines • 43

100

100 Rainfall depth (mm) Wed 26 Aug to Fri 28 Aug 1992

100

120

120 140 180 160 200

200 180 140 012345678km 160

Figure 2.9(a): Total precipitation in the Christchurch area, 26 to 28 August 1992

17 Estimated snow depth (cm) 12 Fri 28 Aug 1992

37 47 57 012345678km

Figure 2.9(b): Estimated snow depth in the Christchurch 1992 snowstorm (Christchurch City Council) 44 • Risks and Realities

ingly more likely for higher elevations. Consequently city especially in the hill suburbs. However, most the ratio of snowdepths between the Port Hills summit power supply was restored by midday on Friday 28 and the plains area ranges as high as 5:1. However, the August. Power supply to the Sugarloaf transmitter was snow on the plains probably attained greater densities lost and fuel for the emergency generator was soon than that on the Port Hills. Anecdotal evidence indi- depleted. It took two days before this could be replen- cates that, as expected, there was greater drifting on the ished. hills with drifts of over 2 m reported.

This second effect would definitely have to be built Surface conditions into a snowstorm scenario for the Christchurch area. Although many people heeded radio broadcasts sug- There is little basis to assume that any particular pattern gesting that they not try to go to work, many vehicles of snow depth will occur on the plains area except that became involved in mishaps as a result of attempting to depths are likely to decrease near the coast because of drive in difficult conditions. the influence of the sea on temperatures. However, as the 1992 storm showed, significant depths were re- Flooding corded quite near the coastline. Surface flooding occurred to a limited degree but is not investigated in detail here. Forecasting the 1992 storm During 1992, there was some criticism of MetService’s Conclusions performance in forecasting low elevation snowstorms, Because of the more readily available recent informa- particularly in relation to the mid-July storm. This was tion, it seems better to use the August 1992 storm as a attributed to loss of personnel and appropriate mete- basis for a snowstorm scenario. However, further re- orological observations arising from funding short- search may reveal more information about the 1945 ages. However, since then two developments have event. occurred to improve the capability of forecasting low elevation snowstorms in the South Island. Firstly, the There is little reason to suggest anything other than a policy of centralising all forecasting staff in Welling- reasonably uniform snow depth of about 30 cm on the ton has been partly reversed and a forecast office is now plains of metropolitan Christchurch with an increase to maintained in Christchurch. Secondly, MetService, in approximately a metre on average on the Port Hills. response to criticisms of its performance, has established a network of snow observers which should allow More work is needed to enumerate the impacts of such early verification when snow is forecast or warnings snowfalls. when it is not. Acknowledgements Snow loads Arthur Tyndall provided valuable insight into design The maximum snow depth recorded on the plains was standards relating to snowfall in New Zealand and 30 cm. Even if a conservatively low density of 200 Andrew Nichols (Christchurch City Council) provided kgm-3 is used, equation (1) gives a load of 0.6 kPa information on August 1992 precipitation and snow which suggests that the New Zealand standard may be depth. somewhat low for Christchurch as densities were undoubtedly higher, particularly once precipitation turned References to rain after noon on Friday August 28. Chinn, T J (1981). “Snowfall variations, hazards and snowmelt”, p 1-21 in Mountain Lands Workshop, Impacts of the August 1992 snowstorm Water and Science Centre, Christchurch, M.W.D.

Damage to buildings Fitzharris, B B, I F Owens and T J Chinn (1992). “Snow and Glacier Hydrology”, p. 75-94 in Mosley, M P (ed) The most spectacular failure was the collapse of the Waters of New Zealand, N.Z. Hydrological Society. roof of the skating rink in Opawa. Other smaller damages occurred particularly as a result of snow creep Griffiths, G A and C P Pearson (1993) “Distribution of onto guttering but no widespread damage was re- high intensity rainfalls in metropolitan Christchurch”, ported. N.Z. Journ. Hydrology, 31(1):5-22.

Power supplies Harris, D W (1981). “Incidence of wind and snow on rural reticulation”, Weather and Climate 1 (2): 53-60. Lines and poles were down over significant areas of the Hazards to Engineering Lifelines • 45

Hughes, J C (1973). “The snow of August 1973”, ple, buried services well keyed into rock in the high risk Tussock Grasslands and Mountain Lands Institute zone may not be vulnerable to damage from soil Review, 30:64-100. sliding, debris flows or falling rock debris.

Kingsland, S (1992). A study of the meteorological Conversely a stable low risk area for housing may conditions accompanying the August 1992 snowstorm, contain locally unstable cut road batters of no risk to unpub. report GEOG 614, Geography Dept., Univer- existing houses but close enough to threaten services. sity of Canterbury. However, in the absence of an actual examination of the critical services corridors in the field, this map is the Neale, A A and G Thompson (1977). “Meteorological best current method to carry out a desk study. (Note: conditions accompanying heavy snowfalls in southem the studies undertaken by the Task Groups were based New Zealand”, N.Z. Met. Serv. Tech. Info. Circ., on the map but supplemented by a drive-over of the 155,18p. likely critical areas by minibus. Representatives of each of the service authorities were on the bus which stopped at critical areas and the likely situation was 2.8 Slope Hazard and Damage discussed with the engineering geologist who was to Services on Hills responsible for the map preparation.) The estimates of the real extent of damage in the three zones is of necessity arbitrary and subjective. Introduction The slope hazard to lifelines in Christchurch is essentially confined to the Port Hills, and is likely to be The event significant only during a severe rainstorm or earth- The triggering event is assumed to be a 1 in 100-year quake. This hazard map is therefore to be read in local rainstorm coming late in a relatively wet winter conjunction with the earthquake or local flooding when soil ground water levels are already high. Alter- hazard maps. The hazard information is based on a natively a similar pattern of damage could result from slope hazard zoning project for the Christchurch City a 1 in 100- or 150-year earthquake occurring in later Council recently completed by Soils and Foundations winter. Ltd. This work was based on aerial photographic analysis and field observation and divided the selected Slope hazard zones with respect to hill areas into three simple zones of low, moderate and services high risk. Each zone broadly corresponds to a level of See Map 5, page 288. “slope hazard” and recommends various degrees of site investigation for future subdivision or building. For existing hill subdivisions, the information is useful Zone 1 for PIMS (Project Information Memoranda) and LIMS Low risk. Likely damage to services negligible. (Land Information Memoranda) applications. Zone 2 The Map Moderate risk. Likely damage may affect 10% to 20% No attempt was made in the map (Map 5, page 288) to of area, normally where zone 2 is closest to zone 3. A separate the various types of potential slope hazard similar proportion (10% to 20%) of roads and overhead (e.g. soil landslides, rockfall or rockroll, tunnel erosion services could be affected. For buried services damage etc.) because this sort of site specific information can is likely to occur only from undermining or deep- only be adequately addressed at the site investigation seated failure reducing the extent of damage to 5% to stage. It also complicates the final map. However, the 10% of service length. three simple categories separate the difficult ground from areas that are generally more stable and much Zone 3 more straight forward. High Risk. Likely damage may affect up to 40% of In the absence of any other rational way to address the area. A similar proportion of the length of roads and risk posed to services by slope processes in the hills, the overhead services could be affected. Once again buried slope hazard zoning map has been used to highlight services are likely to suffer less damage, say 10% to areas where damage to services may be more concen- 20% of the service length. trated.

The difficulty in doing this reflects the difference in vulnerability between houses and services. For exam- 46 • Risks and Realities Seismic Liquefaction and Lifelines • 47

Chapter 3 Seismic Liquefaction and Lifelines

With so much of the eastern side of Christchurch being information which was presented at the workshop and potentially subject to liquefaction, an understanding of extended it to describe in more detail some prediction the potential effects of liquefaction on the various engi- procedures with particular reference to tests that were neering services is particularly important for Christ- undertaken by the University of Canterbury as part of the church. Lifelines Project. Over 30 tests are at present (1997) being undertaken in critical areas so the explanation gives some At the Project Workshop, Dr John B. Berrill, Reader in detail on the cone penetrometer test which is used. Civil Engineering at the University of Canterbury, gave a presentation on liquefaction for the non-geotechnical The 1993 Wellington Lifelines Earthquake Group Report engineers dealing with the causes and effects of liquefac- Section 1.5 dealt with the consequences of liquefaction for tion to help the engineer considering the likely effects on Lifelines and since this was of considerable use in the a service located in a sandy deposit. The presentation Christchurch Project it has, with the permission of the dealt with the gross mechanism of liquefaction, some Wellington Earthquake Lifelines Group, been reproduced slides showing the effect, the soil conditions that are in this section. necessary to provoke liquefaction, some of the methods of predicting level ground liquefaction, and then what the A team from the Wellington Earthquake Lifelines Group University of Canterbury did to help look at some key sites and the Christchurch Engineering Lifelines Project went in the lifelines network of Christchurch. to Kobe in August 1995 and those who saw the considerable damage caused to lifelines’ facilities as a result of It has not been possible to include in this book the full text liquefaction during the 1995 Hyogoken-Nanbu (Kobe) of the presentation, or the slides presented, but Dr. Berrill earthquake are well aware of the potential for very exten- has, in the paper that follows, included much of the sive damage in eastern Christchurch.

Abstract failure of foundations, slopes and embankments, for example, or indirectly through damage to lifelines. An overview of the liquefaction problem is presented, Loss of water for fire fighting, leading to uncontrolled with emphasis on effects on engineering lifelines. The spread of fire is a common example of indirect damage. distinction is drawn between liquefaction flow failures Although evidence of liquefaction can be found in and deformation failures, corresponding in the labora- most major historical earthquakes, and research into tory to the difference in behaviour of loose and dense both its causes and effects had been pursued well sands in undrained tests. Methods for predicting level- before 1964, it was the Niigata, Japan and Anchorage, ground liquefaction potential using in-situ tests are Alaska earthquakes of that year which compelled wide- reviewed. The CPT test is favoured over the SPT. The spread recognition of the importance of liquefaction problem of flow failure versus cyclic mobility is dis- and gave impetus to the large amount of research that cussed and an approach to estimating the stability of followed and continues to the present day. Liquefac- dam slopes against flow failures is sketched in outline. tion-induced damage to many different sorts of life- The success of the Newmark sliding-block method in lines in Kobe in the 1995 Hyogo-ken-Nambu Earth- modelling deformation failures is noted. References quake confirms the devastating potential of seismic are given to studies of a number of other aspects of liquefaction. liquefaction not covered in this review. It is instructive to review the effects of liquefaction at Niigata and Anchorage. We see the same sorts of 3.1 Introduction damage occurring to greater or lesser extent in most earthquakes with magnitude of 6 or more. The city of Liquefaction of fine-grained, cohesionless soils has Niigata is located on the west coast of Japan and is been a major cause of damage in earthquakes. Lique- founded on 30 m or so of alluvial sand deposited by the faction damage has occurred either directly, through Shinano River (Figure 3.1). The M7.5 earthquake of 48 • Risks and Realities

Remarks

Flooding Area

Damaged area

Area ravaged by fire Agano River Collapse of quay wall

Coast sand dune

Shinano River

Niigata City

Niigata St 0 1000 2000 m S Shinanogawa Showa h Echigo Line in Bridge et Bridge su Hakusan St L in Sekiya St e

Figure 3.1: City of Niigata, Japan, which suffered widespread liquefaction damage in the Magnitude 7.5 earthquake of 1964. Epicentral distance was about 50 km (From Yamada, 1966)

16 June, with epicentre about 50 km offshore to the Lateral displacement of the piers of the Showa Bridge north, caused extensive liquefaction in loose sands, (Figure 3.3), for example, caused five simply sup- especially in low-lying fill and old river channel mate- ported spans to fall. Other bridges suffered less dra- rial along the lower reaches of the Shinano River. matic but nevertheless important damage. Lateral Damage included the settlement and tilting of build- spreading also caused severe damage to embankments ings, with some structures settling by a metre or more and to railway yards. and tilting several degrees off vertical. In one case an apartment building (Figure 3.2) tilted as much as 80˚ as soil beneath its foundation liquefied. Lateral spreading on shallow slopes of just a few degrees caused widespread damage to buried services and to bridge and building foundations. Light-weight buried structures floated upwards in liquefied sands. Settlement resulted in inundation of already low-lying areas, and the ejected sand itself proved to be a great nuisance, clogging pipes and hindering recovery operations. Liquefaction was much less prevalent in the denser dune sand along the shore.

Figure 3.3: Effect of lateral spreading on the Showa Bridge, Niigata, 1964.

Lateral spreading is probably the principal cause of damage to lifelines. It may occur in slopes of a few degrees or in level ground in the presence of a bank or river channel. The mechanism is quite simple. Con- sider the free-body diagram shown in Figure 3.4, in which below the watertable we have liquefied soil exerting a pressure of ρgz on the left hand side of the body, where ρ is bulk density of the soil and g is the acceleration due to gravity. This is opposed on the

Figure 3.2: Result of liquefaction-induced founda- right hand end by a pressure distribution going as ρwgz, tion failure on apartment buildings at Niigata where ρw is the density of water. Since the bulk density Seismic Liquefaction and Lifelines • 49

Mechanism of lateral spreading of level river bank Consider free body: River channel

z ρwgz ρgz Liquefied soil

ρ = Bulk density ≈ 1.7t/m3

3 ρw = density of water = 1.0t/m

Figure 3.4: Free-body diagram of soil mass undergoing lateral spreading. Out-of-balance forces towards river channel drive lateral motion. of soil is about twice that of water, there is an out-of- balance force towards the channel, driving the lateral bridges were damaged in the Alaska earthquake. The spreading. Because of dilatancy effects, the soil prob- class of bridges suffering the greatest damage were ably stiffens and reliquefies a number of times during those founded on piles driven into loose to medium- the shaking; thus the spreading takes place in a series dense fine sands and coarse silts. Here, lateral spread- of increments and its magnitude is limited. ing of foundation soils towards stream channels caused displacements of abutments and pier structures, failure Above the watertable, the soil is partially saturated, not of piles and settlements, often damaging the bridge liquefied, and generally somewhat brittle. As it rafts superstructure as well (Figure 3.6). Spreading of along on the underlying liquefied soil, it cracks into approach fills was also common. Much of this damage blocks. Relative movement of these blocks is very was attributed to liquefaction of fine sands (Ross et al., damaging to buried services, and the unliquefied crust 1973). Bridges on piles driven into medium or coarse can impose large forces (in the limit, passive forces) on sands suffered considerably less damage, while those piled foundations and other restrained structures. structures founded directly on rock were hardly damaged at all. In addition to lateral displacements, liquefaction usually results in ground settlement. To give an idea of the magnitudes involved, let us consider the Landing Road Bridge site at Whakatane, New Zealand, in the 1987 Edgecumbe earthquake. Here, the ground surface settled by about 300 mm over an area about one km long and about 300 m wide parallel to the river (Figure 3.5). Lateral movement at the left river bank was about 1.5 m. This movement was due to liquefaction of a layer of very recent loose sand, 4 m thick, during the quite modest magnitude 6.3 earthquake (Berrill and Christensen, 1995). At the Showa Bridge in Niigata, about 8 m of lateral displacement occurred. In larger earthquakes, such as the Alaskan Earthquake, greater settlements and spreading have been observed. Figure 3.5: Landing Road Bridge, Whakatane, New Zealand, looking downstream The M8.4 Alaska earthquake of 27 March, 1964 caused several landslides in and around the city of Anchorage. Lateral spreading of the left bank extended 300 m back from the river channel. Horizontal dis- While there is some doubt about the role of liquefaction placement was as much as 1.5 m; settlement in the large Turnagain Heights slide which occurred was about 300 mm. The buried bridge founda- principally in clay soil, there were several lateral spread- tions resisted the spreading, inducing passive ing failures that have been attributed to liquefaction of failure in the layer of unliquefied soil above the sands as well as flow failures due to liquefaction, most watertable. Lateral forces of about 1 MN were notably, the Potter Hill slide (Long, 1973). Numerous generated at each pier. 50 • Risks and Realities

For Soils of Uniform Grading 100

75 Very easily liquefy (A) 50 Easily liquefy

25 (B) (B)

Percentage finer by weight 0 0.01 0.1 1.0 10 Clay Silt Sand Gravel Particle Size (mm)

For Well Graded Soils 100

75 Figure 3.6: An effect of lateral spreading in the Very easily liquefy (A) 50 Alaskan Earthquake of 1964 Easily liquefy

25 (B) (B) The damage seen in 1964 at Niigata and in Alaska, and

Percentage finer by weight 0 in most other earthquakes before and since, illustrates 0.01 0.1 1.0 10 the principal classes of liquefaction problem and brings Clay Silt Sand Gravel out the importance of soil density and particle grading. Particle Size (mm)

Figure 3.7: Tsuchida’s curves for grading ranges 3.2 Fundamentals of liquefiable soils (from Iwasaki, 1986)

Let us review some fundamental concepts. Firstly, momentarily (twice) during a loading cycle. During under cyclic loading in shear, cohesionless materials the remainder of the cycle effective confining stresses tend to decrease in volume [provided a small shear are significant, and hence considerable strength restrain threshold of about 0.01 percent is exceeded mains. These two different behaviours are illustrated (Dobry et al., 1980)]. This tendency to decrease in in Figure 3.8, taken from Ishihara (1985), showing the volume is much greater in loose than in dense soils. results of undrained cyclic torsional shear tests on a Nevertheless, it is also present in dense soils. When the uniform, medium sand. soil is saturated and drainage of pore water is prevented, the tendency to volume decrease under cyclic A very rapid build-up of excess pore pressure and loading results instead in an increase in pore pressure. subsequent loss of strength, together with the develop- In the laboratory, this effect can be obtained by carry- ment of large strains are characteristics of seismic ing out undrained tests on saturated samples of liquefaction of loose sands. The behaviour of the dense cohesionless soil; in the field, drainage is impeded sand in which strain amplitudes build-up in small naturally by the low permeability of fine-grained soils. increments, is termed cyclic mobility. In the field, Thus cohesionless soils with low permeability, such as liquefaction of loose sands can lead to flow failures of fine sands and silts, exhibit pore pressure increase slopes and large displacements of foundations. Cyclic under seismic loading, and are the most susceptible to mobility, on the other hand, results in limited soil liquefaction. Ranges of most critical gradings, ob- displacement, seen for example in the limited settle- tained from field studies by Tsuchida and Hayashi ment and deformation of some earth dams during (1971), are shown in Figure 3.7. earthquakes.

With loose sands, laboratory tests show that pore Thus liquefaction problems (using the term “liquefac- pressure increase occurs rapidly and that pore pressure tion” in the broad sense) fall into two classes, depend- can approach effective confining pressure in a few ing on whether the soil is loose or dense. The seismic cycles of cyclic loading. In a dense specimen of the behaviour of a soil mass also depends on whether or not same sand, many more cycles of a generally greater shear strength is required for static equilibrium. For amplitude are required to produce a condition of initial example, a sand deposit with a level ground surface liquefaction, where pore pressure u equals initial effec- may loose its shear strength entirely yet still remain in static equilibrium. On the other hand, some shear tive confining pressure σ ο′ . Furthermore, with the strength is always required to maintain a slope or a σ ′ loose sand, u remains near ο during subsequent cy- loaded foundation in static equilibrium. cles of loading. Consequently, effective confining stress, and hence strength, becomes and remains small. The case of liquefaction of loose deposits in level

On the other hand, in dense sand u approaches σ ο′ only ground has been widely studied, and there are many Seismic Liquefaction and Lifelines • 51

Cyclic stress ratio 0.2 Stress-strain curve τd Loose sand 0.1 '

' Torsional shear test

cyclic stress o σ o

/ σ 0 d/σ ' = 0.229 0.6

ratio d τ o τ 0.1 Torsional shear Dr = 47% test 2 σo'= 98 kN/m 0.4 0.2 τd/σ ' = 0.229 10 o Fuji river sand Dr = 47% 5 2 σ '= 98 kN/m 0.2 shear strain 0 o

k = 1.0 Y (%) Y o 5 7 5 3

10 3 5 1.0 γ (%) Shear strain

0.2 '

normalised o 0.5

/ σ 0.4 pore pressure u

0 time (a) Loose sand (relative density Dr = 47%)

Cyclic stress ratio 0.6 Stress-strain curve Dense sand τd

0.3 Torsional shear test σ' ' o

cyclic stress o τd/σ ' = 0.717 0.6

/ σ 0 o ratio d Dr = 75% τ 0.3 2 σo'= 98 kN/m 0.4 0.6 Fuji river sand 0.2 5 3 shear strain 0 5

Y (%) Y 5 2 Torsional shear test d ' = 0.717 D = 75% = 98 kN/m 3 5 τ /σo r σo' γ (%) 1.0 Shear strain

normalised '

o 0.5 0.4

pore pressure / σ u 0.6 0 time

(b) Dense sand (relative density Dr = 75%)

Figure 3.8: Results of cyclic torsional shear tests on (a) loose and (b) dense specimens of Fuji River sand (From Ishihara, 1985) procedures for estimating whether or not a particular liquefaction, and the effect of age was very evident, deposit is likely to liquefy in a given earthquake. Some with lateral spreading, for example, being restricted to involve laboratory testing, others in situ testing. late Holocene deposits.

The widely-used technique of Seed and Idriss (Seed These basic aspects of the liquefaction problem are and Idriss, 1971; Seed et al. 1985) is an example of the discussed in more detail in the remainder of this chap- latter. Procedures for solution of other cases are not as ter, as is an outline of some of the methods of solution well established, and are still largely in the province of proposed. research. For the case of flow failures in loose sands, the work of Castro et al. (1985) and Dobry et al. (1984) offer procedures for estimating whether or not a slope 3.3 Level Ground Liquefaction is stable. Bartlett and Youd (1992) present a method for estimating lateral spreading distances, and the work Although liquefaction of level ground does not cause of O’Rourke and Pease (1992) allows likely damage to as great a threat to life and limb as do flow and bearing buried pipes to be estimated. capacity failures, it is responsible for very costly material losses, chiefly through damage to buried pipelines, Tsuchida’s (1970) grading curves point to the general but also through damage to pavements and other sur- importance of geological considerations. Loose, fine- face works. The main surface manifestation of the grained sediments are deposited only under certain liquefaction of an underlying layer is the formation of geologic conditions, and resistance to liquefaction sand boils, in which sand and water are ejected through increases with age, as weathering, cementation and, fissures or circular vents, leaving shallow cones of certainly, other processes cause the fabric of a soil to sand on the ground surface. Scott and Zuckerman develop. Tinsley and Dupré (1992) find very clear (1973) have studied the formation of sand boils in the correlations between geologic history and liquefaction laboratory. Their experiments indicate that a layer of effects in the Monterey Bay area during the 1989 Loma finer grained soil overlying the liquefiable soil is Prieta earthquake. They note that laterally accreted necessary for the formation of sand boils. The vent structures such as point-bar formations (e.g. at Landing itself is formed by the upwards enlargement of a cavity Road Bridge, Figure 3.5) are especially susceptibly to which begins at the base of the upper layer as it unravels 52 • Risks and Realities

in an unstable, localised fashion into the underlying Whether or not a site will liquefy in an earthquake liquefied sand-water mixture. Once the vent has bro- depends both on the strength of shaking (seismic ken through to the surface, sand and water are ejected, loading) at the site and on the state of the soil. The driven by the hydraulic gradient which is generated as seismic loading can be characterised in a local fashion

the liquefied layer carries the full weight of the over- for example, by peak acceleration amax or modified burden as a fluid pressure. Mercalli intensity I at the site, or by a source description using, for example, magnitude M and epicentral Florin and Ivanov (1961) noted from laboratory ex- distance, re. Liao, Veneziano and Whitman (1988) periments that, given uniform density, liquefaction have made a rigorous evaluation of both types of model begins at the top of a layer and propagates downwards. and conclude that in the study of a specific site, local They also observed that the tendency to liquefaction characterisation fits case history data better. But the decreases with increasing overburden or confining advantage is lost when the value of the local ground pressure. This observation has been made independ- motion parameter must be estimated separately by an ently by Seed and Lee (1966) and by many other empirical attenuation expression. Thus, for regional researchers. hazard mapping, say, it is better to go straight to a source-type liquefaction model. Both Scott and Zuckerman and Florin and Ivanov also observed that subsequent solidification (the opposite By tradition in the US, Japan, and many other seismic of liquefaction) begins at the bottom of the liquefied countries, the state of cohesionless soils has been layer and proceeds upwards, as particles settle out of characterised in situ by the Standard Penetration Test suspension. Scott and Zuckerman also observed that a (SPT). This is somewhat fortuitous, since the SPT N- denser granular layer overlying a loose layer may be value (the blow count) is related to relative density induced to liquefy as support for its solid skeleton is which, together with effective confining pressure, pro- lost as the underlying layer liquefies, even though this vides a measure of the tendency of the soil to dilate or layer would not liquefy on its own. They term this contract under shearing. However, the SPT has some secondary liquefaction. Here, a liquefaction front serious disadvantages especially in loose sands and propagates upwards. Some of the Niigata foundation silts. These will be discussed later. Furthermore, a lack failures have been attributed to secondary liquefaction, of adequate standardisation has led to a wide range of provoked by the initial liquefaction of a deeper layer. hammer efficiencies and to the need for corrections to a reference efficiency, usually 60% (Seed et al., 1985). In general, an intact surface layer overlying a liquefied Correction for overburden pressure is also required if stratum floats on a fluid of quite similar density to its the N-value is to correspond to soil density. own. Therefore it has a tendency to sink to establish equilibrium. Any departure from uniform density, Liao and Whitman (1986) review the overburden cor- thickness or surface loading will tend to induce bend- rection problem and recommend correcting the meas- ing stress in the surface layer which, if brittle, may ured value of N to a value normalised to an effective crack. Fissures and cracks are commonly associated overburden pressure of 1 ton/sq.ft (100 kPa) by the with level-ground liquefaction, and are the cause of formula great damage to buried pipes. Youd (1984) terms this disruption ground oscillation (as blocks of cracked N1 = CNN (1) surface material move relative to one another) and where N denotes the corrected value of N, states that ground oscillation together with lateral 1 spreading have caused more property damage in earth- C = 100 /σ ′ (2) quakes this century than flow and bearing capacity N ο failures. and where σ ο′ is effective overburden stress in kPa.

Prediction Procedures Seed’s Procedure Procedures for predicting the liquefaction potential of The first and still most widely-used procedure for level ground sites fall into two classes. Those based on evaluating whether or not a site is likely to liquefy is laboratory testing of field specimens, and those based that of Seed and Idriss (1971), which has been modified on in situ testing. Because of the difficulty of obtaining successively over the years. A local characterisation of undisturbed samples of cohesionless soils, methods the earthquake loading is employed. based on in situ test results have become more common, and we will focus on them. These methods are all, Seismic loading on the soil layer is characterised by an to some degree, empirical. average cyclic stress ratio τav/ σ ο′ given by the expression: Seismic Liquefaction and Lifelines • 53

τ a σ r The soil is characterised by its SPT N-value, corrected av = max o d 0. 65 (3) for overburden pressure using equation (1) and to a σ ο′ g σ ο′ CM hammer efficiency of 60%, using the information in where τav is equivalent average shear stress, σo is total Table 3.2. The resulting corrected SPT blow count is overburden stress, a /g is peak ground acceleration as max denoted by the symbol (N1)60. a fraction of g, rd is a factor to account for soil flexibil- The final step in the procedure is to check whether or ity, and CM is a magnitude correction factor. not the seismic loading given by equation (3) exceeds This expression is derived by considering the equilib- a threshold value obtained for that soil state (N1)60. The rium of a rigid soil column under a horizontal accelera- threshold value of τav/ σ ο′ is found in the chart repro- tion amax. The factor of 0.65 allows for an average shear duced in Figure 3.9, derived empirically from liquefac- stress somewhat less than the peak stress correspond- tion case histories (Seed et al., 1985). ing to peak acceleration. Allowance for flexible rather than rigid response of the overlying soil mass is made Figure 3.9 contains three curves, for different fines contents. In their appraisal of liquefaction prediction by the term rd, which is conveniently obtained, following Japanese practice (Iwasaki, 1986), from the expres- methods, Liao et al. (1988) investigated how well their sion: comprehensive set of case history data supports such a marked influence of fines content. They conclude that r d = (1− 0.015z) (4) while the presence of more than a moderate percentage where z is depth in metres. The factor CM takes the of fines does have an effect on probability of liquefac- value unity for M = 7.5 (reflecting the procedure’s tion, the effect is not nearly as marked as Figure 3.3 origins in Niigata data); its value for other magnitudes indicates. The data does not suggest a progressive may be found from Table 3.1. increase in resistance with increasing fines content. It does, however, support a division into two classes: Earthquake Number of Correction clean sand and silty sand, with a fines content of 12% Magnitude Representative Factor Cycles at 0.65 ττmax CM as the dividing line. Liao et al. point out that the curves 8.5 26 0.89 for soils with fines in Figure 3.9 are based on laboratory 7.5 15 1.0 tests on specimens at constant relative density, Dr, not 6.75 10 1.13 at constant (N1)60. The field data in terms of (N1)60 does 6.0 5-6 1.32 not show such a strong effect. Liao et al. note further 5.25 2-3 1.5 that when seismic loading is represented by magnitude and distance rather than peak acceleration at the site, the uncertainties associated with attenuation over- Table 3.1: Magnitude Correction Factor C M whelm the distinction between clean and silty soil. Thus when a source characterization of the earthquake Originally, Seed and Idriss applied the CM factor at a later stage; this chapter follows Liao et al. (1988) in is employed, it is not worth the trouble of distinguish- incorporating it in the loading expression, since it does ing between silty and clean sand deposits. indeed represent a loading effect.

Country Hammer Hammer Estimated Correction Type Release Rod Energy Factor for (Percent) 60 Percent Rod Energy Japana Donut Free-fall 78 78/60 = 1.30 Donut Rope and pulley 67 67/60 = 1.12 with special throw release United Safety Rope and pulley 60 60/60 = 1.00 States Donutb Rope and pulley 45 45/60 = 0.75 Argentina Donut Rope and pulley 45 45/60 = 0.75 China Donut Free-fallc 60 60/60 = 1.00 Donut Rope and pulley 50 50/60 = 0.83

a Japanese SPT results have additional corrections for borehole diameter and frequency effects. b Prevalent method in the United States today. c Pilcon-type hammers develop an energy ratio of about 60 percent.

Table 3.2: Summary of Energy Ratios for Various SPT Procedures (from Seed et al., 1985) 54 • Risks and Realities

Finn (1992) raises a further point with regard to the and Shokooh (1979) that pore pressure increase is effect of fines. The greater liquefaction resistance for proportional to the density of seismic energy dissi- silty sands implied in Figure 3.9, and observed by Liao pated. et al., is seen when comparisons are made at similar Its derivation, which uses well-established results from values of (N1)60. However, Troncoso (1990) found that cyclic strength decreased with increasing silt content seismology and soil mechanics, and seeks to keep when he compared samples at constant void ratio. empirical steps to a minimum, proceeds as follows. Furthermore, Kuerbis and Vaid (1989) tested a particu- Combining the expression of Gutenberg and Richter lar sand at constant sand-skeleton void ratio. This (1954) for total radiated energy with a simple geomet- sand-skeleton could accommodate up to 20% fines. ric spreading rule, yields the density of seismic energy They found that for fines contents of less than 20%, the arriving at the site. Hardin (1965) found that energy −1/2 specimens had the same cyclic strength. Finn observes dissipation is proportional to ()σ ο′ . Using this result that from another point of view, these results imply that and the assumption that Δu is proportional to the the penetration resistance of a silty sand is somewhat density of dissipated energy yields the following ex- less than that of a clean sand with the same cyclic pression for seismic pore pressure increase: strength. Thus the presence of silt does not increase λ(N )101.5M liquefaction resistance; it simply reduces the SPT N Δu= 1 2 1/2 (5) value. Ishihara (1993) has found that it requires the r ()σ ο′ addition of cohesive fines to truly increase liquefaction where λ is an unknown function of the corrected SPT resistance. He found that if the plasticity index, I , is p value N , characterising the state of the soil. The less than about 12 there is no increase in resistance. 1 function λ(N1) is then found from case history data using linear discriminant theory to give the final result: Energy Dissipation Approach 450 101. 5M When good estimates of ground motion intensity are u Δ = 2 2 1/ 2 (6) not available for the site, it is more appropriate to use r (N1 ) 60 ()σ ο′ a procedure based on magnitude and distance to the Here, r is expressed in metres, σ ′ in kPa, and (N ) is earthquake source. One such model is that of Davis & ο 1 60 substituted for N which was used in the original Berrill (1982), which performed well amongst a number 1 of procedures of that type tested by Liao et al. (1988). derivation. Its simple functional form makes equation The model is based on the suggestion of Nemat-Nasser (6) particularly suited to probabilistic hazard analysis,

0.6 37 29

' Percent Fines = 35 15 ≤5

o av σ τ 0.5

0.4 20

0.3 20 12 50+

17 27 18 80 50+ 60 20 11 20 0.2 10 10 10 50 52 40 10 20 25 12 10 10 80 20 10 30 20 25 12 12 22 13 75 12 27 67 30 FINES CONTENT ≥5% 75 50+ 0.1 60 10 20 Modified Chinese Code Proposal (clay content = 5%) 13 30 27 Marginal No

Cyclic stress ratio causing liquefaction Liquefaction Liquefaction Liquefaction 31 Pan-American data Japanese data Chinese data 0 0 10 20 30 40 50 (N ) Corrected SPT blow count 1 60

Figure 3.9: Relationship between Stress Ratios Causing Liquefaction and (N1)60 Values, from Seed et al. (1985) Seismic Liquefaction and Lifelines • 55

and an example is worked in the original paper. Liao SPT, spaced at 1 m or greater centres. Some care is et al. (1988) observe that this model does not perform required in interpreting cone resistance in thin layers well with respect to their data set for dense sands, and (Vreugdenhil et al., 1993) but generally, the CPT is a caution against using it for N1 > 20. Furthermore, robust test. because of the simple 1/r2 attenuation rule, it does not work well in the near field of the earthquake source, A cutaway drawing of a standard cone penetrometer is where the rupture cannot be idealized as a point source. shown in Figure 3.10 (the cross-sectional area of the 2 However the energy-dissipation approach has much standard cone is 10 cm , giving a diameter of 35.7 mm). merit, and the authors are continuing to develop it The basic cone has two instrumented sections: the 60˚ (Davis and Berrill, 1996). conical point and the sleeve immediately behind the point. The load carried by each is expressed as a stress;

Liao et al. themselves devised an expression for prob- the cone resistance qc in the case of the point and sleeve ability of liquefaction PL which employs the same function fs in the case of the sleeve. seismic loading term, namely: Cone resistance qc gives a measure of the overall 101. 5M strength or, more fundamentally, density in the case of Λ = e 2 3/ 2 (7) cohesionless soils. Sleeve friction is more usefully r e ()σ ο′ but which was calibrated to their larger and more complete data set. Their expression is:

P L =1/{} 1+ exp[12.922 − 0.87213 n(Λe ) +0.21056(N1 )60 ] (8) e Sleeve friction sleeve force Here, re is epicentral distance. They present a corre- f s = sponding expression in terms of hypocentral distance, sleeve area which fits the data a little better. Cone resistance cone force Remarks on the SPT and CPT q c = projected area The shortcomings of the SPT for use in liquefaction analyses have been discussed at length in the literature, especially in connection with energy standardisation Friction ratio f R s (by Seed et al., 1985; Liao and Whitman, 1985, for f = q example). Apart from the problem of standardising c energy input, the test has two major difficulties when employed in loose sands and silts. The first arises from Pore pressure parameter the discrete nature of the blow count. For a perfectly Sleeve u − u B = o executed test yielding a blow count of 5, for example, q q s c − z o the resolution of the discrete scale is no better than ± 10%. With N then raised to the power 2, as in equation (6) for example, an even greater uncertainty is introduced. The second objection comes from the difficulty in obtaining a clean drill hole, without disturbing the material in the test region at the bottom of the boring. It has been the writer’s experience that even with careful Pore rotary boring with mud, it is often difficult to avoid pressure disturbance at the bottom of the hole. Even experi- filter enced drillers operating under research conditions have had difficulty in very loose sands. The writer is quite Cone sceptical of N-values of much less than 6. 35.7 mm These considerations, together with the problem of energy standardisation, have led us at Canterbury Uni- versity to adopt the cone penetration test (CPT) in our studies of liquefaction sites in New Zealand. Not only does the CPT offer more precision, but also it is Figure 3.10: Cone penetrometer, showing repeatable and gives a continuous measurement of soil instrumented conical point and sleeve and resistance rather than the discrete measurements of the pressure transducer 56 • Risks and Realities

40 expressed as a friction ratio, Rf = fs/qc. It serves as an Sands indicator of soil type. Cohesionless soils have a small 20

friction ratio of one or two percent or less; silts and Silty sands clays have friction ratios in the vicinity or exceeding 10

) 8 Sandy silts

five percent, depending on the value of q . Several 2 and silts c 6 interpretation charts have been proposed (Meigh, 1987). 4

(MN/m Clayey silts

One simple one due to Robertson and Campanella is c and silty clays shown in Figure 3.11. 2

Clays Some penetrometers also have a pore pressure trans- 1 ducer, usually placed between the cone and the friction 0.8 0.6 Peat

sleeve. These cones are called piezocones and the test Cone resistance, q 0.4 denoted by CPTU rather than CPT. At the filter (Figure

3.10), the pore pressure u deviates from its hydrostatic 0.2

value uo depending on the type of soil, its permeability and its density. Interpretation of excess pore pressure 0.1 01 2 3456 Δu = u - u is not yet perfected, but at very least it gives o Friction ratio, R (%) a finer definition of strata interfaces than the other two f

quantities (qc or fs). However, it is difficult to obtain Figure 3.11: Soil identification chart (after good hydraulic contact between the ground water and Robertson and Campanella, 1985) the transducer; the necessary de-airing of the filter circuit and maintaining saturation above the water table is not simple, and adds considerably to the cost of Cone Penetration Methods the test. For everyday use at the time of writing (1997) Some procedures originally based on SPT have been the added complication of the CPTU is generally not converted to CPT using the qc-N relationship. Other warranted. procedures have been formulated directly in terms of

qc, using CPT field data. These include the method of Returning to the liquefaction problem, we note that Shibata and Teparaska (1988) and the well-known since there is a well established relationship between Chinese method (Zhou, 1980). These procedures

them, CPT cone resistance values, qc , can be converted benefit from the greater precision and sensitivity of the to SPT N-values for use in SPT-based procedures such CPT, but they do not fully exploit its potential to

as the two described above. The qc-N correlation of estimate grain size and, in the case of the piezocone test Robertson & Campanella (1985), shown in Figure (CPTU), drainage conditions. 3.12, is widely used for this purpose.

qc, bars; N, blows/foot (1 bar = 100 kPa)

CLAYEY SILTS SANDY SILT CLAY & SILTY CLAY & SILT SILTY SAND SAND 10 4 9

9 8 16 7 13

6 14 /N 5

c 9 12 8 5 3 13 111 10 2 9 7 15 4 12

Ratio, q 5 9 4 12 3 7 13 4 6 11 12 9 5 2 10 8 11 1

0 0.001 0.01 0.1 1.0

Mean Grain Size, D50, mm

Figure 3.12: Variation of qc/N with mean grain size (from Robertson and Campanella, 1985) Seismic Liquefaction and Lifelines • 57

In an attempt to make better use of the diagnostic The analysis was a simple one, based on CPT testing to capability of the piezocone, Dou and Berrill (1991, characterise the site soils, and three published proce- 1993) have employed the pattern recognition tech- dures for estimating liquefaction potential. A typical nique from information theory, together with case CPT log is shown in Figure 3.14a, together with plots history data, to develop a procedure for estimating the of threshold values of cone resistance qc for liquefac- probability of liquefaction, using all three CPTU meas- tion under the scenario earthquake (McCahon et al., urements. So far, the procedure has been implemented 1997) according to the three procedures (Zhou, 1980; using a fast mainframe computer and a limited set of Shibata and Teparaska, 1988; Davis and Berrill, 1982). calibration data, to demonstrate its viability. With Note the divergence between the results of the different Pentium computers and the larger set of piezocone data procedures, illustrating the uncertainty associated with now available, it should be possible to develop a these simple models of a complex phenomenon. Nev- practical procedure for use in engineering practice. ertheless, taken together they give a clear indication of whether or not there is likely to be a problem at a particular site.

3.4 Christchurch Lifelines In a study of various liquefaction models, Vreugdenhil Study (1995) found that the criterion of Davis and Berrill is the most likely to correctly predict liquefaction or not, As part of the Christchurch Engineering Lifelines whereas the other two are more conservative. There- Study, estimates of liquefaction potential were made at fore, the following interpretation was adopted, illus- key sites (Figure 3.13) in the Christchurch lifelines trated in Figure 3.14b. If the cone resistance of the soil networks. The 16 sites examined represented principal falls below all three criteria, indicated by single cross- nodes in the networks, such as water and sewage hatching in Figure 3.14b, the layer is deemed to be pumping stations, telephone exchanges and electrical liquefiable in the scenario earthquake; if it fails the two substations. Of the 16 sites, it was found that 12 might conservative criteria but not the third, it is deemed to be liquefy in the scenario earthquake.

KEY

SHAKING ZONES

N ZONE 1 Bedrock at shallow depth WINTERS ROAD

ZONE 2 Sediments less than 50m deep HAREWOOD ROAD C TRAVIS ROAD R A N F O ZONE 3 R Sediments 50-800m depth D S D T A R O E R E S D E LIQUEFACTION ZONES UD T N LE IN Y C REEK ZONE 2A Predominantly sands 2-10m depth

ZONE 2B Predominantly silts and sandy silts 2-5m depth

ZONE 3A Predominantly sands 2-10m depth P A P A ZONE 3B Predominantly silts and sandy N AD silts 2-5m depth U O I R R ES O G A PA D CPT probes: likely to liquefy

CPT probes: non-liquefiable

TON RD ICCAR AVON R CPT probes: liquefaction uncertain RIVER D A O R S R AD E HEIM RO Y BLEN D LINW OOD AVE

F ER RY RO ESTUARY AD

D R L L E W S 012 L A H H EA TH CO TE

Figure 3.13: Plan of Christchurch, showing the 16 key sites studied for liquefaction potential, as part of the Engineering Lifelines Project. Note that 12 of the 16 sites might liquefy under the scenario earthquake 58 • Risks and Realities

at some risk but not necessarily liquefiable under the 3.5 Sustained Shear Stress present scenario earthquake. In the case of level-ground liquefaction discussed Note that the lines in Figure 3.14a showing the three above, shear strength is not required for static equilib- liquefaction criteria are not continuous. If the combi- rium. But in most other cases (retaining-wall backfill

nation of cone resistance and friction ratio, Rf, in a is a possible exception), the soil mass must resist given stratum show that the soil is too fine to liquefy sustained shear stresses to remain in equilibrium. Slopes according to the criterion of Robertson and Campanella and foundations are two common examples. (1985), then the line is not plotted in that layer. Thus, The behaviour of the soil mass depends, in the large, on even though the measured qc falls below the criteria, the soil at that depth is deemed not susceptible to liquefac- whether it is loose and contractive or dense and dila- tion. tive. If the soil is dense, then any perturbation by the earthquake will cause it to dilate, thereby increasing its Although the study was a fairly broad-brush one, both undrained, or short term, strength. A typical static in the density of sites and in the sophistication of the undrained stress strain curve for a dense sand is shown analysis, for a modest cost it gave a good general in Figure 3.15. At large strains, a steady state of picture of the city-wide liquefaction hazard, of regions deformation is reached at which shear strain continues (and geologic formations) in which further investiga- at a constant shear stress. This undrained steady state tions should be made, as well as showing up hazard to shear strength has been denoted by SUS. A characteris- particular sites. tic of a dense granular soil is a large value of SUS. On the other hand, loose soils have static, undrained stress- The study has prompted utility owners to commission strain curves typically like that illustrated in Figure a number of more detailed investigations of both the 3.16. In this case the curve drops off with increasing hazard and possible strengthening works. Further strain to a relatively small value of S . information about this study may be found in the report US by Guilhem and Berrill (1993). Clearly, a slope or foundation soil composed of dense material should remain stable under seismic loading. On the other hand, the stability of a loose soil mass depends, to a first analysis, on whether or not the steady

Knights Drain Flood Gate — Waitaki Street KNF001.CPT, 06-17-1993

Low value of Rf indicates clean sand (candidate for liquefaction if qc small enough) 0 5 Friction ratio, Rf 10

High value of Rf indicates cohesive fines: little risk of liquefaction 25 even though Shibata & Teparaska qc is low 20 15

10 Cone

(MPa) Rf (%) resistance, qc

c 5 q 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Zhou's criterion Depth (m) Davis & Berrill criterion Low density (low qc) and low R indicating loose f qc below two of three Although q is low, sandy material: q below c c criteria: reasonably combination of q and R all three criteria: liquefaction high likelihood of c f highly likely indicates high fines liquefaction content; liquefaction unlikely

Figure 3.14a: Cone Penetrometer (CPT) log from a typical lifelines site in the Christchurch Study, with the three criteria used in estimating whether or not the site would liquefy in the scenario earthquake Seismic Liquefaction and Lifelines • 59

Knights Drain Flood Gate KNF001.CPT, 06-17-1993 qc (MPa) 0 5 10 15 20 0

2 Highly likely Too cohesive to liquefy to liquefy despite low density 4

Moderate likelihood of liquefaction 6 in scenario earthquake

8 Depth (m) 10

Low friction ratio Rf and moderate 12 density (qc): Likely to liquefy in stronger earthquake

16 50 Rf (%)

Figure 3.14b: Interpretation of liquefaction hazard. Single cross-hatching indicates layers highly likely to liquefy in scenario earthquake; double hatching indicates layer might liquefy

state strength exceeds the static driving stress, τs. there are two consecutive stages. The first comprises These two cases are now examined in more detail. the build up of pore pressure, depending mainly on the

amplitude and duration of cyclic shear strain γc, in- Liquefaction Flow Failures duced by the earthquake. This corresponds to the period of loading up to the peak in Figure 3.16a or the Dobry et al. (1984) note that in liquefaction failures cyclic part in Figure 3.16b. The second stage com- involving flow of material and large displacements,

S Before us Sus cycling During cycling

{ Monotonic loading after Collapse

cycling Sus Sus SHEAR STRESS SHEAR STRESS

After {

reaching σ' = 0 SHEAR STRESS SHEAR STRESS condition During cycling caused by cyclic load STRAIN STRAIN

STRAIN STRAIN (a) (b) (a) (b) Figure 3.16: Stress-strain behaviour for Figure 3.15: Stress-strain behaviour for undrained loading of loose sand. From Whitman undrained loading of dense sand From (1987) Whitman (1987) 60 • Risks and Realities

prises flow driven by the static shear stresses τs, and still be large enough to impair the function of the

proceeds only if τs > SUS. Thus to analyse the stability structure. Whitman (1987) terms such occurrences of a slope or foundation against flow failure, we need deformation failures, and states that their analysis first to check the static stability, using appropriate presents “one of the present-day frontiers of soil dy-

values of SUS in regions of contractive soil. If the namics”. Whitman discusses some computational pro- structure is stable under these conditions, the analysis cedures that have not yet become part of everyday can stop there. However, if it is not, then it is necessary engineering practice. He emphasises the need to test to check whether the build up of pore pressure which computation against experiment, and suggests the use triggers the strength reduction, will indeed occur under of centrifuge model tests because of the infrequency of the design earthquake. Dobry et al. present this ap- earthquakes for full scale tests. Use of the Newmark proach for earth dams, and give details of how it might (1965) sliding block analogy has been suggested for be applied to the dam problem. Their procedure the calculation of limited, permanent displacements. includes a novel laboratory test in which a torsional This suggestion has been taken up by, among others, cyclic shear stress is applied to an undrained triaxial Baziar, Dobry and Alemi (1992) who study lateral specimen which has been consolidated and is main- spreading at the Wildlife, California site in the 1987 tained under an anisotropic stress system representing earthquake, and by Byrne, Jitno and Salgado (1992) the static in situ stress state, simulating the two aspects who apply it to the upper San Fernando dam in the 1971 of the problem. However, this general approach could earthquake. The success of both modellings suggest equally well be applied to the seismic stability of that this is a fruitful approach. shallow foundations and to static liquefaction flow failures such as the Nice Airport failure (Schlosser et Intermediate Cases al., 1985). We have considered the two extreme cases where the While the approach is simple in concept, the determi- soils were either clearly contractive or clearly dilative, and completely undrained. Partial drainage could lead nation of in situ values of SUS is far from trivial. to a reduction in S in a dilating soil. For a soil mass Laboratory determination of SUS depends on the very US difficult task of obtaining “undisturbed” samples. that is in equilibrium before the earthquake, partial Poulos et al. (1985) describe a procedure they have drainage should pose a problem only during shaking

developed over the years. It involves undrained tests and then only if SUS drops below the sum of the seismic on both field and reconstituted samples, with an allow- and static shear stresses. Here, because the seismic ance for sample disturbance. Their method is based on component of shear stress is cyclic, displacements should remain limited. the premise that for a given soil, SUS depends uniquely on void ratio, e. But Vaid et al. (1989), for example, Whitman (1985, 1987) points out two other subtle find that preshearing has a marked effect on the dilatancy variations to the simple cases. The first occurs when a behaviour of sand and thus on S and Konrad et al. US cohesionless soil remains globally undrained, but un- (1991) question the uniqueness of the steady-state line. dergoes local changes in void ratio which cause a loss Seed (1987) and Seed et al. (1988) have presented in strength and thus, possibly, a flow failure. The second concerns high excess pore pressures generated correlations between corrected SPT values, (N1)60, and in a non-critical region, which lead to a critical loss of values of SUS obtained by back analyses of flow failures. However, Finn (1993) points out that there are strength when they diffuse into a more sensitive region. difficulties with this approach, and it suffers from the This could explain several delayed flow failures that general shortcomings of the SPT, mentioned above. have been observed. A local example of delayed failure Because flow failures usually involve quite loose and can again be found at the Landing Road Bridge, where often silty materials, the CPT should be the more the 4 to 5 m high west approach embankment failed appropriate in situ test. Ishihara et al. (1990) present about 15 minutes after the earthquake, presumably as a front of high pore pressure propagated slowly up- correlations between qc and SUS obtained by back analyses of flow failures in Japan. wards from the liquefied soil into the unliquefied crust of silty sand on which the embankment was founded. Deformation Failures In a dilating (dense) rather than contractive (loose) soil, permanent displacements may occur during momen- 3.6 Counter Measures tary strength reductions, as discussed in the introduc- Countermeasures against liquefaction fall into two tion, but these are intermittent, of limited magnitude, classes: and cease when the shaking stops. While permanent displacements in dilating soils are limited, they may a) Those which strengthen the soil. Seismic Liquefaction and Lifelines • 61

b) Those which modify the structural design to cope lifelines engineers, as should the work of Tokida and with effects of liquefaction. his colleagues on drag loads imposed on piles (Tokida et al., 1993). Further case histories of lifeline damage Because of shortage of land and high seismic activity have been assembled by Hameda and O’Rourke (1992) in Japan, Japanese engineers are advanced in the devel- and O’Rourke and Hameda (1992), Brunsden (1996) opment of countermeasures, and the reader is referred and Keenan et al. (1995). These are recommended for to an excellent text by staff of the Port and Harbour the additional insights they provide to the behaviour of Research Institute which has recently been published engineering lifelines. in English (PHRI, 1997). Figure 3.17 from this work summarizes the various approaches that may be taken. The main points made in this review may be summarised as follows:

1 Damage due to liquefaction effects has been exten- 3.7 Conclusion sive and costly in past earthquakes. Lifelines are The aim of this paper has been to give a broad view of particularly susceptible, with pipelines buried in the liquefaction problem without excessive detail. laterally-spreading soil most probably the principal Therefore many significant results have been omitted. victim. I should have liked to discuss results from centrifuge 2 Fine sands and coarse silts are the most susceptible tests, such as those of Lambe and Whitman (1985), soils to liquefaction. From a geological viewpoint, Hushmand, Scott and Crouse (1988) and Lin and laterally-accreted late Holocene deposits are par- Dobry (1992). Centrifuge tests on shallow foundations ticularly liable to liquefy. show that smaller excess pore pressures tend to develop beneath foundations than in the free field. This 3 Liquefaction problems can be separated into two implies that if a reliable level-ground analysis shows classes: Those where static shear stresses must be there should be no significant pore pressure increase in sustained, as in slopes or foundations, and those the free field, then any foundations should also be safe involving level ground, where static equilibrium from liquefaction. does not require any shear strength.

The empirical study of lateral spreading distances by 4 Procedures for predicting the liquefaction potential Bartlett and Youd (1992) should be of interest to

LIQUEFACTION REMEDIATION

Soil Structural Improvement Design

Improve the soil Achieve rapid Maintain Relieve external so that the soil dissipation of stability grain skeleton will forces by softening excess pore by reinforcing or modifying not collapse under water pressure structure earthquake loading structure

EXAMPLES EXAMPLES EXAMPLES • install drains • strengthen pile • adjustment of • replace sand with foundation (increased bulk unit weight gravel number and thickness of buried structures of piles; install • reduce slope angles bracing members • reinforcement of soil Increase the Reduce • prevent deformation liquefaction earthquake- with sheet piles or strength of induced shear underground wall the soil stress ratio

EXAMPLES EXAMPLES • compaction • adjustment of • consolidation bulk unit weight • preloading of buried structures • replacement (refilling • reduce slope angles with material which will not liquefy)

Figure 3.17: Various basic strategies for liquefaction remediation (from PHRI, 1997) 62 • Risks and Realities

of level-ground sites are well-established and a Bazier, M.H., R. Dobry and M. Alemi (1992). “Evalu- number of methods have been presented here. ation of Lateral Ground Deformation Using Sliding Block Model”, Proc 10th World Conf. 5 In cases where shear stresses must be resisted for Earthq. Eng, Madrid, Vol. 3, pp 1401-1406. static equilibrium, behaviour after initial liquefaction depends on whether the residual strength is Berrill, J.B., J. Canou, P. Foray and J.-L Pautre, (1992). sufficient to resist the driving stress. This in turn “Piezocone Testing of Liquefaction Sites: depends principally on whether the soil is in a dense Normalization of Excess Pore Pressure”, Proc. or a loose state. If the soil is loose, then a flow 10th World Conf. Earthq. Eng, Madrid, Vol. 3, failure, with large displacements, is possible. If it pp 1421-1424. is dense, then deformation should be limited, but may still be damaging. Berrill, J. B. and S. A. Christensen (1995). “The Effect of Lateral Spreading on the Landing Road Bridge 6 Determination of the residual or steady state shear in the 1987 Edgecumbe, New Zealand Earth- strength is a very difficult problem because of quake”, Proc 7th Canadian Conf. Earthq. Eng., sample disturbance. Simple methods for measur- Montreal, pp 139-146.

ing SUS have not yet been found. For rough, preliminary assessments, in situ test methods may be used. Bienvenu Véronique (1988). Studies of Liquefaction in 1929 Murchison and Inangahua (1968) New 7 The Newmark sliding-block approach appears Zealand Earthquakes, Master of Engineering promising for estimating displacements in limited- Thesis, University of Canterbury, Christchurch, deformation problems. New Zealand, March 1988.

8 Between the two extreme cases of liquefaction Brunsdon D. R. et al. (1996) “Lessons for New Zealand flow failures and deformation failures, there are Lifelines Organisation from the 17 January many intermediate cases influenced by secondary 1995. Great Hanshin Earthquake”, Bull. NZ effects such as partial drainage, diffusion of excess Nat. Soc. Earthq. Eng., Vol. 29, pp 1-55. pore pressure and redistribution of void ratio. Byrne, P.M., H. Jitno and F. Salgado (1992). “Earth- Great progress has been made in our understanding of quake Induced Displacements of Soil-Struc- liquefaction phenomena, in understanding the me- ture Systems” Proc. 10th World Conf. Earthq. chanics of the various aspects of the general problem Eng, Madrid, Vol. 3, pp 1407-1412. and in formulating analysis procedures. However, further work is required, for example, in finding more Canou J., (1989). Contribution à l’étude et à l’évaluation robust methods for determining steady-state strength. des propriétés de liquéfaction d’un sable, Finally, let us note the importance of case histories, Mémoire de thèse de Doctorat de l’ENPC, which have played a central part in the past. They will Paris, 338 pp. continue to be important, with emphasis directed more Castro, G. (1975) “Liquefaction and Cyclic Mobility to flow and deformation failures than to level ground of Saturated Sands”, Journal of the Geotechnical cases. Engineering Division, ASCE, Vol. 101, GT6, pp 551-568. Acknowledgements Castro, G., S.J. Poulos and F.D. Leathers (1985), “A Re-examination of the Slide of the Lower San The author wishes to acknowledge and thank the many Fernando Dam”, Journal of Geotechnical En- colleagues and former students who have stimulated gineering Division, ASCE, Vol. 111, GT9, pp his interest in liquefaction and improved his imperfect 1093-1107. understanding of it. Special mention should be given to Dr R O Davis and Professor Pierre Foray. Davis, R.O. and J.B. Berrill (1982) “Energy Dissipa- tion and Seismic Liquefaction in Sands”, Earth- quake Engineering and Structural Dynamics, 3.8 References Vol. 10, pp 59-68. Bartlett, S.F. and T.L. Youd (1992). “Empirical Pre- Davis, R.O. and J.B. Berrill (1996) “Liquefaction diction of lateral Spread Displacements”, Proc Susceptibility Based on Dissipated Energy: a 4th Japan-US Workshop on Earthquake Resist- Consistent Design Methodology”, Bulletin. NZ ant Design of Lifeline Facilities, Honolulu, pp. Nat. Soc. Earthq. Eng. Vol 29, pp. 83-91. 351-366. Seismic Liquefaction and Lifelines • 63

Dobry, R., R.S. Ladd, F.Y. Yokel, R.M. Chung and D. Husmand, B., R.F. Scott and C.B. Crouse (1988) Powell (1982). Prediction of Pore Water Pres- “Centrifuge Liquefaction Tests in a Laminar sure Buildup and Liquefaction of Sands During Box”, Geotechnique, Vol. 38, pp 253-262. Earthquakes by the Cyclic Strain Methods, Ishihara, K. (1985) “Stability of Natural Deposits Building Science Series 138, National Bureau During Earthquakes”, Proc. of the Eleventh Int. of Standards, US Department of Commerce, Conf. on Soil Mechanics and Foundation Eng., US Government Printing Office, Washington, Vol. 1, pp 321-376. DC. Ishihara, K. (1993) “Liquefaction and Flow Failure Dobry, R., R. Modhaman, P. Dakoulas, and G. Gazetas during Earthquakes”, Géotechnique, Vol 43, pp (1984) “Liquefaction Evaluation of Earth Dams 351-415. - A New Approach”, Proc of the Eighth World Conf. in Earthq. Eng, Vol. 3, pp 333-348. Ishihara, K., S. Yasuda and Y. Yoshida (1990) “Lique- faction-induced flow failure of embankments Dobry, R., D.J. Powell, F.Y. Yokel and R.S. Ladd and residual strength of silty sands,” Soils and (1980), “Liquefaction Potential of Saturated Foundations, Vol. 30, No. 3 pp 69-80. Sand - The Stiffness Method”, Proceedings 7th World Conf. on Earthq. Eng, Istanbul, Turkey, Iwasaki, T. (1986) “Soil Liquefaction Studies in Japan: Vol. 3, pp 25-32. State-of-the-Art”, Soil Dyn. Earthq. Eng, Vol. 5, pp 2-68. Dou, Yiqiang and J.B. Berrill (1991) “Evaluation of Liquefaction Potential by Pattern Recognition”, Keenan, R. P., J. B. Berrill and J. R. Pettinga (1995) Proc Pacific Conf. Earthq. Eng, Auckland, Vol. “Foundation Loads due to Lateral Spreading at 1, pp 301-311. the Landing Road Bridge, Whakatane”, Proc. Ann. Conf. NZNSEE, New Plymouth, pp 169- Dou Yiqiang and J.B. Berrill (1993) “A Pattern Recog- 176. nition Approach to Evaluation of Soil Lique- faction Potential Using Piezocone Data”, Soil Kuerbis, R.H. and Y.P. Vaid (1989) “Undrained Be- Dyn. Earthq. Eng, (in press). haviour of Clean and Silty Sand”, Proc. 12th Int. Conf. on Soil Mech. Found. Eng, Rio de Finn, W.D. Liam (1993) “Evaluation of Liquefaction Janeiro, Brazil, August. Potential”, Proc. Seminar on Soil Dyn. and Geotech. Earthq. Eng, Lisbon, Balkena, pp Konrad, J.-M., E. Flavigny and M. Meghachou (1991), 127-158. “Comportement non drainé du sable d’Houstun lâche”, Rev. Franç. Géotech. No. 54, pp 53-63. Florin, V.A. and P.L. Ivanov (1961) “Liquefaction of Saturated Sandy Soils”, Proc. 5th Int. Conf. Lambe, P.C. and R.V. Whitman (1981). Dynamic Soil Mech. Found. Eng, Paris, Vol. 1, pp 107- Centrifuge Modelling of a Horizontal Sand Stratum, Research Report R82-14, Dept. Civil 111. Eng, Mass. Institute of Technology. Guilhem, O. and J. B. Berrill (1993). Cone Penetrometer Lambe, P.C. and R.V. Whitman (1985) “Dynamic Results and Estimates of Liquefaction Potential Centrifugal Modeling of a Horizonal Dry Sand at some key Christchurch Lifelines Sites, Can- Layer”, Journal of Geotechnical Engineering, terprise Report CP/2245, University of Canter- ASCE, Vol. 111, 3, pp 265-287. bury, Christchurch, New Zealand, 53 pp. Liao, S.C. and R.V. Whitman (1985) “Overburden Gutenberg, B. and C.F. Richter (1954). Seismicity of Correction Factors for SPT in Sand”, J. Geotech. the Earth, Princeton Univ. Press. Eng, ASCE, Vol. 112, No. 3, pp 373-377.

Hameda, M., and T. O’Rourke eds (1992) Case Studies Liao, S.C., D. Veneziano and R.V. Whitman (1988) of Liquefaction and Lifeline Performance Dur- “Regression models for Evaluating Liquefac- ing Past Earthquakes, (Tech. Report NCEER- tion Probability”, J. Geotech. Eng, ASCE, Vol. 92-0001 and 0002) NCEER, Buffalo, NY, Vols 114, pp 389-411. 1 & 2, Lin, L. and R. Dobry (1992), “Centrifuge Study of Hardin, B.O. (1965) “The Nature of Damping in Sands”, Shallow Foundations on Saturated Sand Dur- J. Soil Mech. Found. Div. ASCE, Vol. 91, SM1, ing Earthquakes”, Proc. 4th Japan-US Work- pp 63-97. shop on Earthq. Res. Design of Lifeline Facili- ties, Honolulu, pp 493-508. 64 • Risks and Realities

Long, E.L. (1973) “Earth Slides and Related Phenom- Schlosser, F. (1985) “Liquefaction de Veines de Sable ena”, The Great Alaska Earthquake of 1964: Lâche dans des Talus Sous-Marins”, Proc. 11th Engineering, National Academy of Science, Int. Conf. Soil Mech. Found. Eng, San Fran- Washington, D. C., pp 644-773. cisco, Vol 3, pp 1713-1716.

Luong, M.P. (1980) “Phénomènes Cycliques dans les Scott, R.F. and K.A. Zuckerman (1973) “Sand Blows Sols Pulvérulents”, Rev. Française and Liquefaction,” The Great Alaskan Earth- Géotechnique, No. 10, pp 39-53. quake of 1964 - Engineering Volume, Commit- tee on the Alaska Earthquake, Division of Earth McCahon et al. (1997). “Hazards to Engineering Life- Sciences, National Research Council, National lines in Christchurch”, In Risks and Realities: A Academy of Sciences, Washington, D.C., pp Multi-disciplinary Approach to the Vulnerabil- 179-189. ity of Lifelines to Natural Hazards, Centre for Advanced Engineering, University of Canter- Seed, H.B. (1987), “Design Problems in Soil Liquefac- bury. tion,” J. Geotech. Eng, ASCE, Vol. 113, No. 7, Aug. pp 827-845. Meigh, A.C. (1987). Cone Penetration Testing, Butterworths, London, 141 pp. Seed, H.B. and I.M. Idriss (1971), “Simplified Proce- dure for Evaluating Soil Liquefaction Poten- Nemat-Nasser, S. and A. Shokooh (1979) “A Unified tial,” J. Soil Mech. Found. Eng Div., ASCE, Approach to Densification and Liquefaction of Vol. 97, SM9, pp 1249-1273. Cohesionless Sand in Cyclic Shearing”, Can. Geotech. J. Vol. 16, No. 3, pp 659-678. Seed, H.B. and K.L. Lee (1966), “Liquefaction of Saturated Sands During Cyclic Loading”, J. Newmark, N.M. (1965) “Effects of Earthquakes on Soil Mech. Found. Eng Div., ASCE, Vol. 92, Dams and Embankments”. Fifth Rankine Lec- SM6, pp 105-134. ture, Geotechnique, London, Vol. 15, No. 2, pp 139-159. Seed, H.B., R.B. Seed, L.F. Harder and H.-L. Jong (1988). Re-evaluation of the Slide in the Lower O’Rourke, T.D. and J.W. Pease (1992). Large Ground San Fernando Dam in the Earthquake of Feb- Deformations and their Effects on Lifeline Fa- ruary 9, 1971, Report No. UCB/EERC-88/04, cilities : 1989 Loma Prieta Earthquake, Tech- University of California, Berkeley. nical Report NCEER-92-0002 Nat. Center for Earthq. Eng, Buffalo N.Y., Vol. 2, pp 5-1 to 5- Seed, H. Bolton, K. Tokimatsu, L.F. Harder and R.M. 85. Chung (1985), “Influence of SPT Procedures in Soil Liquefaction Resistance Evaluations”, J. Peck, R.B. (1979) “Liquefaction Potential: Science Geotech. Eng, ASCE, Vol. 3, No. 12, Dec. Versus Practice”, Journal of Geotechnical En- gineering, ASCE, Vol. 105, GT3, pp 393-398. Shibata, T., and W. Teparaksa (1988), “Evaluation of Liquefaction Potentials of Soils Using Cone PHRI (1997) Handbook on Liquefaction Remediation Penetration Tests.” Soils and Foundations, Vol. of Reclaimed Land, Balkema, 312 pp. 28, pp 49-60. Poulos, S.J., G. Castro and J.W. France (1985) “Lique- faction Evaluation Procedures”, Journal of Tinsley, J.C. and W.R. Dupré (1992), “Liquefaction Geotechnical Engineering, ASCE, Vol. 111, 6, Hazard Mapping, Depositional Faces and Lat- pp 772-791. eral Spreading Ground Failure in the Monterey Bay Area, Central California, during the 10.7.89 Robertson, P.K. and R.G. Campanella (1985) “Lique- Loma Prieta Earthquake”, Proc. 4th Japan-US faction Potential of Sands Using the CPT”, J. Workshop on Earthq. Res. Design of Lifeline Geotech. Eng, ASCE, Vol. 109, GT11, pp 1449- Facilities, Honolulu, pp 71-86. 1459. Tokida, K., H. Iwasaki and T. Hameda (1993), “Lique- Ross, G.A., H.B. Seed and R.R. Migliaccio (1973) faction Potential and Drag Force Acting on “Performance of Highway Bridge Foundations”, Piles in Flowing Soil”, Proc. 6th Int. Conf. Soil The Great Alaska Earthquake of 1964: Engi- Dyn. Earthq. Eng, Bath, U.K. (in press). neering, Nat. Acad. Sci., Washington DC, pp 190-242. Seismic Liquefaction and Lifelines • 65

Troncoso, J.H. (1990), “Failure Risks of Abandoned Vreugdenhil, R.A., R.O. Davis and J.B. Berrill (1995). Tailings Dams”, Proc. Int. Symp. on Safety and “Liquefaction Potential and Piezocone Re- Rehabilitation of Tailings Dams, ICOLD, Syd- sponse”, Bull. NZ Nat. Soc. Earthq. Eng., Vol. ney, Australia, May, pp 82-89. 28, pp 106-112.

Tsuchida, H. (1970), “Prediction and Countermeasure Whitman, R.V. (1985), “On Liquefaction,” Proc. 11th Against the Liquefaction in Sand Deposits,” Intl Conf. Soil Mech. Found. Eng, Vol. 4, pp Abstract of the Seminar in the Port and Harbor 1923-1926. Research Institute, pp 3.1-3.33 (in Japanese). Whitman, R.V. (1987) “Liquefaction - the State of the Tsuchida, H. and S. Hayashi (1971), “Estimation of Art”, Bull. N.Z. Nat. Soc. Earthq. Eng, Vol. 20, Liquefaction Potential of Sandy Soil”, Proc. pp 145-158. 3rd Meeting, U.S. Japan Panel on Wind and Yamada, G. (1966) “Damage by Earth Structures and Seismic Effects, Tokyo. Foundations by the Niigata Earthquake, June Vaid, Y.P. and J.C. Chern (1983) “Effect of Static 16 1964”, Soils and Foundations, Vol 6, pp 1- Shear on Resistance to Liquefaction”, Soil and 13. Foundations, Vol. 23, No. 1, pp 47-60. Youd, T.L. (1984) “Geologic Effects - Liquefaction Vaid, Y.P., E.K.F. Chung and R.H. Kuerbis (1989). and Associated Ground Failure”, pp 210-232 in “Preshearing and Undrained Response of Sand”, Proceedings of the Geologic and Hydrologic Soils and Foundations, Vol. 29, No. 4, pp 49- Hazards Training Program. Open-File Report 61. 84-760, US Geological Survey, Menlo Park, California. Vreugdenhil, R.A. (1995). Interpretation of Piezocone Data and its use in Estimating Liquefaction Zhou, S. (1980). “Evaluation of the Liquefaction of Potential, Research Report 95-7, University of Sand by Static Cone Penetration Test.” Proc. Canterbury, 184 pp. 7th World Conf. Earthq. Eng, Istanbul, Vol. 3, pp 156-162. 66 • Risks and Realities

3.9 Summary of Liquefaction Effects These diagrams are taken from the 1993 report of the Wellington Lifelines Group and are reproduced with permission as a graphic illustration of the likely effects on engineering lifelines.

Post Liquefaction Ground Surface

Eruptions of Sand/Water at Irregular Spacing

Original G.L.

LIQUEFIED MATERIAL

Pipe Lowered

(a) Level field liquefaction

Ground Fissures Post Liquefaction Ground Surface

LIQUEFIED MATERIAL

Pipe lowered and laterally displaced

(b) Lateral spreading due to liquefaction

Pipe raised and displaced

Structure Tilts and Sinks Post liquefaction Ground Surface

LIQUEFIED MATERIAL

Figure 3.18: Ground movement due to liquefaction Seismic Liquefaction and Lifelines • 67

E Span Drops

Pier Rotates

LIQUEFIED SOIL (Provides no support)

Moment capacity Failure

(a) Deep foundation failure

Chamber 'floats' in Liquefied Zone breaking Connections

LIQUEFIED SOIL

(b) Flotation of buried structures

’Full Conduit Less Effected’

Above Liquefied Zone - ’Empty Conduit Floats’ Follows Ground Settlement

Settlement

DENSE SAND

LIQUEFIED SOIL LOOSE SAND

Figure 3.19: Structural displacements

68 • Risks and Realities

Post Liquefaction Toe Post Liquefaction

Original Toe

Lateral Spreading

Ground Surface after

Post Liquefaction Crest

Post Liquefactio Position of Pipe

Original Crest

Plan

Cross-section A Cross-section A - A

LIQUEFIED MATERIAL

Ruptures in Pipe

GROUND CRACKS A

Figure 3.20: Effect of lateral spreading on buried pipes Seismic Liquefaction and Lifelines • 69

FILL FILL FILL

LIQUEFIED MATERIAL

Failure Surface

(a) Shallow lateral failure

FILL

NON LIQUEFIABLE MATERIAL

LIQUEFIED MATERIAL

Failure Surface

(b) Deep seated lateral failure

LIQUEFIED MATERIAL

Failure Surface

Figure 3.21: Embankment Stability 70 • Risks and Realities Civil Services • 71

Chapter 4 Civil Services

4.1 Introduction this activity the manager will need to place before decision makers a sensible compromise that balances The Civil Services Group has concentrated on the land the cost of mitigation against the risk of the event and drainage, sewerage and water supply systems serving the cost of damage that the mitigation work is designed the Christchurch area, including some comment on to reduce or remove. Another balance must be estab- Lyttelton services. Also included in the brief was fuel lished between spending now compared with the cost supplies and a short report on this aspect is included, of putting extra effort into the response. It may, for although the view of the industry is that in the event of example, be wiser to plan for additional crews to repair damage to Christchurch installations a number of alter- breaks than to spend now on renewing vulnerable native means of bringing fuel to the city will be pipelines. available. Because most of a utility manager’s assets are buried, While flooding presents the most frequently occurring the opportunity for mitigating pipeline damage by emergency for Christchurch’s civil service lifelines, retrofitting is limited. Instead, the focus must be on flood damage to facilities and underground services is new work, including renewal, where material and joint localised and minor. Likewise, the effects of extreme choice, design and location can have a significant snow and wind events cannot be shown to pose signifi- influence. Fortunately, factors that improve a pipe- cant threats. Only the seismic hazard carries the poten- line’s performance during an earthquake are well es- tial for significant widespread disruption and damage tablished, involving ductility, flexibility and the ability to the city’s civil services. to accommodate tension, compression, shear and rota- For this study, an earthquake has been defined having tional movement at structure entries. Materials and a return period of 150 years and a shaking intensity jointing systems are available that provide these fea- lying between MM VIII and IX. Coupled with the tures. Mitigation at design stage is also essential for liquefaction potential of much of the city’s eastern Christchurch’s civil service facilities — reservoirs, districts and the magnification effects that can be pump stations, treatment works, retention basins, expected, this earthquake has the potential to cause stopbanks, tidal control structures, etc. Awareness of significant damage to pipelines and facilities, particu- liquefaction potential requires that more attention is larly where design or detailing has not included present given to foundation design, particularly when recent day seismic requirements and considerations. It is fair earthquakes of moderate size are providing evidence of to say that Christchurch is at significant risk. ground accelerations in excess of current code requirements. An exercise in Christchurch that assessed lique- The lifelines project reflects a belief that much can be faction potential at 16 important lifeline facilities (in- done in advance to mitigate or eliminate likely damage cluding nine in the civil services area) using cone and also to prepare for the event in such a way that penetrometer soundings, found that 12 of the structures disruption and restoration times are minimised. It may were founded on soils at risk of liquefying under the seem there is no sensible alternative to this belief but, chosen earthquake scenario. This must ring alarm bells in fact, where it is known that central government for utility managers. Careful investigation and design funding will be generously available after an event, a can overcome this problem. For example, at the Halswell more cynical view can prevail which resists spending water supply reservoir site, additional excavation plus mitigation dollars now on the understanding that some- the installation of a synthetic barrier mitigated the one else’s money will fix the damage after the earth- liquefaction potential that site investigation had re- quake. vealed.

A prudent utility manager will endeavour to under- Problems of inaccessibility do not apply as much to stand the importance and vulnerability of the system’s civil services facilities and structures and here manag- elements, prioritise and recommend mitigating meas- ers have wide scope for mitigation by retrofitting. ures and ensure thorough emergency response plan- Costs are often not great and can usually be incorpo- ning. In seeking to win a share of scarce resources for rated in maintenance budgets. Much more difficult is 72 • Risks and Realities

the state of mind amongst plant operators that looks for On the Port Hills, the natural valleys are the primary any evidence of instability or lack of security under drainage system. No baseflows exist. During storms seismic loading and programmes work to remove it. hill catchments can produce high flows that can scour This awareness must be broad enough to gauge the their easily erodible loess soil mantle. security of other services (water, power, telecommunication, drainage) on which the operator’s own service Many of the larger waterways on the flat are bordered or facility may depend. Raising staff awareness will by trees. In fact, much of the waterway system has a thus be an important part of mitigation. high environmental value. The land drainage system also includes retention basins, stop banks and several In this study, officers with a day-to-day involvement pumping stations of various sizes. with the particular lifeline have assessed both the vulnerability and importance of each element, and its The sewerage system is interconnected at various dependence on other services and recommended miti- locations to provide emergency sewage overflows gation measures for the particular hazard. The work (normally during storms) under controlled conditions. has served to open up the subject of the performance of However, it is anticipated that following a major earth- Christchurch’s lifelines during severe natural events. It quake the land drainage system will also need to have now remains for the managers responsible for those the ability to convey raw sewage from areas where the services to ensure that the work of assessing and sewerage system is inoperable. prioritising the many mitigating recommendations is pursued and appropriate measures implemented. Heathcote River Catchment The Heathcote River drains a catchment area of approximately 10,500 hectares to the Avon/Heathcote 4.2 Land Drainage Estuary. Spring-fed headwaters sustain the river’s baseflow. Seismic risk The upper portion (upstream of the Cashmere Stream System description confluence) drains the predominantly industrial areas of Hornby and Islington via large diameter concrete Christchurch’s land drainage system is centred around pipelines. These drain to three open waterways and three small coastal rivers that flow from west to east to then to the river. the sea. These are the Heathcote, Avon and Styx Rivers and they are maintained by the Christchurch City Below its junction with the Cashmere Stream the river Council. Two similar rivers drain some small areas of generally follows the toe of the Port Hills, draining Christchurch but are maintained by the Canterbury mainly residential suburbs and the Woolston industrial Regional Council. These are the Halswell River and area that discharges via a mixture of open waterways the South Branch Waimakariri River. and pipe networks. Cashmere Stream is the river’s largest tributary. It serves a large rural area drained by The Christchurch City Council maintains roughly 400 a system of open waterways. Within the Heathcote kilometres of waterway (i.e. rivers, dug drains, con- River catchment is the recently constructed Wigram crete channels, timbered drain, etc.) and over 500 retention basin. kilometres of pipeline ranging in diameter from 100 mm to 2,100 mm. Most of the pipes are reinforced Other land drainage structures include the Woolston concrete rubber ring jointed (RCRRJ) but also in- Cut and the new Woolston barrage. Only one of cluded are perforated subsoil pipes, ceramic, ac, and Christchurch’s 18 stormwater pump stations lies within single and double skin “brick barrel” pipelines. the Heathcote River catchment. Unless otherwise stated, piping mentioned in this report or shown on the map is 450 mm diameter or greater Avon River catchment (see Map 18, p 301). The Avon River drains a catchment area of approximately 8,400 hectares to the Avon/Heathcote Estuary. The rivers and their main tributaries on the flat carry a Spring-fed headwaters sustain the river’s baseflow. continuous baseflow that is spring fed in their upper reaches. The baseflow also regulates the inflow and There are six tributary streams branching out above level of shallow groundwater in the eastern half of the Mona Vale (in the vicinity of Fendalton Road/ city. During major storms that generate high rainfall Holmwood Road). These serve mostly residential ar- and low barometric pressures, low-lying urban areas eas by a mixture of open and piped networks. Below are vulnerable to flooding. Mona Vale the river flows through the central business Civil Services • 73

district which is drained mostly by pipelines. There are Sea outfalls some brick barrel pipelines within the business district. The suburb of Parklands is drained directly to the sea Downstream of the central business district the river via a 1,575 mm diameter RCRRJ pipeline. Two other flows through residential areas to the sea. On the way smaller pipelines outfall to the sea at Beach Road and it collects flows from major tributaries such as St Pacific Road, North New Brighton. Albans Creek, Dudley Creek, Shirley Stream and Horse- shoe Lake. Horseshoe Lake is the outlet for the 1,800 Halswell River tributaries and 2,100 mm diameter RCRRJ Dudley Creek diver- Nottingham Stream drains most of the residential sub- sion pipeline and some semi-rural open waterways. urb of Halswell to the Halswell River. Other drains and Knights Stream also drain to the Halswell River. These, From approximately Porritt Park to the Estuary there are sections of stopbanking to prevent tidal flooding of and the river itself, are maintained by the Canterbury low areas. Fourteen stormwater pumping stations are Regional Council. within the Avon catchment. Most are alongside the The 162 ha Halswell Junction Road industrial area is a Lower Avon. subcatchment of Knights Stream. It contains the Halswell Junction Road retention basin. Styx River catchment The Styx River drains a catchment area of approxi- Templeton mately 5,000 hectares to near the mouth of Brooklands The Templeton township is to the south-west of the Lagoon (north of Christchurch). The uppermost part of main Christchurch urban area. Stormwater disposal in the Styx River drains the residential suburbs of this area is by ground soakage. Casebrook and Redwood. The river then flows for approximately 15 km through rural land to the General Brooklands Lagoon, which is located near the Waimakariri River mouth. Most of the pipelines greater than 450 mm diameter are reinforced concrete with rubber ring joints. Within the The lower reaches receive drainage from the small central business district there are some brick barrel settlements of Spencerville, Brooklands and Kainga. pipelines. Many of the smaller pipelines (less than 450 The major Styx River tributary is the Kaputone Stream mm diameter) are located beneath the street side chan- which drains most of Belfast township including a nel and are a mixture of concrete and asbestos cement small industrial area. and are either fully encased or capped with concrete.

Other systems Generally, open waterways are either timbered, concrete lined or earthen banks. South Branch Waimakariri River (Otukaikino Creek) System vulnerability The Otukaikino Creek serves almost entirely rural The normal functions for the land drainage system, land. The exception is the Belfast township. Kaputone together with the additional emergency function of Creek in the Styx River catchment serves about half of conveying raw sewage, would be made vulnerable by Belfast but the rest of the township is drained to the following earthquake effects. Otukaikino Creek via Wilsons and Johns Drains. Waterway bank failure Coastal hillside drains This is particularly the case in liquefaction zones. In the These drain the suburbs of Mount Pleasant, Redcliffs larger waterways, the consequent loss of low flow area and Sumner and are generally a watercourse down the and raised invert levels is likely to have a minimal valley discharging to the Estuary or sea via a pipeline. effect, possibly raising groundwater levels significantly in some low coastal areas. There would still be Outfall drain reasonable storm flow capacity and sewage flows should be uninterrupted. Outfall Drain runs adjacent to Linwood Avenue. It is concrete lined for its upper length and the rest is earth Smaller waterways could be partially blocked. banks. It drains some of the city centre and the suburbs of Linwood and Bromley. It discharges to the Avon/ Pipe damage Heathcote Estuary via a multi-cell box culvert fitted with flapgates at Humphreys Drive. In general, it is likely that most pipelines will still 74 • Risks and Realities

function but with a reduced capacity and a tendency for as the tide would get to this reach more quickly and the blockages to occur. Specific pipelines of greatest vul- level would be higher. nerability are identified below. At the upstream end of the Woolston Cut is the Woolston Barrage which is in zone 3A, a high risk area for Trees liquefaction and ground shake. This structure is new Waterways generally will be vulnerable to the effects (1993) and has been designed in accordance with of trees falling into them either through bank slumping current standards. The structure is built on caissons or breaking of the trunks. founded on non-susceptible soils. There should be no major problems in an earthquake. It is unfortunate that as Christchurch’s main drainage systems flow towards their outfalls and their impor- The Woolston Cut is also in a high risk area for ground tance increases, so too does the liquefaction risk and shake and liquefaction, however it is expected that therefore the possibility of serious damage. Flooding while some displacement of the present wall units will of low lying areas is likely. The likelihood and severity occur, the cut would not suffer significant structural of flooding will increase should a storm occur after the damage. It is a relatively new structure (1986). earthquake. Below the Woolston Cut there is a length of the river In the catchments upstream of the river lengths the that is very close to Ferry Road and is prone to lique- damage will be minor and if a storm occurs the poten- faction. Should the river banks collapse in this area the tial for flooding in these areas will not be high. road could be affected. There is a risk of damage to major service lines which run along Ferry Road at this point. Heathcote River catchment In the upper catchment area of Hornby and Islington, Downstream of this high area the river flows through large reinforced concrete pipes are the main system of low lying land, past Ferrymead Historic Park to the drainage. This area is prone to ground shake but should Estuary. This section of the river is prone to liquefac- not liquefy. It is expected that these pipelines will stay tion but bank failure would cause minimal reduction in basically intact with some minor differential move- hydraulic capacity. ment in joints. While these pipelines may remain in service, it may take several years to bring the system to Avon River catchment a repaired condition. The piped and small open waterways that drain into the From Wigram Road through to the Cashmere Stream upstream tributaries of the Avon River are within zone junction the river passes mainly through zone 3B so it 3 which is prone to high ground shake. It is expected is subject to ground shake and has a moderate risk of that these systems will not suffer major damage and liquefying. Open waterways draining to the river, and should be able to continue to function. However, it may take some years to check the piped networks and repair the river itself, could sustain some slumping of banks. if necessary. If a storm occurs there is a low to moderate risk of some flooding of low-lying residential areas next to the river. The tributary streams above Mona Vale are also in the earthquake zone prone to high ground shake. Most of The river flows in and out of different earthquake risk these streams have a spring-fed base flow and are zones from the Cashmere Stream junction to Opawa generally shallow and wide. In an earthquake they Road. Generally it is vulnerable to high ground shake could be vulnerable to minor bank damage but there and moderate liquefaction. At some points it is close to should not be major restrictions to the flows. The river the hills and could also be affected should a large all the way through the central business district and to landslide occur. The river banks are susceptible to the Dudley Creek junction is in the same category as slumping. This could reduce the area of the main the tributary streams described above. Trees along the floodway and should a storm occur there is a moderate river banks could cause problems if they fall into the risk of some flooding of low lying residential areas waterway. next to the river. The upstream end of St Albans Creek runs through a From Opawa Road to Ferry Road there are stopbanks pocket of land vulnerable to a moderate to high risk of to prevent tidal flooding. At this point the river runs liquefaction. The creek through here is intermittently around the edge of a liquefaction zone, so there is a low timbered and has a permanent low flow. Some bank chance of these banks slumping. This could cause tidal collapse is possible and there is a low chance of the flooding of low areas, with the chance of this happen- creek being severely restricted. In a storm this would ing increasing if the Woolston barrage cannot be closed cause extensive residential property flooding. Civil Services • 75

Most of Dudley Creek and Shirley Stream are in the The Styx River tidegates on the line of the Waimakariri same category as the tributary streams above Mona River stopbanks is an important flood control structure Vale. for the extensive floodplain adjoining the lower reaches. Designed and constructed in the early 1980s it is The large diameter Dudley Creek Diversion pipeline is intended to remain intact in a major earthquake, but within a zone prone to high ground shake and a high could be significantly displaced because of the high liquefaction risk. This pipeline is founded on a crushed liquefaction risk in the area. Complete or partial loss of metal bedding. It could suffer significant displacement gate function would cause saline flooding of pasture downstream of Marshland Road but should stay basi- land and elevated groundwater levels. cally intact apart from some joint failure. Long term it could be expensive and time consuming to get this This length of the river has been classified as having the pipeline back to its original alignment and capacity. greatest vulnerability.

From the Dudley Creek junction at the Estuary, the river is within an area prone to high ground shake and Other Systems also has a high risk of liquefying. Horseshoe Lake also Otukaikino Creek falls within this category. Through this length of the river there is a high chance of the river banks slumping. From Dickeys Road to the Waimakariri the river is The intermittent stopbanking could also be affected in within a moderate liquefaction risk zone. There could this way. Should a storm occur there could be flooding be some bank slump but this should not seriously of low-lying residential land. impair the flow.

However, the major risk will be tidal inundation of low Coastal hillside drains lying residential areas usually protected by the stopbanking. The suburb of Bexley is particularly at These drains are all vulnerable to moderate ground risk. shake and a moderate risk of liquefaction. There could be damage to the valley watercourses, redirecting the There would probably need to be large-scale evacua- flow, and bank slumping is possible in the downstream tions. Repair work could be difficult and time consum- reaches. Some of these drains have been identified as ing due to the repetitive nature of tidal flooding. This having a high risk of damage to property should a storm length of the river has been classified as the most occur. This is because the drains do not have a perma- vulnerable to damage. nent base flow so any damage to the waterway will not be causing immediate problems and therefore may not Styx River catchment be discovered immediately. Therefore, if a storm occurs there could be widespread flooding of the flat The uppermost portion of the Styx River down to Main North Road is prone to ground shake but should not residential land at the base of the watercourses. incur any major damage. Outfall drain Kaputone Creek and the river from Main North Road Outfall Drain is in the zone vulnerable to a high risk of to Radcliffe Road are both vulnerable to a moderate liquefaction and ground shake. The length of drain risk of liquefaction and bank slump could occur. Be- with earth banks could suffer from some bank slump. cause the river flows through rural land any damage The precast concrete lined portion could be some would be confined mainly to crops and pasture. displacement of the precast units. Any possible dam- From Radcliffe Road to the Brooklands Lagoon there age in either lengths should not cause major flooding if is a high risk of liquefaction and high ground shake. a storm occurred. The river is tree lined for most of this length and there The tidegates at the Estuary are at risk of settling due is a risk of trees snapping or slumping into the river and to liquefaction and saltwater intrusion could be a impeding the river’s flow. There is also a high chance problem in the event of tidegate failure. of bank slump. These factors could cause adjoining low-lying rural land to flood from baseflow back up and elevated groundwater levels. It would be difficult Sea outfalls to drain this land quickly and crops and pasture could These outfalls are prone to high ground shake and a be extensively damaged. high risk of liquefaction and also damage from the sea There is a high chance that these systems, particularly Near the Brooklands Lagoon is the small settlement of the Parklands Outfall, could be extensively damaged. Brooklands. Low-lying properties could suffer flood- If a storm occurs, surface flooding is highly likely. ing. 76 • Risks and Realities

Halswell River tributaries respective outfalls. Surface flooding would be inevita- The Nottingham Stream flows through land vulnerable ble and this could hamper property access and repair of to high ground shake but not to liquefaction. The services. channel should not be seriously damaged. Mitigation measures General It is clear that in a major earthquake event in Christ- church, serious damage and widespread disruption is Pumping stations likely to occur to the land drainage system in some The 14 stormwater pump stations have been inspected areas. Damage to property due to flooding is extremely and their seismic vulnerability assessed as being in the probable should a storm occur after the event. Mitiga- low risk category. However, piped connections to the tion measures for the land drainage system fall mainly stations were classified as having a low to moderate under the category of organising response measures. risk of damage. That is, checking and repairing the various elements in order of priority so that we are not caught out should a storm occur, and that the passage of baseflows (includ- Brick barrel pipelines ing sewage) is maintained. These are all within areas at risk of liquefaction and high ground shake. Brick pipelines are particularly It is essential that planning be undertaken to identify vulnerable to ground shake and it is expected that the likely responses after an earthquake especially in significant damage will occur. These pipelines have the following areas: been classified as having the highest vulnerability to damage. • Maintaining the passage of baseflows and/or sewage flows together with public health precautions.

Retention basins • Checking the flood channels are clear in the three These are in areas subject to high ground shake. Wave major rivers. action could cause minor damage. • The Woolston Barrage has to be checked. If it gets jammed shut then a storm could cause flooding Water quality upstream or if jammed open saltwater damage is The entry of raw sewage into the land drainage system possible. will occur via existing overflows. These are generally • Checking stopbanks along the Avon River and located close to the rivers and adequate baseflow responding to damage to prevent tidal inundation. should be available for conveyance. However, there is likely to be ponding in areas of significant bank slump- • Making sure hillside waterways and outlets are ing or pipe failure where no baseflow exists. clear.

All of the rivers are vulnerable to hazardous spills due • Ensuring that personnel and equipment are avail- to storage tanks failing in an earthquake. Some of the able for the above. possible contaminants could cause extensive damage to wildlife and the river biota. It could take a long time Other mitigation measures that can be completed prior for the rivers to recover from such spills. to an earthquake include:

A significant pollutant will be silt due to mini land- • Providing adequate fixing and bracing of plant in slides and the river banks collapsing. Once again this pumping stations. type of pollutant could cause serious long-term damage to the river’s environment. • In areas subject to liquefaction/lateral spreading use flexible pipeline materials with full joint strength. For key mains, traditional rubber ring Kerb channels and sumps jointed pipes may not be adequate in these condi- In the Edgecumbe earthquake in 1987 it was reported tions. that 30% of all stormwater sumps were either destroyed or displaced. This is caused by longitudinal • Use maximum strength couplings or joint lengths movement of kerb and channel and/or upward dis- when connecting pipelines to structures. This may placement of sumps. Christchurch can expect similar be particularly appropriate for large diameter pipes damage. If a storm occurred, stormwater would not be where joints normally have less capacity to absorb able to drain efficiently to damaged sumps and their rotation. Special joints may be required. Investi- Civil Services • 77

gate the feasibility of allowing space between joints — Priority list of drainage facilities to be to permit pipeline to compress without damage. checked, and a standardised format for reporting damage. • Insert hazard maps as layers in Geographic Infor- mation Systems for planning and location of serv- — Procedures for thorough debriefing. ices. Snowstorm Risk • Ensure equipment sourced from overseas meets seismic detailing requirements. System description • Embody seismic thinking into the planning and Refer to Land Drainage Seismic Risk, Section 4.2. design process. Ensure that seismic risks are covered in the revision of the Design Manual. Vulnerability • Put in place management plans and provide train- The normal functions for the land drainage system, ing to ensure controlled and intelligent response to together with the emergency function of conveying disaster situations. Prepare a database of plant and sewage overflows, would be made vulnerable by the materials, both within and outside areas likely to be following snow storm effects. affected. Weather pattern • Identify spare parts and sandbag requirements and The low barometric pressure usually required to pro- hold in stock or identify ready sources of supply. duce a significant snowfall in Christchurch also has the • Identify and prepare cost/benefit estimates and effect of increasing the tide levels. High runoff from priorities of the specific mitigation measures avail- snow and associated heavy rainfall combined with able to protect the integrity of the drainage system. these tidal effects can produce flooding, particularly in Factors to be taken into account include cost, rela- the lower reaches of the rivers (see Section 2.7 for tive risk and extent of damage, public safety and details on the weather patterns required for heavy health, receiving water quality and property dam- snowfall). age. Hillside slips • Investigate methods of improving brick barrel pipelines by relining or other methods. If this is not Saturated hillsides can be made more vulnerable to possible, programme replacement of pipelines with slipping by the weight of heavy snow fall. Slips can block off the usual natural flow paths and hillside other suitable materials. drains, causing stormwater to find an alternative way • Establish where hazardous spills are possible and down the hill. This can sometimes be through residen- encourage owners to install bunds to contain spills tial sections and buildings. During the August 1992 in an earthquake. snow storm, hillside slips were a major concern.

• Study the performance of existing materials and Loss of electricity and telecommunications investigate possible design improvements. During the initial phase of snow storm it can be • Prepare a comprehensive earthquake response expected that electricity and telephone lines could be management plan covering: cut off for a number of hours to most areas of Christ- church. — Response structure such as staff requirements, duties, decision making delegation, Loss of telephone communications would prevent the communications liaison with Civil Defence public from reporting drainage problems. However, and other service authorities, etc. internal communications by radio telephone (or cellphone) would be less vulnerable and still likely to — Emergency response centre facilities. be functioning.

— Measures for staff rescue, accommodation, Stormwater pumping stations could be affected by loss safety and support. of electricity but this should not have any major consequences if electricity is restored within a day. Some — Identification of essential plant, equipment pumping stations do not have gravity overflows so will and materials, and sources of emergency need to be given priority for reconnection. The large supply. pumping station at Horseshoe Lake has a standby power generator. 78 • Risks and Realities

At the sewer pumping stations that do not have diesel drainage personnel to gain access to sites where backup, sewage overflows to the rivers would occur. assessments or operations are required.

• Encourage considering four-wheel drive when look- Blockages ing at future vehicle purchases. Heavy snow can build up on inlets to the land drainage system causing partial or full blockages. Snow loads Pumping stations would cause branches to break off trees, which could also contribute to blockages. • Investigate which pumping stations are at risk of flooding if electricity supply is cut off. Install This can occur in open drains, usually at inlets to pipes gravity outfalls where possible and carry out any causing backing up and possible localised flooding. other works that would prevent the pump stations Street sumps could be blocked causing street flooding, from failing. thereby aggravating already difficult traffic conditions. Snow blockages are easily cleared. However, Tsunami, Local Flooding and Slope due to staff limitations, problems will still occur. Hazard Risk

Transportation System description During the initial stages when roads are covered with Refer to Land Drainage Seismic Risk, Section 4.2. snow there can be problems with staff getting to emergency and operation control centres. Inspecting Vulnerability trouble spots in response to public complaints is also difficult without the availability of four-wheel drive Tsunami vehicles or chains. River banks Mitigation measures During the scenario described, waves would propagate up the Avon, Heathcote and Styx Rivers, overtopping Weather pattern the stopbanks along the lower reaches. River banks are • Ensure that all weather warnings, flood procedures vulnerable to damage from the scouring effect of the and responses are up to date. fast moving wave and slumping by rapid draw down effects as the wave recedes. Stormwater pipe outfalls to the river could be seriously damaged in the lower Hillside slips reaches. • Require maintenance staff to report possible problems on the hillsides and include an ongoing report- Sea outfalls ing procedure for signs and causes of possible The sea outfalls, including the hillside drain outfalls, future slips. Where necessary, do preventative work. are vulnerable to quite serious damage and wave effects could cause damage and flooding further up- Loss of electricity and telecommunications stream. • Provide priority list to Southpower for reconnection of important pumping stations. Pumping stations Two of the stormwater pumping stations are vulner- • Seek extra safeguards for radio telephone/cellphone able to flooding from a tsunami. They are: systems. • Pumping station number 209 — Beachville Road. Inlet blockages Subject to inundation from Estuary.

• If resources are seriously stretched and the weather • Pumping station number 217 — Bridge Street. allows it, encourage capable members of the public Subject to inundation from Estuary and beach. to clear snow away from roadside sumps. Local flooding Transportation Local flooding occurs in those areas that are either • Ensure that at least one four-wheel drive vehicle is permanent natural floodplain and ponding areas or available and also chains for other vehicles for land small in extent and frequency such that further improvement in the drainage system is difficult or expensive. Civil Services • 79

Slope hazard lems on the hillsides and include an ongoing reporting procedure for signs and causes of possible Earthquake scenario future slips and debris problems and, where neces- All waterways on the hillside are vulnerable to moder- sary, carry out preventative work. ate ground shake and a moderate risk of liquefaction. Slips and falling debris could cause blockages in wa- • Consider possibility of secondary flows and their terways. If a storm occurs during or following the effects when designing hillside drainage systems. earthquake, flows could be redirected through residential sections and buildings. 4.3 Sewer System Heavy rainfall scenario As for the earthquake scenario, slips could cause block- Seismic hazard ages in waterways. It is highly likely that property System description damage will occur due to secondary flows through residential sections and buildings. Secondary flows The Christchurch sewer system comprises: and/or inundated overland flows could also induce • 175 km of trunk gravity sewers (300 mm and larger slips. diameter) and pressure mains;

Mitigation measures • approximately 1,135 km of minor gravity mains reticulation (225 mm or smaller diameter); Tsunami • 74 pumping stations; and • Establish communication linkage with early warning system. • one major treatment plant at Bromley and two minor satellite plants at Belfast and Templeton. • Investigate the possible effects of inundation of the two vulnerable pump stations and carry out preventative works if necessary. Christchurch sewer mains analysis The lengths, types and construction materials of Christ- • Consider scour and draw-down effects when de- church sewers are shown in Table 4.1. signing new bankworks. Instigate response measures to ensure that following the event all beach The city’s first sewers were constructed in 1880 with outfalls and affected waterways are inspected for most early construction occurring in three boom peri- damage and any repairs carried out in an appropri- ods, 1880 to 1884, 1900 to 1910 and 1924 to 1931. In ate order of priority. the first period, 51 km of ceramic sewer was laid, mostly within the four avenues but also branching into • A greater understanding of effects is required. St Albans, Sydenham, Addington and Linwood. The Model the tsunami scenario using MIKE II so that city’s 7 km of brick barrel sewers and the old number the extent of flooding can be assessed. Mitigation 1 pumping station were also constructed during this measures for flood affected area might include: period. From 1900 to 1910, a further 120 km of ceramic — determination of areas that should be sewerage was laid, and pumping stations 2 and 3 evacuated where human safety is at risk. constructed. This extended the system radially an average of 2 km. Parts of Cashmere and Sumner were also — gates that provide for greater flow of flood separately reticulated around this time, although most water into waterways and rivers when the of these Sumner sewers have now been renewed. wave receding. The years 1925 to 1930 saw an amazing 250 km of sewer and 23 pumping stations built, extending the Local flooding system into the suburbs of St Albans, Richmond, • Identify on flood awareness maps. Avonside, Linwood, Woolston, Opawa, St Martins, Beckenham, Somerfield, Spreydon, Riccarton, • Control flood damage through floodplain manage- Fendalton, Bryndwyr and Papanui. During this period ment plans prepared jointly by Christchurch City a mixture of cement jointed ceramic and reinforced Council and Canterbury Regional Council. concrete pipes were used. Ceramic pipes continued to be used until around 1950 before being superseded in Slope hazard all but 100 mm diameter laterals by rubber ring jointed • Require maintenance staff to report possible prob- reinforced concrete pipes. From 1973 to 1987, cheaper 80 • Risks and Realities

Length Gravity Pressure Other Average Age range (m) main main age Cast Iron 4040 290 3730 - 73 16-83 Asbestos 19,217 1966 16,115 1,136 17 4-30 Cement Reinforced Concrete 137,921 114,758 20,363 2240 31 4-66 (rectangular) Earthenware 4033 3770 283 95 60-110 Brick Barrel 5145 5145 - - 110 110

Total 170,356 125,929 40,491 3376

Note: 3,796 metres not included — material type not known

Table 4.1: Christchurch city sewers

asbestos cement pipes were used in all but the larger Vulnerability diameters (450 mm upwards). Since 1987 PVC piping has been the predominant sewer material used. Reticulation In this section reference is drawn to the Seismic Hazard The city’s two most important sewers, the northern and map (Map 19, p 302) on which the trunk sewer system southern relief sewers were constructed in the 1950s has been superimposed. along with pumping station 1 which was commissioned in 1956. Most trunk reticulation is less than 40 Most vulnerable to earthquake damage are the brick years old, the main exceptions being the brick barrels, barrel sewers, lime cement mortar jointed ceramic the two pressure mains from terminal pumping station sewers, and all sewers in the 2A and 3A liquefaction 11, laid in 1907 and 1924, and ceramic sewers in the zones (the 3A zone covers most of east Christchurch) central city area. Widespread damage of these sewers can be expected, with numerous broken junctions, broken collars, dis- More than 80% of trunk gravity reticulation is con- placed joints, and possible losses of grade. structed of rubber ring jointed reinforced concrete pipes, the rest being (in decreasing order of length) The traditional use in Christchurch of short pipes on ceramic, brick, asbestos cement, PVC, concrete lined either side of manholes, will help rubber ring jointed steel and cast iron. gravity sewers remain operational, but ceramic and brick barrel sewers are almost certain to suffer col- Most pressure mains are either asbestos cement or lapses. reinforced concrete, PVC and cast iron being the other materials used. Many of the pressure mains laid in liquefiable ground will be sufficiently damaged to make them unusable Age of Christchurch city sewage treatment until repaired. facilities Of the 74 pumping station catchments covering the The Bromley Wastewater Treatment Plant was com- urban area, 30 have no emergency overflow with 15 of missioned in 1962 with later expansions. Treatment these having catchments of more than 50 ha. currently consists of screening, primary sedimentation, biological filtration, secondary sedimentation and Sewage overflows will be unavoidable and may create oxidation ponds. a serious health hazard in catchments without bypasses to the stormwater system. The Belfast oxidation ponds were constructed in 1972, and the Templeton Plant in 1958. Most river crossings are siphons, important exceptions being pressure mains suspended from the Pages Road The reticulation has an estimated replacement value of and Ferrymead Bridges. These are likely to break, $475 million, and the Bromley Wastewater Treatment cutting off South Brighton and all suburbs east of Plant $200 million. Ferrymead Bridge. Civil Services • 81

Pumping stations priate alterations. However a pumping station with a An inspection of all 74 pumping stations was carried high electrical demand could cause problems. Only out to check the seismic vulnerability. The findings are pumping stations 1, 11 ,15, 78 and 80 have standby summarised below. generation.

Pumping station buildings Alarm/Scada equipment The majority of buildings are likely to suffer moderate Alarms and Scada equipment generally rely on the damage with cracking of brittle components such as Telecom Network to operate but in general are not concrete/brick/block walls, but will remain substan- critical to the functioning of the network. As noted tially intact. Fifteen percent could have walls collaps- above, a new Scada control and alarm system is being ing or failing completely. The cracking of basement or installed in all pumping stations which is radio based. well walls could allow the ingress of water and sewage into basements and dry wells. Treatment works

Pumping stations 2, 3, 4, 5 and 51 are in the “may Bromley wastewater treatment plant collapse” category. Of these, only pumping station Buildings — these were found to be of low risk of number 5 has a catchment area of any significance (51 damage. However at Bromley, should cracks appear in ha). walls below ground level or substantial leaks occur in Pumping stations 6, 7, 9, 11, 12, 18, 26, 27, 28 and 59 the basement reticulation systems, the sump pump are in the next category up of “Moderate damage, some system could be unable to cope. walls fail in face loadings.” Of these stations pumping Pumping equipment — pumping equipment was also station number 11 is of vital importance to the network, found to be of low risk. followed by pumping station number 6, then pumping station numbers 18 and 28. Pipework and connections— as for pumping stations, highly vulnerable at pipe/structure interfaces. Many stations have manual sump pumps which may not be activated in time to possibly prevent the dry well Electrical services — generally low vulnerability. from being flooded by seepage from the wet well. The Emergency generating equipment operates on site and Scada alarm system, now being installed in Pumping is able to supply the fuel power needs of the plant. station number 15 and ultimately planned to serve all Alarms and control equipment — VDUs and computer important pump stations, will, if still working after the equipment were found to be unrestrained. earthquake, warn of water rising on the dry well floor.

Templeton plant Pumping equipment Minimal damage likely to buildings, a small possibility The majority of pumping equipment (90%) falls in the of damage to pumping equipment and pipes/connec- small possibility of damage category. The remaining tions and damage to electrical services and alarm 10% is in the “damage highly unlikely category”. equipment unlikely.

Pipes/connections Proposed mitigation measures The majority of pipe connections were found to be inflexible and therefore likely to be damaged where Reticulation piping enters wells and basements, particularly where The following mitigation measures are recommended: liquefaction/settlement is likely to occur (65% of installations). • Construct controlled overflows to the stormwater system at critical points in those catchments where Electrical services none exist at present. This work to be prioritized on Most installations were found to have little risk of catchment importance. Estimated cost $10,000 per damage to electrical services (91%). However the overflow with possibly 20 new overflows required. majority of pump stations rely on electricity supplied • Strengthen structurally suspect brick barrels iden- by Southpower with no emergency generator backup. tified from CCTV(closed circuit television). The Supply is usually via a Southpower kiosk substation likely need for this work has already been recog- and these are highly vulnerable. In some cases the nised and it will be funded from the sewer renewal failure of one kiosk may mean others in the area could budget. Estimated cost is between $1 million and take up the load once Southpower has made the appro- $3 million. 82 • Risks and Realities

• Take all opportunities presented by major road- cient distance between the valve box and the pumping works to renew remaining trunk ceramic sewers station to install a flexible joint, the valve box could and also those serving the central city commercial instead be rigidly connected to the pumping station and area. Possible cost $300,000 per year on average, just one flexible joint installed in the pipework on the again funded from the sewer renewal budget. other side. Estimated cost is $5,000 per station. Flex- ible joints could be achieved using a length of • Investigate earthquake protection measures as part polyethylene pipe between long gibault joints or a of upcoming review of the Design Manual. length of rubber piping either flange or steel band • Use flexible pipeline materials with full joint connected. There are also manufactured couplings strength for future pressure main work, e.g. welded suitable for the purpose. polyethylene. A further mitigation measure is to identify spare parts • Prepare an earthquake response plan. It should requirements and hold in stock or identify ready sources include a prescribed procedure for prioritising re- of supply. pair work. Electrical equipment Pumping stations Recommended mitigation measures are to: • liaise with Southpower as regards security of sup- Pumping station buildings ply to key pumping stations in the network; Recommended mitigation measures are: • investigate standby generation for pumping station • Seismically strengthen buildings that are most at numbers 20, 35, 36, 42, 46, 63 and 20, these risk and pose greatest threat to network as whole. pumping stations being most important to the net- Only pumping station number 5 and, possibly, work (excluding pumping station numbers 1, 11 pumping station number 2 have been identified in and 15 which already have standby generation); this category. Estimated cost $100,000. • ensure all electrical cabinets are bolted down; and • Ensure sump pumps are adequately maintained, secure, and in important stations have backup elec- • ensure all emergency battery banks are restrained. tricity supply. Providing backup electricity supply for sump pumps will require a small standby gen- Alarm Scada equipment erator, sometimes housed outside the pump station because of space limitation, and associated electrics. No mitigation measures recommended. New radio Some 14 important pumping stations have been based scada system currently being installed will be identified for this work. Estimated cost $20,000 less vulnerable. each. Miscellaneous • Establish a stock of petrol-driven pumps, stored in Restrain all miscellaneous loose equipment capable of key locations around Christchurch that could be causing damage, e.g. diesel tank stands and drums, used to back up sump pumps in an emergency or spare parts, fluorescent lights (often hung from hooks), where no pumps drain dry wells. water cisterns, etc. • Two pumps are currently reserved for this purpose but more are considered necessary. Treatment works

Pumping equipment Bromley wastewater treatment plant • Ensure equipment is properly restrained. Buildings — review sump capacity and ensure operation in the event of a power failure.

Pipes/connections Pumping equipment — no mitigation measures recom- • Establish a programme of fitting flexible joints to mended. pipe entries at high liquefaction risk sites that are also important to the network as a whole. Pipework and connections — establish a programme to upgrade reticulation system at key/critical points. These This work may be needed at possibly 20 pumping critical points will first need to be identified. stations still to be identified. Depending on the distance between the valve box and the pump station, up to three Alarms and control equipment — restrain VDUs and flexible joints may be needed. Where there is insuffi- computer equipment with inexpensive hold down straps. Civil Services • 83

Wind Mitigation measures Proposed mitigation measures are to: Vulnerability The wind storm will not affect buried reticulation and • install overflows to the stormwater system where should cause only minimal damage to sewer pumping necessary; station buildings. Sewer pumping stations will how- • liaise with Southpower regarding security of sup- ever be vulnerable to both loss of power and telephone ply to key pumping stations in the network, supply- (used by the alarm system). This could result in dry ing them with a prioritized list; wells being flooded in stations that are not protected by overflows to the stormwater system. • install automatic sump pumps with reasonable capacity in all stations (currently underway); At the treatment works, some scour of the oxidation pond banks is possible but serious damage is consid- • seal all high level ducting into wet and dry wells; ered unlikely. • ensure availability of four-wheel drive transport; and Mitigation measures Proposed mitigation measures mainly relate to power • ensure sufficient stocks of submersible pumps, failure. They are to: generators, lights and diesel.

• install sewer overflows at or close to unprotected Tsunami pumping stations;

• liaise with Southpower regarding security of sup- Vulnerability ply to key pumping stations in the network, supply- Areas affected by the tsunami hazard are Brooklands, ing them with a prioritized list; Spencerville, North Brighton, New Brighton, Brighton Spit, Ferrymead, and east of Ferrymead Bridge. • check pumping station sites for possible damage to buildings and overhead telephone wires from trees; A total of 13 pumping stations in these areas will be and inundated, with few likely to remain operational at present. The stations affected, in descending order of • ensure aerials for the new radio-based Scada sys- catchment size are pumping station numbers 35, 37, tem will not be damaged by trees. 57, 31, 38, 34, 52, 55, 77, 78, 33, 59, 48, and 30. The failure of these stations will not affect sewage disposal Snow in other pumping station catchments.

Vulnerability Mitigation measures In the August 1992 snow storm, power failure lead to Proposed mitigation measures to minimize damage to the Sumner pump station flooding and two other pump pumping station plant are to: stations east of Ferrymead Bridge being nearly flooded. The pumping station alarm system did not help much • waterproof pump station buildings below inunda- as it only indicated high water at most stations and did tion levels, replacing louvre doors with solid ones not warn of the more serious problem of water on the (installing higher vents where necessary), install- dry well floor. ing waterproof seals around all doors, wet well covers, and other openings below flood level; This meant that the most critical stations needing urgent attention could only be identified by visiting • seal all high level ducting into wet and dry wells; each station in turn, a job that was hindered by lack of and four-wheel drive transport. (The new Scada alarm • ensure electrics in outside electrical cabinets are system currently being installed will warn of water on above flood level. the dry well floor.)

At the treatment works, build up of snow on the Local flooding trickling filter covers caused concern and the oxidation ponds were close to overtopping. Vulnerability Pumping stations with door and other openings below The snow storm caused three hill slips which exposed predicted flood levels will be vulnerable to dry well minor sewers but did not break them. 84 • Risks and Realities

flooding. Also vulnerable are electrics below flood Christchurch city is served by 84 pumping stations and level in outside cabinets. 37 service reservoirs. The pumps draw water from five underground aquifers at depths between 22 and 190 Heathcote River catchment metres. Working up the river, pump stations situated in the Approximately 50% of the primary pumping stations floodplain area are pumping station numbers 15, 12, have a standby diesel generator installed in case of 13, 18, 19, 20, 21 ,22, 23, 43, 68 and 42. Of these, power failure. Secondary pumping stations do not have pumping station number 15 will be inundated 400 mm, standby generators. Of the total pumping stations, 25% pumping station number 18, 250 mm, pumping station have standby generators. There is also a mobile diesel number 19, 330 mm, pumping station number 21, powered pump available. 1.4 m, pumping station number 22, 650 mm, pumping station number 23, 600 mm, pumping station number The city has five supply zones based on the system 43, 400 mm. Pumping station numbers 12, 13, 20, 68 operated by the various authorities that merged in 1989 and 42 are above flood level. to form the new Christchurch city. Those trunk mains analysed in this study totalled approximately 160 km in length (total length of mains is 1,300 kms). Avon River catchment Pumping stations numbers 45, 35, 46, 63, 39, 54 and 28 These zones operate at differing pressures but in an are situated in the Avon River floodplain. Pumping emergency can be interconnected by opening valves station numbers 35, 54 and 28 should escape flooding, that have been installed at zone boundaries. but estimated inundation depths at the other stations Because of the climate in Christchurch a very small are pumping station number 45, 300 mm, pumping proportion of the supply is used as potable water — the station number 46, 200 mm, pumping station number majority used for irrigation. 63, 350 mm and pumping station number 39, 300 mm. A small stand-alone system serves the small settlement Styx River of Kainga/Stewarts Gully. This system does not have storage, but does have a standby generator. Only pumping station number 77 in Brooklands is located in the floodplain. This is a submersible pump Trunk mains linking the pumping stations and reser- station with an above ground electrical cabinet which voirs as well as major distribution pipes are shown on will be inundated at approximately 500 mm. Map 20 (p 303) and range in size up to 600 mm. The length of water supply mains in the city is approxi- Mitigation measures mately 1,300 kilometres. The mains range in size from 80 mm to 600 mm diameter and are constructed with Proposed mitigation measures are to make affected widely differing materials. Table 4.2 shows the ap- pumping stations watertight below inundation levels. proximate length of mains for each size and material. This work will include replacing louvre doors with solid ones (installing vents higher up if necessary), The first watermains were installed in the city during concrete boxing around doors and electrical cabinets, the period 1900 to 1905 and some of these old mains are ensuring wet well covers seal well and sealing high still in use. In fact, some 10% (131 kms) of the mains level ducting into the pump station. were installed before 1920. Table 4.3 summarises the length of mains and materials installed in each 20 year period since 1900 (note that the last period is 12 years). 4.4 Water Supply Also in the city are two small systems that are not controlled by the Council’s Water Supply Unit. These Seismic risk are:

System description • Spencer Park recreation area which is a popular Christchurch is situated partly on a flat alluvial plain camping area where the supply is owned and oper- and partly on the Port Hills to the south. The city is ated by the city’s Parks Unit. The system incorpo- bounded on the east by the Pacific Ocean and is rates a small elevated tank but there is no standby underlain with aquifers. generator.

Some pumping station suction tanks, particularly in the • Christchurch International Airport has a stand alone eastern part of the city, are fed by artesian wells. system with four of its seven wells serviced by standby generators. Civil Services • 85

Material Size of Pipe (mm) 80 100 150 175 200 225 250 300 375 450 600 Total A C 3 350 220 1 135 2 8 54 3 776 CI 25 136 42 27 3 233 CLS2551 3 31233358 DI 1111 3 7 PVC 20 15 3 12 50 ST13512 5 24561181 EVER 1 65 20 1 8 95 Total 55 612 311 2 175 10 13 88 11 9 14 1300

Length (kms) for each size (mm) AC: Asbestos Cement, PVC: Poly Vinyl Chloride, CI: Cast Iron, ST: Spiral Welded or Riverted Steel, CLS: Concrete Lined Steel, EVER: Asbestos Cement (Everite), DI: Ductile Iron.

Table 4.2: Summary of sizes and materials of mains

Outside the city, the Banks Peninsula District Council System vulnerability has a system in the Lyttelton harbour basin served by The whole area would be subject to earthquake dam- water drawn from seven wells in Heathcote, piped to age, some flat areas would be subject to liquefaction, settlement tank at Heathcote and then distributed to ground settlement, flooding and, near the coast, tsu- Lyttelton by two systems. nami.

Firstly the water is pumped to the Heathcote Reservoir The Sumner, Redcliffs, Mt Pleasant area east of the above the Tunnel Road from which it is piped through Ferrymead Bridge over the Heathcote River is particu- the rail tunnel in a 200 mm main to supply lower larly vulnerable. If the link across the bridge is cut then Lyttelton and the Exeter Street pumping station. Water the area will be entirely dependent on the water that is can then be pumped to the main reservoir in Lyttelton. in the reservoirs at the time. There are no wells in the area. The second supply is pumped directly from the Heathcote pumping station to the Cornwall Road res- With regard to the reticulation system, the main prob- ervoir via a 200 mm main passing through the ventila- lem areas appear to be where the main is attached to or tion shaft of the Lyttelton Road Tunnel. built into bridges crossing Christchurch’s rivers. While the more modern bridges should not be too vulnerable Water is distributed to Diamond Harbour via a 150 mm to damage themselves, most could suffer slumping of steel submarine main and to Governors Bay via a the bridge approaches with the distinct possibility that uPVC trunk main. the mains could shear. The distribution lines consist predominantly of 75 mm Damage to pipes will also occur due to differential cast iron with the more modern mains being uPVC. movement at soil transition zones and in areas of Apart from uPVC, the pipes are 70 to 100 years old.

Material Period 1900-1920 1921-1940 1941-1960 1961-1980 1981-1992 Total AC 1 2 158 424 191 776 CI 88 13 98 18 16 233 CLS 21 1 13 13 10 58 DI 77 PVC 1 4 45 50 ST 21 28 9 11 12 81 EVER 3 88 1 3 95 Total 131 47 367 471 284 1300

Length (kms) shown for each period (years) AC: Asbestos Cement, PVC :Poly Vinyl Chloride, CI: Cast Iron, ST: Spiral Welded or Riverted Steel, CLS: Concrete Lined Steel, EVER: Asbestos Cement (Everite), DI: Ductile Iron.

Table 4.3: Summary of length and materials in periods 86 • Risks and Realities

liquefaction. Landslides may well cause displacement • A number of the pumping stations and reservoirs or lack of support to pipelines. are of modern earthquake resistant design.

Forty percent of the pumping stations are at low risk, • A number of the stations are artesian fed. 40% of very low risk, with the remaining 20% high risk (these being substantially constructed of unreinforced • Since 1994, a duplicate McCormacks Bay reser- masonry). A few small reservoirs are laterally unre- voir has been constructed and this substantially boosts storage east of the Ferrymead Bridge. strained and therefore highly vulnerable. Pumping equipment is of low risk. The Banks Peninsula District system is critically dependent on the mains through the rail and road tunnels Seventy percent of the pipework and connections at and on its old substandard reservoirs and pump houses. reservoirs is of low risk with 30% likely to be substan- However, since the commencement of the lifelines tially damaged. There is a general lack of flexible project the Heathcote settlement tank, pumping station connections. Ninety percent of electrical services are and the Cornwall Road reservoir have been replaced. of little risk. For reticulation, the main mitigation measures pro- Details of the vulnerability of the reservoirs and pump- posed are the installation of sufficient additional valves ing stations are shown in the vulnerability charts pre- either side of vulnerable bridge crossings so that the pared for the Council by Kingston Morrison. loss of water can be minimised. Extra valves at major Those pumping stations that do not have a standby intersections will reduce areas required to be shut diesel generators are of course doubly vulnerable and down to repair broken mains. Residents should be shearing of well pipes could be a problem. encouraged to have on-site storage sufficient for at least three days and it should be mandatory for business Vulnerable components within the Banks Peninsula relying on water to have on site storage. Toilet cisterns District reticulation are numerous. The 300 mm trunk and hot water cylinders hold a certain amount but roof main from the Dyers Road well is strapped to the side storage tanks should be installed. Garden irrigation of the bridge on the Tunnel Road with no means of should be banned after any extensive damage to the shutting it off quickly. water supply systems.

The Heathcote pump house is a very old building and Mitigation measures proposed for the reservoirs and likely to be damaged during an earthquake. All reser- pumping stations include: voirs in Heathcote and Lyttelton are old and not de- • seismically strengthen buildings that are vulner- signed to meet current seismic standards. Both tunnel able and considered important enough; portals present likely damage to the watermains if ground movement in the area is large. • restrain all reservoirs that are presently unrestrained;

Pump houses and reservoirs are 70 to 100 years old. • initiate an audit to check on the strength of all reservoirs against current code requirements and The trunk mains to Diamond Harbour and Governors strengthen where appropriate; Bay are likely to be damaged due to landslip. • liaise with Southpower as regards security of sup- Mitigation measures ply; At first glance it would appear that the Christchurch • restrain emergency batteries, cabinets, caustic tanks, water supply system is rather vulnerable to earthquake spare motors, trolleys, sump pumps, computer damage but the following points need to be made equipment, switchboards, etc.; and which show the system in a somewhat better light: • initiate a programme of installing flexible connec- • The Christchurch system is a massive grid network tions to critical installations. and is not dependent on long isolated trunk mains that are not duplicated. With judicious operation of Where standby generators are installed, there should valves there are numerous routes available to get be a minimum diesel storage capacity sufficient for water from one point to another. As well as the 160 three days operation. kms of mains analysed, there are 1,140 kms of An earthquake response plan should be prepared to subsidiary mains that can be utilised. There is cover the following points: considerable redundancy built into the system. • the amount and nature of stores that should be held • Twenty five percent of the pumping stations have including stores held by neighbouring local au- standby diesel generators. thorities; Civil Services • 87

• staff requirements and duties, delegation authority, Strong winds will cause problems with loss of power. liaison with Civil Defence and other authorities; Falling tree branches and trees will block access and possibly damage buildings, reservoirs and aerials. • availability of equipment and manpower with pos- Standby diesel generators and battery backup for com- sible assistance from neighbouring local authori- munications will considerably lessen the effects of ties; wind. • response priorities with a list for checking of facilities using a standardised damage report form; Mitigation measures Damage to pump houses, reservoirs and aerials should • available option; and not be great. However, mitigation measures should • facilities to be available at emergency response include: centres. • increasing battery capacity for communications to 24 hours; Snow • checking all reservoir and pump house sites and Vulnerability access tracks on the Port Hills for trees that could be Snow in Christchurch is most likely in June, July, removed to minimise the problem; and August and based on the August 1992 snow storm • liaise with Southpower as regards security of sup- approximately 30 cm can be expected on the flat and up ply. to 1 metre on the Port Hills. Heavy snow could cause problems with loss of power, possible aerial damage and access problems to reservoirs. The effects would Tsunami be greater on the Port Hills (particularly if strong winds accompanied the snow). Vulnerability A typical tsunami that could affect Christchurch would However, with 25% of the pumping station having be caused by a large earthquake centred on coastal standby diesel generators and virtually all sites having South America. Widespread inundation would occur backup battery systems available for all radio commu- along the coast with water depths ranging from 0.2 m nication channels the effects should not be great. One and 1.2 m at New Brighton and South Brighton respec- further point is that water consumption during winter is tively. Moncks Bay area could have water depths up to low. 1.7 m with 0.7 m at Sumner.

Structural damage to reservoirs and pump houses is not It is not anticipated that pump houses will be damaged likely to be a problem. but pumps and electrical equipment could be put out of action at Effingham Street (1), Palmers Road (2), Mitigation measures Estuary Road (3), Moncks Spur number 1 (2) and The problems caused by heavy snow should not be Clifton number 1 (2). great and mitigation measures proposed for the reser- The Banks Peninsula District water supply could be voir and pumping stations include: affected by damage to pumps and electrical equipment • increasing battery capacity for communications to at the Dyers Pass Road pump station (2). 24 hours; Water supply could also be affected by:

• ensuring that at least one four-wheel drive vehicle • inundation from beach overtopping; with chains is available and that chains are available for other vehicles; and • inundation from Estuary; and

• liaise with Southpower as regards security of • combined inundation from beach and Estuary. supply. Mitigation measures Wind With the extensive grid network of the reticulation system it will still be possible to supply water to Vulnerability residents even if the above stations are out of action. Strong winds can be experienced in Christchurch at all Mitigation measures proposed are: times of the year and typically are northwest or southwest, but strong northeast winds are also possible. • ensuring that all electrical equipment is, if possible, 88 • Risks and Realities

mounted above possible inundation levels; and of any one of these is usually every nine to ten days. Each of the tankers bring in the same product grades, • investigating the feasibility of waterproofing the although Taiko is more black oil based (diesoline, pumping stations. heavy and light fuel oil, bitumen), Kuaka and Kotuku are more white oils (motor spirit, jet A1). Other over- Local flooding seas vessels (imports) bring in lesser quantity products (avgas 100, mineral turpentine - HAWS, white spirits Vulnerability - LAWS) and generally arrive to replenish shore stocks Local flooding of the Avon, Heathcote and Styx Rivers every two to three months. and their tributaries would not seriously disrupt the city The average discharge from the refinery-based vessels water supply. Access could be somewhat restricted but amounts to approximately 16 to 17 million litres which is not considered to be a serious problem. is generally made up of (say) eight million litres motor Pumps and electrical equipment could be affected at spirit, five million litres diesoline, and the balance jet Chapmans Road, St Johns and Bexley. A1. These product quantities are pumped ex ship at a rate of 700,000 litres per hour via one or both 200 mm marine loading arms to 250 mm wharf lines (hardpipe) Mitigation Measures and into industry storage tanks. The loss of the above stations will not seriously affect the system’s ability to supply water to the residents. Jet A1 (dual purpose Kerosine), HAWS, LAWS, Avgas Mitigation measures proposed are: 100, LFO, HFO and bitumen is permanently stored at Lyttelton and distributed by tank truck from there. • ensuring that all electrical equipment, if possible is mounted above possible inundation levels; and Motor spirit and diesoline are transferred from any Lyttelton storage tank through to the Mobil storage • investigating the feasibility of waterproofing the facility at Woolston (and distributed by tank truck from pumping stations. there) via the Lyttelton-Woolston pipeline which is of 100 mm diameter. The pipeline which is in operation General review of standby capacity 18-20 hours per day Monday through Friday, transfers A total review of the water supply standby diesel at a rate of approximately 80,000 litres per hour and is capacity is required and this review is scheduled to be operated solely to provide day-to-day requirements for undertaken in the next 12 to 18 months. The review will all industry product drawoff at Woolston. include location, total capacity and suitability of exist- Total port storage covering the products mentioned in ing equipment. still-in-use tanks include:

The review is required not only for emergency situa- • motor spirit 27,900,000 litres (7,300,000 litres at tions but also for factors such as load shedding to Woolston); ensure that the best use is made of energy tariffs. • diesoline 16,300,000 litres (2,700,000 litres at Woolston); and

4.5 Petroleum Products • jet A1 (DPK) 17,300,000 litres (2,900,000 litres at Woolston). Description There are three coastal shipping tankers that provide Vulnerabilities bulk stock replenishment into the Lyttelton tank farm The petroleum industry is very concerned at the vul- storage. nerability of its systems to fire and efforts to ensure Kotuku, Kuaka and Taiko are each contracted to freight safety from a fire viewpoint are very much in the petroleum products to all New Zealand ports from the forefront. New Zealand Refining Co at Marsden Point in The industry is in the process of a full redevelopment Whangarei. Kuaka and Kotuku were built in 1975 and of the Lyttelton tank farm (see Chapter 11, Section carry a capacity of 34.8 million litres. Taiko was built 11.9) and consultants have been employed by the in 1984 and has a storage capacity of 34.6 million litres, various oil companies in designing for this work. of which 5.6 million litres are bitumen dedicated. Although the Hazards Group is concerned at the pos- Approximately half of each tanker capacity is trans- sible liquefaction potential of the tank farm area, such ferred from ship to shore at Lyttelton and the frequency tests as have been undertaken have convinced the oil Civil Services • 89

companies to their satisfaction that such concern is the progressive redevelopment of the tank farm is unwarranted, at least in the foundation areas for the making the area very much safer. tanks. However there is less certainty regarding the stability of the bunds (the “stopbanks”) around the tank Mitigation Measures farm and there is a possibility that the stability of the Most of the measures relate to automatic shutting off of seaward edge of the tank farm is in question. valves in the event of anything unusual and the newer Detailed examination of the vulnerabilities requires structures (the petroleum pipeline and the installations more skills than are presently available in the industry at Woolston) are considered by the industry to be in Christchurch, and because of the redevelopment adequate. proposals the industry has chosen not to participate in In the event of the Lyttelton port becoming unusable the formal vulnerability analysis. This is not from any then petroleum products could be provided to Christ- lack of concern but simply an acknowledgement that church from other ports, particularly Nelson or Timaru.

Lyttelton tank farm 90 • Risks and Realities Electrical and Communications • 91

Chapter 5 Electrical and Communications

5.1 Telecom New Zealand Calls between parties connected to different local exchanges in the greater Christchurch area are gener- Limited ally connected via a direct single junction link. (Small rural exchanges are the exception.) Description of the Telecom network

Telephone network Tandem exchanges There are three types of tandem exchange in Christ- Telecom New Zealand Limited (TNZL), together with church. a number of other telecommunication operators, provides telecommunication services in the greater Christ- • The local tandem feature of Christchurch central church area. TNZL operates a very wide range of switch which hosts a small number of rural ex- services and technologies as a single network through- changes. out New Zealand. • The two toll tandem exchanges at Christchurch This network is one of the most modern in the world which provide access to the rest of New Zealand with considerable investments and operating systems and overseas and are thus referred to as “toll ex- to support a high level of robustness and diversity. changes” These have dual routes to all the LXs in the greater Christchurch area and are configured as Telephone Exchange Types a ‘mate’ pair to reduce the impact of a junction or switch failure (see Network Diversity). The network consists of the following types of exchanges in a broad form of hierarchy: • Tandems performing hosting for the RLUs within the Christchurch free calling area. • Remote line units (RLU);

• Local exchanges (LX); Junction routes • Tandem exchanges (TX); and Except for circuits between toll exchanges, circuits provided between exchanges at different sites are re- • Toll exchanges (SX). ferred to as “junctions”. The majority of the junction routes in the Christchurch area are provided using fibre Remote line units optic transmission system (FOTS). The fibre optic cables are drawn into PVC ducts and terminal equip- A remote line unit (RLU) is dependent upon another ment is located at each exchange to derive the junctions switch (its host) to provide service. All calls to or from required. The layout of the Christchurch metropolitan a RLU are switched via the host exchange, even when area has enabled Telecom to construct a network of calling and called parties are connected to the same FOTS links between the major exchange sites in a RLU. configuration that provides two or three independent All RLUs in Christchurch have either diverse dedi- routes out of each site. cated links to their host switch, or have stand-alone control modules (SACMs) which will provide local Trunk routes switching functions if the link to the host exchange is Circuits between Christchurch’s toll exchanges and lost. toll exchanges elsewhere in New Zealand are referred to as “trunks”. Major trunk routes from Christchurch, Local exchanges which use either microwave radio or fibre optic sys- By contrast, a local exchange operates as a stand-alone tems, are as follows. unit. When both the calling and the called parties are connected to the same local exchange, a call can be North Island established without assistance from another exchange. Digital Microwave Radio, (DMR) and Fibre Optic 92 • Risks and Realities

Transmission systems (FOTS). pleting calls, but leave the rest of the network operating normally. More widespread events can cause a general South Island network blockage with a collapse of its ability to function through overloading. Two FOTS as far south as Waimate, then a mix of FOTS and DMR to Dunedin. Traffic management systems are used to control these situations by restricting the ability to originate calls. Network Management However, it is vital that radio and television can be used The Network Management Centre at Hamilton is staffed to discourage unnecessary use of the telephone net- 24 hours per day to monitor the state of the network, work following a major disaster. Failure to do so and is able to implement traffic control measures means that load shedding will have to be activated with within minutes of problems being detected. a concomitant risk of disallowing emergency calls. Currently the cellular network could suffer even greater overloading if many of its cells are out of service. Network diversity Within the Christchurch metropolitan area a high level An example of this process occurred during the of bearer diversity has been achieved by creating a Edgecumbe earthquake where, because of high levels ‘ring’ system with ‘spokes’ to the central hub. of calls incoming to the area, it was necessary to apply 100% restriction to calls from selected exchanges such The major exchanges in the Christchurch area are all as International and Auckland for several days after the directly connected to each other over this bearer sys- event. tem with the maximum diversity that current technologies permit. A minimum of two (and often three) This behaviour will be repeated in every civil emer- independent routes are established between each pair gency, i.e. network overloading because of customer of exchanges, so that the impact of the ‘normal’ failure behaviour, even though no physical network failure mechanism, i.e. single damaged cable or FOTS failure, has occurred. is minimised. In the event of such a failure there would be an immediate impairment of service but not a total Mobile radio network loss. Procedures for restoration require manual reconnection of bearer plant and may involve a visit to a The mobile radio network (both VHF and UHF) pro- suburban site. vides facilities for communication between fixed sites and portable radio equipment, which can be either Trunk diversity both north and south of Christchurch is mounted in vehicles or hand-held. It does not normally available as parallel dual digital FOTS links (in com- provide access to or from the public telephone net- bination with digital microwave radio). These dual work, although the new trunked dispatch service does links traverse the length of the South Island and then provide such access. continue through the North Island.

The network is structured to provide alternative rout- Fixed radio sites ing of traffic should failure of exchanges, trunk or VHF radio repeater equipment, comprising both trans- junction routes occur. For customers other than those mitters and repeaters, is provided for the mobile radio connected to a particular exchange, any such failure service at elevated sites. would result in impaired service, not total loss. Total loss generally requires the simultaneous loss of several Each transmitter and receiver unit work as a set, on exchanges, trunk or junction routes. separate transmit and receive frequencies called a “Channel”. The particular channel transmitter is only Customer Behaviour active when switched on by the user’s base station or mobiles. The base station is linked to the radio site by Since the introduction of SPC exchanges Telecom has either land-lines or radio. had the capability of monitoring calling behaviour on a real-time basis. Experience with monitoring has shown that normal network performance is greatly Mobile radio privacy affected by customer calling patterns and these are in Mobile radio operates by broadcasting calls and recipi- turn affected by calendar events or local events. Expe- ents need to identify their own calls. Although this rience has also proved that small-scale local events can results in a lack of privacy, this method of operation considerably increase calling rates. Localised events can have its own advantages, e.g. when making enquir- can result in network congestion in a local area and ies as to availability or location of mobiles. cause difficulties for the affected customers in com- Electrical and Communications • 93

Trunked dispatch (Fleetlink) Cellular radio sites in the Christchurch area Many of the mobile radio systems described above Radio equipment, comprising both transmitters and have been upgraded to a new system called trunked receivers, are provided for the cellular radio service at dispatch. This system dynamically allocates channels several sites, spread throughout the greater Christ- to customers under control of the repeaters. Simplex church area. As cellular telephone traffic grows, addi- operation is also possible if mobiles are equipped for it. tional cellular radio sites will be provided, with the area covered by existing sites reducing accordingly. System channel sharing Although some mobile radio services use dedicated Links between cellular radio sites and the MSC channels many users share channels, and perhaps trans- All cell sites in Christchurch are connected to the MSC mitter/receiver equipment. Thus a number of remote via fibre optic transmission systems and a limited control units may be controlling a particular transmit- amount of digital microwave radio. ter. Sharing of channels may result in congestion should calling patterns increase (as is likely following Assessment of the Effect of a Major a major disaster). This applies particularly to cellular Earthquake and trunked dispatch systems. This section defines the expected impact that a moderately large to large earthquake of 150 year return period Fixed radio equipment failure would have on the Telecom networks. Owing to the nature of the operation of mobile radio, all radio communication normally passes through the Vulnerability fixed radio site repeater unit. This has the disadvantage Vulnerability of the telecommunications network was that the failure of the repeater associated with a particu- assessed in accordance with the methodology used for lar channel usually means that channel is no longer able other lifelines (see Figures 5.1, 5.2 and 5.3). The to be used to communicate between any other radio following points emerged from the review. transceiver on that channel. However by using “Simplex” mode, it may be possible for these mobiles • All telecommunications equipment is critically re- to communicate to other similarly-equipped mobiles, liant on the power supply. Standby power genera- but limits to range will be imposed by the transmitter tion equipment (to provide an essential supply) is power available and the terrain. provided at most locations. The adequacy of fuel reserves at individual sites is being reviewed, giv- Cellular radio network ing consideration to the possibility of a prolonged power outage and likely difficulties in obtaining The cellular radio network provides facilities for com- access to the sites for refuelling. This underlines the munication between the public telephone network and interdependence of lifelines and the need for mu- portable telephones, or between portable telephones. tual understanding of both resilience and expecta- Selection of the desired distant party is by dialling. tions.

Method of operation • Switching equipment relies on air conditioning equipment to ensure the temperature does not ex- Portable telephones using cellular radio can be either ceed certain levels. While air conditioning equip- mounted in vehicles or hand-held. ment is not operating, temperatures need to be It operates on the basis of radio terminals being allo- monitored to ensure they do not rise to damaging cated a specific area of coverage known as a “cell”. As levels. Air conditioning equipment is reliant upon a cellular telephone moves from one cell to another, the power and, in some cases, a water supply (replen- second cell’s radio terminal automatically takes over ishment only), and adequate seismic restraint of all the provision of service without the user being aware of fans, ducting, pipework and chillers. the change. This operation is controlled by a central- • The telecommunications network makes good use ised cellular radio switch (MSC), which also dynami- of diversity by means of alternative physical routes cally assigns one of a limited number of radio fre- and alternative media. However, there are still quency channels associated with each cell to the user at some parts of the network where improvements can that particular time. be made. All exchanges in the Christchurch area As the cellular radio service is connected to the public are stored program control. This has resulted in switched telephone network (PSTN), access is avail- some centralisation of switching operations with able to and from the rest of the PSTN. 94 • Risks and Realities

UTILITY: TELECOM NZ LIMITED REGIONAL LOCAL NETWORK: TELECOMMUNICATIONS

VULNERABILITY IMPACT

TO HAZARD OF DAMAGE

IMPORTANCE 1 - 5 IMPORTANCE EARTHQUAKE HAZARD GROUND SETTLEMENT RETURN TO NORMALITY RETURN TO LIQUEFACTION GROUND SHAKE DURING EARTHQUAKE LANDSLIDE PERIOD FOLLOWING NETWORK COMPONENTS AFTER IMMEDIATELY COMMENT ELEMENTS

EXCHANGE A Site 2 3B Building 1 1 2 0 1 00 0 3 South Christchurch Telecom Equipment 2 1 1 00 0 0 22Check top restraint of transmission equipment Mechanical Services 1 2 1 00 0 0 0 2 Check restraint of air conditioning duct: roof space Electrical Services 1 1 0000 0 0 2 Check restraint of exchange batteries EXCHANGE B Site 2 1 Bank at back of site Building 1 01 1 0 00 0 0 Southeast Telecom Equipment 2 2 0 00 0 0 22Check top restraint of transmission equipment Christchurch Mechanical Services 1 1 0 0 0 0 0 0 2 Re-support cable runways below ceiling of 61K Electrical Services 1 1 0 0 0 0 0 0 2 NEAX exchange area EXCHANGE C Site 2 3 Building 1 01 0 0 00 0 3 Building strengthening recommended by recent North Christchurch Telecom Equipment 2 1 0 00 1 1 22detailed structural assessment of building Mechanical Services 1 1 0 0 0 0 0 0 2 Fluorescent luminaires could swing and impact RLU Electrical Services 1 2 0 0 0 0 0 0 2 during earthquake EXCHANGE D Site 2 3 Building 1 02 0 0 00 0 3 Brick end wall of building (south end) may Southwest Telecom Equipment 2 2 0 00 1 1 22collapse. Provide top fixing for rack in the Christchurch Mechanical Services 1 2 0 0 0 0 0 0 2 transmission equipment row RLU Electrical Services 1 1 0 0 0 0 0 0 2 Provide restraint for cable air drier EXCHANGE E Site 2 3 Building 1 00 0 0 00 0 0 West Christchurch Telecom Equipment 2 1 0 00 1 1 22 Mechanical Services 1 1 0 0 0 0 0 0 2 Provide restraint for hose reel water pressure tank RLU Electrical Services 1 1 0 0 0 0 0 0 2 Check restraint of exchange batteries EXCHANGE F Site 2 3 Building 1 01 0 0 00 0 0 Southwest Telecom Equipment 2 1 0 00 1 1 22 Christchurch Mechanical Services 1 1 0 0 0 0 0 0 2 Fluorescent luminaires could swing and impact RLU Electrical Services 1 1 0 0 0 0 0 0 2 during earthquake EXCHANGE G Site 2 3 Building 1 02 0 0 00 0 2 Possibility of failure of brick veneer, otherwise Northwest Telecom Equipment 2 2 0 00 1 1 22building grading = 1 Christchurch Mechanical Services 1 1 0 0 0 0 0 0 2 Provide additional top restraint for MDF RLU Electrical Services 1 1 0 0 0 0 0 0 2

Figure 5.1: Sample vulnerability chart 1

remote line units being dependent upon their hosts ing discontinuities in soil types, are likely to be for switching of calls. damaged. Cables at bridge crossings would be especially vulnerable. • Local reticulation (exchange to customer) is mainly by underground cables using copper conductors. • Junction and trunk cables between exchanges are By its widespread nature this reticulation is suscep- by fibre optic cables installed in ducts. These are tible to damage, with certain common types of vulnerable in areas subject to significant ground cables, now obsolete, more likely to give problems. movement, e.g. crossing a fault, in areas subject to Any cables affected by severe shaking or signifi- liquefaction, or where a landslide carries both ducts cant differential ground movement, such as at cross- and cables with it. Electrical and Communications • 95

SAMPLE VULNERABILITY CHART

UTILITY: TELECOM NZ LIMITED REGIONAL LOCAL NETWORK: CABLING

VULNERABILITY TO HAZARD IMPACT OF DAMAGE

LANDSLIDE

COMMENT

LIQUEFACTION

ZONE BOUNDRY

GROUND SHAKE

IMMEDIATELY AFTER IMMEDIATELY

IMPORTANCE 1-5 IMPORTANCE

PERIOD FOLLOWING

COMPONENT ELEMENT

DURING EARTHQUAKE

GROUND SETTLEMENT RETURN TO NORMALITY RETURN TO

CH-MH954A MANCHESTER 1 1 1 1 1 1112 CABLE, 79 2000

CH-MH957A HIGH ST 1 11 1 1 1 11 2 CABLE, 78 2000

CH-MH985B HIGH ST 1 11 1 1 1 112 CABLE, 53 3200

CH-MH960A FERRY RD 11 1 1 1 1112 CABLE, 77 2000

CH-MH968 FERRY RD 1 11 1 1 1112 CABLE, 80 2000

CH-MH310 MANCHESTER 1 33 333333CABLE, 44 2000

CH-MH291 COLOMBO 11 1 111112 CABLE, 50 4000

CH-MH291 COLOMBO 11 1 111112 CABLE, 74 2000

CH-MH101 HEREFORD 1 1 2 222223 CABLE, 48 2400

CH-MH105 COLOMBO 11 1 111112 CABLE, 67 2000

CH-MH105 COLOMBO 11 1 1 1 11 1 2 CABLE, 64 2000

CH-MH303 HEREFORD 1 11 1 1 1 1 1 2 CABLE, 72 2000

CH-MH968 FERRY RD 1 11 1 11 11 2 CABLE, 68 2000

Figure 5.2: Sample vulnerability chart 2

• Where ducts and cables enter buildings in areas intact. However, modern systems are reasonably subject to liquefaction, damage may occur due to tolerant of movement. differential movement between the surrounding ground and the building. • Exchange buildings have varying degrees of earthquake resistance, mainly depending upon their age • Aerial systems for microwave radio could become and the attention given to detailing for seismic misaligned even when the supporting structures are 96 • Risks and Realities

SAMPLE VULNERABILITY CHART

UTILITY: TELECOM NZ LIMITED REGIONAL LOCAL NETWORK: CABLING

VULNERABILITY TO HAZARD IMPACT OF DAMAGE

LANDSLIDE

COMMENT

LIQUEFACTION

ZONE BOUNDRY

GROUND SHAKE

IMMEDIATELY AFTER IMMEDIATELY

IMPORTANCE 1-5 IMPORTANCE

PERIOD FOLLOWING

COMPONENT ELEMENT

DURING EARTHQUAKE

GROUND SETTLEMENT RETURN TO NORMALITY RETURN TO

HTN-HSL 3 1 2 2 2 2 2 2 3 342

HTN-HSL 1 1 2 2 2 2 2 2 3 308

HTN-MH249 3 1 1 1 1 1 1 1 2 337

MH-BKM249 3 1 2 2 2 2 2 2 3 337

BKM-CAS 3 1 2 2 2 2 2 2 3 314

ISL-WILMERS RD 3 1 1 1 1 1 1 1 2 341

WILMERS RD-HSL JNT RD 3 2 2 2 1 2 2 2 341

HSL JNT RD-HSL EXCH 3 1 1 1 1 1 1 1 2 341

ISL-TRANS 4 1 1 0 1 1 1 524

HTN-HSL 2 1 2 2 2 2 2 2 3 309

Figure 5.3: Sample vulnerability chart 3

capacity during their design. Some older buildings South Island are completely dependent on the cen- need to be brought up to an acceptable standard. tral switch. Failure of this switch or the links to it would completely disable these services. • The operation of some telephone exchanges may be unstable due to congestion caused by repeated • The fleetlink trunked dispatch service is normally redialling attempts. Dial tone may be slow to dependent on a central switch. However, failure of appear. this switch or the links to it would merely disable the linking between repeater sites. Many users • The cellular and paging networks for most of the would be unaware of such a switch failure. Electrical and Communications • 97

Impact on the public-switched have proven to be an extremely useful and speedy telephone network (PSTN) way for utilities to augment their communications • Little effect on the majority of buildings and equip- systems. ment. • The management of cellular network congestion • Little disruption to customer services from damage by Telecom should be in favour of the utility and that is incurred by plant and equipment at network emergency organisations, and the restriction of sites. other Christchurch cellular customers should be considered to establish that preference. • Widespread calling congestion of all exchange switches. Dial tone may be slow to appear as a Impact on the land mobile radio result. services • Widespread damage to copper cables between cus- • Only a minimal loss of services overall (e.g. 10%), tomers and the exchange (“local access cables”). providing the timber masts at hill top radio stations This is due to water penetration through cracks in do not fail. the sheathes of lead-sheathed cables, which are in turn caused by severe ground shaking and differen- • Little effect on the land mobile radio equipment at tial earth movement in liquefaction zones. these stations.

• Prolonged restoration repairs to local access lead- • Loss of all physical links to the radio stations, sheathed cables over many months, but with virtu- resulting in the loss of access to the PSTN for fleetlink users. ally all other elements of the PSTN being restored to full capacity within one week. • Little disruption to the conventional land mobile channel services due to traffic congestion. • Little damage (except at some bridge crossings) to the route diverse fibre optic junction cable system, • Congestion may occur on fleetlink channels, but and hence little loss of junction links although there can be overcome by installing extra shared chan- may be some reduction in junction capacity. nels (within say two days). • A strong possibility of the failure of both trunk links north of Christchurch due to lack of geo- Impact on paging network graphic diversity. • Loss of the entire Canterbury paging network due to the trunk links north being lost, or due to loss of Impact on the cellular network vulnerable cable links between the paging switch and the hill top radio stations. • Very heavy cellular network congestion for the first two hours, with some relief after three to six • Paging network should be fully restored within two hours (as service demands are impeded through to four days by using temporary radio links. individual cellphone batteries depleting and the probable lack of mains power for re-charging). Mitigation measures • Disruption of service at some cell sites due to the Telecom is undertaking a number of mitigation meas- depletion of battery reserve following the loss of ures which have been identified as a result of this mains power to sites. lifelines study. These measures are additional to the normal earthquake and seismic control practices which • Disruption of service from some cell sites due to the provide a good standard of protection to plant inside loss of their fibre optic cable links to the cellular buildings. Some of the mitigation measures identified switch. in the lifelines study are:

• Little effect on the equipment and masts at the cell • Telecom to establish pre-plans with broadcasting sites. media so that the public can be advised against the • Little effect on the connection of the cellular net- use of their telephones or cellphones after major work to the PSTN or the trunk network. disasters, unless necessary for their safety or to obtain supplies. • The cellular network should be restored to virtually full capacity within four days. • Restoration of internal communication to be given highest priority. • In recent major international disasters cellphones 98 • Risks and Realities

• Where several mobile cell sites have overlapping riri Flood Hazard”. A 500-year flood of the Waimaka- coverage, investigate the feasibility of carrying the riri River could break out on the south bank at: links to half the sites on alternate bearers. • Halkett (15% probability); • Telecom to keep back-up databases of critical plant records. • Crossbank (16% probability); and

• All electronic databases holding copies of network • McLeans Island (3% probability). records have back-up power supply to ensure these Flooding at Christchurch Airport, or in the upper records remain available. reaches of the Avon, for example, could be caused by • Lead-sheathed cables to be recognised as having a breakout at either Halkett or McLeans Island. The higher risk of damage than modern plastic types. probability of this happening in any one year is (15% Cables to important lifeline sites such as Civil + 3%) x 1/500 = 0.036% (equivalent of an event with Defence, schools, Southpower, hospitals, fire, po- a 3,000 year return period). lice to be reviewed with this in mind. Map 12 (p 295)shows the location of key nodes and • Some minor building structures to be strengthened. cable routes in the Telecom network with reference to the general flood plain expected for the Waimakariri • Some switching cabinets to be given additional top 500-year flood scenario, if breakout occurs at all three support. of the above locations. Note that the flood plain shown is slightly offset from its proper location due to slight • Casters to be removed from computer tables and differences in the street/river mapping databases used replaced with flat footing. by Telecom and the Christchurch City Council. The • Bookcases and filing cabinets to be secured to more accurate Christchurch City Council plans (CCC avoid toppling. flood plain shown on a CCC street/river mapping database) have been used to assess the vulnerability of • Flexible connections to be provided on engine specific Telecom sites. alternator fuel lines to storage tanks where required. There should be some advance warning of this event, as rainfall monitoring in the Waimakariri catchment • Fuel supplies to be reviewed for extended periods can give in excess of 12 hours warning of a threatening of mains power supply unavailability (also trans- situation. If a breakout occurs, it will take four hours port of fuel to sites after a disaster). for the floodwaters to reach the outskirts of Christ- • Some power plant (batteries and rectifiers) require church, eight hours to reach the Square, and over 12 additional bracing. hours to reach the Estuary.

• Some roof-mounted air conditioning condensers to Except at “bottlenecks”, flow rates should not exceed be checked for adequacy of bracing. 4 km/hr. The average depth of flood channels in the western part of Christchurch will vary from 200 mm to • Continue with the installation of diverse bearer 350 mm. Slightly greater depths of flooding can be routes. expected at “low spots”. Much greater depths of flooding can be expected within the Avon and Heathcote Assessment of the Effect of Flooding river terraces. The duration of the flooding could be up Two flooding scenarios were considered: to two days west of the Square, and up to three days east of the Square. • a 500-year flood in the Waimakariri River causing the southern stopbanks to fail; and The flooding is expected to cause:

• sustained extremely heavy rainfall event in Christ- • some roads to be impassable due to the depth of church causing local flooding. flood water; • major surface flooding in many suburbs (see Map Waimakariri River 500 year flood 12, p 295);

Scenario • the flood waters to contain silt, and possibly sewer- For details of the scenario, see Section 2.3 “Waimaka- age; and • possible loss of the road and rail bridges across the Electrical and Communications • 99

Waimakariri River, and/or washing out of their cessfully salvaged and restored to full service within approaches, with consequential loss of services days. carried by these bridges. Exchange cable wells If heavy rainfall falls on Christchurch prior to, or during, this flood, these effects will be further aggra- High water levels in Telecom ductlines/manholes may vated. cause exchange cable wells to fill with water at a rate faster than the permanently installed sump pump can handle. In many cases, the sump pump outlet pipe Impact of flood levels discharges into a gully trap below the “extreme” flood Buried cables level — hence it may not be able to work in these “extreme” flood situations. If the water level rises to Telecom buried cables are designed to function in wet within one metre of the exchange floor level, in many environments, so the effect on them should be mini- exchanges this could result in unfilled unpressurised mal. Surface water flooding levels will increase the plastic cables, installed between the air dams in the static water pressures on cable sheaths, and in some cable well and the main distribution frame on the floor cases this will result in some water penetration, bring- above, becoming partially submerged. Faults affecting on faults within one to two days after the flood. ing many customers could result. These will be relatively randomly distributed in the “flooded” areas of Christchurch, and should only affect a small proportion of Telecom customers. Other sites Both local circuit area (LCA) equipment buildings Cabinets may be flooded. The general comments made above about exchange equipment also apply to the electronic These are almost always located in the road reserve at transmission equipment at these sites. The one differ- the fenceline. Flood levels up to 200 mm in the cabinet ence is that, given sufficient warning, it may be possi- will have no effect. Beyond this, up to 100 customers ble to relocate/remove the equipment and back plane could lose service for each additional 40 mm of flood wiring at the foot of the racks, and keep the equipment level. working at full, or near full, capacity, even with water up to 200 mm above floor level in the building. Pillars Are located in the road reserve at the fenceline. Flood Toll links levels up to 400 mm will normally have no effect. There are two diverse toll links north, and two diverse Beyond this level, faults may occur, typically affecting toll links south. There is a possibility that flooding of only one to four customer lines. the Waimakariri River could affect each of the toll The only cabinets and pillars expected to be signifi- links north. Since the probability of both these events cantly affected by the flood levels are those few that are happening is negligible, and either link can support located in river terraces, where the deeper flood levels 99% of normal peak toll traffic, minimal effect on the (up to 1.0 m to 1.5 m) could occur. Only a very small toll traffic north is expected. proportion of customers would be affected. Both of the diverse toll links south are carried by buried optical fibre cables, and will be unaffected by flooding Exchange equipment from the Waimakariri River. Exchange equipment will be at risk as soon as flood levels rise more than 50 mm above the exchange floor Cellular network level. Of those looked at, two Telecom exchanges are Two of the twelve Christchurch cell sites could suffer at risk of this. These are both outlying exchanges that from flood immersion to varying extents, but loss of are not essential to the Telecom network. both these sites would only have a minimal effect on If the equipment is powered up at the time, irreversible the coverage and service level of the surviving cellular damage could occur, requiring replacement of the network. immersed equipment (and possibly other equipment as well). If the equipment is powered down before Land mobile radio/Fleetlink/paging immersion, it is expected that once the flood levels These networks should be completely unaffected. The have subsided the immersed equipment can be suc- only impact that seems possible, if the Colombo Street bridge and/or the Ferrymead bridge over the Heathcote 100 • Risks and Realities

River is lost, or its approaches are washed away, is loss one suitable bridge crossing of the Heathcote River and of the Telecom cables carried by the bridge(s). This the Avon River, Wairarapa Stream and Hewlings seems extremely unlikely since the floodwater veloc- Stream. ity is only expected to be little more than 2 km/hr. Even if this happened, this would only result in the loss of the Summary fleetlink access to the public switched telephone network (PSTN), some land mobile radio landline con- Public switched telephone network nections, and the loss of all paging signals to one of the A relatively small number of customers may lose two Christchurch paging transmitters. Since all paging service. Otherwise, except for some initial congestion, signals are broadcast from both the transmitters, loss of this network will be virtually unaffected. one bridge will only result in the Christchurch paging coverage being slightly decreased (i.e. more “black spots”). However, loss of both bridges (or their ap- Cellular network proaches) will result in the total loss of the Christ- Virtually unaffected. church paging network. The basic land mobile and fleetlink radio operations would continue essentially Land mobile/fleetlink/paging networks unaffected. Unaffected. Traffic overloading Repair work required In common with all of the scenarios, some initial congestion on the telephone, cellular and Fleetlink Much of the repair work will not be able to begin until networks is expected. If people are at home during the the flood levels have partially or fully subsided. This flood, it is expected that the telephone network will be is expected to take between one and three days (pre- very busy, but shouldn’t be catastrophically over- sumably from the time the rainfall eases off in the upper loaded. Some exchanges may be subject to minor load Waimakariri River catchment). shedding. Loss of some cell sites may increase traffic on other cells, but since most of the PSTN will still be Cable network functional this is not expected to be a problem. Any connectors in pillars, connection blocks in cabinets and telephone sockets in houses that have been Mains power immersed will probably have to be replaced. Associ- If it is assumed that all Telecom sites in, or near the ated cabling and mounting hardware will have to be flood plain (see Map 12, p 295), lose mains power, the hosed down (to remove silt, etc) and dried. Unfilled only sites affected that do not have their own engine cables suffering from water penetration will have to be alternator set mains back-up are: opened up and dried, or replaced. Service should be restored to most customers within one week, and to all • one exchange; customers within two weeks.

• four cell sites; and Exchange equipment • both LCA equipment buildings. Assuming any exchange equipment which suffered All these sites have many hours battery reserve. immersion was powered down before the event, sal- Telecom has several transportable engine alternator vage and full restoration is expected to be possible. sets in Christchurch so, at the very worst, only a few of However this could take up to three to five days after the least important cell sites or LCA equipment build- flood waters subside to below floor level. ings would suffer total power loss. This would not If the equipment is still powered up when immersed, have a significant impact on either the cellular network irrecoverable damage will result. The electronic cards or the telephone network. will need to be replaced, and the backplane wiring and connectors cleaned out. Depending on the availability Road access of spare cards, this could take from one to four weeks. Some roads may be impassable for virtually all vehicles, and others may only be useable by high wheel Toll links base vehicles such as Hi Luxes or trucks. However, If both toll links north are lost, a temporary limited this should not present any great difficulties to the capacity radio link bypassing the affected area could be Telecom repair effort, so long as there remains at least Electrical and Communications • 101

installed within two to three days. Restoration of a full in floodwater (after flood levels have sub- capacity link north will depend on the lesser of the sided). At least one “kit” should be estab- times needed to restore either link north. lished in Christchurch for this purpose.

• Make provision for a temporary 140 Mbit radio link Mitigation measures to bypass the vulnerable section in one of the toll • In exchange cable wells, seal all incoming ducts/ links north. holes/slots (including the sump pump discharge pipe hole). If the cable wells have external manhole • Make provision for external connection of a mains lids, these should be either sealed, or fitted so that supply from a transportable engine alternator set at when immersed, water will only “trickle” through all cell sites, LCA equipment buildings. them. This should stop water entering the cable • Ensure the circuit breakers protecting any power well at more than a trickle, even under severe flood points in exchange cable wells (i.e. below floor conditions. level), are not also protecting other important power The seal must be able to withstand a reasonable feeds above floor level. head of water (say 3 m) and should preferably be fire retardant. Duct caps are not an acceptable Sustained extremely heavy rainfall in means of sealing ducts. Note that ALL ducts/holes/ Christchurch (local flooding) slots must be sealed — flood water must not be allowed to “pour” into the cable well. Scenario If there is sustained heavy rainfall in Christchurch, • Sump pump discharge pipes should end well above street flooding on the flat, flooding in the vicinity of the any anticipated flood levels. Typically they should Avon, Heathcote and Styx Rivers, and slips on the Port end at least 500 mm above ground level. Hills are likely to occur. • Small electronic equipment buildings that have no cable well should have all incoming pipes or holes Street flooding (full or empty) sealed (to withstand a 3 m head of Figure 5.1 shows a typical road cross-section to which water). ALL open pipes must be sealed, including new roads and many existing roads are currently de- those containing other services like the incoming signed. The formed road “kerb-to-kerb” is designed to mains power. contain rainfall from a five year (return period) storm. • Prepare preplans for: Rainfall from a 10 year storm should be contained within the road reserve, “fenceline-to-fenceline”. — Sandbagging or fitting shutters to the ex- Building floor levels 150 mm above fenceline ground change/LCA equipment building doors (and level should remain dry in a 50 year storm, and floor any other “water entry” weak points) to levels 250 mm above fenceline ground level should ensure that even when flood levels are above remain dry in a 150 year storm. floor level, the flow of water into the building is restricted to a trickle. Figure 5.1 does not apply to many of the (older) roads/ footpaths in Christchurch. Also, the dimensions shown — Bringing in a bigger pump, if the sump on it are fairly approximate — they should be taken as pump can not cope. Ensure there is some an “indication” only of susceptibility of properties to window/hole/vent above the maximum flood street flooding. level, through which the pump discharge hose can be fed. All roads in Christchurch have a minimum grade of 1:500, to channel rainfall into stormwater collection — Relocating/removing equipment cards and points. These could be rivers, creeks or major backplane wiring from the bottom 200 mm stormwater pipes. Christchurch City Council (CCC) of racks in both LCA equipment buildings. 1:2000 stormwater plans shows the street water (down- hill) flow direction, the stormwater collection points, — Removing all power, DC or AC (including and their associated catchment area. engine alternator generated mains) from all equipment, once the water level rises above Local street flooding levels can rise to more than the floor level (if the first two measures fail). levels shown on Figure 5.1 if:

— Cleaning and salvaging all electronic cards, • The stormwater drains, pipes, kerbing, culverts etc, wiring and connections that were immersed become blocked with leaves, debris, etc. 102 • Risks and Realities

150 yr STORM FENCELINE 100 10 yr 50 yr STORM 150 STORM 5 yr STORM 100

125

3 m SEWER MANHOLE

Figure 5.1: Typical road cross-section

• Rainfall rates are so heavy that the stormwater pipe Impact of local flooding system can’t cope, causing water to “back-up”. All the general comments made in the section relating The worst hit areas will be those close to the to the impact of the Waimakariri 500 year flood (except stormwater collection points. for the comment on toll links), apply equally to this “local flooding” scenario. • The local area (or building) is in a “low spot”, relative to the general ground level in that area. Toll links Christchurch City Council 1:2000 sewer plans give the Both diverse toll links north will be unaffected by local absolute levels of sewer manhole lids (typically lo- flooding. The two diverse toll links south are carried cated in the crown of the road). These can be used to by buried optical fibre cables, which will only be measure the absolute levels of the kerb, fenceline affected by washouts of bridges and/or bridge ap- ground level, and floor level of particular sites of proaches at rivers south of Christchurch. Since the two interest. If accuracy is required, the sewer manhole lid routes south are reasonably geographically diverse, it should be lifted and the distance to the bottom of the is extremely unlikely both routes will be affected. sewer pipe invert measured. The absolute level of the invert is noted on the sewer plans, and this should be Specific Telecom buildings very accurate. The relative levels of the kerb, fenceline (ground level) Sewer manhole lids however are often raised or low- and floor were measured at: ered when the road is resealed, and these changes in level are not always entered on the CCC sewer plans. • eight of the twenty-two Christchurch exchanges; • both LCA equipment buildings; and Flooding in the vicinity of the Avon, Heathcote and Styx rivers • three of the nine cell sites (that are not in one of the above buildings). The expected local flooding in the vicinity of the Avon, Heathcote and Styx Rivers for a 1000 year storm, is These comprised the Telecom sites in, or near, the shown on Map 10. Most of these flooding areas and Waimakariri 500 year flood plain, the exchange in the associated flood levels are similar to, or less than, those Tsunami “flood plain”, and the three most important associated with the Waimakariri 500 year flood. Those Telecom exchanges. few local flooding areas not in the Waimakariri 500 Of these, two exchanges seem quite vulnerable to year flood plain are not in significant areas for any of flooding, having floor levels only 50 mm and 80 mm the Telecom networks. respectively above the fenceline level.

Slips on the Port Hills Five of the six other exchanges measured, both LCA equipment buildings and one cell site may also be This issue is covered in Section 2.8 “Slope Hazard” vulnerable, having floor levels between 150 mm and (page 45). 305 mm above the fenceline level. Electrical and Communications • 103

Since the actual levels of street flooding are likely to from street flooding and hence have to be powered vary considerably from site to site, and cannot be down. readily predicted, more exact determination of which sites will flood to above floor level is not possible. Cellular network If all incoming ducts/holes into the exchange cable Virtually unaffected. well are sealed, the exchange doors and other water entry weak points are sandbagged, and a fairly substan- Land mobile/fleetlink/paging network tial pump is working in the building, it should be possible to sustain outside flood levels of up to 400 mm Unaffected. above floor level, without the “internal” flood levels exceeding the critical 50 mm height. Repair work required All the comments made in the section relating to the Only one of the Telecom sites has a basement. If street impact of the Waimakariri 500 year flood on the cable flooding levels exceed 50 mm above kerb level, water network and exchange equipment, apply equally here, will flow down the vehicle access ramp into the base- except that if the basement of the central exchange ment. The roller door at the basement entrance is not building is flooded, and remains flooded for some time expected to withstand more than a metre or two of (with consequential failure of the air driers and subse- water pressure from water “pouring” down the ramp. quent water entry into some high pair count cables) Important equipment housed in this basement includes then it may take many months to fully restore service the Southpower 11kV/400V transformer supplying to the central business district. this site, and the air driers keeping up air pressure in (and hence water out of) most of the high pair count Telecom cables servicing the central business district. Mitigation measures If the mains supply is lost, the engine alternator sets All the comments relating to the impact of the (sited at ground level) will automatically start up and Waimakariri 500 year flood apply. Additional mitiga- carry the essential load. If the air driers fail, after some tion measures include: time water could enter major pair count cables, causing major loss of service throughout parts of the central • making provision at the top of basement access business district. down ramps for a 400 mm high barrier to be installed, when required. Store the barrier (and sand and sandbags) in the basement of the building. Mains power It is not anticipated that street flooding and local river • considering shifting air driers in basements up to flooding will have significant effects on the mains ground floor level. supply in Christchurch. Only a few isolated areas • measuring the relative kerb, fenceline, and floor could be affected. Since all but one of the twenty two levels at the remaining fourteen Christchurch ex- exchanges, and four of the twelve cell sites have engine change sites. alternator set mains back-up, only one exchange site, eight cell sites, and the two LCA equipment building sites could lose all mains power. With only a few sites Assessment of the effect of a major (if any) expected to be affected, the battery reserve at windstorm each of these sites should be ample to maintain service until one of the several Christchurch Telecom trans- Scenario portable engine alternator sets can be brought to the For details of the scenario, see Section 2.6 “Extreme affected sites to supply mains power. Wind Storm Scenario”. A 150 year north-westerly windstorm, causing gusts on the plain to 53 m/s to 57 m/s (180 km/hr to 205 km/hr) with lulls of 26 m/s Summary (94km/hr). Speedup over hills will accelerate the wind Public switched telephone network by 15% to 30%. Thus we have to consider peak gusts of 205 x 1.3 = 270km/hr at the radio sites on the Port If no exchange has to be powered down (in response to Hills, and 205 km/hr on the plains. The peak of the a threat of partial immersion), only a relatively small storm would only last for (say) three to four hours, but number of customers will lose service. Except for this would probably be followed by heavy rains com- some initial congestion, this network will be virtually pounding the damage. The following are expected: unaffected. However, it is possible that in extreme circumstances a few of the less critical of the twenty- • Flying objects to cause a lot of damage. two Christchurch Telecom exchanges could suffer 104 • Risks and Realities

• Some roofs to be lost and many windows to be the E/A sets can in fact be used to restore cellular broken. service.

• Some roads to be impassable due to fallen trees. Aerial wire • Widespread loss of power as fallen trees disrupt Many customers will lose service due to loss of their overhead power wires. The 220 kV feeds to the city aerial wire service connections. Flying debris and may also be down, thereby causing a total power falling trees pose the greatest threat. These customers black-out. will be scattered in pockets around the city. As the • Fires started by arcing power lines which the Fire programme to convert aerial service leads to under- Brigade will find difficult to control as they might ground leads progresses, this will become less of a not have road access and wind speeds will still be problem in the future. high making fire fighting both difficult and hazardous. Toll links north

• Nor’westerly windstorm to be followed by a south- Both of the two toll links north could be affected by a erly change bringing rain. windstorm. There will be little impact if only one link is lost. This scenario involves wind speeds only 20% higher than the 1 August 1975 storm. Toll links south These should not be affected as there is buried fibre for Probable extent of damage the whole route between Christchurch, Timaru, Dunedin and Invercargill. Towers and masts

There are four Telecom radio stations and twelve cell Roofs sites. The wind ratings of towers are based on historical data for the area and certain derating factors. Many The roof of one radio station is vulnerable due to its sites are designed to handle gusts of 320 km/hr (90 lightweight construction. Loss of roofs at mountain m/s) and have safety margins over and above this. tops is not unknown, as in November 1982 one radio Other sites will have lower ratings. Cellular towers are site lost a section of its roof. If roofs are lost, equipment designed to handle wind speeds of 180 km/hr (50 m/s) will probably still function, but rain following the with safety margins of over 30%. Hardwood masts windstorm could then damage/disable the equipment. have historically been designed by making them Roofs at other sites are expected to remain intact. stronger than anything that has previously been blown over. It is probable that masts and towers will survive Windows the wind loadings, but that some dipoles and other Many windows would be blown in or be broken by antennas on these masts could be destroyed. flying objects. Rain or snow following the windstorm would be able to enter the buildings through the broken In summary, this means that land mobile, paging, and windows and damage the equipment. Broken glass trunked dispatch networks could be partially disabled. would have to be extricated from the equipment during the clean up after the windstorm; this would pose a risk Loss of most cell sites to the availability of the network as during the work Eight of the twelve Christchurch cell sites do not have equipment could be accidentally disturbed/disrupted. standby power. Many of these could lose their mains Most critical network sites either do not have windows power supply. Those that do will completely fail after or they have been filled in. the five hour battery reserve has been used unless a portable engine alternator (E/A) is connected to the cell Traffic overloading site. As Telecom has several transportable E/A sets in If people are home from work during the windstorm, it Christchurch it should be possible to keep most cell is expected that the telephone network will be very sites functional unless road access is blocked by fallen busy, but shouldn’t be catastrophically overloaded. trees. It is concluded that it will be harder to get cellular Some exchanges may experience minor load shedding. calls through, but major difficulties are not expected Loss of some cell sites may increase traffic on other throughout most of Christchurch. cells, but since most of the PSTN will still be functional A prioritised pre-plan is required to determine whether this should not be a problem. Electrical and Communications • 105

Telephone Network the storm. There may be some rain after the storm Most customers will have service, but toll links north compounding the flood damage. The following are may be lost or only half available, and many customers expected: with aerial service connections will lose service. They • Flooding as the snow melts. will be relatively easy to repair once the winds have died down. It will take some time to repair these • Some roads to be impassable due to deep drifts connections if damage is widespread. (especially on the Port Hills) or localised flooding.

• Access to the radio sites on the Port Hills to be Time to repair blocked. Helicopter access will be impossible due The time to repair depends upon: to wind or white-out conditions.

• The difficulty of gaining access to sites. There will • Widespread loss of power as the weight of snow be many fallen trees blocking access to critical either directly or indirectly through broken tree sites. Repair crews will need chainsaws so that they branches breaks overhead power wires. Flooding can cut access ways through fallen trees and winches may also disable some underground power serv- so that they can pull trees out of the way. ices.

• The time taken for the winds to subside to a safe • The Fire Brigade to be busy dealing with localised working level. This could take quite some time at flooding. exposed radio sites. • Some roofs to be heavily loaded and possibly Mitigation measures collapse. • Review the safety margins used when specifying • Increased wind loadings due to surface area in- ties to hold roofs down. crease following snow and ice build up. • Remove windows on outside walls by filling in or barricading them from equipment rooms at key Probable extent of damage sites. Towers and masts • Have chainsaws available so that repair crews can There are four Telecom radio sites and twelve cell sites. clear paths through fallen trees. The combined snow and wind ratings of masts and • Develop prioritisation pre-plans showing the allo- towers are expected to survive the storm, but some cation priorities for engine alternators. dipoles and other antennas on these masts could be damaged. This means that the land mobile, paging, and • Provide tarpaulins and sand bags (empty) so that trunked dispatch networks may be partially disrupted. temporary covers can be arranged if a roof is blown off. Loss of most cell sites • Review the wind ratings of antennas, associated Eight of the twelve Christchurch cell sites do not have fittings and feed cables. standby power. Many of these could lose their mains power supply. Those that do will completely fail after • Review the wind ratings of masts at strategic life- the battery reserve has been used, unless a portable lines radio sites in Canterbury. engine alternator (E/A) is connected to the cell site. As Telecom has a number of transportable E/A sets in Assessment of the effect of a major Christchurch it should be possible to keep most cell snowstorm sites functional unless road access is blocked by snow drifts. Major difficulties are not expected throughout Scenario most of Christchurch. For details of the scenario, see Section 2.7 “Snowstorm Scenario”. A snowstorm drops snow to a depth of 0.3 Aerial service leads m throughout most of the city with much greater depths Many customers will lose service due to loss of their on the Port Hills. (The 1992 storm averaged 0.2 m deep aerial service leads through the weight of snow break- on the flat.) The peak of the storm would probably last ing wires or tree branches overloaded with snow fall- for (say) two days, but this would probably be accom- ing and breaking wires. Many telephone poles will be panied by some wind making access difficult during broken. These customers will be scattered in pockets 106 • Risks and Realities

around the city. As the programme to convert service Mitigation measures leads to underground leads progresses this will become less of a problem in the future. • Develop prioritisation pre-plans showing the allocation priorities for engine alternators.

Toll links north • Continue the conversion programme for convert- Both of the two toll links north could be affected by a ing aerial service leads to underground. snowstorm. If only one link is lost, this will have very • Review the combined snow and wind ratings of: little impact. — aerials, dipoles, waveguides, and coaxial cables Toll links south etc., at the critical land mobile sites; Should not be affected as there is buried fibre for the — the radio station building roofs as they are whole route between Christchurch, Timaru, Dunedin unknown; and and Invercargill. — the land mobile radio masts.

Telephone network (PSTN) Assessment of the effect of slope Many customers with aerial service connections might hazard lose service. Although they will be relatively easy to repair, it will take some time to repair all these connec- Scenario tions if damage is widespread. For details of the scenario, see Section 2.8 “Slope Hazard and Damage to Services on Hills”. The slope Traffic overloading hazards considered at each site/route were: If people are home from work during the snowstorm, it • debris landing (or flowing) onto the site/route from is expected that the telephone network will be very above (i.e. soil falls and flows, and rock falls); and busy, but shouldn’t be catastrophically overloaded. Some exchanges may experience minor load shedding. • collapse of the ground at the site/route, due to loss Loss of some cell sites will increase traffic on other of foundations or formation by mass movement of cells, but since most of the PSTN will still be functional either the subgrade or the underlying natural soil this should not be a problem. and rock.

The triggering events are either: Roofs The roofs at all Telecom sites are expected to carry the • an extreme rainfall event with a return period of snow, although the strength of those at some radio approximately 100 years; or station buildings need to be reviewed. • a serious earthquake occurring when groundwater levels are high, again with a return period of ap- Interdependence proximately 100 years. Most Telecom telephone poles support power service The worst impacts would result from an earthquake leads from the opposite side of road, as well as occurring when groundwater levels were high (e.g. at telephone service leads. The priority that the replace- the end of winter). An earthquake in dry soil conditions ment of broken telephone poles is given may be dic- would cause some rock failures, but very little soil tated by the need to restore power to houses, rather than would “slip”, and hence this would not present a great the need to restore telephone service. slope hazard problem.

Time to repair Impact of slips The time to repair depends upon: To assess the likely impact of slips on key services on • The difficulty of gaining access to sites. Repair the Port Hills, a visual inspection was carried out with crews will be reliant upon heavy machinery to Mark Yetton, Engineering Geologist and other service remove large snow drifts for gaining access to sites. providers.

• The time taken for whiteout conditions and/or any Telecom equipment sites strong winds to subside to a safe working level. This could take quite some time at exposed radio Telecom has only three equipment sites on the hills sites. Electrical and Communications • 107

near Christchurch. These are all radio stations and are • Loss of links to some exchanges in the Banks located on the “brow” of a hill. At each of these sites Peninsula area. there is very little chance of any problems occurring. Slope angles are very gentle, the rock appears strong • Loss of one cell site, with consequential loss of and there are no significant cuts to undermine the cellular coverage in the Lyttelton Harbour area, and strength of the adjacent slopes. a major degradation in off-shore cellular coverage.

The only structure at risk is the structure supporting the • Loss of all broadcast links to the BCNZ Gebbies aerials for the radio links to exchanges on Banks Pass AM radio transmitters (3YA, 3YC). Peninsula. This is located on moderately steep ground. • Loss of land mobile radio landline connections to The rock supporting one of the bottom feet could fail in one of the Telecom radio stations This would only a very strong earthquake, causing the structure to have a minimal impact on the operation of land buckle. If this occurs, all links to these Banks Penin- mobile radio channels at this site. sula exchanges would be lost. Heavy rainfall is not expected to have any effect on this structure. • Slight loss of paging network coverage over Christ- church. Key cable routes If both the vulnerable ductlines failed, in addition to the There are four cable routes on, or at the foot of, the Port above impacts, the Christchurch Paging network would Hills that are of particular importance to Telecom. be severely affected. After visual inspection of these routes, it was concluded there is likely to be very little overall impact on Telecom plant buried at the foot of the Port Hills is not these cable routes. However, cable breaks resulting expected to be affected by any slope failure, even from road shoulder failure are distinctly possible at two though the roads themselves may become covered with places. slip material.

One of these places has a history of failure, failing as Mains power recently as 1992 in the “Big Snow” and leaving Telecom pipes exposed with no support, i.e. “high and dry”. On In a significant disaster it is probable that slope failure that occasion there were no breaks in the ducts, and no will cause mains power to be lost to several radio sites. damage to the cables. All these radio stations have engine-alternator sets to provide back-up mains power, so loss of mains power These two weak spots could be overcome by the should not cause any problems, so long as fuel supplies construction of appropriately designed retaining walls. to the engine-alternator sets can be replenished in time. Other vulnerable locations may exist on these routes, but these could not be ascertained by surface inspec- Road access tion only. While substantial damage to roads on the Port Hills can A complete break in the cables in ducts at one of these be expected, road access to each of the radio sites known vulnerable places, would result in: should not be a problem, after allowing twelve hours for debris to be cleared from the roads. In many places • Loss of fleetlink access to the public switched roads may be reduced to one lane, reducing traffic flow telephone network (PSTN). rates. • Loss of land mobile radio landline connections to a Telecom radio site. This would only have a mini- Summary mal impact on the operation of land mobile radio channels at this site. Public switched telephone network Scattered pockets of customers on the Port Hills may • Slight loss in paging network coverage over Christ- lose service. Also, links to several exchanges in the church. Banks Peninsula area could be lost. • Loss of telephone service to most Cashmere customers. Cellular network • Loss of various FM radio station links to Sugarloaf. Cellular coverage of the Lyttelton Harbour area, and much of the off-shore area could be lost. The rest of A complete break in the cables in ducts at the other of Christchurch would not be significantly affected. these known vulnerable places, would result in: 108 • Risks and Realities

Land mobile of one hour from the peak to the trough of the first Virtually unaffected. wave. • Wave not breaking in Lyttelton Harbour, with Fleetlink maximum water levels of +5 m (MSL) at the port Could lose access to the public switched telephone and +3 m (MSL) at the head of the harbour. The network. Otherwise unaffected. total port area will be covered to a depth of between 1 m and 2 m for up to 30 minutes. The water level will not reach the township. Paging Very little effect, unless several independent ductlines • Wave breaking in the Avon and Heathcote estuary, all fail. with no discernible wave trough. The wave speed in the estuary will be about 20 km/hr. The initial wave height at the estuary mouth will be +3 m Repair work required above normal tide level. This wave height will decrease by 0.3 m/km of travel beyond the estuary Aerial support structure mouth. At Dyers Road the wave height will be Although the structure may buckle if the rock under about +1.5 m above normal tide level (i.e. 2.5 m one of the front feet collapses, the aerials should not be above MSL at high spring tide). damaged. To restore service, some sort of temporary bracing for the support structure would be required, • Water receding from the estuary will scour the and the aerials would have to be realigned. This should bridge abutments at the Ferrymead bridge and the be completed within two days. Bridge Street bridge and scour the sea wall along the Moncks Bay and Redcliffs foreshore. This may lead to possible wall failure with collapse of the Failure of vulnerable ductlines associated road and services. Temporary repairs to the optical fibre cables should be completed within two days. This will restore the • Bores initially 3 m above high water level will cellular, land mobile, fleetlink, and paging networks to travel up the Avon and Heathcote rivers. The bore full service. This will also restore all links to Banks height will reduce to 0 m (above high water level) Peninsula exchanges. about Fitzgerald Avenue on the Avon River, and about Opawa Road on the Heathcote River. Areas Temporary repairs to the copper cables to restore of inundation from the first tsunami wave are telephone service to individual customers on the Port shown on Map 11 (p 294). Hills will take longer. Service will be restored progressively over a two day to five week period. • Scour through the beach dunes is expected to increase the length of each entry point by 20%, and Mitigation measures reduce its height by 20%. Scour at the estuary mouth is similarly expected to widen it by 20% and • Consider building appropriately designed retain- deepen it by 20%. Consequently, although the ing walls at identified vulnerable points on key second wave will be at a lower height, it is expected ductlines. to result in greater levels of inundation than the first wave. (Map 11 shows inundation areas for the first Assessment of the effect of a major wave only.) tsunami Inundation levels over ground at estuary margins will Scenario be: For details of the scenario, see Section 2.5 “Tsunami • Sumner, 0.7 m Hazard”. A tsunami with a return period in excess of 100 years. Assumed features of the tsunami are: • Moncks Bay/Redcliffs, 1.7 m

• Total water level variation of 5 m above mean sea • McCormacks Bay, 1.1 m over causeway level (MSL) to 5 m below MSL, i.e. 4 m above spring high tide to 4 m below spring low tide. • Ferrymead, 0.3 m

• Water level disturbances reducing over a three to • South Brighton, 0.5 m (Bridge St) increasing to five day period. 1.65 m (South Brighton spit).

• Tsunami wave period of 3 hours, with a minimum Average inundation levels and flow velocities along Electrical and Communications • 109

the east Brighton coast (from “overtopping” of dunes/ connect cabinets, and 400 mm at Telecom pillars, sea wall): customers’ telephone circuits will start short circuiting. To minimise damage to the Telecom cable circuits, • From Rawhiti St to Rodney Street —1.2 m, 4 km/ they should be powered down prior to immersion. To hr (i.e. from 700 m north of the New Brighton achieve this, it may be necessary to disconnect tel- clock tower to 700 m south of the New Brighton ephone service from blocks of customers in much clock tower) larger areas than the inundation areas shown in Map 11 • Elsewhere, from Rothesay Road — 0.2m to 0.55 m, (p 294). 0.5 km/hr to 1 km/hr (Waimairi Beach) to South Brighton spit except for Jellicoe Road (0.7 m) and Summary Heron Street (1.0 m). Public switched telephone network Impact of tsunami Large numbers of customers in the eastern suburbs may lose telephone service. Equipment sites Two exchanges, and one cell site are located in areas Cellular network expected to be inundated by the tsunami. Of these, only Cellular coverage of the Lyttelton Harbour area, and one exchange is expected to have any problems. It is much of the off-shore area could be lost. The rest of expected to be inundated to a depth of 1.2 m, with water flow velocity of 1 m/s, or about 4 km/hr. If seawater Christchurch would not be significantly affected. cannot be kept out of the exchange, it is vital that it be powered down with consequential loss of telephone Land mobile service to its customers. Virtually unaffected.

Cable routes Fleetlink Scouring of roadways and bridge abutments could Unaffected. cause breaks in the Telecom cables buried along these roads. This could result in: Paging • loss of links to some exchanges in the Banks Virtually unaffected. Peninsula area; • loss of the link to one cell site, with consequential Repair work required loss of cellular coverage in Lyttelton Harbour area, If important Telecom cables are broken due to scouring and a major degradation in off-shore cellular cov- of roadways and bridge abutments, work could begin erage; immediately on installing a temporary radio link to • loss of all broadcast links to the BCNZ Gebbies restore some of the lost circuits. This could be com- Pass AM radio transmitters (3YA, 3YC); pleted within two days, restoring:

• loss of Land Mobile Radio landline connections to • links to all exchanges in the Banks Peninsula area; one of the Telecom radio sites (this would only • full (normal) cellular coverage to the Lyttelton have a minimal impact on the operation of land Harbour and off-shore areas; mobile radio channels at this site); • all broadcast links to the BCNZ Gebbies Pass AM • slight loss of Paging network coverage over Christ- radio transmitters (3YA, 3YC); church; • all land mobile radio landline connections; and • loss of all junction links to three exchanges to the south of the estuary; and • full (normal) paging network coverage over Christ- church. • loss of telephone service to all of Moncks Bay, Redcliffs, Ferrymead, and much of Woolston and All remaining repair work, including restoring junc- Bromley. tion links to the three exchanges, and telephone service to large pockets of customers in Sumner, Moncks Bay, Other plant Redcliffs, Ferrymead, Woolston, Bromley, Aranui, Once water levels exceed 200 mm at Telecom cross- Avondale, Bexley and New Brighton, would not be 110 • Risks and Realities

able to start until the continuing tsunami waves sub- slots must be sealed — water must not be allowed side. The waves could take two to four days to subside to pour into the cable well. sufficiently. • Sump pump discharge pipes should end well above any anticipated flood levels. Typically they should Cable network end at least 500 mm above ground level. At the Any connectors in pillars, connection blocks in cabi- worst affected exchange the sump pump discharge nets, telephone sockets in houses, and internal house pipe should end at least 1500 mm above ground wiring that have been immersed, will probably have to level. be replaced. Associated cabling and mounting hardware will have to be hosed down (to remove silt, sand, • Prepare preplans for responding to a tsunami alert by: etc.) and dried. — External sandbagging, or fitting shutters to Repair of Telecom cables broken due to scouring of the exchange’s external doors and any other roadways and bridge abutments may have to wait until water entry weak points (e.g. ventilation repair of the associated road/road shoulder is suffi- louvres), to ensure that even when outside ciently advanced. Customers’ service affected by water levels are 1.2 m above floor level, the either immersion of plant or cable breaks, should be flow of water into the building is restricted progressively restored over a period of two days to two to a trickle. months, after the tsunami subsides. The junction cable — Bringing in a bigger pump, if the sump links should be restored within two days of the tsunami pump can’t cope. Ensure there is some subsiding. window/hole/vent above the maximum water level, through which the pump dis- Exchange equipment charge hose can be fed. Assuming any exchange equipment which suffers im- — Disconnecting telephone service to those mersion is powered down before the event, salvage and blocks of customers expected to be inun- full restoration is expected to be possible. However, dated by the tsunami, by removing links on this could take up to five days after waters subside to the main distribution frames at the appropri- below floor level. ate exchange. If the equipment is still powered up when immersed, — Removing all power at an exchange (in- irrecoverable damage will result. The electronic cards cluding engine-alternator generated mains) will need to be replaced, and the backplane wiring and from all equipment. connectors cleaned out. Depending on the availability of spare cards, this could take from two weeks to six — Prepare preplan for cleaning and salvaging months. all electronic cards, wiring and connections that were immersed in floodwater (after This will restore telephone service to all those customers connected to the exchange who were outside the water levels have subsided). At least one tsunami inundation areas (or were in areas where the “kit” should be established in Christchurch inundation level was less than 200 mm) — probably for this purpose. about 70% of the exchange’s customers. — Move sump pump power point at the exchange to above floor level. Mitigation measures — Ensure the circuit breaker protecting the • In exchange cable wells, seal all incoming ducts/ power points in the exchange cable well is holes/slots (including the sump pump discharge not also protecting other important power pipe hole). If the cable wells have external manhole feeds above floor level. lids, these should be either sealed, or fitted so that when immersed, water will only “trickle” through them. This should stop water entering the cable well at more than a trickle, even under severe 5.2 New Zealand Fire Service “flood” conditions. Communications Network Description • The seal must be able to withstand a reasonable head of water (say 3 m) and should preferably be Introduction and seismic vulnerability fire retardant. Duct caps are not an acceptable The New Zealand Fire Service is divided into six means of sealing ducts. Note that all ducts/holes/ Electrical and Communications • 111

Regions with two of these being in the South Island. electronic devices are DC-powered with battery back- Christchurch is the location of No. 5 Region Headquar- up. ters, which is responsible for all areas north of a line drawn along the Waitaki River, across the Southern Control room evacuation Alps, and along the Haast River. This area is divided The New Zealand Fire Service has numerous fall-back into four areas (Christchurch, Nelson, Timaru and positions in the event of an evacuation being necessary. West Coast). In the first instance, the service has a mobile command The Christchurch Area Headquarters is located in the unit which is capable of accepting all incoming 111 main central fire station at 200 Kilmore Street, and is calls by using cell phones and then activating the responsible for the operation of over 50 fire brigades appropriate brigade by VHF radio. from Hinds in the south, to Springfield in the west, and Kaikoura to the north. The urban Christchurch area, A secondary control room has been established at the which this report is particularly concerned with, in- Woolston Training Centre in the event of a more cludes nine fire stations of which six are permanently prolonged evacuation being necessary. manned, and the other three are staffed by volunteers (Sumner, Lyttelton and Brooklands). There is a further secondary control facility based at Nelson which also has the capability of turning out all The New Zealand Fire Service relies on two main brigades in the region. methods of communications — VHF radio and telephone. Vulnerability The resources of the Fire Service are widely separated Radio with each brigade being totally self-reliant to handle all The New Zealand Fire Service primary communica- emergencies within its area of responsibility. In the tion is by VHF (AM) radio-telephone network which event of any natural disaster, it would be a relatively covers the entire region. All brigades are activated by simple process to utilise surrounding brigades to take a selective tone callout system (Secal) based in the responsibility for any station which may become inop- Regional Control Room located at 200 Kilmore Street. erative. In the event of any repeater failure, all radios in this region are multi-channelled which allows access to any Mitigation measures other repeater still in operation. Should there be a total repeater failure, all radios have simplex crystals fitted • To undertake a review of all electronic and compu- allowing point-to-point communications. ter-based equipment for security.

All brigades communication equipment is DC-pow- • To check on all seismic bracing of electrical and ered with adequate battery back-up to enable continu- electronic cabinets and stand-by generators. ity of operation in the event of any mains failure for an • To assess all fuel lines for flexibility and adequate indefinite period. The Regional Control Room has all fuel requirements. its main electronic equipment securely bolted down, with the taller racks being bolted together and top- • To review the risks of “topping” hazards. braced. All sub-stations and volunteer brigades have their equipment in standard cabinets in secure loca- • To continue the policy of region-wide standardisa- tions. tion of all communications equipment, and maintaining an adequate level of spares to enable immediate restoration of services. Telephone The sole responsibility for accepting emergency 111 Flooding Vulnerability calls throughout this region is in the Regional Control Room in Christchurch. All telephone circuits, private Waimakariri flood hazard fire alarms, the PABX and any other miscellaneous This event will have the least effect on the New circuitry is provided by Telecom or their subsidiaries. Zealand Fire Service’s ability to maintain an effective A degree of diversity has been provided by having firefighting force and will certainly have no effect on alternate cable routes from two Christchurch Telecom communications networks. It may flood some of the exchanges for incoming 111 lines. There is a diesel nine urban fire stations such as Brooklands, but would stand-by generator and an UPS to maintain continuity have to be extremely severe to put communications out of operation in the event of any mains failure. All main of order, or be of sufficient depth to prevent appliances from being operated effectively. 112 • Risks and Realities

Local river flooding hazard (Avon/ Zealand Fire Service’s ability to effectively communi- Heathcote) cate with its fire fighting force. This event could cause some concern, not so much for However, this event could create problems if interde- the ability to maintain communications which will be pendency issues become involved, specifically the largely unaffected, but more so for the ability to move issue of damage to Telecom lines, or radio systems, as appliances around the city if bridge approaches are the Service’s communication system is dependent on washed out, etc. these to operate in a normal state. It is most unlikely that any flooding would occur with Damage to stations may mean that some resources are such speed that the Service would not have time to unobtainable, but because of the spread of stations the reposition appliances to either higher ground at strate- loss of all vehicles would be extremely unlikely. gic sites around the city or to other, unaffected urban fire stations. Snowstorm Vulnerability Tsunami flooding hazard This event will have little or no effect on the New Zealand Fire Service’s ability to effectively communi- There are several sites which are close to the coastline cate with its fire fighting force. such as New Brighton, Sumner and Woolston. However, this event could create problems if interde- Depending on the warning period of such an event, pendency issues become involved, specifically the these appliances could be rapidly moved to more issue of damage to Telecom lines, or radio systems, as inland sites. the Service’s communication system is dependent on It is most unlikely that any adverse effect would be these to operate in a normal state. noticed on the communications network, as all sites are Damage to stations may mean that some resources are self-sustaining on several days battery back-up sys- unobtainable, but because of the spread of stations the tems. loss of all vehicles would be extremely unlikely.

Communications system — mitigation Slope Hazard Vulnerability measures available This event will have little or no effect on the New Principal issues and actions Zealand Fire Service’s ability to effectively communicate with its fire fighting force. • The only fixed site which has a unique function is the main “mobilising centre” located at the Christ- However, this event could create problems for the church Central Fire Station in Kilmore Street. Service if the interdependency issues become involved, specifically the issue of damage to Telecom lines, or • The flood level of this centre is well above the 50- radio systems, as the Service’s communication system year flood plain and is most unlikely to be affected is dependent on these to operate in a normal state. in the event of the Avon River flooding. Damage to stations may mean that some resources are • Should any unexpected event happen which pre- unobtainable, but because of the spread of stations the vented this centre from remaining operational, the loss of all vehicles would be extremely unlikely. Service has the ability to carry on the functions of this site from literally anywhere in the city at minimal notification (minutes rather than hours). 5.3 New Zealand Police • It is undoubtedly to the Service’s advantage that resources are spread very evenly around the city Introduction and suburbs, with each Brigade being totally inde- Although there has been a recent review of the way in pendent of the others to carry out their operational which the Police are organised since the time of the tasks. Lifelines study, this change has not affected the sys- • It is similarly of considerable advantage that all the tem’s vulnerability. Police staff in Canterbury were Service’s equipment is highly mobile and can be divided into two separate districts: readily relocated to safer areas. • Christchurch City District is responsible for policing of the city and suburbs. Staff from the Central Windstorm Vulnerability Police Station cover the inner city area, but a recent This event will have little or no effect on the New move towards community-oriented policing has Electrical and Communications • 113

seen the introduction of major suburban police redundancy in the event of some network components stations at Hornby, New Brighton, Papanui and being damaged by a natural disaster. Sydenham, and Community Constable offices in numerous other suburbs. Interdependency on other providers does exist — e.g. Southpower for mains power. Wherever possible, • Canterbury Rural District is responsible for polic- emergency mains power is supplied from diesel standby ing all areas outside the city boundaries. Police generators. A requirement that radio equipment on all stations at Ashburton, Kaiapoi, Lyttelton and sites is DC powered with battery back-up is also rigidly Rangiora are supplemented by one or two man adhered to. stations throughout the rural area.

Police rely on three main areas of communication — Data telephone, radio, and data. Police stations in Canterbury are served by a computer data network providing access to the Wanganui Com- puter Centre, and to host computers in Christchurch Telephone and Wellington, running a variety of applications. The Police control room at Christchurch Central provides the sole answering point for 111 emergency and Critical components in this network are data switches *555 circuits in Canterbury. Direct telephone circuits and protocol equipment housed at the Central Police to other emergency services are also provided to this Station. Once again, high interdependency exists on location. Police have embarked on a programme of Clear Communications and Telecom for provision of Mitel PABX installations at all major Police stations. digital data circuits into police buildings and, to a lesser These PABXs are networked countrywide via digital extent, Southpower for mains power. or analogue circuits. At Christchurch Central and major suburban Police There is a large degree on interdependency on other stations, diesel standby generators and UPSs provide providers — Telecom for 111, *555, direct and trunk emergency mains supply for data equipment. lines, Clear Communications and Telecom for major tie-lines and, to a lesser extent, Southpower for mains Control room evacuation power. Police have considered the need for possible evacua- Diversity for critical circuits such as 111 lines and tion of Christchurch Central in an emergency. A major tie-lines has been provided through alternative suburban police station has been fully equipped with Telecom exchanges. At Christchurch Central and an emergency control room. The 111 lines from two major suburban police stations, diesel standby genera- separate Telecom exchanges and the Central Police tors and UPSs provide an emergency mains supply for Station trunks can be diverted to this room. An alter- PABX units. native method of accessing all police radio channels is also provided. This room is also equipped with access System-fail bypass circuitry ensures trunk circuits are to the data network, although this is reliant on equip- automatically routed through to Police control-rooms ment at the Central Police Station remaining intact. and watch-houses in the event of PABX failures. At Christchurch Central the control-room consoles pro- Seismic Vulnerability viding integrated telephone and radio facilities are DC powered with battery back-up. Moderate event Police stations in Christchurch are relatively modern Radio with the Central Police Station being 25 years old, and Police operate a VHF FM radiotelephone network major suburban stations being built within the last serving mobiles throughout Canterbury, and a UHF seven years. They should, therefore, be capable of FM radiotelephone network serving portables through- withstanding an earthquake of this magnitude with out Christchurch City. A series of VHF and UHF hill only moderate damage. Equipment within is, there- top repeaters are linked via microwave and UHF links fore, likely to remain functional or require little reme- to the Central Police Station control room. dial action. Similarly, all repeater site buildings are solidly constructed with good foundations. Frequency diversity in link paths exists between some sites, as does automatic activation of back-up trigger Major items of electronic equipment for each network equipment. Generally though, reliance on the quantity are housed in rack cabinets complete with doors. These of channels available and the physical separation of the cabinets, ranging in height from 1 m to 2.1 m are bolted various repeater sites is considered to offer sufficient individually to wooden plinths, and to each other. 114 • Risks and Realities

It is expected that damage sustained in an earthquake • damage to freestanding electronic equipment, e.g. of this type and magnitude would be as follows: data terminals etc.

• interruption of primary mains supplies to Police The vulnerability chart which assisted in the investiga- stations; tion is shown in Figure 5.4.

• interruption of primary mains supplies to radio repeater sites; Major event Under this scenario the following damage may be • loss of some telephone and data circuits into Police expected to occur: stations causing a variety of outages dependent on alternate circuit routing; • structural damage to police stations;

• possible misalignment of microwave and UHF • structural damage to repeater site buildings; aerials causing interruption of radio circuits without back-up facilities; and • structural damage to transmission towers, masts, coaxial feeders and aerials;

UTILITY: NEW ZEALAND POLICE REGIONAL NETWORK: Station 2

VULNERABILITY TO HAZARD IMPACT OF DAMAGE

COMPONENT COMMENT

ELEMENT

LANDSLIDE

LIQUEFACTION

GROUNDSHAKE

IMPORTANCE 1 - 5 IMPORTANCE

PERIOD FOLLOWING

IMMEDIATELY AFTER IMMEDIATELY

DURING EARTHQUAKE RETURN TO NORMALITY RETURN TO EARTHQUAKE HAZARD GROUND SETTLEMENT

Building 5 1 0 1 3 3

Control System 4 2 0 1 2 3

Rented speech/data circuits 5 3 0 1 3 3

PABX 5 2 0 1 3 3

Data equipment 4 3 0 2 3 3

Microwave links 5 3 0 1 2 3

UHF repeaters 2 1 0 0 1 2

UHF links 3 2 0 0 1 2

VHF standby triggers 4 1 0 1 2 2

Mains supply 4 3 0 0 1 2

Standby mains supply 4 2 0 0 1 2

UPS supplies 5 2 0 223

DC supplies 5 2 0 223

Figure 5.4: Police vulnerability chart Electrical and Communications • 115

• loss of primary and possibly standby mains sup- this is located on a raised terrace with fuel tank inlet plies; pipe and vents at terrace tarmac level.

• loss of telephone and data circuits irrespective of alternate routing; Major suburban police stations These have differing communications equipment room • misalignment of microwave and UHF link aerials; locations with Hornby on the ground floor and New and Brighton and Papanui on the first floor. Diesel standby plants at Hornby and New Brighton would be moder- • damage to electronic equipment both freestanding ately vulnerable to Waimakariri flooding and tsunami and rack-cabinet mounted. respectively with generators at ground level and in- The police consider the following when designing ground tanks possibly subject to fuel contamination. their communications networks to maximise their reliability particularly during disasters: Police VHF and UHF repeater sites • constantly review police station buildings and re- These sites are not expected to be affected by flooding peater sites to identify communications equipment or by the effects of excessive rainfall, with the excep- and cabling which is unacceptably vulnerable to tion of a site of lesser importance in the New Brighton seismic activity; area with high vulnerability to a tsunami.

• meeting standards for seismic bracing of equipment cabinets and battery stands in equipment Mitigation measures rooms; The police have considered the following to minimise the vulnerability of their communications networks to • ensuring there is route diversity and equipment flooding: redundancy when installing new equipment in each network; • new suburban Police stations planned for the Christ- church area are to have communications equip- • ensuring sufficient spares are held for major net- ment located on the first floor; work nodes so that service can be restored without delay; and • diesel standby plants to be installed at Papanui and Sydenham Police Stations are to be equipped with • ensuring that sufficient trained staff are locally above ground fuel tanks; and available to maintain the critical elements of the network. • sufficient spares, including battery-powered emergency radio repeaters and PABX units, have been Flooding Vulnerability purchased for rapid deployment should flooding cause equipment failures. The police communications networks would be only minimally affected by flooding hazards whether these originate from the Waimakariri river, the Avon or Heathcote rivers, or from a tsunami. 5.4 Trans Power New Zealand Ltd Electrical System Christchurch Central The control room, the various communications equip- Description ment rooms and the electrical switchboards are located at least one floor above ground level. Introduction The electrical distribution system comprises a regional The basement and an adjacent loading dock for the supply from the national grid to a local distribution cell-block are situated below street level. However, network. The regional supply network is owned by electrical or communications cabling which runs in Trans Power New Zealand and the distribution net- this area is sealed with terminations being accom- work is owned and operated by Southpower. plished on upper floors. Two sump pumps, alarmed for pump failure, are set into the basement floor with outlets two floors above. Regional network Southpower’s Christchurch network is supplied with The transformer kiosk feeding the building may be electricity from four Trans Power substations, namely: vulnerable to flooding. Flooding problems are not, however, expected with the diesel standby generator as • Islington; 116 • Risks and Realities

• Addington; on the Port Hills or lateral spreading of the banks of the Lower Heathcote River (affecting one tower). • Bromley; and

• Papanui. Moderate regional event A moderate to severe regional earthquake of up to MM Of these, Islington is the most important, followed IX (peak ground acceleration of approximately 0.4g, closely by Bromley, then Addington and Papanui. 28% probability in 50 years) is less than Trans Power’s Islington is a focal point for the South Island section of design requirements for new buildings and equipment. the national grid. Although it is part of a network, it is Therefore, it is likely that most significant damage will principally supplied by 220 thousand volt (kV) alter- occur with older equipment and buildings. nating current (AC) from Twizel (three transmission In addition to building surveys that have been carried lines) and Livingston (one line). These transmission out in the past, a walk-through inspection of all of lines are carried by steel lattice towers. Islington Trans Power’s Christchurch facilities was carried out supplies Bromley at 220kV, Addington at 66kV and to assess the current vulnerability of equipment and Papanui at 66kV via overhead towered lines. buildings. Bromley Substation, although being supplied from The assessed vulnerability of principal facilities is as Islington, is also connected directly from Twizel at follows. 220kV. This has built an element of redundancy into the system, should a problem occur at Islington. Addi- tionally there is a Southpower 66kV transmission line Islington Substation between Islington and Bromley which could be uti- Islington Substation (220kV and 33kV) was built in the lised by Trans Power in an emergency. early 1950s. A significant amount of equipment has subsequently been replaced. The most vulnerable The above substations supply the Southpower network areas are transformer restraint (due to long lead times at 66kV, 33kV and 11kV. for major repair or replacement), equipment mounted on brittle insulators, e.g. reactors, circuit breakers etc., Electrical system vulnerability and bus work where there is no allowance for differen- See Figures 5.5, 5.6 and 5.7. tial movement.

Some brittle components have been base isolated using Regional system a shock absorbing system which should greatly increase the chance of surviving intact. A significant System seismic design amount of spares are held on site. Liquefaction in The equipment and buildings at the Trans Power sub- unlikely to occur on this site. stations in the Christchurch Region vary in age from approximately 65 years old (Addington Substation Bromley Substation Building) to relatively new (South Island Control Cen- tre at Islington). Most equipment in switchyards was Bromley Substation (220kV and 66kV) was partly sourced from overseas and has been designed to with- built in the early 1950s and extended in the early 1970s stand a variety of earthquake loadings from 0.25g or (220kV switchyard). Again, the most vulnerable areas less for older equipment to 0.7g for newer equipment. are transformer hold down, brittle equipment and bus Some brittle equipment mounted on porcelain insula- work. Some spares are held on site. This site was tors now has to withstand up to 1.5g because it is unable specifically checked for liquefaction potential and to absorb earthquake energy in a ductile fashion. found to be low risk.

In the last few years, Trans Power has retrofitted Addington Substation improved seismic restraint to critical equipment including transformers and circuit breakers, and has Addington Substation (66kV) was started during World carried out building strengthening work. War I and extended in the early 1920s. The building has subsequently been strengthened and most, if not Transmission lines and their supporting towers nor- all, of the equipment subsequently replaced. The mally perform well in earthquakes as they are designed major transformer banks have been retrofitted with to resist wind and snow loads which tend to exceed seismic hold down brackets to current Trans Power earthquake induced loading. The main potential prob- Standards. The areas of greatest vulnerability are the lems could be foundation failure due to land slippage insulator mounted equipment and the internal 11kV Electrical and Communications • 117

Islington Utility: Trans Power Regional/Local Network: 220 kV Sub Station

Vulnerability to Hazard Impact of Damage

Component Size Ground Shake Comment

Landslide Immediately After Period Following Ground Settlement During Earthquake Return to Normality Zone Boundary Liquefaction Importance

Transmission lines 5 1 0 0 1 321 2 days Temporary towers available

Termination towers 5 10 0 1 321 2 days Temporary towers available Gantries bus work 5 1001 23 1 2 days Spares available Circuit breakers 2 0 0 1 3 1-2 1 2-4 days Spares available (external) 4

CTs/VTs 4 2 0 01 3 1-2 1 2 weeks Spares available

Reactors 4 2 0 0 1 3 1-2 1 1 week Spares available Oil storage tanks 1 3 001 111 4 weeks

1 Spare may be avialable —need Power transformers 5 1-2 0 0 333>18 mths structure check

Underground cabling 5 2 00 113 2 1-2 days

Communications 5 1 001122 >2 mths tower Scada 5 1 0 0 1 333>1 week Spares available

Overall building 4 1 001 1-3 1-3 1 >2 weeks

Control cabinets Some cabinets have HD panels 5 2 00 1 3 2 1 1-2 days bolts missing Emergency generator 3 1 001 211 3 mths

Suspended ceilings 1 1 001 1-3 11

Cranes 1 1 001 1 11 3 mths

Batteries 5 1 001 3 2 1 1 day Distilled water will tip over

Condensors 4 1-2 001 3 2 1 >12 mths Storage spares 4 3 001 1 1-3 1 Varies Spares not held down 33 kV yard Buswork 52 0 0 1 321 2 days

Figure 5.5: Trans Power vulnerability chart 1 switchgear. Previous testing suggests that this site has equipment. Sites tested nearby indicate liquefaction moderate liquefaction potential. potential is low.

Papanui Substation South Island Control Centre, Islington Papanui Substation (66kV) was initially built in the The South Island Control Centre was constructed in the early 1950s. Some seismic strengthening has been early 1980s and is unlikely to sustain anything more carried out including transformer hold down. Vulner- than superficial damage. The most vulnerable areas able areas are mainly bus work and insulated mounted are computer terminals in the control room and com- 118 • Risks and Realities

UTILITY: TRANSPOWER REGIONAL/LOCAL NETWORK: ELECTRICITY TRANSMISSION

VULNERABILITY TO HAZARD IMPACT OF DAMAGE

COMPONENT ELEMENT COMMENT

SLOPE STABILITY IMMEDIATELY AFTER IMMEDIATELY IMPORTANCE 1 - 5 IMPORTANCE PERIOD FOLLOWING

DURING EARTHQUAKE RETURN TO NORMALITY RETURN TO ISLINGTON-BROMLEY Access to most towers difficult TRANSMISSION LINE on Port Hills TOWER NUMBER 16 0 1333On flat ground 17 1 1333 18 1 1333 19 1 1333Rapaki Track 20 1 1333 21 1 1333

22 1 1333 23 1 1333Huntsbury 24 1 1333 25 2 1333 26 1 1333 27 1 1333Victoria Park 28 1 1333Dyers Pass Road 29 2 1333 30 2 1333 31 1 1333 32 2 1333Worsleys Spur 33 1 1333 34 1? 1333 35 1 1333Kennedy's Bush Road 36 1 1333Kennedy's Bush Spur 37 1 1333 38 1 1333 39 1 1333 40 1 1333 41 1 1333 42 0 1333 43 0 1333Landsdowne

Figure 5.6: Trans Power vulnerability chart 2 Electrical and Communications • 119

UTILITY: TRANSPOWER REGIONAL/LOCAL NETWORK: ELECTRICITY TRANSMISSION

VULNERABILITY TO HAZARD IMPACT OF DAMAGE

COMPONENT ELEMENT COMMENT

IMMEDIATELY AFTER IMMEDIATELY PERIOD FOLLOWING IMPORTANCE 1 - 5 IMPORTANCE

WAIMAK FLOODING WAIMAK LOCAL FLOODING LOCAL

DURING EARTHQUAKE TSUNAMI FLOODING

ISLINGTON SUBSTATION 5

Switchyard 5 100 1100On edge of flooding Control building 5 1 00 0011

SOUTH ISLAND SYSTEMS CONTROL 5 Building 5 2300 200Electrical D.B. vulnerable, then computer floor vulnerable

BROMLEY SUBSTATION 5 Switchyard 5 0 01 1 10 0 Site well above rivers Control building 5 001 1 1 0 0

ADDINGTON SUBSTATION 5 Switchyard 5 0 1 0 11 0 0 Control building 5 0 1 0 12 0 0 66 kV circuit breaker

PAPANUI SUBSTATION 5 Switchyard 5 0 1 0 110 0 Near Dudley Creek Control building 5 001 11 0

ADDINGTON STORE 4 Buildings 4 0 1 0 1 2 0 Levels vary Yard 4 0 1 0 1 2 0

TRANSMISSION LINES 5 ISL-BROM 5 1 11 3 2 0 Lower Heathcote ISL-ADD 5 1 00 2 1 0 ISL-PAP 5 1 00 2 1 0

Figure 5.7: Trans Power vulnerability chart 3 120 • Risks and Realities

puting equipment in the computer room. As for the led to the identification of specific concerns and items substation, liquefaction is unlikely. requiring attention in order that objectives be met.

Addington South Island Warehouse Principal issues and actions The Addington South Island Warehouse was recon- Principal issues to emerge were: structed in the 1970s. Older buildings have been retrofitted with seismic strengthening. Large spares • Some major items of equipment do not fully com- that are required to be stored upright are held in ply with Trans Power’s earthquake standards and a specially designed support frames. Other components programme is in place to rectify the situation. are stored horizontally outside. As for the substation, • The availability of spares for damaged equipment this area has moderate liquefaction potential. is dependent on some spares held on site and others held at the Trans Power South Island Warehouse at Transmission lines Addington. It is important that all critical spares are Transmission lines are not very vulnerable to seismic stored in such a manner that damage during the attack and, therefore, are likely to sustain only minor event is highly unlikely. This could be achieved by damage. The portion that passes over the Port Hills is ensuring that spares are securely held down or well founded and, therefore, unlikely to be damaged by stored lying horizontally on the ground to eliminate landslip. Only one tower could be affected by liquefac- the chance of tipping over. tion/lateral spreading of the Heathcote River banks. There • Subsequent to the Wellington Lifelines Study, a is also a 66kV Southpower line running parallel with this new national system of equipment and spares inline and in an emergency could take over from it. ventory is in place and will greatly assist the reinstatement operation. Summary • All Trans Power substations in the Christchurch The Trans Power Christchurch region installations are region have a degree of redundancy and it is consid- likely to withstand a moderate to severe regional event ered likely that sufficient equipment would with- with minor damage. The damage is likely to be stand the earthquake to be able to restore some repaired sufficiently to reinstate full service within a power supplies in a day or two, if not sooner. few days. • Trans Power continually review design standards Electrical system — mitigation in line with experience gained from Edgecumbe measures and overseas earthquakes.

Regional Network • Continue the development of operational procedures and training to respond to a major earth- Background quake. Responsibility for this network rests with Trans Power • Continue the development of procedures for re- Limited which has a long-standing record of attention routing lines and by-passing damaged equipment to seismic aspects of its facilities. This has included for the various scenarios expected following a surveys and strengthening of buildings and vital equip- major earthquake. ment.

Trans Power Limited has a policy in place with the Mitigation summary following objectives: Key mitigation issues to be addressed are:

• to maintain power supplies during and after an • continue with the seismic upgrading of transmis- Edgecumbe size earthquake (MM IX); sion equipment and buildings where appropriate;

• to restore power supplies to earthquake damaged • continue to review maintenance contract condi- areas within three days; tions to ensure the availability of experienced repair personnel (self sufficiency); • to ensure safety of the public and personnel; and • continue review of seismic design standards; and • to minimise the cost of repairs. • ensure adequate contingency plans are in place for Involvement in the lifelines project has provided a the operation and the repair of the transmission broader insight into earthquake related issues, and has system damaged by a major earthquake. Electrical and Communications • 121

Electrical system flooding vulnerability Local rivers flooding hazard All of the substations are outside the local river flood Regional system plains and, therefore, primarily vulnerable to local flooding due to blocked drains etc., and would be Waimakariri flood hazard unlikely to result in a loss of service. A similar This event is the least likely flooding event, but is the situation applies for the South Island Systems Control one most likely to have the greatest effect on Trans and the Addington Warehouse. Power’s electricity transmission system. Transmission lines Islington Substation As for the Waimakariri flood hazard, the transmission The switchyards and control buildings are located on towers are unlikely to be vulnerable to significant the edge of the potential flood plain, and would not be damage from floating debris etc., and are normally highly vulnerable to damage from shallow local flood- well founded to avoid problems with erosion, etc. ing. All vulnerable equipment is at first floor level. Tsunami flooding hazard Bromley Substation Only the Islington to Bromley transmission line ap- The substation is approximately 0.7 km from the near- pears to be affected by flooding as a result of a tsunami est point of the flood plain and is on relatively high and again, is unlikely to receive sufficient damage to ground and, therefore, unlikely to be vulnerable. result in a loss of service.

Addington Substation Principal issues and actions The site is approximately 1.5 km from the nearest flood The principal issues and actions to emerge were: plain and so not vulnerable. • Some items of critical equipment in South Island System Control building could be damaged in the Papanui Substation very unlikely event of a Waimakariri flood hazard. This substation is 1.0 km from a flood plain and, Appropriate mitigation measures would be to en- therefore, also very unlikely to be vulnerable. sure that existing sump pumps are adequately maintained. The provision of equipment for sand bag- South Island Control Centre, Islington ging around doorways would also be prudent. The control centre is within a flood plain and so is potentially vulnerable to flooding. However, the floor Electrical system windstorm level is approximately 300 mm above surrounding vulnerability ground and, as the centre is on the edge of the flood Windstorm plain, flooding is unlikely. Should it occur, 100 mm depth of water inside the building may enter the main The 150-year return period wind (28% probability is electricity distribution board and 200 mm would be 50 years) consisting of up to 200 km/hr gusts on the above the false computer floor. plains with peak gusts on exposed portions of the Port Hills up to 270 km/hr, is greater than the assumed design wind speeds on Trans Power buildings and Addington South Island Warehouse equipment. The site is approximately 1.3 km from the nearest flood plain and is, therefore, not likely to be vulnerable. Transmission lines Most transmission towers are not highly vulnerable as Transmission lines their as-built strength generally exceeds design strength. Only the Islington to Bromley transmission line passes The conductors have a factor of safety in excess of two through any potential flood plains, namely south of based upon design loads, the critical load case being Halswell and in the Heathcote Valley area. Tradition- high winds combined with extremely low tempera- ally, transmission towers have not been highly vulner- tures. Such temperatures (freezing or below) are able to damage as a result of flooding. The velocity of unlikely as part of the windstorm scenario. Most of the floodwaters in the flood plains is not likely to be transmission lines are on towers and, therefore, above sufficient to damage towers with floating debris. the height of toppling trees. 122 • Risks and Realities

Substations • Transmission lines could be made less vulnerable Substation buildings and equipment, where little or no to outages due to severe snowstorms by either using flying debris is likely to enter the sites are not highly heavier conductors or closing up the spacing be- vulnerable to damage, except perhaps for some win- tween towers. Both measures would involve con- dows in substations. The greatest risk is from flying siderable expense and, therefore, be difficult to debris (such as roofing iron, etc.) hitting equipment justify for such an unlikely event. and causing outages due to flashovers. Also, debris • Trans Power now have their own all-terrain vehicle entering buildings through windows could cause trip to provide access to difficult sites during adverse outs and damage to control equipment. conditions.

Electrical system windstorm mitigation Electrical system slope instability Regional network vulnerability The principal issues and actions to emerge were: Regional system

• practically, it would be difficult to prevent debris Port Hills slope instability from entering the site; Slope instability in the vicinity of transmission tower • substation windows could be made less vulnerable foundations could result in tower damage and, in the by the use of Georgian wired glass or by blocking worst case, tower failure. Transmission of electricity up non-essential windows; and could only be resumed once temporary towers had been erected. • temporary transmission towers are held at Islington should a number of permanent towers fail. Islington-Bromley 220 kV Transmission Line: Electrical system snowstorm vulnerability This is the only transmission line in the Christchurch region passing over the potentially unstable sloping Snowstorm ground on the Port Hills. Of the 25 towers on the Port The 150-year return period snowstorm (28% probabil- Hills, 21 have low vulnerability and the remaining four ity is 50 years) consisting of approximately 300 mm of moderate vulnerability to damage. snow on the ground. This event is greater than the usual design criteria for the design of transmission line. Electrical system — mitigation measures

Regional network Transmission lines Transmission lines are the weak link in the system as The principal issues and actions to emerge were: far as vulnerability to outages due to snowstorms are • The four moderately vulnerable towers could be concerned. During the August 1992 snowstorm (ap- investigated in detail to establish the feasibility and proximately 80 year return period or 46% chance in 50 cost of modifying foundations, to lower vulnerabil- years) the Islington to Bromley transmission line was ity. out of service for a number of hours. This was likely due to flashovers caused by frequent clashing between • Such an investigation may also indicate that actual conductors when snow suddenly fell from a lower risk is lower than anticipated or that the cost of conductor allowing it to spring up and hit the conductor mitigation could not be justified in relation to the above. In addition, as access was a problem, the risk. reconnection of the line was delayed for safety reasons until it could be inspected. 5.5 Southpower Distribution Substations Network Substations buildings and equipment have low vulnerability to damage during a severe snowstorm. Introduction Southpower receives power from the following nine Electrical system snowstorm mitigation Trans Power substations:

Regional network Islington; The principal issues and actions to emerge were: Electrical and Communications • 123

Addington; been laid in a common trench spaced 300 mm apart at a minimum depth of 750 mm. Papanui; Over their full length they have been encased in weakmix Bromley; cement-bound sand (ratio 1 to 15) to improve their Springston; thermal rating, especially in dry soil or sand. The concrete has cross-sectional dimensions of 600 mm Hororata; (w) by 300 mm (h), and is capped by a 50 mm layer of hard concrete dyed red which provides some physical Arthurs Pass; protection from mechanical diggers.

Castle Hill; and Most 66 kV District substation buildings meet latest Coleridge. building codes, except for Fendalton (1953) and Brighton (1957). Most of the power is distributed from 45 district (zone) substations fed at either 66 kV, 33 kV or 11 kV from the 33 kV district substations Trans Power substations. The 66 kV network includes All of these substations, except Moffett, are supplied three outdoor switchyards located at Halswell, by a 33 kV overhead line network, and have either one Heathcote and Pages. Several Southpower district or two 33 to 11 kV transformers. Moffett is fed by two substations are outside the lifelines study area and are 33 kV underground cables. not listed. The larger substations have two independent trans- The district substations feed a comprehensive 11 kV formers fed from separate overhead lines with each network of approximately 7,500 small substations transformer and line rated to carry the full district which in turn provided a 230/400 volt street supply. substation load should the other fail. The 11 kV switchgear feeds up to 8 x 11 kV cables and is housed District substation network in two switchrooms linked by a buscoupler. There are 30 district substations (Table 5.1) within the The 33 kV overhead line network has the flexibility to lifelines study area are and these are supplied from four allow inter-connection between substations via differ- Trans Power substations. ent routes. Some interconnection is also possible using All of these substations can be remotely controlled and the 11 kV overhead line network. monitored from Southpower’s Armagh Street control Middleton, Moffett and Shands substation buildings room using a computerised supervisory control and are of modern design and constructed from a series of data acquisition (SCADA) control system. Manual large, rectangular, reinforced concrete pipe sections control can be exercised locally if necessary. connected together to form a room.

66 kV district substations Harewood, Hornby and Sockburn substations are of Most of these substations are supplied by two radial 66 concrete block construction, and were built prior to kV cables connected to two 66 to 11kV transformers. 1970. Each 66 kV cable and associated transformer has an emergency rating equivalent to the full load of the 11 kV district substations district substation (normally 40 MVA) and can main- These substations are directly supplied at 11 kV by tain supply should the other cable or transformer fail. either 3 or 4 radial 11 kV cables, and do not require supply transformers. The cables have usually been laid The transformers supply 11 kV switchgear housed in along the same route and have sufficient capacity to either two, three, or four, fire- and explosion-resistant supply the full District substation load should one switchgear rooms. The switchgear may feed up to 20 cable fail. The 11 kV switchgear may feed up to 12 x x 11 kV cables and can be sectionalised using 11 kV cables and is housed in either two or three buscouplers between the switchgear rooms. switchgear rooms links by buscouplers.

There are very few interconnections between substa- These substations can be interconnected during a fault tions at 66 kV but ties are provided by a comprehensive using the comprehensive 11 kV underground cable 11 kV network. network. Most 11 kV district substation buildings For each district substation the two 66 kV cables have meet latest building codes, except for Grimseys Win- ters (1953) and Woolston (1940). 124 • Risks and Realities

Trans Power Southpower Substations Supply Rated Substations Voltage Capacity (kV) (MVA) Addington Armagh 66 40 Fendalton 66 40 Milton 66 40 Oxford/Tuam 66 40 Foster 11 20 Knox 11 20 Montreal 11 20 Spreydon 11 20 Plus 1 single cable feed 11 - Bromley Barnett Park 66 20 Brighton 66 40 Dallington 66 40 Heathcote (incl switchyard) 66 40 Pages (switchyard) 66 - Portman 66 20 Linwood 11 20 Pages Kearneys 11 20 Plus 5 single cable feeds 11 - Woolston (indirect supply) 11 20 Papanui McFaddens 66 40 Bishopdale 11 20 Grimseys Winters 11 20 Harris 11 20 Plus 2 single cable feeds 11 - Islington Halswell (switchyard) 66 - Hawthornden 66 40 Hoon Hay 66 40 Harewood 33 15 Hornby 33 20 Middleton 33 23 Moffett 33 23 Shands 33 23 Sockburn 33 20

Table.5.1: District substations within the lifelines study area

Primary 11 kV system outer city areas. The substations are ring connected Each district substation feeds a primary 11 kV under- where possible with the mid-point switches left open. ground cable distribution system which supplies a The circuits are protected at the network substations by number of network building substations usually con- overcurrent relays. Normally the faulted portion can nected as closed rings. The ring cables have “differen- be isolated and the 11 kV system switched to quickly tial” protection which automatically isolates them dur- restore power via another route. There are approxi- ing a cable fault without loss of supply. These substa- mately 2,300 kiosks and 300 small building substations house several units of 11 kV metal clad switchgear tions. Although Southpower has approximately 5,000 and a local distribution transformer and, in the inner pole-mounted 11 kV to 400 volt transformers, only 300 city, are often within “high rise” building complexes. are within the urban area. Several are within building basements and most have pumping systems and water level alarms. Altogether Kiosk substations are of sheet metal construction and there are approximately 260 network substations. house switchgear and a distribution transformer to provide a street supply of 230/400 volts. The transformer sizes used range from 200 kVA to 500 kVA. Secondary 11kV system This restriction in size ensures the load may be easily Each network substation feeds a secondary 11 kV transferred to surrounding substations in the event of a underground cable distribution system which supplies system fault. a number of kiosk substations and small building substations. Overhead 11 kV lines are used in some In industrial and commercial areas, the transformer Electrical and Communications • 125

size may need to be increased to 1000 kVA or several and Sockburn substations which have unfilled con- transformers used. Either small building substations, crete block walls and Woolston substation which has or kiosks with externally mounted transformers, may concrete columns and brick infill panel walls. These be used to accommodate the larger transformers. have a higher risk of moderate structural damage but this is not expected to cause a loss of supply.

230/400 volt system At Sockburn and Moffett substations, sound attenua- The 230/400 volt circuits are a mixture of overhead tion walls require checking for strength and stability as lines and underground cables with much of the inner they could damage switchyard equipment if they fell. city supplied by underground cables. Southpower has a long-term programme for undergrounding most of The 66 kV to 11 kV transformers at district substations the city overhead reticulation, and for several years all weigh approx 45 tonnes and have a separate radiator new subdivisions have been designed using under- bank mounted on a common continuous foundation ground cables. pad. These have adequate mounting strength. The 33 kV to 11 kV transformers at district substations weigh Most of the 230/400 volt system has been designed slightly less but in some cases the radiator bank has with interconnections to allow the quick restoration of been mounted on a separate, non-continuous founda- power during a fault. tion pad. Some of these also have inadequate fixing to their foundation pads. The cooling radiator banks and Electrical system vulnerability oil conservator tanks on Ferranti transformers needs to be checked for strength and may require extra bracing. See Figures 5.8 to 5.19. The 11 kV indoor switchgear at district substations is District and local networks mounted on concrete floors and appears to be adequately fixed. The 11 kV auxiliary and voltage trans- Earthquakes formers mounted on rails on the top of some switchgear The study of the electrical system has largely been units may need to be more securely fixed. confined to 30 district (zone) substations as they sup- Both 66 kV and 33 kV outdoor switchgear is used at ply the bulk of the power to the lifelines study area of some of Southpower’s district substations. At Hornby urban Christchurch and Lyttelton. and Sockburn substations, the switchgear support District substations supply approximately 260 net- frames are very flexible and may require further brac- work substations and these form the next most impor- ing. tant distribution layer. A detailed study of these has not Control cabinets within these substations have been been carried out because their area of supply can checked and found to be securely mounted. usually be fed from surrounding network substations using the large number of 11 kV interconnections that normally exist. Network substations A small percentage of these buildings were built in the To assist with this study, Southpower has employed a 1920s and 1930s, and some may be sufficiently dam- consultant (Kingston Morrison) to carry out a survey of aged to disable the equipment inside and cause a loss of all district substations and some selected smaller sub- supply. stations. These very old substations were constructed from Earthquake vulnerability charts have been prepared for double brick, unusually with a reinforced concrete district substations and important Southpower office band at roof level, but without reinforced concrete and stores buildings. columns. Fortunately they are normally surrounded by more modern ones which can pick up their load via the The study has also highlighted Southpower’s risk of 11 kV system. There are six areas of the city where exposure to claims of contravention of the provisions several older network substations (prior to 1940) are in of the Resource Management Act due to oil spill. close proximity. These are in the suburbs of Addington, Beckenham, North Linwood, Papanui, St Albans and District substations Woolston. District substation buildings are mostly well designed The 11 kV switchgear in network substations has been using reinforced and filled concrete block construction upgraded and well maintained over the years and and should withstand an earthquake without much should not cause many problems. damage. The most vulnerable are Harewood, Hornby 126 • Risks and Realities

UTILITY: Southpower DISTRICT NETWORK: 66 kV District Subs (& 66 kV cables)

VULNERABILITY TO HAZARD IMPACT OF DAMAGE

COMPONENT ELEMENT COMMENTS

PERIOD FOLLOWING

LANDSLIDE LIQUEFACTION GROUND SHAKE IMPORTANCE GROUND SETTLEMENT RETURN TO NORMALITY RETURN TO DURING EARTHQUAKE EARTHQUAKE HAZARD IMMEDIATELY AFTER IMMEDIATELY QTY Armagh Building x1 5 3B 1 110 1 1 1 Built? Strong 11kV swgr x36 1 1 1 1 66kV tsfmrs x2 1 1 1 1 66 cables x2 3/3B 1 2 2 3 1 Two old bridges (Armagh) Barnett Park Building x1 5 2 1 001 1 1 1 Built 1977 11kV swgr x12 1 1 1 1 66kV tsfmrs x1 1 1 1 1 66 cables x2 2/2B 1 11 2 2 2 1 Possible pylon damage Brighton Building x1 5 3A 22 0 1 1 1 2 Built 1957 11kV swgr x23 1 1 1 1 66kV tsfmrs x2 1 1 1 1 66 cables x2 3A 1 1 1 1 Dallington Building x1 5 3 1 110 1 1 1 Built 1972 11kV swgr x26 1 1 1 1 66kV tsfmrs x2 1 1 1 1 66 cables x2 3/3A 1 1 1 1 1 Modern foot bridge (Snell) Fendalton Building x1 5 3 2 001 1 1 2 Built 1953 11kV swgr x23 1 1 1 1 66kV tsfmrs x2 1 1 1 1 66 cables x2 3/3B 1 1 1 1 1 Halswell Building x1 5 3B 1 110 1 1 1 Built 1969 and 1989 Switchyard 66kV swgr x7 2 2 2 2 Designed to 0.75g 66 lines x4 3/3B 1 221 2 2 1 Possible pylon damage Bldg-Ripple x1 3 3B 1 110 1 1 1 Built 1989 Ripple eqpt x1 1 1 1 1 Hawthornden Building x1 5 3 1 001 1 1 1 Built 1970 11kV swgr x26 1 1 1 1 66kV tsfmrs x2 1 1 1 1 66 cables x2 3 1 2 2 1 Possible pylon damage Heathcote Building x1 5 3B 1 1 0 1 1 1 1 Built 1971 Switchyard 11kV swgr x26 1 1 1 1 66kV swgr x6 2 2 2 2 Designed to 0.75g 66kV tsfmrs x1 1 1 1 1 66 cables x4 3A/ 1 2 1 2 2 2 1 Possible pylon damage 3B Hoonhay Building x1 5 3B 1 1 0 1 1 1 1 Built 1972 11kV swgr x26 1 1 1 1 66kV tsfmrs x2 1 1 1 1 66 cables x2 3B 1 1 1 1

For each component: Assess importance 1 to 5 Assess vulnerability on hazard map 0 to 3 Assess impact of damage 0 to 3

Figure 5.8: Southpower vulnerability chart 1

Kiosk and building substations satisfactorily because they are connected to flexible The transformers installed within kiosk substations HV and LV cables and restrained from moving far by have not been bolted down because of the difficulty of the sides of the Kiosk housing. Most have metal cable gaining access to their mountings. These substations boxes over the HV bushings which should protect are occasionally hit by vehicles and it has been found them from damage by impact. With a severe earth- that the transformer damage is reduced by its ability to quake the damage may be more substantial. A suitable move. method of mounting these transformers is being investigated. It is likely that they will survive a moderate earthquake Electrical and Communications • 127

UTILITY: Southpower DISTRICT NETWORK: 66 kV District Subs (& 66 kV cables)

VULNERABILITY TO HAZARD IMPACT OF DAMAGE

COMPONENT ELEMENT COMMENTS

PERIOD FOLLOWING

LANDSLIDE LIQUEFACTION GROUND SHAKE IMPORTANCE GROUND SETTLEMENT EARTHQUAKE HAZARD QTY RETURN TO NORMALITY RETURN TO DURING EARTHQUAKE IMMEDIATELY AFTER IMMEDIATELY McFaddens Building x1 5 31001 111Built 1972 11kV swgr x26 1 111 66kV tsfmrs x2 1 111 66 cables x2 3/3B 1 1 111 Milton Building x1 5 3 1 00 1 111Built 1981 11kV swgr x28 1 111 66kV tsfmrs x2 1 111 66 cables x2 3/3A 1 1 111 /3B Oxford-Tuam Building x1 5 3B 1 110 111Built 1974 11kV swgr x24 1 111 66kV tsfmrs x2 1 111 66 cables x2 3B 1 111 Pages Building x1 5 3A 1 2 0 1 111Built 1989 Switchyard 66kV swgr x1 2 111 66 cables x2 3A 1 111 Bldg-ripple x1 3 3A 1 110 111 ripple eqpt x1 1 111 Papanui Bldg-ripple x1 3 3/3B 1 110 111Built 1989 (ripple) ripple eqpt x1 1 111 Portman Building x1 5 3B 1 110 111Built 1986 11kV swgr x18 1 111 66kV tsfmrs x1 1 111 x1 3A/ 66 cables 1 111 3B

For each component: Assess importance 1 to 5 Assess vulnerability on hazard map 0 to 3 Assess impact of damage 0 to 3

Figure 5.9: Southpower vulnerability chart 2

All of the transformers in smaller building substations possible for the cable tension to also break the should be bolted down, but a check is required to transformer bushings; confirm this. The mounting brackets on some transformers appear to be insufficient and should be re- • at the base of power poles supporting cables con- placed. nected to the overhead system; and • in kiosks where underground cables connect di- Cable systems rectly onto low and high voltage switchgear.

It is expected that most underground cables will re- These problems are less likely to occur at building spond well to an earthquake although damage can be substations as the cables are supported in an open expected where cables are stretched as a result of trench system and have freedom to move if required. ground subsidence. Some district substations are fed by 66kv cables which Some cable strain may occur where underground ca- have been encased in concrete to improve their thermal bles rise directly from the ground without slack and are rating and mechanical protection. They are expected to supported or terminated on ground mounted equip- withstand an earthquake satisfactorily except in areas ment. Locations where this can occur are: where differential ground settlement is possible. This • at external pad mounted transformers which are may crack the unreinforced concrete sleeve and cause connected directly to underground cables. It is localised crushing or shearing of the cable. 128 • Risks and Realities

UTILITY: Southpower DISTRICT NETWORK: 11 kV District Subs (& 11 kV supply cables)

VULNERABILITY TO HAZARD IMPACT OF DAMAGE

COMPONENT ELEMENT COMMENTS

PERIOD FOLLOWING

LANDSLIDE LIQUEFACTION GROUND SHAKE IMPORTANCE GROUND SETTLEMENT RETURN TO NORMALITY RETURN TO DURING EARTHQUAKE EARTHQUAKE HAZARD IMMEDIATELY AFTER IMMEDIATELY QTY Bishopdale Building x1 531 001 111Built 1966 11kV swgr x18 1 111 11kV cables x3 3 1 0 1 1 1 Cashmere Road Building x1 521 0 0 1 111 Built197? - strong (network centre) 11kV swgr x11 1 111 11kV cables x3 2/3B 1 0 111 Foster Building x1 53B1 110 111Built 1992 11kV swgr x20 1 1 1 1 11kV cables x3 3B 1 0 0 1 1 1 Grimseys- Building x1 532 110 1 1 2 Built 1953 Winters 11kV swgr x23 1 1 1 1 11kV cables x6 3/3B 1 0 0 1 1 1 Harris Building x1 5 3B 1 110 1 1 1 Built 1966 11kV swgr x18 1 1 1 1 11kV cables x3 3B 1 0 0 1 1 1 Knox Building x1 531 0 0 1 1 1 1 Built 1968 11kV swgr x21 1 1 1 1 11kV cables x3 3/3B 1 0 1 11 Linwood Building x1 53A1 110 1 11Built 1962 11kV swgr x16 1 1 11 11kV cables x4 3A 1 0 0 1 1 1 Montreal Building x1 53B1 1 0 1 1 11Built 1964 11kV swgr x18 1 1 11 11kV cables x3 3B 1 0 0 1 11 Pages Building x1 53A1 1 0 1 1 11Built 1977 Kearneys 11kV swgr 16 1 111 11kV cables x3 3A 1 00 111 Simeon Building x1 5 2 1 0 1 1 221 Built ? - swg room OK (network centre) 11kV swgr x11 1 1 1 1 11kV cables x2 1/2/ 1 1 2 22 1 Possible line damage 2B Spreydon Building x1 5 3B 1 1 0 1 111 Built 1966 11kV swgr x18 1 1 11 11kV cables x3 3/3B 1 0 0 11 1 Woolston Building x1 532 1 10 112Built 1940 (network centre) 11kV swgr x16 1 1 1 1 11kV cables x4 3A/ 1 0 0 111 3B

For each component: Assess importance 1 to 5 Assess vulnerability on hazard map 0 to 3 Assess impact of damage 0 to 3

Figure 5.10: Southpower vulnerability chart 3

Of particular concern is the pair of 66kv cables that run Overhead line systems between the Addington and Armagh substations and Southpower’s overhead line systems operate at 66 kV, pass over two Armagh Street bridges. An investigation 33 kV, 11 kV and 230/400 volts. Only the outer areas into the vulnerability of these bridges indicates that of Christchurch rely on supply from the overhead high damage in the form of cracking of the spandrels and voltage system. The 66 kV overhead lines are sup- wing walls is possible. Ground settlement at the ap- ported by pylons rather than poles. proaches is possible which may damage the cables. Although most of Southpower’s damage is likely to Their repair requires specialist skills. Electrical and Communications • 129

UTILITY: Southpower DISTRICT NETWORK: 66 kV District Subs (& 33 kV cables)

VULNERABILITY TO HAZARD IMPACT OF DAMAGE

COMPONENT ELEMENT COMMENTS

PERIOD FOLLOWING

LANDSLIDE LIQUEFACTION GROUND SHAKE RETURN TO NORMALITY RETURN TO GROUND SETTLEMENT IMPORTANCE DURING EARTHQUAKE IMMEDIATELY AFTER IMMEDIATELY EARTHQUAKE HAZARD QTY Harewood Building x1 532 00 1 1 1 2 Built 1972 - unfilled blk 11kV swgr x9 1 1 11 33kV swgr x2 1 2 2 2 33kV tsfmr x2 3 3 3 2 Inadequate hold down 33kV lines x2 3 2 2 2 2 1 Possible line damage Hornby Building x1 5 3 2 100 1 1 2 Built? - unfilled blk 11kV swgr x11 1 1 1 1 33kV swgr x7 1 2 2 2 33kV tsfmr x2 3 3 3 2 Inadequate hold down 33kV lines x3 3 2 2 221 Possible line damage Middleton Building x3 4 3 1 00 1 1 1 1 Built 1972 11kV swgr x11 1 1 1 1 33kV swgr x2 1 2 22 33kV tsfmr x2 3 332 Inadequate hold down 33kV lines x2 3 2 2 2 2 1 Possible line damage Moffet Building x3 4 3 1 010 1 11 Built 1972 Sound wall x1 1 1 11 11kV swgr x11 1 1 1 1 33kV swgr x3 1 222 33kV tsfmr x2 3 3 3 2 Inadequate hold down 33kV lines x2 3 2 2 2 2 1 Possible line damage Shands Building x3 3 3 1 00 1 1 11 Built 1977 11kV swgr x11 1 1 1 1 33kV swgr x2 1 2 22 33kV tsfmr x2 3 3 3 2 Inadequate hold down 33kV lines x2 3 2 2 2 2 1 Possible line damage Sockburn Building x1 4 3 2 00 1 1 1 2 Built? - unfilled blk Sound wall x1 2 2 2 1 11kV swgr x10 1 1 1 1 33kV swgr x3 1 22 2 33kV tsfmr x2 3 3 3 2 Inadeuate hold down 33kV lines x2 3 2 2 2 2 1 Possible line damage

For each component: Assess importance 1 to 5 Assess vulnerability on hazard map 0 to 3 Assess impact of damage 0 to 3

Figure 5.11: Southpower vulnerability chart 4 occur to the overhead reticulation system, it is expected • Overhead line and insulator damage. The lines and that it will be easily repaired. Failure of supply can be insulators may break if poles move out of align- caused by: ment.

• Poles moving out of alignment. This is likely to Emergency spares occur with poles subjected to a sideways strain and where the ground is unstable or may liquefy. Southpower has a works depot at Packe St where distribution equipment spares are stored. Work is in • Pole mounted transformer damage. This can occur progress to catalogue these items and identify addi- when crossarm hung transformers and platform tional items to be ordered. Attention has been given to mounted transformers shift position. The out of bracing the storage racks and providing restraints to balance weight of larger transformers may cause prevent items falling off shelves. The possibility of some poles to lean or break. restraining stored transformers is being investigated. 130 • Risks and Realities

UTILITY: SOUTHPOWER LIMITED REGIONAL/LOCAL NETWORK: ELECTRICITY DISTRIBUTION

VULNERABILITY TO IMPACT OF HAZARD DAMAGE

COMPONENT COMMENTS

ELEMENT

Period following Local flooding Immediately after During flooding Return to normality Tsunami flooding Tsunami Waimak flooding Waimak Importance 1 - 5 Quantity

Armagh Building x1 5 2 1 0 1 1 0 0 11 kV Sw/ gr x36 5 2 1 0 1 1 0 0 66 k V Trans x 2 5 2 1 0 1 0 0 0 66 Cables x2 5 1 1 0 2 0 0 0 Edge of Flood plain Barnett Park Building x1 5 1 1 0 1 1 0 0 11 kV Sw/ gr x12 5 1 1 0 1 1 0 0 66 k V Trans x1 5 1 1 0 1 0 0 0 66 Cables x2 5 1 1 0 2 0 0 0 Brighton Building x1 5 3 3 0 3 3 2 2 11 kV Sw/ gr x23 5 3 3 0 3 3 2 2 66 k V Trans x2 5 2 2 0 1 0 0 0 Within flood plain 66 Cables x2 5 1 1 0 2 0 0 0 Dallington Building x1 5 1 1 0 1 1 0 0 11 kV Sw/ gr x26 5 1 1 0 1 1 0 0 66 k V Trans x2 5 1 1 0 1 0 0 0 66 Cables x2 5 1 1 0 2 0 0 0 Fendalton Building x1 5 1 1 0 1 1 0 0 11 kV Sw/ gr x23 5 1 1 0 1 1 0 0 66 k V Trans x2 5 1 1 0 1 0 0 0 66 Cables x2 5 1 1 0 2 0 0 0 Halswell Building x1 5 1 1 0 1 1 0 0 Switchyard 66kV Sw/gr x7 5 1 1 0 1 0 0 0 66 Lines x4 5 1 1 0 2 0 0 0 Bldg - Ripple x1 5 1 1 0 1 1 0 0 Ripple Eqpt x1 5 1 1 0 3 2 1 1 Hawthornden Building x1 5 1 1 0 1 1 0 0 11 kV Sw/ gr x26 5 1 1 0 1 1 0 0 66 k V Trans x2 5 1 1 0 1 0 0 0 66 Cables x2 5 1 1 0 2 0 0 0 Heathcote Building x1 5 2 2 0 1 1 0 0 11 kV Sw/ gr x26 5 2 2 0 1 1 0 0 66kV Sw/ gr x6 5 1 1 0 1 0 0 0 66 k V Trans x1 5 1 1 0 1 0 0 0 66kV Lines x4 5 1 1 0 2 0 0 0

For each component: Assess importance 1 to 5 Assess vulnerability on hazard map 0 to 3 Assess impact of damage 0 to 3

Figure 5.12: Southpower vulnerability chart 5

Office buildings kVA standby generator provides a backup supply for Southpower’s new office buildings at the corner of the control room. Southpower is currently looking into Manchester and Armagh Streets, which have some the possibilities of relocating the control room from important functions, are of sound construction. this old building (since done). Southpower’s old Armagh Street buildings, which The Packe St works depot buildings were checked for have other important functions, are of inadequate adequate strength. strength. The control room which monitors and controls the operation of the power distribution system is located within this old building complex and is manned Radio communication links by system controllers/operators 24 hours per day. A 35 A number of Telecom transmitter sites are used for Electrical and Communications • 131

UTILITY: SOUTHPOWER LIMITED REGIONAL/LOCAL NETWORK: ELECTRICITY DISTRIBUTION

VULNERABILITY TO IMPACT OF HAZARD DAMAGE

COMPONENT COMMENTS

ELEMENT

Period following Local flooding Immediately after During flooding Return to normality Tsunami flooding Tsunami Waimak flooding Waimak Importance 1 - 5 Quantity

Hoon Hay Building x1 5 1 2 0 1 1 0 0 11kV Sw / gr x26 5 1 2 0 1 1 0 0 66kV Trans x2 5 1 1 0 1 0 0 0 Edge of local flood 66 Cables x2 5 1 1 0 2 0 0 0 plain

Mc Faddens Building x2 5 1 1 0 1 1 0 0 11kV Sw / gr x26 5 1 1 0 1 1 0 0 66kV Trans x2 5 1 1 0 1 0 0 0 66 Cables x2 5 1 1 0 2 0 0 0 Milton Building x1 5 1 1 0 1 1 0 0 11kV Sw / gr x28 5 1 1 0 1 1 0 0 66kV Trans x2 5 1 1 0 1 0 0 0 66 Cables x2 5 1 1 0 2 0 0 0 Oxford - Tuam Building x1 5 2 1 0 1 1 0 0 11kV Sw / gr x24 5 2 1 0 1 1 0 0 66kV Trans x2 5 2 1 0 1 0 0 0 66 Cables x2 5 11 0 2 0 0 0 Edge of flood plain Pages Building x1 5 2 2 0 0 0 0 0 Switchyard 66kV Sw / gr x1 5 1 1 0 1 0 0 0 66 Cables x2 5 1 1 0 2 0 0 0 Bldg - Ripple x1 5 2 2 0 1 1 0 0 Edge of flood plain Ripple Eqpt x1 5 2 2 0 3 2 1 1 Papanui Bldg - Ripple x1 5 1 1 0 1 1 0 0 (Ripple) Ripple Eqpt x1 5 1 1 0 3 2 0 0 Portman Building x1 5 1 1 0 1 1 0 0 11kV Sw / gr x18 5 1 1 0 1 1 0 0 66kV Trans x1 5 1 1 0 1 0 0 0 66 Cable x1 5 1 1 0 2 0 0 0 Harewood Building x1 5 1 1 0 1 1 0 0 11kV Sw / gr x9 5 1 1 0 1 1 0 0

33kV Sw / gr x2 5 1 1 0 1 0 00Edge of Waimak food plain 33kV Trans x2 5 1 1 0 1 0 0 0 33kV Lines x2 5 1 1 0 2 0 0 0

For each component: Assess importance 1 to 5 Assess vulnerability on hazard map 0 to 3 Assess impact of damage 0 to 3

Figure 5.13: Southpower vulnerability chart 6

day-to-day vehicle communication. Most vehicles have Electrical system mitigation measures radiotelephones which can switch channels to different — earthquakes repeaters if required. General A radio link at Mt Sugarloaf is also used to control and Southpower is able to respond promptly to electricity monitor the operation of some remote district substa- outages caused by a wide range of emergencies as part tions. Within most of the Christchurch urban area of its routine operations. This method of operation and remote control is achieved using Southpower’s under- preparedness will be of assistance in the event of a ground communication cable network. major system disruption. 132 • Risks and Realities

UTILITY: SOUTHPOWER LIMITED REGIONAL/LOCAL NETWORK: ELECTRICITY DISTRIBUTION

VULNERABILITY TO IMPACT OF HAZARD DAMAGE

COMPONENT COMMENTS

ELEMENT

Period following Local flooding Immediately after During flooding Return to normality Tsunami flooding Tsunami Waimak flooding Waimak Importance 1 - 5 Quantity

Hornby Building x1 5 3 x1 0 1 1 0 0 11kV Sw / gr x11 5 3 x1 0 1 1 0 0 Within Waimak 33kV Sw / gr x7 5 x2 x1 0 1 0 0 0 flood plain

33kV Trans x2 5 x2 x1 0 1 0 0 0 33k V Lines x3 5 x2 x1 0 2 0 0 0 Middleton Building x3 5 x1 x1 0 1 1 0 0 11kV Sw / gr x11 5 x1 x1 0 1 1 0 0 33kV Sw / gr x2 5 x1 x1 0 1 0 0 0 33kV Trans x2 5 x1 x1 0 1 0 0 0 33k V Lines x2 5 x1 x1 0 2 0 0 0 Moffett Building x3 5 x1 x1 0 1 1 0 0 Sound Wall x1 5 x1 x1 0 1 1 0 0 11kV Sw / gr x11 5 x1 x1 0 1 1 0 0 33kV Sw / gr x3 5 x1 x1 0 1 0 0 0 33kV Trans x2 5 x1 x1 0 1 0 0 33k V Lines x2 5 x1 x1 0 2 0 0 0 Shands Building x3 5 x1 x1 0 1 1 0 0 11kV Sw / gr x11 5 x1 x1 0 1 1 0 0 33kV Sw / gr x2 5 x1 x1 0 1 0 0 0 33kV Trans x2 5 x1 x1 0 1 0 0 0 33k V Lines x2 5 x1 x1 0 2 0 0 0 Sockburn Building x1 5 x1 x1 0 1 1 0 0 11kV Sw / gr x10 5 x1 x1 0 1 1 0 0 33kV Sw / gr x3 5 x1 x1 0 1 0 0 0 33kV Trans x2 5 x1 x1 0 1 0 0 0 33k V Lines x2 5 x1 x1 0 2 0 0 0 Bishopdale Building x1 5 x1 x1 0 1 1 0 0 11kV Sw / gr x18 5 x1 x1 0 1 1 0 0 11kV Cables x3 5 x1 x1 0 1 1 0 0 Cashmere Rd Building x1 5 2 x1 0 1 1 0 0 (Network Centre) 11kV Sw / gr x11 5 x2 x1 0 1 1 0 0 Edge of flood plain 11kV Cables x3 5 x2 x1 0 1 1 0 0 Foster Building x1 5 x1 x1 0 1 1 0 0 11kV Sw / gr x20 5 x1 x1 0 1 1 0 0 11kV Cables x3 5 x1 x1 0 1 1 0 0

For each component: Assess importance 1 to 5 Assess vulnerability on hazard map 0 to 3 Assess impact of damage 0 to 3

Figure 5.14: Southpower vulnerability chart 7

Widespread damage to the distribution system would District substations place enormous demands on resources, and it may be • Strengthen the mounting system for some 33/11 kV necessary to approach other electrical supply authori- supply transformers. ties for assistance. In the event of a major emergency, it is likely that Trans Power’s system will also be • Check the strength and stability of 33/11 kV supply affected with the consequent loss of supply to most or transformers foundation pads which were not con- all of Christchurch. structed as one continuous unit.

• Check the strength of the radiator banks and oil Engineering measures conservator tanks on the Ferranti 66/11 kV supply The following mitigation measures have been identi- transformers and fit extra bracing if necessary. fied. Electrical and Communications • 133

UTILITY: SOUTHPOWER LIMITED REGIONAL/LOCAL NETWORK: ELECTRICITY DISTRIBUTION

VULNERABILITY TO IMPACT OF HAZARD DAMAGE

COMPONENT COMMENTS

ELEMENT

Period following Immediately after Local flooding Return to normality During flooding Tsunami flooding Tsunami Waimak flooding Waimak Importance 1 - 5 Quantity

Grimseys Winters Building x1 5 1 1 0 1 1 0 0 11kV Sw / gr x23 5 1 1 0 1 1 0 0 11kV Cables x6 5 1 1 0 1 1 0 0 Harris Building x1 5 1 1 0 1 1 0 0 11kV Sw / gr x18 5 1 1 0 1 1 0 0 11kV Cables x3 5 1 1 0 1 1 0 0 Knox Building x1 5 1 1 0 1 1 0 0 11kV Sw / gr x21 5 1 1 0 1 1 0 0 11kV Cables x3 5 1 1 0 1 1 0 0 Linwood Building x1 5 1 1 0 1 1 0 0 11kV Sw / gr x16 5 1 1 0 1 1 0 0 11kV Cables x4 5 1 1 0 1 1 0 0 Montreal Building x1 5 1 1 0 1 1 0 0 11kV Sw / gr x18 5 1 1 0 1 1 0 0 11kV Cables x3 5 1 1 0 1 1 0 0 Pages Kearneys Building x1 5 1 1 0 1 1 0 0 11kV Sw / gr x16 5 1 1 0 1 1 0 0 11kV Cables x3 5 1 1 0 1 1 0 0 Simeon Building x1 5 0 0 0 1 1 0 0 (Network Centre) 11kV Sw / gr x11 5 0 0 0 1 1 0 0 Lyttleton 11kV Lines x2 5 0 0 0 1 1 0 0 Spreydon Building x1 5 1 1 0 1 1 0 0 11kV Sw / gr x18 5 1 1 0 1 1 0 0 11kV Cables x3 5 1 1 0 1 1 0 0 Woolston Building x1 5 1 1 0 1 1 0 0 (Network Centre) 11kV Sw / gr x16 5 1 1 0 1 1 0 0 Edge of Flood plain 11kV Cables x4 5 1 1 0 1 1 0 0

For each component: Assess importance 5 to 5 - 5 most important Assess vulnerability on hazard map 1 to 3 - 3 most vulnerable Assess impact of damage 1 to 3 - 3 most impact

Figure 5.15: Southpower vulnerability chart 8

• Check the strength of the 33 kV outdoor switchgear old network substation buildings if they have a support frames and strengthen if necessary. critical role in the security of supply. This could be achieved by bracing the building at roof level to • Check the stability of the sound attenuation walls at transfer the wall face loads to the opposite side Sockburn and Moffett substations and strengthen walls. The walls themselves could be strengthened or replace as necessary. by bolting steel members to them at appropriate • Check the stability of the auxiliary and voltage positions to provide structural ribs to reduce face transformers mounted on the top of some units of loads. 11 kV switchgear and improve the method of fixing if necessary. Kiosk and building substations • If possible, establish a suitable method of securely Network substations mounting transformers in kiosk substations despite • Consider either strengthening or replacing selected the difficult access. 134 • Risks and Realities

UTILITY: SOUTHPOWER LIMITED REGIONAL/LOCAL NETWORK: ELECTRICITY DISTRIBUTION

VULNERABILITY TO IMPACT OF HAZARD DAMAGE

COMPONENT COMMENTS

ELEMENT

Importance 1 - 5

During Wind

Period following Wind Exposure Immediately after Return to normality

66kV TOWER LINES Access to most towers difficult on Hawthornden - Islington 66kV 5 3 3 3 2 0 Port Hills Islington - Halswell 66kV 5 3 3 3 2 0 Halswell - Heathcote 66kV 4 3 3 3 2 0 On hills Heathcote - Bromley 66kV 5 3 3 3 2 0

Heathcote - Barnett Park On hills 66kV/11kV 53 3320

33kV POLE LINES 5 3 3 3 2 1

11kV POLE LINES Heathcote - Lyttleton Line 5 3 3 3 3 1 On hills Other 11kV Lines 5 3 3 3 3 1

400V POLE LINES CITY Urban 4 2 3 3 2 1 City Rural 4 3 3 3 2 1

SUBSTATION STRUCTURES Heathcote 5 2 3 3 3 0 66kV Halswell 5 2 3 3 3 0 66kV & Ripple Plant Pages 5 2 3 3 3 0 66kV & Ripple Plant Papanui 5 2 3 3 3 0 66kV & Ripple Plant Sockburn 5 2 3 3 3 0 33kV Hornby 5 2 3 3 3 0 33kV & Ripple Plant Middleton 5 2 3 3 3 0 33kV Moffet 5 2 3 3 3 0 33kV & Ripple Plant Harewood 5 2 3 3 3 0 33kV Shands 5 2 3 3 3 0 33kV

SUBSTATION BUILDINGS 5 1 11 0 0

POLE MOUNTED 43 22 SUBSTATIONS 0 0

Vulnerability Chart: Define components and elements of network at Regional and District level For each component Assess importance 1 to 5 - 5 most important Assess vulnerability 1 to 3 - 3 most vulnerable Assess impact of damage 1 to 3 - 3 most impact

Figure 5.19: Southpower vulnerability chart 9

• Check the mounting for building substation trans- these should be investigated to determine if extra formers and strengthen any which are inadequate. support is needed.

Note: There is a small risk that HV and LV cables Cable systems could be stressed where they rise directly from the • The 66 kV cables could be damaged by excessive ground and connect to ground mounted equipment. differential ground settlement during an earthquake. The number of locations where this is possible is Sites identified where this could occur are the very large, and mitigation measures are not practical. approaches to the two Armagh Street bridges, and Electrical and Communications • 135

UTILITY: SOUTHPOWER LIMITED REGIONAL/LOCAL NETWORK: ELECTRICITY DISTRIBUTION

VULNERABILITY TO IMPACT OF HAZARD DAMAGE

COMPONENT COMMENTS

ELEMENT

Importance 1 - 5

Period following Immediately after Snow Loading During flooding Return to normality

Hawthornden - Islington 66 kV 5 2 3 3 1 0 On flat

Islington - Halswell 66kV 5 2 3 3 01 On flat

Access to most towers Halswell - Heathcote 4 3 3 3 20difficult on Port Hills

Heathcote - Bromley 66kV 5 2 3 3 10On flat

Heathcote - Barnett Park 5 3 3 3 20Access to most towers 66kV/11kV difficult on Port Hills

Lyttelton - Heathcote 11kV Access to most towers Line 5 3 3 3 20difficult on Port Hills

Sugarloaf 11kV Line 4 3 2 22 0 On hills but has standby

Marleys Hill 11kV Line 4 3 2 220 On hills but has standby

Mt Pleasant 11kV Line 4 3 2 2 2 0 On hills but has standby

33kV Lines 5 2 3 2 1 0 Alternative routes exist

Some alternatives Other 11kVLines 4 2 3 2 1 0 available

Some alternatives Low Voltage Lines 3 2 2 2 1 0 available

Some alternatives 3 Streetlighting 2 1 1 1 0 available

Substation Buildings 5 1 1 0 0 0

Switchyards 5 1 1 0 00

Figure 5.17: Southpower vulnerability chart 10

Overhead line systems need to be erected to accelerate the restoration of • The mounting strength of transformers on two pole power to some underground supply areas. structures should be checked and upgraded if necessary. General measures Note: Damage may be widespread on the overhead • The battery banks in some substations have inad- system, but should be repaired without too much equate support and require further bracing. difficulty as only standard components are used • Some of the essential equipment in the main office and access is easy. Temporary overhead lines may building is not adequately fixed. This includes 136 • Risks and Realities

UTILITY: SOUTHPOWER LIMITED REGIONAL/LOCAL NETWORK: ELECTRICITY DISTRIBUTION

VULNERABILITY TO IMPACT OF HAZARD DAMAGE

COMMENTS

COMPONENT ELEMENT

Importance 1 - 5 Slope Stability Period following Immediately after During Earthquake Return to normality

Access to most towers Halswell - Heathcote difficult on Port Hills 32 Towers Transmissions Line Tower Number 30 4 0 1 3 3 3 On flat ground 29 4 1 1 3 3 3 28 4 1 1 3 3 3 27 4 1 1 3 3 3 Rapaki Track 26 4 1 1 3 3 3 25 4 1 1 3 3 3 24 4 1 1 3 3 3 23 4 1 1 3 3 3 Huntsbury Track 22 4 1 1 3 3 3 21 4 1 1 3 3 3 20 4 1 1 3 3 3 19 4 1 1 3 3 3 Dyers Pass Road 18 4 1 1 3 3 3 17 4 1 1 3 3 16 4 2 1 3 3 3 15 4 2 1 3 3 3 14 4 1 1 3 3 3 13 4 1 1 3 3 3 12 4 2 1 3 3 3 11 4 1 1 3 3 3 Worsleys Spur 10c 4 1 1 3 3 3 10b 4 1 1 3 3 3 Westmorland 10a 4 1 1 3 3 3 9 4 1 1 3 3 3 Off Cashmere Road 8 4 1 1 3 3 3 7 4 2 1 3 3 3 6 4 3 1 3 3 3 5 4 0 1 3 3 3 On flat ground

Vulnerability Chart: Define components and elements of network at Regional and District level For each component: Assess importance 1 to 5 - 5 most important Assess vulnerability 1 to 3 - 3 most vulnerable Assess impact of damage 1 to 3 - 3 most impact

Figure 5.18: Southpower vulnerability chart 11

most computer centre equipment (computer, pe- prepared, and will be analysed to determine if any ripherals) and some control room equipment (VDUs increases are necessary. Some components can etc.). easily be sourced but the spare parts for some equipment such as older 11 kV switchgear are no Planning to manage the impact longer available. If some old equipment is seriously damaged, it may have to be completely Issues to consider are: replaced rather than repaired. • The maintenance of all spares in stock is being Electrical and Communications • 137

UTILITY: SOUTHPOWER LIMITED REGIONAL/LOCAL NETWORK: ELECTRICITY DISTRIBUTION

VULNERABILITY TO IMPACT OF HAZARD DAMAGE

COMMENTS

COMPONENT ELEMENT

Importance 1 - 5 Slope Stability Period following Immediately after Duringearthquake Return to normality

Access to most towers Heathcote-Barnett difficult on Port Hills

19 Towers Transmission Line Barnett Park Tower Number 19 5 1 1 3 3 3 Dyers Pass Road 18 5 2 1 3 3 3 17 5 2 1 3 3 3 16 5 3 1 3 3 3 Mt Pleasant Road 15 5 1 1 3 3 3 14 5 1 1 3 3 3 13 5 1 1 3 3 3 12 5 2 1 3 3 3 11 5 2 1 3 3 3 Bridle Path Road 10 5 0 1 3 3 3 On flat ground

Heathcote-Lyttleton Access to some poles difficult Transmission Line Tower Number* 1 5 1 3 3 3 2 5 1 3 3 3 3 5 1 3 3 3 4 5 3 1 3 3 3 Above vulnerable bank 5 5 1 1 3 3 3 Bridle Path track 6 5 1 1 3 3 3 7 5 1 1 3 3 3 8 5 1 1 3 3 Summit Road. New 9511 33 3 stucture 10 5 1 1 3 3 3 11 5 3 1 3 3 3 12 5 2 1 3 3 3 Off Harmans Road 13 5 3 1 3 3 3 Off Harmans Road 14 5 3 1 3 3 3 Adjacent Harmans Road 15 5 1 1 3 3 3 Harmans Road 16 5 2 1 3 33Harmans Road

* (Refer to Southpower layout 28-07-94) HV System Diagram sheet 15.2

Figure 5.19: Southpower vulnerability chart 12

• Ensuring as many staff as possible are well trained Southpower’s exposure to flooding risk and skilled in the restoration of services. General • Continuing with the present planning policy of In general, the distribution network is not exposed to ensuring alternative supply routes are provided for any great risk because of moderate flooding. Most HV consumers wherever possible and commercially equipment is constructed such that flooding in excess viable. 138 • Risks and Realities

of one metre would be required before catastrophic Other District Substations within the city are not af- failure of equipment would occur. fected by the flood plain.

Events such as the August 1992 storm, its associated snow melt and high spring tides, have already shown Local rivers flooding hazard the network to be quite robust with only localised Brighton District Substation falls within the local river flooding around substations close to the Heathcote and flood plain. However, as identified previously the Avon rivers. It would be possible, where localised Avon stopbank is expected to prevent major damage flooding in excess of one metre occurs, to electrically occurring, as past history has shown. isolate those substations on a substation-by-substation Hoon Hay District Substation appears to be on the edge basis as the need arises before major damage occurs to of this flood plain, but is not expected to be affected by electrical equipment. it. Heathcote District Substation is also on the edge of this flood plain, but is higher than the surrounding Control centre ground and is not expected to be affected by it. Pages This is located on the second floor of the Armagh site Switchyard is on the edge of the flood plain and, as and is not considered at risk. above, should not be affected.

Other district substations all appear to be unaffected by Waimakariri flood hazard flooding of the Avon and Heathcote rivers. Hornby District Substation appears to be the most exposed district substation to a Waimakariri flood and Tsunami flooding hazard falls within the potential flood plain. This substation may be vulnerable to damage from local flooding. This event is not expected to affect any district substa- However, this event is considered the least likely of the tions or major plant. flooding events. It is possible to transfer the load from Some local loss of power may occur at distribution this substation on to neighbouring substations should substations located close to the coast. However, this is the need arise. considered minor and it should be relatively easy to Brighton District Substation appears to be the next restore power. most exposed district substation which would be affected by the Waimakariri flooding as it appears to lie Localised heavy rainfall within the flood plain. Past history has shown that in Localised heavy rainfall in the central city can cause actual fact this substation has been unaffected by any flooding in basement substations situated below ground peak flooding which has occurred in recent years, and level in the basements of central city buildings. is protected by the Avon stopbank.This stopbank appears to be designed for a 50-year return period. How- It is expected they may not cope with the volume of ever, if this level was to be exceeded then the Brighton water that may flow down access ways because of District Substation would be exposed. flooding in the streets. These substations either have limited pumping facilities or rely entirely on the Heathcote District Substation is on the edge of the stormwater systems associated with the basements in flood plain, but on slightly higher ground than the which they are situated. surrounding area. There is not expected to be any major problems here. In the past twenty years there has been flooding in the central business district caused by localised heavy Armagh District Substation is located on the edge of rainfall. None of these occasions have caused major the flood plain and no major problems are expected to flooding of basement substations. However, the streets occur here. have not been flooded for long periods. Oxford-Tuam District Substation is on the edge of the Some of these substations are fitted with high water flood plain and has some diversification with half of level alarms. the switchgear located at ground level and the balance at the first floor level. Mitigation measures Pages Switchyard is on the edge of the flood plain. The subtransmission network (66 kV and 33 kV) Water entry into buildings could damage ripple equip- would generally not require any special mitigation ment. This is not expected to happen since the building measures since most of this equipment is mounted foundations are about 300 mm above ground level. higher than the distribution network. Electrical and Communications • 139

The electricity distribution network at the 11 kV level debris may cause conductor damage as they have in the is made up of many lines and cables able to provide a past. Hardware on towers can fail or bend, but this does multitude of options and alternatives for at least limited not usually disrupt the power supply and repairs can be power supply for the many flooding scenarios that carried out when the wind subsides. exist.

Specific flooding mitigation measures are as follows: 33 kV and 11 kV lines These lines are the most exposed to wind damage. Pole • control centre standby — shift standby power plant lines are designed to withstand wind speeds based on from below ground level to a position above ground NZS4203:1984. Poles are selected on their mechani- level to provide the control centre with a more cal load rating to meet Canterbury conditions. secure standby power supply; Lines are also exposed to falling trees and flying debris, • Hornby district substation — prepare a plan to however it is expected most damage would be limited transfer the load of this substation of nearby substa- to conductor damage since a conscious effort has been tions; made to reduce line spans to limit pole damage as a mitigation measure learnt from the 1975 wind storm. • Brighton district substation — identify the essential services that could be supplied from alternative Reducing the span size to 55 m to 60 m has proved nearby substations. beneficial in limiting damage as was proven in the high winds during April 1994 where some gusts were re- • Ensure pumps and hoses are on hand if needed; and ported to 130km/h.

• Establish a supply of sand bags. Heavy emphasis has been placed on having trees To achieve this the following is required: maintained near overhead lines in the last few years with the result that the number of power cuts have been • informed and competent operational staff; reduced.

• reliable UHF radio systems for communicating with operational staff; 400 V lines The low voltage network is usually less affected by • competent emergency contractors to provide resto- wind other than trees toppling or branches breaking off ration of services as soon as possible; and bringing down lines. The average pole span in the urban area is 40m. These poles usually have conduc- • reliable UHF radio and/or cell phone network for tors attached in a number of directions making failure communicating with contractors. less likely.

Southpower’s exposure to wind hazard Lines that fail are easily repaired, once the wind has reduced to a level that worker safety is not compro- General mised. Wind damage is considered high risk to the overhead line network. The most devastating winds in Canter- Pole mounted substations bury have been from the north west. History has shown These substations are at risk of failing in high winds lines crossing the north west wind suffer more damage because of the transformer’s large surface area to wind that others. and the relatively high centre of gravity of this struc- Events such as the big blow in 1975 caused major ture. The rural area distribution transformers are damage in the rural area, however the city urban area predominantly mounted on single pole structures and was less affected. two pole substation structures still exist in some of the older areas of Christchurch. The majority of the city Falling trees and flying debris are responsible for area now has transformers mounted on the ground and causing damage to the overhead distribution system. these are not expected to be affected by strong winds. Repairs cannot usually be made until the wind subsides to a safe level. Substation buildings and structures The greatest risk to substation buildings and structures 66 kV lines is caused by flying debris. Substation structures are It is expected that these lines will withstand damage more at risk than their adjacent buildings and any from high winds, however falling trees and flying damage could have a significant impact. Much of this 140 • Risks and Realities

equipment is older with the availability of spares located alongside the Southpower Halswell to Heathcote limited. line for much of its length. It is probable that both lines will be affected by these conditions as they were in Mitigation measures 1992, the result of which is loss of power to much of the eastern side of the city for some hours. • instigate emergency storm contingency plan; Snow and ice falling off conductors can cause me- • an active campaign to encourage tree owners to chanical failure of line hardware. keep trees clear of lines;

• hold some replacement switchgear as an emer- 33 kV lines gency spares; These lines are exposed to damage from snow and • hold emergency lines spares; and movement because of the damp ground conditions. Past experience has shown precast concrete poles to be • temporary concrete blocks. more at risk than wooden poles because of shock loading on the lines caused by falling trees and snow To achieve this the following is required: and ice falling from conductors. These lines are • informed and competent operational staff; located on the outskirts of the city and have alternative routes available to supply various district substations • reliable UHF Radio systems for communicating as the need arises. with operational staff;

• competent emergency contractors to provide resto- 11 kV lines ration of services as soon as possible; and The most exposed 11 kV lines are located on the Port Hills and some loss of power can be expected in this • reliable UHF radio and/or cell phone network for area. Repair work will be hampered by access to the communicating with contractors. site due to heavy snow fall.

The 11 kV line to Lyttelton is expected to be affected Southpower’s exposure to snow hazard by heavy snow fall but was unaffected by the 1992 General storm because the snow depth was minimal on the Lyttelton side of the Port Hills. In general the distribution network is not exposed to any long term risk because of a severe snow storm. Important communication sites on the Port Hills appear to be backed up by standby power plants. It is expected that any snow fall will be heavier on the Port Hills than on the plains and, as such, those lines located on the hills will be most exposed, with the Low voltage lines effect that access for repair will almost be impossible These lines are exposed, however the low voltage during the storm and immediately after. network is very robust and major problems tend to be limited to poles moving causing lines to hang low, The August 1992 snow storm tested the network with barge boards pulling off houses and leaning poles. heavy snow loading and damp ground conditions. A Generally loss of power on this part of the network small number of poles failed, however the major dam- would be minimal. age was due to poles moving with the extra weight causing lines to hang dangerously low or even break in The LV network relies on Telecom poles on the alter- some cases. native side of the road for its road crossings. These poles are generally not as robust as those in the electric- 66 kV tower lines ity network and some failures can be expected. This would result in loss of power supply at the local level The Halswell to Heathcote line and the Heathcote to only. Barnett Park lines are most exposed because of their location predominantly on the Port Hills. To aid the flow of traffic initially it may be necessary to cut road crossing supplies free of the network. Later These lines are generally very robust and constructed repair is relatively simple. on steel towers. In severe snow conditions ice build up on the conductors causes them to sag and sometimes break. Substation buildings and switchyards Substation buildings and switchyards are expected to Trans Power’s supply to their Bromley Substation is be relatively unaffected by a heavy snow fall. Electrical and Communications • 141

Mitigation measures a pole structure in this line a power supply to Lyttelton It is expected that some loss of supply will occur in the would be lost. city due to heavy snow loading on overhead lines. Islington-Bromley 66 kV line As much of the city gets its supply via the underground cable, network only parts of the city should be affected The most vulnerable section at risk to slope hazard is by heavy snow fall. the Halswell-Heathcote portion. This section has 32 towers in total. Virtually all the towers have a low to During snow fall only limited repairs will be achieved, medium exposure to slope stability. One tower with a however once it has subsided it is expected repairs will high rating adjacent to Cashmere Road is a new tower be made relatively quickly subject to site access avail- installed for the Westmorland deviation in 1991. This ability. has concrete belled foundations on each leg 4 m deep with a 1.2 m top diameter to 1.8 m bottom diameter. Specific mitigation measures are as follows: Blasting was not necessary for this foundation, how- • instigate emergency storm contingency plan; ever considerable effort was required to obtain the depth of 4 m. This work was carried out before any • prepare plans to shift load from Bromley Trans disturbance to the subsoil. Power supply to the Islington Southpower supply via the Southpower 66 kV network; Heathcote to Barnett Park 66 kV and 11 kV • prepare plans to provide a limited supply to the lines eastern part of the city via the 11 kV network should This line also has one tower with high exposure to slope the Heathcote/Halswell 66 kV line fail; stability. The other 18 towers all have low to medium exposure to slope stability. • identify essential services that should have priority within the city; Tower 16 is located on rock adjacent to a creek. The foundations for this tower are of the grillage type. The • ensure emergency levels of poles, line hardware grillage is 6’x4' and the four holes were blasted and dug and replacement conductor are available; and out to suit the grills. The lower two foundation holes have been keyed into the rock. • establish an emergency contact with appropriate helicopter companies. Lyttelton to Heathcote 11 kV line To achieve this the following is required: This line consists of 16 two-pole structures installed 2 • informed and competent operational staff; m apart bolted together with two horizontal crossarms. Four structures on this line have high exposure to slope • reliable UHF Radio systems for communicating stability. This line is to have maintenance performed with operational staff; on it in the next few years. Position of structures will be a necessary consideration at this time. • competent emergency contractors to provide restoration of services as soon as possible; and Replacement of the in-line structures will require specialist advice to achieve the best location. This pole • reliable UHF radio and/or cell phone network for line has been in this position for 40 years and appears communicating with contractors. to have escaped any major slips.

Southpower’s exposure to slope hazard Mitigation measures The Halswell to Heathcote line is operated in the General unloaded mode (i.e. open) and can be used to provide The risk of slope hazard affecting distribution towers a limited power supply to the eastern side of the city on the Port Hills is considered relatively low. Access should the Trans Power tower lines fail. This line has in wet weather can be difficult to some locations. The a lower importance because it is an alternative supply. Heathcote-Halswell section of this line is normally run as a backup supply thus reducing the importance of this Tower 6 is of concern. It is one of the most recently section of line. constructed towers with a lot of effort being placed on foundation construction. As a mitigation measure 15 The Heathcote-Lyttelton distribution line is consid- m hardwood poles are held as an emergency spare to ered vulnerable to slope hazard. Should a slip take out replace a tower should one fail. 142 • Risks and Realities

The Heathcote to Barnett Park tower line has a high Television Broadcasting facilities in Christchurch are level of importance and tower 16 has been identified as operated by Television New Zealand ( a State-Owned- the most vulnerable. Enterprise ) and by several private broadcasting organisations. There are network stations relaying pro- Further checks of this foundation have revealed that it grammes from Auckland, and local stations originat- is constructed on rock. As a mitigation measure 15 m ing programmes from studios in Christchurch city. hardwood poles are held as an emergency spare and as Television transmitter sites are located on the Port above should a tower fail these poles would be used as Hills, the major site is Sugarloaf, to the south of the a substitute. Christchurch metropolitan area, with some smaller The Lyttelton to Heathcote 11 kV line is to have pole translator sites serving topographically isolated areas. maintenance and replacement performed on it within the next few years. As a mitigation measure slope Sound Radio Broadcasting hazard will be a consideration at the design stage of this project. Radio New Zealand Radio New Zealand operates three sound radio broad- A further problem on the hills has been identified as cast stations in Christchurch, all with transmitters access to the location of individual towers or poles. As remote from the studios. These are: a mitigation measure a bulldozer is on standby as a contingency measure. • 675 kHz - 3YA, 10 kW transmitter at Gebbies Pass, broadcasting the “National Radio” programme from To achieve this the following is required: studios in Wellington, via the studios in Kent • emergency spares, 15 m poles; House, Christchurch. This is a network programme.

• establish an emergency contact; • 963 kHz - 3YC, 10 kW transmitter at Gebbies Pass, broadcasting the “AM Network” programme from • heavy lift helicopter, bulldozer or large backhoe, studios in Wellington, via the studios in Kent 4x4 vehicles; and House, Christchurch. This is a network programme.

• temporary concrete blocks/stays. • 89.7 MHz - 3CCP, 50 kW transmitter at Sugarloaf, broadcasting the “Concert FM” programme from To achieve this the following is required: studios in Wellington. Programme is carried by • informed and competent operational staff; B.C.L. circuits.

• reliable UHF Radio systems for communicating Private Radio with operational staff; The majority of sound radio broadcast stations in • competent emergency contractors to provide resto- Christchurch are privately owned. Almost all have ration of services as soon as possible; and their transmitters remote from the studios. These are:

• reliable UHF radio and/or cell phone network for • 612 kHz - 3XG, 2 kW transmitter in Winters Road, communicating with contractors. Mairehau, broadcasting the “Radio Rhema” programme from studios in Upper Queen Street, Auck- land. This is a network programme. Some local programming is included during the day from stu- 5.6 Broadcasting System dios in Birmingham Drive, Christchurch.

Introduction • 1098 kHz - 3ZB, 5 kW transmitter at Ouruhia, Sound Radio Broadcasting facilities in Christchurch broadcasting the Christchurch “Newstalk ZB” programme from studios in Kent House, Christchurch. are operated by several private broadcasting organisations, along with two publicly owned networks run by • 1260 kHz - 3XA, 2.5 kW transmitter in Hills Road, Radio New Zealand. There are local stations originat- Mairehau, broadcasting the “Classic Rock, C93” ing programmes from studios in Christchurch city, and programme from studios in Kilmore Street, Christ- network stations relaying programmes from elsewhere. church. This is an off-air re-broadcast of the “C93- Radio transmitter sites are located on the fringes of the FM” transmission. Christchurch metropolitan area as dictated by technical requirements. Most of the FM sound radio transmit- • 1413 kHz - 3XP, 100 W transmitter at the Ferrymead ter stations are co-sited at the Sugarloaf television Historic Park, Bridle Path Road, Christchurch. transmitter site, located on the Port Hills. (Studio and transmitter co-located.) Electrical and Communications • 143

• 90.5 MHz, 1.5 kW transmitter at Sugarloaf, broad- programme circuits, others use UHF radio, only a few casting the “90.5 Tahu FM” programme from stu- have standby circuits. dios at the Nga Hau E Wha National Marae in Pages Road, Christchurch. This is part of the Iwi network. Transmitters — the locations, structures, and equipment are all solidly constructed with good foundations, • 91.3 MHz - 3ZM, 50 kW transmitter at Sugarloaf, and are considered to be capable of withstanding broadcasting the “91ZM” programme from studios moderate earthquakes and other hazards without inter- in Kent House, Christchurch. ruption to service.

• 92.1 MHz, 25 kW transmitter at Sugarloaf, broad- Antennas — at all AM sites, there is no alternative to casting the “92 More FM” programme from studios the designated antenna system. At Sugarloaf, and at in Victoria Street, Christchurch. Ouruhia, the FM antenna systems are constructed in two identical units, so that if one fails, the other can • 92.9 MHz, 25 kW transmitter at Sugarloaf, broad- carry programmes at reduced efficiency. casting the “Classic Rock, C93” programme from studios in Kilmore Street, Christchurch. Television Broadcasting • 93.7 MHz, 25 kW transmitter at Sugarloaf, broadcasting the “Radio Pacific, (T.A.B.)” programme Television New Zealand from studios in Ponsonby, Auckland. This is a Television New Zealand operates two television broad- network programme. Some local programming is cast stations in Christchurch, transmitters are fed from included during the day from studios in Manchester studios in Auckland. These are: Street, Christchurch. • TV-1, Channel 3, 50 kW transmitter at Sugarloaf, • 94.5 MHz, 40 kW transmitter at Ouruhia, broad- broadcasting the “TV One” programme from stu- casting the “ i 94.5 Easy Listening” programme dios in Auckland. This is a network programme. from studios in Kilmore Street, Christchurch. • TV-2, Channel 8, 300 kW transmitter at Sugarloaf, • 96.1 MHz, 600 W transmitter at Huntsbury Hill, broadcasting the “TV 2” programme from studios broadcasting Polytech Broadcasting Students pro- in Auckland. This is a network programme. grammes from studios in Madras Street, Christ- church. Transmission times are irregular to suit Private Television course timetables. In addition to the Television New Zealand stations, • 96.9 MHz, 3.5 kW transmitter at Sugarloaf, broad- there are eight private television broadcast stations in casting the (Access Radio) “Plains FM” programme Christchurch, with transmitters at Sugarloaf (or at from studios in Madras Street, Christchurch. Marleys Hill). These are:

• 97.7 MHz, 40 kW transmitter at Ouruhia, broad- • CTV, Channel 48, 30 kW transmitter at Sugarloaf, casting the “Classic Hits” programme from studios broadcasting the “Horizon TV” programme from in Kent House, Christchurch. studios in Christchurch.

• 98.3 MHz, 100 W transmitter at Ilam Road, broad- • TV-3, Channel 6, 300 kW transmitter at Sugarloaf, casting the (Students Association Radio) “98 broadcasting the “TV 3” programme from studios R.D.U.” programme from studios in Ilam, Christ- in Auckland. This is a network programme. church. (Studio and transmitter co-located.) • Sky TV, Channel 30, 160 kW transmitter at • 99.3 MHz, 5 kW transmitter at Winters Road, Sugarloaf, broadcasting the “Sky Movies TV” pro- broadcasting the “Life FM” programme from stu- gramme from studios in Auckland. This is a dios in Birmingham Drive, Christchurch. network programme.

• Sky TV, Channel 34, 160 kW transmitter at Sound Radio Broadcasting Backup Sugarloaf, broadcasting the “Sky Sport TV” pro- Facilities gramme from studios in Auckland. This is a Power Supplies — all Radio New Zealand, and all network programme. Radio Network, and most of the other studios and transmitters are equipped with standby diesel genera- • Sky TV, Channel 46, 160 kW transmitter at tors. Sugarloaf, broadcasting the “Sky News TV” programme from studios in Auckland. This is a Programme Links — some stations use Telecom network programme. 144 • Risks and Realities

• Sky TV, Channel 54, 160 kW transmitter at have access to portable linking equipment, for use if the Sugarloaf, broadcasting the “Sky Four TV” pro- main programme bearers are damaged. gramme from studios in Auckland. This is a network programme. The “Big One” — Earthquake • Sky TV, Channel 58, 160 kW transmitter at With an earthquake of this type, it is expected that Sugarloaf, broadcasting the “Sky Five/Action TV” broadcasting stations will sustain the following dam- programme from studios in Auckland. This is a age: network programme. • structural damage to studios; • Cry TV, Channel 56, 0.9 kW transmitter at Marleys • damage to studio equipment; Hill, broadcasting the “Cry TV” programme from studios at Marleys Hill. (Studio and transmitter • loss of power supplies to the studios; together.) • loss of power supplies to the transmitters;

Television Broadcasting Backup • interruption of programme circuits between stu- Facilities dios and transmitters; Power Supplies — All Television New Zealand studios and transmitters, and most private studios and • interruption of programme circuits between studio transmitters are equipped with standby diesel genera- centres; tors. • structural damage to transmission buildings;

Programme Links — Stations use Telecom and/or • damage to transmission equipment; B.C.L. protected programme circuits. • damage to antenna systems; and Transmitters — The locations, structures, and equipment are all solidly constructed with good foundations, • damage to transposer/translator systems. and are considered to be capable of withstanding moderate earthquakes and other hazards without inter- Tsunami And Flooding ruption to service. Television House — in a flooding event, it is expected Antennas — At Sugarloaf, the TV antenna systems are that the only broadcasting building that has a basement constructed in two identical units, so that if one fails, and likely to be affected would be Television House the other can carry programmes at reduced efficiency. (TVNZ). Ouruhia transmission site — it is estimated that The Broadcasting System Vulnerability Radio Network transmission site at Ouruhia would have water lapping at the base insulator during a worst- Moderate Earthquake case Styx River flooding event. “Newstalk ZB” would It is expected that with an earthquake of this type and still be on the air at that level of water. magnitude, the only damage sustained by (sound radio and television) broadcasting stations would be: Many transmission sites could be affected by the loss of Telecom programme circuits. • the interruption of power supplies to studios; The Southshore Translator (and Transposer) site could • the interruption of power supplies to transmitters; suffer substantial damage from a tsunami. This site and serves Redcliffs and Sumner with:

• the interruption of programme circuits between • TV One Ch 5 and Ch 50 studios and transmitters. • TV2 Ch 1 All stations would be unaffected, for a limited time, by the loss of power supplies as the studios and all the • Concert FM 99.9 MHz transmitters are equipped with standby diesel genera- • More FM 95.3 MHz tors.

Sound Radio Broadcasting Stations, NewstalkZB (3ZB) Windstorm and National Radio (3YA) both have a standby (UHF) Sugarloaf — The antenna tower is designed to with- programme circuit, in addition to the normal Telecom stand a 60m/s windspeed at its top. programme circuit. Television Broadcasting Stations Electrical and Communications • 145

Ouruhia — It is “guestimated” that the antenna mast Specific mitigation measures identified during the will withstand a 70 m/s windspeed. This should be course of the project include: confirmed by an engineering review of the entire structure. • A review of the overall robustness of each network and the establishment of a plan to manage with a reduced system is required, (i.e. set up a disaster Landslide And Slope Instability response plan). This review should be ongoing and Sugarloaf — The antenna tower and the equipment repeated on an annual basis. building are solidly constructed with good foundations, the site is considered not to be susceptible to any • Detailed reviews of equipment, studios, buildings, landslide. and masts and towers should be undertaken to identify specific weak links, including an assess- Gebbies Pass — The antenna mast and the equipment ment of the cost-benefit of the resulting mitigation building are solidly constructed with good founda- measures. Normal office equipment and furniture tions, the site is considered not to be susceptible to any should not be overlooked. Particular attention to landslide. cable entry/exit points is required. These reviews should be repeated on an annual basis. All of the other Christchurch broadcasting buildings and facilities are located away from the hills, except for • Review the design standards for new and existing some minor translator and transposer installations. facilities.

• A review of Kent House in its role as a common Broadcasting System — Mitigation distribution point should be carried out. Plan to Measures increase diversity. Although floodwaters from the The above investigations show a consistently high Avon River and excess stormwater is not expected impact, particularly in the periods during and immedi- to rise above floor level, a review of the location of ately following most types of disaster. This points to PABX and Telecom equipment could be under- the need for critical examinations of the robustness of taken. all sections of the broadcasting services vital to recovery.

Slimline microwave tower Sugarloaf television transmitter 146 • Risks and Realities

• A review of the CTV building in it’s role as a the after-effects of a disaster. This training must be common distribution point for CTV, TVNZ, and ongoing. the Sky TV services should be carried out. Plan to increase diversity. A review of the heavy duty • A review of the Ouruhia and Gebbies Pass trans- pumping requirements or the relocation of PABX mission sites should be carried out. Acquire re- and other equipment within TV House should be sources and capabilities to reduce the time to re- carried out. instate antennas and transmission equipment.

• Investigate and/or improve adequacy, reliability • It is estimated that The Radio Network transmis- and robustness of standby generating plant. Atten- sion site at Ouruhia would still be on the air during tion to fuel supply lines is required. Also, ensure a worst-case Styx River flooding event, however, a that fuel supplies are adequate until access is re- means of keeping floodwater away from the mast stored. base insulator should be studied.

• Initiate and continue staff training in dealing with • An engineering assessment of the Ouruhia antenna structure to determine the actual failure windspeed should proceed.

• A review in conjunction with BCL of the Sugarloaf transmission site should be completed. Particular attention should be paid to the RF feeder routing.

• Translator and transposer systems are generally small installations and do not warrant extensive mitigation measures.

Telecom exchange

Power transformer hold-down (Trans Power)

Main Telecom building Transport • 147

Chapter 6 Transport

6.1 Introduction — the All emergency services and immediate resources are located in the urban area mostly close to or in the city Christchurch Transport System centre. The city centre itself contains the principal Figure 6.1 shows the Christchurch urban area and the operational and control centres for fire, ambulance, main physical elements of the city’s transport system. Police and Civil Defence, the principal hospital, and for other services such as power, communications and This system is made up of: water.

• a road network; The external transport linkages are provided by road, air, sea and rail services. These are needed both by • a rail network; Christchurch and, because it is the South Island’s • the Port of Lyttelton; and largest centre, by much of the island.

• the International Airport and Wigram Aerodrome. Road and rail Other components such as vehicles, fuel, control and Both systems beyond Christchurch are, for much of personnel are part of the transport system but their their lengths, close together, particularly where they vulnerability and possible mitigation measures are cross major rivers or topographical features. Whereas beyond this report’s scope. this could allow alternative or transfer facilities it is also likely that both systems would be disrupted at the Except for the Port Hills, the urban area is flat, with the same place. All these long-distance lifelines are ‘long Avon and Heathcote Rivers passing through the city. and thin’ and thus vulnerable to a break at any point The coast forms the eastern edge with the Port Hills on which renders the whole route unserviceable. Few the southern edge between the city and the Port of alternative road routes are available and they usually Lyttelton, the Waimakariri River system the northern, cross the same rivers or physical features as the ‘nor- and the Canterbury plains the western. mal’ route. The rivers are crossed by some forty bridges and the Port Hills by a road tunnel, a rail tunnel and three road Road passes. These crossings, particularly the bridges, have Roads link to the north principally to the Waimakariri been identified as the most vulnerable parts of the District, a significant urban area in its own right (and transport system which otherwise has many alternative just on the edge of the Christchurch urban area, but routes available throughout the dense road network. outside the area considered in this project), and to the North Island via the Cook Strait Ferries. The northern Beyond Christchurch, road and rail cross major rivers part of the West Coast, and Nelson are also served by and pass through rugged alpine or coastal terrain which the northern road links. Thus, beyond Rangiora the can, in an emergency make maintenance of services resources and centres of population are remote from difficult and full reinstatement time consuming. Christchurch.

The Main South Road connects Christchurch to the 6.2 Strategic Role of Transport urban areas of Timaru, Dunedin and Invercargill and System in an Emergency their ports, populations and resources, the former a half-a-day’s travel, the latter two a full day. The internal road system of Christchurch serves its resident and visiting population of about 300,000 peo- The roads to Akaroa and Greymouth are of lesser ple and is important to another 50,000 people nearby importance. and generally to the north. This system is thus the most important transport lifeline both in the period immedi- Rail ately following an event and in the reinstatement of The three trunk rail lines connect Christchurch to normality. It is described in more detail in Section 6.3. Picton, Dunedin and Greymouth, and these, in turn, to

148 • Risks and Realities

Rail Road

kilometers

13 SUMNER

v E

iver

n R y

Avo a

ESTUARY Transport System Transport

d PORT

d a a

a o

o o

R R

Heathcote River Heathcote

R s

e s

g e

a D P

Brighton n

s v e

LYTTELTON

v l

r A

d HEATHCOTE Christchurch Engineering Lifelines Study 1994 a o R rt

Road

e d

o IL N o

n i H L

r Road Woodham Marshland r e

Fitzgerald Ave

RWAY

OTO

M d Ave

d d

o oa

a R R

Street o Pass

R et

d t

s S Colombo St o

t rs

n e T

i y h t

r rd D

o Bealey

W o s t N f R

e e

n e

ra n tr

C n I S O

Ro P

Moorhouse Ave

M n

Brougham to

n E PARK i

ui rr

an a

p HAGLEY B

I Pa n

o t R l

BELFAST a d

d d

n a

e o

F R

Northcote Road a o

e r

e m i

u o

n t

e h

v a n

c s A

h R

a HALSWELL r

G ad aimairi Ro d W d R

w d

e i a

e o

a k R

H a

r i l a ia r W o

m d

e I a

M o

R n

R io I t c

WIGRAM u

R J

A AERODROME

K A ll e

M ls

I a

H A

r d W a

u o

R AIRPORT h

d l

i r

YALDHURST a

PREBBLETON

TEMPLETON

R R

C u

i e

W M

Figure 6.1: Christchurch transport system Transport • 149

Lyttelton. They mainly handle freight, with a few with blocks about 250 x 125 metres, surrounded by intercity passenger services. There are no suburban four dual carriageways (Deans, Moorhouse, Fitzgerald passenger services. and Bealey Avenues), which distribute traffic around the city centre and between the principal suburban Sea roads. Madras and Barbadoes Streets act as the main north-south distributor from Bealey Avenue to Lyttelton is close to Christchurch and is an important Brougham Street and act in effect as a dual six-lane lifeline for heavy equipment, food and bulk supplies. It road. Brougham Street and its eastward extension to is within one days sailing of Timaru and Wellington, Lyttelton and westward to Sockburn is the main East- and within a few days sailing of all of New Zealand’s West industrial distributor with many industrial land ports (and some overseas ports). Timaru is also linked uses along or near it. to Christchurch by road and rail within a few hours travel; other ports such as Nelson, Picton and Dunedin In the city commercial centre the roadways are gener- are more distant and their road and rail connectors ally 14 to 16 metres wide within a 20-metre corridor could also be disrupted. between the building facades. Most buildings are low rise (2 to 4 stories), including a considerable number of Air vulnerable brick buildings. Outside this, most build- The international airport at Harewood is very close to ings are one or two storey houses and are generally set the city and is connected to it by a dense and robust well back from the carriageway. multi-route road network. It has large freight, passen- The central city street system provides the right-of- ger and military operations capacity for national and way for all main utility services, many of which are international operations. It is of importance to the carried on the thirteen road bridges which cross the whole South Island if a major event occurs in any of its Avon River. Typically these bridges have a two-metre urban areas, and possibly important to Wellington. headroom above normal water levels, are 20 metres Wigram aerodrome has had considerable, particularly wide with a similar clear span. The river is generally military, operational capacity, but its future is still less than 1 metre deep and has a hard bottom over its uncertain. Other airports at Timaru, Dunedin, central city length. Greymouth, Nelson and Queenstown are of limited The central street system is controlled by a centralised capacity and are remote from Christchurch. traffic signal control system which also serves some suburban arterials. Whereas disruption to this could Transport interchange make traffic movement difficult, it is flexible enough All systems have good interchange facilities with road. to control most movement, particularly within the The Port of Lyttelton has sea/road/rail facilities as does commercial centre, on an emergency or contingency Timaru. The airports are served only by road. Distant basis. The system is controlled by computers in the ports and airports are dependent on road or road/rail City Council’s Tuam Street building by a series of in- links to connect them to Christchurch if sea or air links ground cables. are unavailable. If these computers go down, individual controllers at each set of signals will allow them to operate independently of each other. Power loss to the signals them- 6.3 Description of the selves would stop them operating — there are no Transport System backup power facilities. Associated with this traffic control system is a series of traffic monitoring cameras, The road system mounted above key intersections. They are able to revolve, zoom in/out and could provide very useful The Christchurch road network can be considered as data on the condition of roads following a hazard event. comprising three main systems, the city centre streets, The monitors are located with the traffic control com- the suburban roads and the external highways. They puters. operate together as one hierarchical network and whether or not a road is a State Highway, City Council or other council’s road is immaterial to the transport Suburban road system needs of the area. The principal radial routes from the city are connected to the four avenues and lead out into the suburbs. Like City centre streets the city centre, a hierarchy of roads exists with arterial and distributor roads carrying larger volumes of traffic The city centre streets are based on a rectangular grid between residential suburbs, district centres, schools, 150 • Risks and Realities

factories and other concentrations. A dense network of General — road network local roads serve needs within these concentrations. The network is operated as one hierarchical system All roads are, however, interconnected giving a very which includes State Highways, City, Banks Penin- wide range of alternative route options. sula, Selwyn and Waimakariri Council roads. The road system is generally built to a high standard and the As with the central city roads most of these provide main elements are suitable for heavy vehicles and rights-of-way for a range of services. Buildings are heavy volumes of traffic. The central streets typically generally one-storey residences well set back from the carry up to 10,000 vehicles per day with some major road. There are some district and local shopping centres where two-storey buildings abut the road. In arterials carrying 20,000 to 30,000+ vehicles per day Sydenham, many of these buildings are built of brick. both in the centre and on the radials and distributors. Many suburban distributor roads carry over 5,000 Important bridges occur on the lower Avon and vehicles per day, whereas of the rural roads remote Heathcote Rivers, the latter providing important links from the city boundary, only Main South Road and to Lyttelton. There are several rail and road overbridges Main North Road carry flows greater than this. and the Lyttelton Road tunnel on major suburban roads. Major high tension electricity cables cross The rural highways (on the outskirts of the city) gener- several major roads and overhead services are wide- ally carry lesser volumes but at higher speeds, and with spread. a high proportion of heavy vehicles. Russley Road at its southern end carries over 13,000 vehicles per day, There are three routes across the Port Hills. Evans, about twice the volume at its northern end (Johns Dyers and Gebbies Passes are linked, at high level, by Road). the Summit Road and at sea level on the Lyttelton side. Mount Pleasant Road also links to the Summit Road. The vehicle fleet is modern with a significant number These routes are all hilly and/or winding and only the of private four-wheel drive cars and recreation vehicles Evans Pass Road through Sumner is suitable for regu- available. A large coach and bus fleet operates from lar use by heavy vehicles, and is so used by vehicles too the city as do most heavy commercial vehicles. Petrol large or otherwise undesirable for the road tunnel. service stations are widely dispersed. Most vehicles are petrol or diesel powered — LPG vehicles are relatively few and many are dual-fuelled. Road repair External road links and maintenance materials are available close to Christ- Beyond the main urban area there are two main corri- church. Emergency service and contractors yards are dors and two lesser ones. To the north, Main North shown on Map 1 page 284 ; these are all served by road. Road and State Highway One link Christchurch to Rangiora, Kaiapoi and Woodend and further north to Picton. To the south, Main South Road (State Highway Rail system description One) connects Christchurch to Ashburton, Timaru and The Railway system serving Christchurch comprises the south. three major lines — Main North Line, Main South Line and Midland Line, and the line to Lyttelton. The other main roads are Yaldhurst Road and State Highway 7 to the West Coast, and State Highway 75 The Main North Line enters Christchurch from the and Halswell Road to Akaroa. These main roads north down the east coast, serving the major freight generally pass through rural areas with few close flows to and from the North Island (via the inter-island buildings. ferries) and the Nelson/Blenheim area and minor intercity passenger flows. The Main South Line enters Their ability to link Christchurch to the outside is Christchurch from the south, also along the east coast, limited by major river crossings and alpine terrain. The serving the freight flows to and from Timaru, Dunedin availability of each route may depend on a river cross- and Invercargill. It also has a minor intercity passenger ing or route characteristic some considerable distance flow. The third major line, the Midland Line, services from Christchurch. For example, the Main South Road the South Island’s West Coast, both in a freight capac- crosses the Rakaia River on a 1.8 km long bridge, the ity and as an increasingly important tourist passenger Main North Road traverses the slip-prone Kaikoura line. It crosses the alps and links with the Main South coastal strip and shares the Seddon Bridge with the Line at Rolleston. railway, the West Coast Road crosses the Alps via Porters and Arthurs Passes and the Otira Gorge. The The Lyttelton line (actually designated as part of the Waimakariri bridges are on the edge of the main Main South Line) connects Christchurch City with the Christchurch area, and separate it from Kaiapoi, Port of Lyttelton and serves the import/export freight Rangiora and Woodend. flows through the Port. A short spur line (of 4.5 km) Transport • 151

running from the Main South Line at Hornby serves as of the Bealey River. It then runs up the Bealey Valley a feeder line for the industrial area in and around to Arthurs Pass, and, via the 5.5 km long Otira Tunnel, Hornby. The terminus of all freight traffic can be taken to the West Coast. as Middleton Shunting Yards, and for passengers the new Christchurch Station, at Addington. Lyttelton Line The section of line between Christchurch and Lyttelton Main North Line commences at Middleton. It runs through the suburbs The single track Main North Line runs from Middleton of Addington, Linwood, Woolston and Heathcote, Marshalling Yard past the new Christchurch passenger before entering the single track 2.6 km long Rail terminal and through the northern suburbs of Riccarton, Tunnel, to the Lyttelton Freight Yard and Port facili- Fendalton, Papanui and Belfast. ties. The length of this line is 14 km and is double track between Middleton and Heathcote. It then crosses the Waimakariri River, on a steel span concrete piered bridge, 610 metres long, downstream The only other major structural features of this line are of the State Highway 1 Bridge and immediately adja- the twin bridges spanning the Heathcote River (steel cent to the Main North Road Bridge. It passes through spans on timber piers with a length of 45 metres), the Kaiapoi and, north of Rangiora, crosses the Ashley Waltham Road overbridge and the tunnel approach River on a 550-metre concrete span bridge with con- embankment and associated structures (including the crete piers. In general, the topography of the land over Martindale Road brick arch underbridge) on the Christ- which the line runs is flat (Waimakariri flood plain or church approach. The tunnel itself was constructed in Marshlands peat marsh) the only significant embank- the mid-1860s and has brick, granite block or natural ment being both approaches to the Waimakariri River rock lining throughout its length. The portal at the Bridge. Lyttelton end is spanned by a complex two-level road bridge. Two 200 mm cast iron water mains, and Main South Line Telecom New Zealand cables run through the tunnel. This line leaves Middleton Marshalling Yard and runs south through the suburbs of Sockburn, Hornby, Is- Railway stations and terminals lington and Templeton running underneath two road The only use current use of the original Moorhouse overbridges and, in general, runs relatively parallel Avenue railway station by Tranz Rail Limited is to with State Highway 1 from Templeton south. It is house train control operations, the railway telephone double track to Islington and single from there south. It exchange, and the signalling relay room (for all signals crosses the Selwyn River on a steel span concrete pier between Wilsons Road and Montreal Street). Train structure, 300 metres long. The line continues parallel control monitors the running of all trains in the South to State Highway 1 and, 60 km south of Christchurch, Island north of Oamaru, while the telephone exchange crosses the Rakaia River, the bridge being a steel span/ covers the Christchurch area, but is linked to all other concrete pier structure, 1750 metres long. The topog- main centres through the Tranz Rail Limited system. raphy along this line is flat over the Waimakariri/ In the event of a mains power failure these facilities are Selwyn/ Rakaia River flood plains with no major powered by a 3.3 kV stand-by diesel generator, located earthwork features. alongside the track just east of Colombo Street.

All passenger services are now handled at the passen- Midland Line ger terminal at Addington, which is located on the This single track line commences at Rolleston and south-east side of the Blenheim Road overbridge on travels across the Canterbury Plains, passing through the old railway workshop site. This is a new facility 48 kms of gently rising alluvial plains with no major constructed in 1993. earthwork or structural features. It passes through Darfield, and Springfield before climbing into the The main freight handling terminal is in Waltham Southern Alps. Yard, between Waltham Road and Wilsons Road. This area also houses the main wagon repair facility for the From Springfield to Arthurs Pass there are sixteen Christchurch region. On the east side of Wilsons Road tunnels with a total length of 4.2 km, four high viaducts is the locomotive stabling, fuelling and servicing depot (steel lattice spans on steel towers) and three standard at Linwood, which is in the process of being relocated bridges in excess of 100 metres length. This section of to the Middleton Marshalling Yard. line generally follows the course of the Waimakariri River some height above it and in rock or earth cuttings The Middleton Marshalling Yard is the main train for much of the distance until it reaches the confluence make-up and break-up facility for the Christchurch 152 • Risks and Realities

Area and is located between Matipo Street and Annex Communications Road on the north side of the Main Line. The communication centre for the port is the Signal Tower which is at the top level of the Lyttelton Con- Port of Lyttelton — description tainer Terminal Administration Building, adjacent to The Port of Lyttelton has a major role servicing South the Container Terminal entrance off Gladstone Quay. Island imports and exports. It is also important for the A backup communication centre is available in the movement of coastal cargo and shipping, and in 1993 Lyttelton Port Company Administration Building situ- handled about 3.5 million tonnes of cargo. The port can ated on the corner of Norwich Quay and Dublin Street. accommodate ships up to 12.3 metres draught.

There are four important areas: Container cranes The port has two container cranes, a Paceco (maximum • the navigation channel; lift 46 tonne) and a Liebherr (maximum lift 70 tonne). • Cashin Quay; Both are designed for combined wind and earthquake loading. Both cranes have storm pins which stop them • the inner harbour; and moving along the wharf in strong winds. They are at their least stable with the boom in the parked up • the Naval Point reclamation. position.

These and other features of the port are shown in Figure The Liebherr crane also has storm tie downs on the 6.2. seaward side to prevent the possibility of it’s bogies lifting under the combined effects of wind and earth- Navigation channel quake. The stability of the Paceco crane is such that The navigation channel is approximately 7 km in storm tie down is not required. In the event of both length and 180 metres wide excluding the batter slopes. container cranes becoming inoperable, there are some The bottom is soft and requires regular maintenance large mobile cranes that could unload cellular con- dredging. tainer ships. Alternatively ships carrying their own container cranes could be used.

Cashin Quay Coastal cargo to Lyttelton is shipped in roll on/roll off The Cashin Quay wharves provide 855 metres berthage vessels (e.g. Pacifica) allowing trucks to be loaded length and are protected by a 180 metre long breakwa- with supplies and equipment which can simply be ter. The breastwork is supported by timber, steel and driven to where they are needed. prestressed concrete piles. The storage area behind the breastwork has been reclaimed using quarried rock and Transport links is faced with a concrete retaining wall. Several jetties and breastworks have capacity for heavy equipment and cargoes. Heavy equipment can be The area includes a container terminal operating with transported by quarter-ramp, stern ramp or lift-on lift- two container cranes, a bulk export shiploader for coal off ships, or as deck cargo on container ships. There is and other products and facilities for general cargo. a heavy-duty ramp at Gladstone Pier which is still Road and rail transfer is provided in the area. operable, but most roll on/roll off ships in service have stern or quarter ramps. Inner harbour The area provides several breastwork and jetty berths No 7 Jetty is an important link for Pacifica ships which in a more sheltered environment. A bulk liquids berth carry their own stern ramps for loading/unloading cargo. Alternatively Pacifica ships can use No. 4 Jetty. is available for ships carrying petroleum products and No. 2 Jetty is another important inner harbour cargo adjacent to this is a dry dock. The bases of the jetties handling berth. Inner harbour heavy duty pads are are founded on bedrock. available at No. 7 Jetty, No.1 Breastwork and Gladstone Pier. The Cashin Quay berths are also heavy duty. The Naval Point reclamation Oil Wharf is located in the inner harbour. Several oil tanks and pipelines are sited on the land adjacent to recreational facilities. The land was formed Christchurch International Airport — from low-strength pumped dredgings from the Lyttel- description ton Harbour seabed. Christchurch International Airport is at Harewood, 11 km north-west of the city centre. The airport is the

Transport • 153

(coal)

Bulk Materials Stockpile Area

Te Awaparahi Awaparahi Bay Te

s a

Workshop

o o

i R

Vehicle

a 500 r

e n

m u

400 S

CQ1

300

Bulk

materials

ship loader

Scale in Metres

CQ2

u n

100 200

s i

h a

C a

CQ3

crane

container

Signal Tower 46 tonne pace

Quay

ne CQ4

to crane

Lyttelton

s 70 tonne

container overhead Police Station t

e d

r r

t a

Medical Centre l

S e

r G P o f Container Terminal

y t

a g

Fire Station n y

o e

h J

c 2

n i o

N y

o t

w t

r e

L J

o 3

t h t

N N

e t

e R J

e r

Lyttelton Port Lyttelton

S e

U n o

l B b N

Company Building u

D Z

y B t t

J R

5 A

y t N t

e H

N R

t E

t e

J N 1

o Point Naval

N IN Rail Tunnel

Q 12 km

n r

o o Oil Wharf To Christchurch To

e M

Road Tunnel

m h

i c y Harbour Lyttelton a

y a

Y S Cattle Jetty

a u

u Q

Q s Dry Dock m ia ll

Area

s r a

u u e r

y Q Fuel Storage T

m C

e e

d n

a t G

S r J

B Coastguard

o l

Reserve

Recreation

c P

y T

e r

Marina

Proposed

Magazine Bay k

a P

Figure 6.2: Port of Lyttelton 154 • Risks and Realities

South Island’s only international airport and is a major Subsidiary areas element of the domestic main trunk system. It is also To the west of the main runway and lying parallel to it the air servicing base for aircraft operations to the Ross is a grass strip used by light aircraft. It is 525 m long and Sea area of Antarctica. There are about 100,000 100 m wide. There are a series of apron areas lying to aircraft movements each year and 3.2 million passen- the east of the main runway and equipped with in- gers. A plan of the airport is shown in Figure 6.3. ground piped refuelling systems for jet aircraft. The fuel is pumped via welded steel pipes from a fuel farm Runways situated on the airport at Wairakei Road. The main 02-20 runway is 3,287 m long and 45 m wide The domestic and international aprons are used by all and is suitable for all aircraft currently flying. The commercial aircraft up to and including B747s and are subsidiary or nor-west 11-29 runway is 1,741 m long clustered around the terminal buildings. The Antarctic and 45 m wide. It is not suitable for B747 or C5A apron is used by a variety of military type aircraft up to aircraft but will accommodate all other aircraft regu- and including Galaxy aircraft engaged in Antarctic larly visiting Christchurch. Both runways have paral- support. Part of this apron is used for overnight freight lel taxiways which could be used as runways in emer- operations. To the south and north of the main terminal gency situations. area there are two smaller apron areas used for domestic and international air freight operations. The runways and associated taxiways consist of pavements of flexible type construction, i.e. an asphalt surface over a compacted gravel base. In general they Buildings consist of 850 mm of compacted gravel with a 50 mm The terminal building consists of a basement and three asphalt surface course. Pre-1960 runway construction upper floors with a control tower equivalent to a height records are incomplete and construction depths are not of five floors. The building is equipped with a stand-by known accurately for the oldest portion of the runway. power generation system. A regular maintenance programme is in place to pro- The principal aircraft maintenance buildings adjoin the tect the sub-grade from water intrusion. main terminal buildings, the Antarctic apron and near The airport is built on gravels laid down by the the northern end of Orchard Road. The Rescue Fire Waimakariri River during the extended erosion proc- Service is situated in a purpose-built building with a ess of the east side of the main divide and lies across radio equipped watchroom on an upper level. This several old flood channels which originate in the coun- watchroom could be used in emergency situations for try area on the south bank of that river some 20 km west on-airport control tower functions. of the airport. Utility services Navigational aids The airport is fully serviced with underground utilities The main 02-20 runway is equipped with an instrument as follows: landing system, the nor’west 11-29 runway is a non- • Water supply is from seven wells fitted with sub- instrument strip. mersible pumps feeding into a ring main system. A Descent guidance is provided by visual approach slope stand-by generator and pump system has been indicator systems on both runways. The runways are installed to maintain essential services. The ring delineated by edge lights and the taxiways by centreline main consists in general of 250 mm diameter asbes- lights. A stand-by power system owned and operated tos cement pipes. by Airways Corporation is available for all navigation • Electricity supply is provided by way of 11 kV and lighting systems in the aircraft operational areas. 33 kV underground cables from Grays-Avonhead Roads and 11 kV from Wairakei Road. Some The radio navigation aids at the airport include an overhead supply is fed on to the airport from two instrument landing system, a very high frequency positions on Russley Road. omni-directional radio range and distance measuring device, several non-directional beacons, and a primary • The airport sewerage system is owned by the Air- surveillance radar system. Standby power is available port Company and is connected to the city system for all on-airport navigation facilities. Where stand-by at the corner of Wairakei and Stanleys Roads. It power is not available, i.e. in some remote locations, consists of a series of concrete mains ranging from limited service is available by way of battery back-up 300 mm to 375 mm in diameter and was laid in 1964 power. to Christchurch Drainage Board specifications.

Transport • 155 20

AD RO

ROAD

Mtc.

Air NZ

Mtc.

Ansett

Fuel

Farm Mtc. JESSONS CIAL

WAIRAKEI RD

RUNWAY

RUSSLEY ORCHARD RD

ROAD

Antarctic Apron

MEMORIAL AVENUE MEMORIAL

Mtc.

Mt. Cook

Terminal

Mtc.

Air NZ

International

POUND

City

Terminal Domestic

SUBSIDIARY RUNWAY 29

11 ROAD

AVONHEAD

ROAD

MAIN

Service

Airport Fire

500

ROAD

Metres

GRAYS

POUND 0 100 200 300 400

Figure 6.3: Christchurch International Airport

Note: This plan does not show the terminal extensions under construction in 1997 156 • Risks and Realities

• There is an extensive network of underground Reticulated services provided at the airport consist of cables servicing the telecommunication needs of a foul sewer which relies on all sewage being collected the airport community. This network is connected at a common point and then being disposed of to the by fibre-optic cable to the Memorial Avenue ex- city system via a rising main. Storm water drainage is change and a similar linkage exists between the provided by way of soakpits. Electricity supply is airport and Airways Corporation Control Centre in provided by way of 1100 volt ring main with a series of the technology park. strategically placed transformers. Water supply for both domestic and fire purposes is provided by five • Three oil companies supply fuel and lubricants to underground wells situated around the existing water aircraft at Christchurch International Airport. They tower. The water is reticulated by way of a ringmain operate from separate depots off Wairakei Road system. The system can be connected to the Christ- and Ivan Crescent. Total stocks of approximately church City supply. Telephone services are provided 4.2 million litres of fuel is held at the airport. Shell by way of an on-base telephone exchange and under- Oil has a pipeline and in-ground hydrant system ground reticulation. servicing the international and domestic passenger aprons, and Mobil Oil has a similar one that serv- The future of the aerodrome is uncertain. ices the Antarctic apron. BP Oil uses the Shell system to service the international apron as required. All companies use tanker trucks on the 6.4 Vulnerability — Earthquake apron where necessary. The fuel depot is resupplied by fleets of tanker trucks that haul fuel from Lyttel- The vulnerability of the transport system has been ton and/or Woolston via Port Hills Road, Brougham assessed for three stages following a major earthquake: Express way, Curletts Road, Yaldhurst Road, 1 Immediately following (what will happen) Russley Road and Wairakei Road. At normal usage rates, reserves of fuel range from seven to fourteen Period: hours days. The oil companies provide all of their own plant, equipment and buildings on land leased from 2 Shortly after (how can the city be put back into the Airport Company. operation) Period: days Wigram Aerodrome Wigram aerodrome is a former Royal New Zealand Air 3 Full return to normality Force facility approximately 8 km west-southwest of Period: weeks or months. Christchurch city centre, which is now closed.

It is an air base with runways (one sealed), taxiways, Road network vulnerability hangars and support areas including training buildings, The road network has two basic components, the roads workshops and residential accommodation. It has its themselves, and the bridges which carry them. own reticulated system for power, water supply and sewage collection and disposal. Immediately following a major earthquake The main hard surfaced runway 03-21 is 1580 m long and 30 m wide. It is capable of handling relatively Road system damage types heavy aircraft such as the P3 Orion and the C130 Roads may be severed by damaged or collapsed bridges. Hercules. The length of 1580 metres would create limitations for its use by some civilian air craft. There There may be localised damage on many streets caused are a series of secondary grass runways (vectors, 18-36 by structural failure, collapse of underground and and 24-90) located to avoid cross wind operations. overhead services, damage from water mains or other flooding, landslides in hillside suburbs or riverbank The airfield is built on gravels laid down by the and swampland slumping. Debris from buildings, Waimakariri River and has one flood channel which poles, bridges, other structures, overhead wires and crosses the airfield in a north-west/south-east direc- abandoned or crashed vehicles may also disrupt roads. tion. Apron areas provide hard standing facilities adja- Kerbs and channels and road surfaces would be dis- cent to the hangar buildings. torted and broken.

Fuel storage facilities are provided in underground Generally, alternative routes would be available nearby storage tanks for both aviation fuels and motor trans- possibly on a restricted speed, width or loading basis. port. Hillside, riverbank and bridge damage may take con- Transport • 157

siderable time to reinstate whereas roads subject to are potentially susceptible to dislocation damage in an liquefaction and other localised damage will be repair- earthquake event. able, albeit temporarily, relatively quickly.

Damage to drainage, sewerage and high-pressure wa- Road network vulnerability ter systems will cause extensive damage to roads. Analysis of the network on a sector and primary route Flooding or intense rain following an earthquake would basis provides a basis for assessing overall network exacerbate the situation. vulnerability, operational capability following an event, and priority for reinstatement and mitigation meas- Damage to telephone and power cables and the exca- ures. vation to allow repairs will also cause damage. Figure 6.5 (page 159) shows the sectors, subsectors, The network’s ability to function will depend on the primary routes and internal sector distributors used for magnitude and extent of the above damage and disrup- the analysis. All points in the city centre are linked by tion and is thus difficult to predict. the city street grid and this gives access to the four There are, however, streets in the city centre which are avenues which distribute traffic around the city centre expected to be disrupted by debris from collapsed and to the suburbs. buildings (particularly brick) following the earthquake Five primary routes linked by the four avenues, serve incident. five primary destinations and the suburbs through Three levels of disruption are shown in Figure 6.4: which they pass:

• roads which can be cleared by a grader within a few • the airport; hours; • the north;

• those which require a bulldozer or similar machine • the eastern suburbs; to create a one-way street within two days; and • the port; and • those where collapse of steel or reinforced concrete buildings would require lifting equipment to create • the south. a one way street within five days. Six sectors are bounded by the primary routes and the There is little documented information as to the effects four avenues, and there are four subsectors as follows: of earthquakes on trees. When trees are brought down by earthquakes it is mostly by way of landslides, earth • in the Southern Sector slumping, etc. There is the question of how trees might — the Hillside suburbs from Heathcote to react to the phenomenon of liquefaction (sand or swamp Westmoreland. ground conditions). If the whole root zone became a morass and the tree began to sway with the movement — the Harbour basin. of the earthquake, then uprooting would be most probable. Trees in the north and eastern parts of the city • in the Eastern and North Eastern Sector would be most at risk. In sand and shingle soils with — east of Ferrymead bridge (a mixture of low shear strength the swaying motion of the tree may coastal and hillside suburbs). cause roots to lift through the loosened soil and the tree to fall. In soils with a high shear strength, roots may — ‘Brighton’ between Travis Swamp, the hold firm but the swaying motion of a tall slender sea and estuary and the Avon River. trunked tree may cause it to break at some point above a quarter of the height of the tree. Plantation and These subsectors have specific features and bounda- shelterbelt conifers would be most at risk. There is ries different to the main sectors which are generally evidence that trees can die following earthquakes due flat and internally robust, with many road alternative to direct root damage and disruption of the soil in the routes available. root zone. Obviously any tree which had its root anchorage significantly weakened by an earthquake Bridge vulnerability would also be more susceptible to windthrow for some Earthquake damage to bridges will vary greatly. Com- years following the event. mon damage to bridges designed within recent years Traffic control computers and monitoring equipment will be controlled yielding of columns and joints.

158 • Risks and Realities

Building debris zone Level 1 Useable (minor damage) Building debris zone Level 2 Building debris zone Level 3 Useability restricted (moderate damage) Severe damage, likely to be unuseable Bridge number Hospital Civil Defence Sector HQ City Council Service Centre Fire station Police Contractors yard

P F C

R122

NOTE: LEVEL 1 LEVEL Brick debris may lie on the street but can be cleared within 6 hours of earthquake LEVEL 2 LEVEL Bulldozer or equivalent required to create a one-way street within 48 hours of earthquake LEVEL 3 LEVEL Collapse of reinforced concrete or steel buildings. Additional lifting equipment needed to create a one-way street within 5 days of earthquake

GROVE

ESSEX

HEYWOOD

ALEXANDRA LONDON ST

R109

ELM

LIVINGSTON MOORHOUSE AVE

R FITZGERALD AVE VE FITZGERALD AVE FALSGRAVE RI N O V A

DORE ST E REES

T BANGOR ST E

CHURCHILL ST G

ALFRED ST C

D C

I T

R N

B CLARKSON E

R S M

O R A

E F C

D N X

A O

KILMORE ST DAWSON ST R

Y ST ASAPH ST

TUAM ST ARMAGH ST R CASHEL ST CASHEL

E HEREFORD ST

WILLOW ST

WORCESTER ST F

GLOUCESTER ST

R701

HURLEY ST HURLEY

CHESTER ST EAST LICHFIELD ST

R110 NOVA PL NOVA

BARBADOES ST BARBADOES ST BARBADOES ST

WILLIAMS COVENTRY

OTLEY ST OTLEY

MELROSE ST ELY ST ELY

LATIMER SQ

R112

MOA PL MOA R111 MADRAS ST MADRAS ST

LATIMER SQ

ASH ST

A POPULAR

ALLEN ST

SOUTHWARK ST SOUTHWARK

LIVERPOOL BEDFORD ROW

ABERDEEN STREET

CARLYLE ST CARLYLE

R113

EATON ST EATON MORTIMER PL MANCHESTER ST

T S

E H NEW REGENT G

C I

CAMBRIDGE TCE

T H

D R CD

O H F

CIVIC

DUNDAS ST

WOMEN'S

HOSPITAL WELLES ST

OFFICES

MOORHOUSE AVE

R703

CHRISTCHURCH

R702 BEALEY AVE BEALEY

COLOMBO ST COLOMBO ST

SQUARE

CATHEDRAL

R114

R115

BATH ST BATH

R116

PETERBOROUGH ST

LICHFIELD ST

GRACEFIELD AVE WINCHCOMBE

OXFORD TCE ALCESTER

DURHAM ST NORTH I DURHAM ST SOUTH

R118 CIVIL DEFENCE CIVIL

HEADQUARTERS CAMBRIDGE TCE

R119

R704 P CD

ARMAGH ST ORBELL

KILMORE ST

TUAM ST

CHESTER ST WEST

ST ASAPH ST

SALISBURY ST SALISBURY WALKER ST WALKER

PEACOCK ST

CONFERENCE ST BEVERIDGE ST

CRANMER SQ GLOUCESTER ST

HEREFORD ST

WORCESTER ST

CASHEL ST CASHEL WILMER ST TELECOM MONTREAL ST MONTREAL ST

CRANMER SQ CAMBRIDGE TCE E C

T ACTON ST ACTON

D ST DAVID ST ST DAVID R120 R O F X

O HALKETT

ROLLESTON AVE ER RIV ON

R E R

I IV R R

R122 N

N O

O V

HORATIO ST HORATIO

A BALFOUR TCE

ST ASAPH ST

A R123

N MOORHOUSE AVE

STEWART ST

CHRISTCHURCH HOSPITAL

Figure 6.4: Central city building debris zones (immediately following a major earthquake)

Transport • 159

kilometers

Major Sector Sub Sector Route Importance KEY Distributor Boundaries Primary Route

SUMNER

13 R

s a

SUB SECTOR s

E HARBOUR BASIN

iver y

n R COASTAL

Avo a

w FERRYMEAD

SUB SECTOR

ESTUARY

d d a

a a o

o o

R R

Heathcote River Heathcote

s e r

a D

SECTOR EASTERN

COASTAL u

BRIGHTON

Brighton n

s v e

LYTTELTON

v l

a i

SUB SECTOR

r A

4 d HEATHCOTE a o R rt

Road

L e d

o I N o

n i H L

r Road Woodham Marshland r e

Fitzgerald Ave HILLSIDE

SUB SECTOR RWAY SECTOR

OTO

M d Ave

d d

o oa

a R R

Street o Pass

R et

d t

s S Colombo St o

r R

NORTH EASTERN NORTH

t rs

n e T

i y h t

r rd D

o Bealey

W o s t N f R

e e

n e

a n

r 5 tr

C n I d S O

R SECTOR P

CENTRAL

3 n i

Moorhouse Ave

M n

R Brougham to

5 5 n

E i i u rr SECTOR

an a V p B

I a

P n SOUTHERN

o t R l

BELFAST a d

d d

n a

e o

F R

Northcote Road a o

e r

e m i

u o

n t

SECTOR e

v a n

c s A

5 o

h R

a HALSWELL r

G ad aimairi Ro

d W

d NORTH WESTERN NORTH R

w d

e i a

e o

a k R

H a

r i l a ia r W o

m d

e I a

M o

R n

R io I t c n

u WIGRAM

R J 3

SECTOR

WESTERN

K A ll e

M ls

I a

H A

r d W a

u o

R AIRPORT h

d l

i r

YALDHURST a

PREBBLETON

TEMPLETON

R R

C u

i e

W M

Figure 6.5: Christchurch transport, sectors, subsectors and primary routes 160 • Risks and Realities

Although repairs will be needed the bridges should traffic (i.e. normal highway traffic) and speed remain useable by vehicles. restricted (i.e. 30 km/hr). For severe damage, or uncertain structural capacity, it may be neces- A number of road and rail overbridges built before the sary to limit the bridge to one heavy vehicle at mid-1960s are expected to sustain considerable dam- a time. This restriction would remain until age such as retaining wall and batter slope failure, more permanent repairs could be completed, yielding, rotation, lateral movement and settlement. one to six months. They are also likely to produce considerable debris and in some cases may collapse. Such damage may disrupt • Where some damage is identified, but the bridge the transport network although ground level or nearby remains useable, the probable restriction would be road diversions are available in most cases. They may, a speed limit of say 30 km/hr until repairs are however, require complete reinstatement. completed.

Most older river bridges are relatively robust, includ- Summary tables ing the central city Avon River bridges which are squat structures built into the river banks. The bridges over Four sets of tables (located at the end of the “Earth- the lower reaches of the Avon appear vulnerable being quakes” section) summarise information on the roading located in the liquefaction zone. system (Tables 6.1, 6.2, 6.3 and 6.4):

The external routes from the city pass over large • Primary route vulnerability (Table 6.1) page 167; bridges such as the Waimakariri and Rakaia bridges, • Internal sector vulnerability (Table 6.2) page 170; and their alternatives (the respective Gorge Bridges) are also large. Their vulnerability has been assessed on • Key bridges — primary routes (Table 6.3) page the same basis as the internal ones but require more 176; and detailed assessment, and extension to include distant bridges such as the Awatere bridge at Seddon. • Bridges summary — internal sectors (Table 6.4) page 179. The use of many bridges may be restricted (even though they may remain structurally sound) as the result of These tables include mitigation measures and priorities. damage to other services using the bridges or the approaches and the work of reinstating these services. Period following earthquake Immediately following an earthquake some bridges The general comments above describe the immediate would be subject to restricted loadings which would effects of the earthquake where establishing emer- need to be imposed for varying periods. In general: gency corridors for services, rescue and movement of people are the priority. Following this is a period when • Where it has been identified that bridge collapse is the city is getting back on its feet before the eventual likely, or that a structure is unusable, then all return to normality. vehicles would be prevented from using the bridge for an extended period (until major repair/rebuild- Activities required include: ing is completed), i.e. in the order of months (or for • opening up routes with priority decided by the a temporary bridge, e.g. a Bailey bridge, or prop- hierarchy of primary routes, sector distributors ping to be installed — one week). followed by remaining arterial, distributor and lo- • Where moderate to severe damage is predicted, cal roads. and/or where temporary propping or similar work • assessing limits, restrictions and long-term rein- is needed, then the probable vehicle restrictions statement priorities. would be: • co-ordinating with road-based services (water, dam- — Immediately after the earthquake, emergency age etc.). vehicles only, one vehicle at a time, and crawl speed (note that it is unlikely that a weight limit • removing debris and re-establishing maintenance would be imposed in the majority of cases). regimes. This restriction would remain until the temporary work is completed — say in the order of • clearing landslides slips and blockages on hillsides three days. and riverside roads.

— Following propping and/or other short term The time reconstruction will take depends to a great repairs the bridge could be restricted to Class I extent on the season in which the earthquake occurs, and weather conditions. Transport • 161

Four Avenues itself is expected to be out of action or isolated by the These roads are robust for most of their lengths, with earthquake for some considerable time. only three bridges on them, at Fitzgerald Avenue, The alternative, Marshland Road, is also vulnerable Carlton Mill and the Colombo Street flyover. because of the likely collapse of the bridge over the The two river bridges are unlikely to sustain significant Styx River. damage and reinstatement would be rapid. The river State Highway One could probably be opened within bank alongside Fitzgerald Avenue just north of the two or three days unless major bridges are out or the bridge may collapse but this would affect at most only Kaikoura slips have moved significantly. Damage due one carriageway. to most slips could be repaired in two to three days but The flyover is predicted to sustain severe damage, the Kaikoura Coast slip areas could take up to three possibly requiring major and thus lengthy repairs. weeks. Debris from its damage could block Colombo Street, For some bridges on this and other rural routes, fords but Gasson Street is available as an alternative. Simi- and other temporary crossings may be available or can larly, Moorhouse Avenue traffic would, eventually, be be rapidly constructed. able to use the ground level alternatives alongside the bridge. Immediately after an earthquake, traffic would have to use the one-way system. Eastern suburbs primary route This route is likely to be affected by road slumping The avenues themselves have large numbers of trees alongside Avonside Drive (which can be bypassed by but few close buildings, and most damage is likely Worcester Street), and by the collapse of the Pages from broken major underground services. The former Road bridge over the Avon. Temporary bridging could be cleared within hours, the latter, days, weeks or should re-establish the route within three days. An months depending on the extent of serious damage. alternative route is available via Travis Road to the Restoration of services could probably be the greatest north or the Bridge Street bridge to the south. source of damage to the roads as a result of a severe earthquake. Adequate reinstatement of the avenues could be completed within days except for the flyover Lyttelton Primary route (if it is replaced). This route has high potential vulnerabilities. The Waltham Road overbridge is likely to be severely Airport primary route damaged but Gasson Street and Wilsons Road effectively bypass it. There are several Heathcote River This route is likely to open within the first day. As well bridges, most of which should remain serviceable with as serving the airport it connects to the McLeans Island some load restrictions. Reinstatement of at least one area, the primary source of roading aggregate and some crossing should be made within two days. plant for reconstruction work. The Road tunnel itself is unlikely to suffer severe Northern primary route damage but both portals and the cuttings and embankment and bridges on the Tunnel Road are vulnerable, as The twin motorway bridges over the Waimakariri is the lighting and ventilation equipment. The Tunnel River are likely to remain useable after an earthquake. Road can be bypassed (Port Hills and Bridle Path The old Main North Road Bridge is likely to be Roads) but adequate reinstatement may take up to three severely damaged, and unusable. The rail bridge could weeks. Collapse of the tunnel harbour end portal could be used for emergency road use, access is available at damage Norwich Quay and access to the wharves — both ends and temporary decking could probably be clearance may take two to three days. provided. If all four bridges are unusable, reinstatement of one crossing could be very lengthy. Fording is If the tunnel is unusable, the Ferrymead Bridge and impracticable, the crossing would require a large number Evans Pass route will be important. This route is of Bailey or temporary span materials and the alterna- vulnerable to landslips and could take several days to tive Waimakarari gorge bridge is very vulnerable to reinstate. The vulnerable Ferrymead bridge could be earthquake damage. Overall, this crossing is a key bypassed via Bridle Path Road. lifeline point of potentially high vulnerability. Sea connections for Wellington would be established long Southern primary route before the road could be opened. This route is most vulnerable at the Blenheim Road and The Styx Mill overbridge could be replaced quickly by Sockburn (two) overbridges, and the Rakaia Bridge a ground level rail crossing immediately; the rail line north of Ashburton. The overbridges have level cross- 162 • Risks and Realities

ing alternatives available (within hours), and other face of the three small bridges across the Avon River, road routes are available nearby. If the Rakaia Bridge Waimairi and Wairarapa Streams. is not totally destroyed this route should be operable within three days. The section from Sturrocks Road to the Waimakariri River is likely to sustain the most significant damage. It has three major bridges; two of which have relatively Further considerations high approach embankments. These are likely to suffer In addition to the vulnerability of the bridges, the roads from severe ground settlement and slumping-out of themselves can be expected to suffer, in dispersed and batters. The bridges are founded below the silt/sand localised areas with damage from stormwater systems layer and are likely to suffer damage through ground disruptions, water main breakages, trench slumping, shake, mainly differential pier settlement and possible and slipping of banks and other earthworks. Reinstate- holding down bolt failure of span/pier linkages. ment of these will depend on resources available and route priorities. Extensive damage over wide areas The remainder of the section will suffer minor distor- could take several weeks to repair given that in that tion (in comparison) of track alignment. case major structures and services will also be under reinstatement. Main South Line This line, within the limits of this study, is located on Return to normality zone 3 alluvial. As such, reasonably severe lateral and This process will benefit from some breathing space, vertical displacement is likely to be the only resultant detailed planning and assessment particularly on damage. The only structures on this line are the whether or not to replace bridges. Road bridges over Curletts Road and Sockburn Overbridges. It is ex- railways (in particular with the lower train numbers pected that some blockage to the line will occur only at and different legislative requirements than when most the Sockburn Structure. were built) may be replaced with level crossings. Reconstruction of major river bridges will take months. Middleton — Lyttelton Line The time taken for the overall return to normality will From the terminus to Ensors Road minor track damage ultimately depend on the total extent of reconstruction is expected in the form of lateral and vertical displace- required and the available resources. ment. Blockages are also expected as a result of Full reconstruction of the Waimakariri or other bridges damage to the Colombo Street and Waltham Road Overbridges. such as the Awatere road/rail bridge at Seddon may take several months. It must be borne in mind that the From Ensors Road to Thompson Road the track crosses transport system is only one lifeline and the same three different types of substructure. Damage will vary limited contracting, machinery and manpower will be from minor to severe displacement as above. How- reinstating all of them. ever, the worst damage on this section is likely to be on the approaches of the twin bridges crossing the Rail system vulnerability Heathcote River. This embankment is on a high liquefaction zone and is likely to suffer from signifi- Immediately following major earthquake cant settlement. This in turn will create severe differential settlement at the abutments. The bridges them- Main North Line selves are likely to sustain light damage through bolt This line would be closed for the majority of the length shear. under consideration in this study. From Thompson Road to the Tunnel Mouth — This is From the Middleton Yard terminus to Riccarton Road on zone 2 and 2B. From the Tunnel Road Overbridge it crosses a liquefaction zone. Because of the low to Martindale Road the track is built on a high embank- surface loading of rail track likely damage will be ment. This section is likely to sustain the most severe minimal, with some horizontal and vertical alignment damage, with significant settlement of the embank- correction being required. The Blenheim Road ment and probable collapse of the masonry abutments Overbridge is likely to obstruct this section. of the Martindale Road underpass. The track is also likely to be blocked by landslip in and around the From Riccarton Road to Sturrocks Road there is likely tunnel portal, with possible failure of the portal struc- to be significantly more damage, with both vertical and ture. horizontal displacement of the track over its entire length, and minor damage to the abutment/span inter- Lyttelton Tunnel and Lyttelton Yard Complex — The tunnel itself is unlikely to sustain any significant dam- Transport • 163

age apart from minor loosening of some masonry circles at the Cashin Quay reclamation. Severe lining. However, the portals, as mentioned above, are shaking could unbalance the equilibrium of the likely to suffer at least partial collapse and/or landslip reclamation causing subsidence of the pavements blockages. This could be worsened at the Lyttelton adjacent to the wharf and uplift of the seabed. The portal with failure of the masonry supports for the piled structure could suffer some consequential Sutton Quay overpass. damage due to movement of the seabed, but complete collapse is not likely. Damage would occur at Cashin Quay yard is likely to be severely damaged the interface between the breastwork and land, i.e. through rock fall blockages and slip circle effects on to tiebacks, deadmen and retaining wall. The slip the reclamation. circle risk also exists at Z Berth, Gladstone Pier Inner and No 1 breastwork. Rail yard facilities • Jetties supported on piles would suffer little or no These would be totally immobilised as a result of damage owing to their flexible behaviour (except at liquefaction and ground settlement effect, disabling the interface with the land). trackwork and derailing locomotives and rolling stock. • The channel is not at risk from shaking as the virgin mud is too cohesive to suddenly slump. The vulner- Period following earthquake event ability of the inner harbour seabed is also low. During this period the rail system would be reinstated Rocks from rubble slopes could, however, roll into from Lyttelton (Priority 1), from the South (Priority 2), some of the berth areas. Slumping of sediment and from the North (Priority 3). These priorities may, accumulating under jetties is also a low risk. however, change, depending on the interrelationship of road/rail to provide primary links to the remainder of • The rock protection barriers are very stable. Cashin the country. Quay breakwater at Sticking Point could suffer limited damage, but is not considered a significant It is difficult to establish a time frame; however within risk. one to two weeks a partial service on all lines could be provided, although this would be dependent on the • Liquefaction of the reclaimed areas is not consid- availability of materials, equipment and manpower ered to be a risk. The quarry rock fill material used for the Cashin Quay reclamation is not susceptible to resources, as well as the priority of providing a rail liquefaction. An oil company report noted that the service in relation to other needs at the time. tank farm at Naval Point, which is on land reclaimed using pumped dredgings, is not considered suscep- Return to normal tible to liquefaction, and slip circle failure of the sea A best guestimate for the full “return to normal” of all wall is unlikely to affect the tanks. However the rail services would be in the region of six months. grading of the marine silts comprising the reclamation is close enough to those of silts vulnerable to There are two key locations that are likely to control the liquefaction to justify further investigation. length of this period. These are: • The central substation is the only link that the (1) The old Christchurch Railway Station, Moorhouse Lyttelton Port Company has with Southpower’s Avenue — This building houses the computerised network and failure would result in a total loss of Train Control system for Oamaru-Picton and the power supply to the port. The substation is situated West Coast, as well as the Rail Telephone Ex- below Simeon Quay adjacent to No. 7 Jetty. There change. The ability to bring these facilities back is a standby generator for essential functions in the into full service will be governed by the amount of Lyttelton Container Terminal Administration Build- damage sustained by the building and the equip- ing, including the communications equipment in ment. the Signal Tower.

(2) Main North Line Bridges — These are the weak • The Southpower link from Heathcote is an over- links in the local system and are likely to require head cable and any problem can be quickly recti- extensive permanent repair. fied. The poles supporting the overhead wires are a low risk for earthquakes.

Vulnerability of Port of Lyttelton • The Signal Tower itself is vulnerable to building damage requiring evacuation and was designed Immediately following earthquake only to the normal code requirements for office • The greatest risk arises from the effect of slip buildings. 164 • Risks and Realities

Several navigation marks are mounted on piles which have recently been upgraded to reduce vulnerability. are not susceptible to earthquake damage. Some of the Some LPG installations are under review now. navigation lights are solar powered and should not be damaged in an earthquake. The main sector lights are Period following and return to normality land based and have standby batteries to cover for a power cut. There would be some risk to the buildings In view of the limited extent of damage expected, the housing these lights from landslides. Loss of some period following the earthquake required for return to lights would restrict movement of vessels at night. normality is expected to be short unless major damage to the main runways occurs. Full runway reinstatement could take some weeks to repair, but reduced runway Period following earthquake lengths (or one of the two main runways) could allow Restoration of facilities in the period following an restricted operations. earthquake would follow the following priorities:

• restore port communications in the Signal Tower; 6.5 Mitigation Measures — • restore power, water and telephones services; Earthquake • reinstate damaged fuel/oil pipelines (assuming these The vulnerability analysis for the different elements of had been isolated); and the transport system has identified a need for many • reinstate damaged links between inner harbour relatively minor mitigation works which could be jetties and land, e.g. temporary bridging. carried out to existing structures, control and power systems and ancillary equipment. Specific measures for consideration are listed below for each transport Return to normality mode. Further detailed investigations are required for After a major earthquake event ten months is estimated these measures. to restore the Cashin Quay breastwork and six months for other wharves. Road network In all but extreme events it is probable that inner harbour jetties (e.g. No. 2 and No. 7) would still be Bridges serviceable. Bridge vulnerability mitigation measures are indicated for bridges in the summary tables (Tables 6.1 to 6.4).

Vulnerability of airport Typically mitigation measures may include:

Immediately following earthquake • Strengthening connections, particularly between Damage to runways, taxiways and aprons is likely to be superstructures and substructures. negligible and easily repaired. The large grass areas would remain useable by light aircraft and military • Increasing column strength and ductility. aircraft such as Hercules. • Strengthening or renewing retaining and approach Most buildings, including the main passenger termi- structures. nals, are likely to remain useable despite suffering • Strengthening lateral/longitudinal restraint mecha- superficial damage such as cracks, displaced linings nisms. and broken windows. • Constructing landing slabs. One area of concern is damage to machinery and equipment within buildings, particularly that associ- • the addition of information plaques to bridges over ated with communication, navigation and power dis- which key lifelines services pass, describing their tributors. This was found to be inadequately secured in location and nature (services authorities). many cases and failure or fire could result from movement. • Further technical assessment of the Sutton Quay rail overbridge/road retaining structure and clarifi- Underground services including fuel reticulation, sew- cation of responsibilities (TNZ, Tranz Rail, BPDC). erage and water supplies are unlikely to sustain dam- • Detailed geotechnical and structural investigation age from liquefaction or settlement but gravel soils of vulnerable bridges located on primary routes, or may be subject to lateral displacement and thus cause significant alternative routes, for which simple damage. The fuel storage and reticulation systems mitigation measures are not immediately apparent Transport • 165

(e.g. Ferrymead Bridge, Pages Road, Marshland Bridge Street (Avon River) — approach works. Road-Styx River). Route to South • Service authorities to consider isolation of their services from the Ferrymead Bridge. Blenhiem Road overpass — check bridge

A priority list for the initial re-establishment of the Main South Road (Sockburn overpass) — clear primary routes allowing bridges to carry heavy vehicle debris to establish ground level crossing. loads is as recommended below: Access to Port Hills

Central City Colombo Street (Heathcote River) — approach Moorhouse/Colombo — clear debris to establish works. ground level bypass. Access to east of Ferrymead

Barbadoes Street (Avon River) — approach works Ferrymead Bridge (Heathcote River) — approach (State Highway). works and propping, services reinstatement critical to eastern areas. Montreal Street (Avon River) — approach works. These are considered the minimum to establish an Airport Route emergency level of access to key areas throughout the Fendalton Road (Avon River) — approach works. city, possibly requiring some use of alternative routes. Other bridges listed in Tables 6.1 to 6.4 should then be Fendalton Road (Wairarapa Stream) — approach progressively checked for damage and temporary and/ works. or permanent repairs programmed.

Port Route It also needs to be recognised that after shocks may further damage already weakened structures. Waltham overbridge — approach works (State Highway) Roads Opawa Expressway (Heathcote River) — approach General measures identified are: works (State Highway). • Clarify ventilation requirements to allow contin- Port Hills Road (Tunnel Road overpasses) — clear ued operation of Road Tunnel to Lyttelton, provide debris (State Highway). backup power facilities for emergency use.

Tunnel Portals — clear debris (State Highway). • Development of integrated response plans for the Norwich Quay/Sutton Quay — check for damage road and bridges network, recognising the accessi- bility needs for evacuations, supplies, fuel sup- (State Highway). plies, and services reinstatement. Route to North • Make a list of services that are likely to require Main North Road (Styx Rail overbridge) — clear extensive (disruptive) remedial work immediately debris to establish ground level rail crossing (State after the hazard event on primary routes and other Highway). high priority access routes, and make it available to emergency response organisations and service au- Main North Road (Styx River) — approach works thorities. (State Highway). • Ensure that the vulnerability of services on primary Northern Motorway (Waimakariri River) — ap- routes and other high priority access routes is proach works, structural repairs (State Highway). minimised. For example, early replacement of older, vulnerable underground pipes (e.g. Northern Motorway (Tram Road, Ohoka Road, stormwater) on key lifelines roads. Smith Street overpasses) — clear debris to establish ground level by passes, approach works on side • Undergrounding overhead services, particularly streets (State Highway). on significant lifelines routes.

Eastern Suburbs Route • Service authorities to consider lifelines issues when planning service installations on key roads. Pages Road (Avon River) — check bridge 166 • Risks and Realities

• Selection of trees which are more likely to survive the Lyttelton Tunnel, and replacement of masonry an earthquake, particularly on key routes. Plant piers on both the Martindale Road underpass and only known rootfirm species in areas where lique- Sutton Quay overpass. faction may occur. The need to secure communication and signal control • Earthquake proofing traffic signal control and sur- equipment, or to make it re-locatable was also identi- veillance systems. fied, as was reviewing the performance of rail system infrastructure between Picton and Christchurch under • Investigation of the need for railway and road earthquake hazard to identify and prioritise mitigation bridges when they become due for reconstruction measures for reducing earthquake vulnerability. or rehabilitation, and safeguarding land for ground level replacements. The elimination of some of those bridges would decrease route vulnerabilities Port of Lyttelton for road and rail. No major physical works for improving the robustness of the existing port’s structures have been identified. • Identification of alternative river crossings (such as fords) including investigation of the use of rail The effects of a major earthquake could be mitigated in bridges (and the tunnel) for road vehicles, and future planning for redevelopment of the inner har- safeguarding the land for this. bour. This would necessitate identifying key structures that would provide links in the lifelines chain and • Review route diversion procedures and informa- designing them to remain serviceable after a major tion systems — these could be used for minor earthquake. The lifelines goals would need to be one of emergencies such as weather induced route the objectives of the redevelopment programme. The severances. Lyttelton Port Company has written action plans for civil and other emergencies. • Investigation and recording of temporary emergency bridging materials and equipment (including The Lyttelton Port Company and Southpower are bailey bridges) and their locations. looking at establishing another substation at the east end of the Lyttelton township in order to reduce the • Securing road tunnel mechanical/electrical serv- possibility of power interruptions caused by damage ices. both outside and within the Lyttelton Port Company • Review standards of construction and maintenance network. However, total loss of power within the Lyttelton Port Company network would not prevent of all transport facilities to reduce the vulnerability ships from delivering supplies and equipment. of the road network, for example trench backfilling and reinstatement requirements. Using the power ability of one of the rail ferries to provide power to the port would be unlikely as it would • Develop a co-ordination strategy with the contract- require the provision of an expensive (and otherwise ing industry to provide for a planned response redundant) 6.6-11 kV transformer. capability for machinery resources. In the event of damage to the base of a jetty where it • Make a list, and make it available to emergency links to the land, a plan for temporary bridging could response organisations, of plant machinery re- shorten the time taken to recommence operations. sources, location, likely availability, contact names Identifying the type of bridging needed and its avail- and means of contact. ability and location should be undertaken. This should • Undertake measures to minimise disruption due to include the possibility of quickly erecting a rail linkspan slips on the Dyers Pass/Summit Road route to in the event of major damage to the rail system on the communications repeater sites and other key sites lines to Picton and to the south. on the Port Hills. Airport Rail system The buildings and infrastructure of the airport are Given the nature of the rail system there are likely to be robust and unlikely to suffer major damage. However few actions that can be taken from a network planning the vital services such as navigation aids, lighting, perspective. power supplies and communications and security systems are vulnerable since they can be displaced, over- The only engineering measures identified are the better turned or disconnected by minor structural damage. securing of bridge beams to piers and abutments, Securing such equipment to its shelves (and the shelves strengthening of batter slopes at the Heathcote portal of themselves) and assessing their portability or back-up

(continued on page 184) Transport • 167

Priority

services strengthening.

Mitigation Measures

- See Bridges table.

- Undergrounding/

- See Bridges table.

- Undergrounding/

Alternatives

vulnerability of Marshland Road bridge over Styx River.

available, although many Avon River crossings disrupted.

Availability of

- Limited because of the

- Other routes generally

Effect of Hazard

Bridges

Rail overbridge, clearance works/temporary rail crossing required.

other bridges.

Roads

effects over parts of route.

overhead services likely to disrupt roads.

channels and road surfaces.

Bridges

Road bridge will disrupt this route.

Roads

effects.

overhead services.

road surfaces.

- Major damage at Styx

- Approach works likely at

- Possible liquefaction

- Damaged underground and

- Damage to kerbs and

- Major damage at Pages

- Anticipated liquefaction

- Damaged underground and

- Damaged kerbs, channels,

(i)

(ii)

(i)

(ii)

of Route

Importance

EARTHQUAKE

most important.

Road Names

Essential and should be implemented as soon as possible.

Cranford; Main North; Motorway

Woodham; Pages

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Primary

Function

Route and

HAZARD EVENT

Northern

Access to northern areas (Rangiora, Woodend) and linkage to North Island and northern South Island

Eastern

Access to eastern Christchurch suburbs.

Importance: 1 to 5 - 5 Priority: A

Table 6.1: Primary route vulnerability 168 • Risks and Realities

Priority

tunnel portal buildings.

services strengthening.

- See bridges table.

- Hillside plantings.

- Review/strengthen

- See Bridges table.

- Undergrounding/

Mitigation Measures

Alternatives

Southern Sectors tables.

southern sectors.

Availability of

- Refer Eastern and

- Refer western and

Effect of Hazard

Bridges

overbridge, Tunnel Road overpass structures, and Norwich/Sutton Quays structure is likely to severely disrupt this route. Roads

parts of route, including services damages.

due to slips on Port Hills Road. Tunnel

ventilation equipment.

would affect tunnel operation.

Bridges

and Sockburn Rail overpasses will partially (at least) disrupt this route. Roads

surfaces and services.

- Major damage at Waltham

- Liquefaction damage on

- Possible partial blockages

- Damage to lighting and

- Portals damaged by slips

- Damage to Blenheim Road

- Damaged kerbs, channels,

(i)

(ii)

(iii)

(i)

(ii)

of Route

Importance

EARTHQUAKE

most important.

Road Names

Essential and should be implemented as soon as possible.

Barbadoes; Brougham; Opawa Expressway; Port Hills; Tunnel; Norwich Quay

Blenheim; Main South

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Primary

Function

Route and

HAZARD EVENT

Lyttelton

Access to Port of Lyttelton.

Southern

Access to south, Dunedin etc.

Importance: 1 to 5 - 5 Priority: A

Table 6.1: Primary route vulnerability (continued) Transport • 169

Priority

undergrounding of power lines.

- Complete

- See bridges table.

- Complete

Mitigation Measures

Alternatives

Availability of

Refer Western and north- western sectors.

Other local roads and central City streets, one way pairs.

Effect of Hazard

Bridges

Roads

services damaged.

Bridges

Moorhouse/Colombo overbridge, may affect route.

Fitzgerald Avenue bridge.

Roads

collapsed services or broken overhead lines.

- Minor effects only.

- Trees, some overhead

- Robust route.

- Severe damage to

- Some damage to

- Some liquefaction effects. - Damage/blockage due to

(i)

(ii)

(i)

(ii)

of Route

Importance

EARTHQUAKE

most important.

Road Names

Essential and should be implemented as soon as possible.

Fendalton; Memorial

Bealey; Fitzgerald; Moorhouse; Deans; Harper

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Primary

Function

Route and

HAZARD EVENT

Airport

Access to airport.

Four Avenues

Distribution around City centre; access to essential services.

Importance: 1 to 5 - 5 Priority: A

Table 6.1: Primary route vulnerability (continued) 170 • Risks and Realities

B/C

Priority

overhead services.

risk underground services at key locations.

of potential slip zones: - Evans Pass - Bridle Path Road

Physical Works

- Refer Bridges table.

- Undergrounding of

- Renew/strengthen at

- Retainment or planting

Mitigation Measures

Alternatives

be by-passed, provided access available Port Hills Road/Bridle Path Road. However, Port Hills Road likely to be affected by Primary Route damage, Bridle Path likely to be closed by slips.

Evans Pass (Tunnel, Gebbies Pass, Dyers Pass).

crossings available.

route over hills is Christchurch Road Tunnel.

Availability of

- Ferrymead Bridge could

- No ready alternative to

- At grade railway

- Nearest alternative

Effects of Hazard

Bridges

Ferrymead Bridge would threaten route availability.

Causeway bridge.

Roads

Linwood Avenue, Ferry Road areas.

other services damage, broken overhead power lines in these areas and adjoining above bridges.

would close road.

Bridges

overbridge unusable.

OK.

- Major damage to

- Moderate damage of

- Liquefaction damage in

- Burst water pipes and

- Slips on Evans Pass

- Colombo Street rail

- Heathcote River crossing

(i)

(ii)

(i)

of Route

Importance

(An alternative route to the Port if Tunnel is closed).

EARTHQUAKE

most important.

Essential and should be implemented as soon as possible.

Route - Road Names

Four Avenues to Lyttelton

Ferry Road/Worcester Street; Linwood Avenue; Causeway; Main Road (Sumner); Evans Pass

Four Avenues to Governors Bay

Colombo; Dyers Pass

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Sector

HAZARD EVENT

Eastern

Southern

Importance: 1 to 5 - 5 Priority: A

Table 6.2: Internal sector vulnerability Transport • 171

B/C

Priority

overhead services.

risk underground services at key locations.

of potential slip zones: - Dyers Pass Road

overhead services.

services at bridges.

Physical Works

Mitigation Measures

- Undergrounding of

- Renew/strengthen at

- Retainment or planting

- Undergrounding of

- Strengthening of

Alternatives

likely to be more robust.

Availability of

- Sparks Road route

Effects of Hazard

Roads

throughout sector - slumping and broken kerbs and channels.

damage to other services, broken overhead power lines.

road blockages due to slips on Dyers Pass Road.

Bridges

Road overpass, readily cleared.

Heathcote River, approach works. Roads

possible.

kerbs and channels etc.

- Liquefaction damage

- Burst water pipes and

- Medium to high risk of

- Some debris from Lincoln

- Some disruption at

- Minor liquefaction damage

- Some damage to services,

(ii)

(i)

(ii)

of Route

Importance

EARTHQUAKE

most important.

Essential and should be implemented as soon as possible.

Route - Road Names

Four Avenues to South-West

Lincoln; Halswell

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Sector

HAZARD EVENT

Importance: 1 to 5 - 5 Priority: A

Table 6.2: Internal sector vulnerability (continued) 172 • Risks and Realities

Priority

Undergrounding. C

Physical Works

Mitigation Measures

Alternatives

Availability of

- Primary routes.

- Fendalton/Memorial. -

Effects of Hazard

Bridges

Roads

possible.

kerbs and channels etc.

through commercial areas.

Bridges

Roads

possible.

kerbs and channels etc.

through commercial areas.

lines.

- Nil.

- Minor liquefaction damage

- Some damage to services,

- Collapsed buildings

- Nil.

- Minor liquefaction damage

- Some damage to services,

- Collapsed buildings

- Broken overhead power

(i)

(ii)

(i)

(ii)

of Route

Importance

EARTHQUAKE

most important.

Essential and should be implemented as soon as possible.

Route - Road Names

Four Avenues to West

Riccarton; Yaldhurst

Four Avenues to Johns Road

Papanui; Harewood

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Sector

HAZARD EVENT

Western

North- Western

Importance: 1 to 5 - 5 Priority: A

Table 6.2: Internal sector vulnerability (continued) Transport • 173

B C

B/C

Priority

Styx River bridge on Marshland Road.

risk undergrounding services at key locations.

Physical Works

Mitigation Measures

- Replace/strengthen

- Undergrounding. - Renew/ strengthen at

- Undergrounding. - Services strengthening.

- Undergrounding.

Alternatives

alternative may be disrupted at rail bridge.

possible.

to cross Waimakariri River.

Availability of

- Main North Road

- Brooklands diverse

- Johns Road route. - Divert onto motorway

- Local roads.

Bridges

Styx River.

Waimakariri River.

Roads

damage likely.

other services damage, broken overhead power lines.

Bridges

most locations in order to open route, no major structural problems anticipated.

Roads

Travis Road, Brougham Street, including services damages effects.

Bridges

Roads

Effects of Hazard

- Closure of road likely at

- Closure likely at

- Significant liquefaction

- Burst water pipes and

- Approach works likely at

- Liquefaction effects

- Nil.

- Overhead power lines.

(i)

(ii)

(i)

(ii)

(i)

(ii)

2-3

of Route

Importance

EARTHQUAKE

most important.

Route - Road Names

Four Avenues to Waimakariri River

Hills; Shirley; Marshland; Main North

Brougham; Southern Arterial; Curletts; Peer; Waimairi; Grahams; Greers; Northcote; Winters; QE II; Travis

Johns; Russley; Carmen (SH1)

Essential and should be implemented as soon as possible.

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Sector

HAZARD EVENT

North-Eastern

Circumferential Routes

Importance: 1 to 5 - 5 Priority: A

Table 6.2: Internal sector vulnerability (continued) 174 • Risks and Realities

Priority

Physical Works

Mitigation Measures

- See Bridges table. - Undergrounding. C

Alternatives

Availability of

- Local roads. - Colombo Street bridge.

Effects of Hazard

Bridges

generally required.

Heathcote River may be unusable.

Roads

south of Riccarton Road.

channels and road surfaces.

Bridges

Cashmere Road.

Roads

south of Riccarton Road.

channels and road surfaces.

- Minor approach works

- Barrington Street bridge at

- Some liquefaction effects

- Services damage. - Some damage to kerbs and

- Temporary repairs at

- Some liquefaction effects

- Services damage. - Some damage to kerbs and

(i)

(ii)

(i)

(ii)

2-3

of Route

Importance

most important.

EARTHQUAKE

Essential and should be implemented as soon as possible.

Route - Road Names

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Main North; Papanui; Blighs; Idris; Strowan; Whiteleigh; Barrington

Curletts; Hoon Hay; Cashmere

Sector

Importance: 1 to 5 - 5 Priority: A

HAZARD EVENT

Table 6.2: Internal sector vulnerability (continued) Transport • 175

B/C

Priority

overhead services.

risk underground services at key locations.

overhead services.

risk underground services at key locations.

Physical Works

Mitigation Measures - Undergrounding of

- Renew/strengthen at

- Refer Bridges table. - Undergrounding of

- Undergrounding and

- Renew/strengthen at

Alternatives

available.

available, at grade railway crossings.

River crossings at Rutherford Street, Radley Street.

Availability of

- Many other local roads

- Other local roads

- Alternative Heathcote

Bridges

Roads

other services damage, broken overhead power lines.

Bridges

Tunnel Road bridges which would require temporary repairs to ensure serviceability.

Roads

throughout sector - slumping and broken kerbs and channels.

damage to other services, broken overhead power lines.

Effects of Hazard

- Nil.

- Liquefaction damage. - Bust water pipes and

- Moderate damage to

- Liquefaction damage

- Burst water pipes and

(i)

(ii)

(i)

(ii)

of Route

Importance

EARTHQUAKE

most important.

Route - Road Names

Ensors; Aldwins

Tunnel; Bexley

Essential and should be implemented as soon as possible.

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Sector

HAZARD EVENT

Importance: 1 to 5 - 5 Priority: A

Table 6.2: Internal sector vulnerability (continued) 176 • Risks and Realities

Comment Priority

easily created

Bridge likely to be severely damaged

5 See note*

3 Ground Level by pass

4 Old Main North Road

3 Gasson Street bypass C

Importance

Route/Structure

Measures

Mitigation

Strengthen structure; or replace with arch structures None required Improve longitudinal EQ restraint. Upgrade superstructure/pier cap connection

(ie replace)

uneconomic to strengthen

3 Remove structure -

3 Remove structure; or

2 2

3 Uneconomic to strengthen

Category

Bridge Name Damage

Styx River Chaneys O'pass

Motorway (Waimak R)

Northern Primary Route

Eastern Suburbs Primary Route

Lyttelton Primary Route

I D

T702 Styx (Rail o/b)

T401 T701

T501

R102 Pages (Avon River)

R701 Waltham (Rail o/b)

R205 Opawa Ex'way (Heathcote R) 1 Nil

Table 6.3: Key bridges — primary routes Transport • 177

Comment Priority

Port Hills Road/Heathcote Valley, however this is also affected by debris from the Overpass structure.

could affect road, and rail access to port and access to wharves.

3 Bypass available

4 By pass available via

4 Bypass available B

3 Bypass available

3 Bypass available B

5 This complex structure

Route/Structure

Mitigation Importance

and ductility. Underpin abutments

structure

abutments

strengthen abutment base lateral resistance

seismic capacity and strengthen if necessary required.

1 Review portal building

3 Enhance column strength

3 Build new "culvert" type

2 Uneconomic 4 " " B

2 Concrete props between

2 Strengthening expensive

3 Further investigation

Category

Bridge Name Damage

o/pass)

o'pass)

Road o'pass)

Lyttelton Primary Route (Cont)

Southern Primary Route Blenheim Road (Rail o/b) 3 Strengthen cribwalls,

(rail o/b)

- Tunnel

- Norwich Quay/Sutton Quay

I D

T711 Porthills Road (Tunnel Road

T712 Horotane Valley (Tunnel Rd

T713 Heathcote Valley (Tunnel

T708 Sockburn (underpass)

T709 Sockburn (Rail o/b)

R705

Table 6.3: Key bridges — primary routes (continued) 178 • Risks and Realities

Mitigation

Comment Priority for

3 Bypass available B

Importance

Route/Structure

Measures

Mitigation

strengthening between super and sub structure and between the two spans

1 Nil

2 Strengthen connections

1 Nil

Category

1. Minor, possible speed restriction, some works required immediately and following earthquake. 2. Moderate, restrictions (load/speed) would be applied, significant amount of works required. 3. Severe, collapse likely/bridge likely to be unusable, or useable for emergency traffic only, major works or bypass required in long term.

Bridge Name Damage

A. Essential and should be regarded as an urgent works or further investigate item. B. Should be programmed for works or further investigation. C. Works desirable, but may require further investigation (e.g. of need for structure). * The Heathcote River Bridge on the Expressway is the most robust crossing (Ferrymead and Tunnel (Roundabout) Rd Bridges are vulnerable). This bridge with use of Tunnel interchange ground level by passes and Port Hills Road would provide route Lyttelton. It could provide access to Evans Pass (via Bridle Path Road) if Tunnel unuseable.

Airport Primary Route

Four Avenues Primary Route Carlton Mill (Avon R) 1 Nil 4

- Fendalton (Wairarapa)

I D

R125 Fendalton (Avon R)

R123 R702 Moorhouse/Colombo (O'pass) 3 Longitudinal seismic R109 Fitzgerald (Avon R)

Damage Category:

Priorities:

Table 6.3: Key bridges — primary routes (continued) Transport • 179

Mitigation

Priority for

Temporary Basis

Comment - Works to re-open on

Temporary joint repairs, approach works and propping.

Temporary reinstatement of approaches.

Temporary bypass could be established.

Likely to require temporary approach work and props to remain usable.

Approach works required.

Temporary minor approach work.

Approach works.

Mitigation Measures

Strengthening to be investigated.

Major structural improvements would be required.

Nil.

Shear keys between superstructure and pier and abutments.

Damage

Category

most important.

Bridge Names

Essential and should be implemented as soon as possible.

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Eastern Sector - City to Port of Lyttelton via Ferry Road/Linwood Avenue and via Evans Pass.

Ferrymead.

Causeway.

Eastern Sector - Cross Sector Routes.

Tunnel Road rail overbridge.

Tunnel Road.

South Brighton.

Eastern Sector - Other Bridges.

Garlands Road.

Rutherford Street

Southern Sector

Opawa Road.

R201

R801

T710

T201

R101

R203

R202

Importance: 1 to 5 - 5 Priority: A

Table 6.4: Bridges summary — internal sectors 180 • Risks and Realities

Mitigation

Priority for

Temporary Basis

Comment - Works to re-open on

Temporary approach works.

Minor repairs possibly.

Minor approach works (long term pile repairs likely).

Minor approach works.

May be damaged beyond repair or temporary use.

Minor repairs and temporary propping.

Minor approach works.

Minor approach and joint repairs. Controlled yielding.

Mitigation Measures

Possible strengthening of connections.

Major strengthening/renewal.

Major strengthening of substructure.

Major strengthening/renewal of culvert structure.

Damage

Category

most important.

Bridge Names

Essential and should be implemented as soon as possible.

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Ensors Road.

Wilsons Road.

Tennyson Street.

Colombo Street (Heathcote).

Barrington Street.

Cashmere Road.

Rose Street.

Sparks Road.

Lincoln Road.

Curletts Road.

Lincoln Road overpass.

Raceway overpass.

Southern Sector Cont

Wrights Road overpass.

R208

R209

R210

R213

R214

R217

R219

R220

R221

T202

T703

T704

T705

Importance: 1 to 5 - 5 Priority: A

Table 6.4: Bridges summary — internal sectors (continued) Transport • 181

Mitigation

Priority for

Temporary Basis

Comment - Works to re-open on

Minor approach works.

Minor approach works and joint repairs. Controlled yielding.

Unusable, unrepairable.

Superstructure settlement repairs.

Minor approach works.

Unlikely to be usable.

Mitigation Measures

Major seismic strenghtening.

Increase bearing seating area.

Major strengthening.

Major strengthening or renewal.

Improve lateral restraint.

Damage

Category

most important.

Bridge Names

Essential and should be implemented as soon as possible.

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Annex Road overpass.

Curletts Road Rail overpass.

Colombo Street Rail overpass.

Durham Street Rail overpass.

Western Sector

Straven Road (north Kilmarnock).

Straven Road (south Fendalton).

Clyde Road.

Waimairi Road.

North-Western Sector

Idris Road.

Harewood Road (Styx River).

Sawyers Arms Road (Styx River).

North-Eastern Sector

Marshland Road (Styx Road).

T706

T707

R703

R704

R131

R159

R146

R150

R412

R411

R405

Importance: 1 to 5 - 5 Priority: A

Table 6.4: Bridges summary — internal sectors (continued) 182 • Risks and Realities

for

Priority

Mitigation

Temporary Basis

Comment - Works to re-open on

Minor approach works.

Minor approach works, temporary propping, repairs.

Apporach works, propping.

Bridge unusable.

May be usable. Extensive propping.

Major approach works, propping.

Minor approach works.

Mitigation Measures

Strengthen connections.

Major strengthening.

Seismic lateral restraint at abutments and strengthen abutment walls.

Strengthen abutments.

Major strengthening, new Expressway bridge proposed.

Shear connections superstructure/abutments.

Damage

Category

most important.

Bridge Names

Essential and should be implemented as soon as possible.

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Northcote Expressway Aluflo Culverts.

New Brighton Road (Dudley Creek).

Stanmore Road (Avon).

Swanns Road (Avon).

Gayhurst Road (Avon).

Avondale Road (Avon).

Wainoni Road (Avon).

Shirley Road (Shirley Stream).

Central City Sector

Barbadoes Street.

Kilmore Street.

R105

R108

R107

R106

R104

R103

R303

R110

R111

Importance: 1 to 5 - 5 Priority: A

Table 6.4: Bridges summary — internal sectors (continued) Transport • 183

for

Priority

Mitigation

Temporary Basis

Comment - Works to re-open on

Minor approach works.

Mitigation Measures

Shear connections superstructure/abutments.

Damage

Category

most important.

Bridge Names

Essential and should be implemented as soon as possible.

B Desirable and should be programmed for works or further investigation. C Desirable but consideration can be deferred.

Madras Street.

Manchester Street.

Colombo Street (Avon).

Armagh Street.

Gloucester Street.

Hereford Street.

Durham Street.

Montreal Street.

R112

R113

R114

R115

R116

R118

R119

R120

Importance: 1 to 5 - 5 Priority: A

Table 6.4: Bridges summary — internal sectors (continued) 184 • Risks and Realities

facilities is a relatively cheap task which should be Rail system vulnerability undertaken. Air conditioning plant is widespread at the Rail operations would be suspended during the peak airport and is similarly vulnerable and its failure or wind event due to the risk of passenger carriages (in displacement can damage more vital equipment, its particular) blowing over during wind gusts. This would security should be also investigated. require sufficient early warning to enable arrangements for passenger safety to be made. Areas most at Beyond the airport the security of communication and risk would be the main trunk lines to the north and south power supply needs further investigation in addition to of the city. investigation on the airport. Other rail infrastructure including buildings (especially large glazed areas, e.g. Addington Station) and 6.6 Vulnerability — Major communications aerials would be at risk. Windstorm Airport vulnerability Road network vulnerability The main impact of the worst scenario wind storm The windstorm peak may last 1 to 1.5 hours, however would be on navigational aids and communications high winds may continue for some time (see Section equipment. The existing aerials and masts associated 2.6). with the communications and glide slope system are not designed for wind speeds in the order of those set Roads will be blocked by trees and overhead services, out in the scenario. traffic signs and roofing, glass and other building debris. Abandoned and crashed vehicles, including The restoration of services after a wind storm could be trucks and trailers, could also disrupt roads. Alterna- achieved relatively quickly with the availability of tive routes may not be available and rain following the spare aerials and mast components from both on site windstorm could occur and may exacerbate the situa- and from other sites within the Airways Corporation tion. network.

Immediate response may be difficult as communica- The vulnerability of radar systems has been assessed tions and power towers, pylons and poles may be and it is felt that the secondary surveillance radar and damaged. Damage is likely to be widespread. Broken radome on Cass Peak should be able to withstand the power lines may prevent an early response to road expected wind strengths (the identical unit in Welling- clearance. Initial priorities would concentrate on es- ton has demonstrated excellent performance in severe tablishing emergency corridors for services, rescue storm conditions). The primary radar mast is felt to be and movement of people. vulnerable.

Activities required include: Other considerations for wind storm conditions are the impact on buildings and, particularly, on structures • opening up routes with priority decided by the such as the control tower which is very exposed and hierarchy of Primary Routes, Sector Distributors may suffer damage. An alternate site for Air Traffic followed by remaining Arterial, Distributor and Services exists in the Fire Service watch room if Local Roads. Determining priorities will need to be significant damage is sustained. Wind blown debris on co-ordinated with service authorities in ensuring runways would need to be cleared prior to operations that the most effective actions from the total City recommencing. point of view are undertaken first. (Service Author- ity work in restoring all overhead services is likely to take several months but this need not impinge Restoration of Service markedly on key transport routes.); The airport would be closed during a wind storm of the order estimated. The airport could be reopened very • assessing limits, restrictions and long-term rein- quickly after a storm has passed but the operation may statement priorities; be limited to daytime visual flight rules if navigation • removing debris from roads, footpaths, drains, be- aids have been damaged. The degree of damage would neath bridges, etc., and re-establishing mainte- define the limitations to operations and a priority nance regimes. would have to set at the time on which navigation aids would be repaired first. Initial road clearance will be followed by detailed clean up programmes and finally a return to normal Full restoration of services assuming the worst case maintenance schedules. damage could take up to three weeks. Transport • 185

Port of Lyttelton vulnerability terms of the worst scenario. If it is practical to strengthen Wind has not caused any known damage of signifi- the aerials and masts, an upgrading programme should cance to port installations. Provided adequate warning be undertaken. CIAL should discuss the writing of a is provided, container cranes and other critical equip- formal restoration plan with the Airways Corporation ment and berthed ships can be secured. Design wind to enable speedy reinstatement of service. speeds for port structures are typically 70 knots (130 A disaster plan for the wind storm scenario should be km/h) to 85 knots (160 km/h), with the container written to provide protection to buildings and people. cranes designed for 92 knot (170 km/h) winds. Veloci- Points to be considered could include advice on open- ties above this level are likely to result in damage to ing windows on the lee side of the building to minimise facilities such as the signal tower and container cranes. windows blowing out etc. A structural engineer should At wind gust velocities of 170 km/h it is possible that be engaged to establish strategies for minimising inter- bollards on older inner harbour finger jetties would nal wind pressure differentials. pull away if ships were berthed there.

Trees 6.7 Vulnerability — Major Wind storms present the biggest threat to trees in Canterbury with winds of the severity of the 1975 Snowstorm storm anticipated to recur on average approximately every 30 years. In most years, however, high winds are Road network vulnerability experienced which cause significant damage to planta- The snowstorm may last for several days. Roads may tions and amenity trees. In Canterbury forest situations be completely blocked by snow, and in some locations in 1975, 64% of all trees uprooted or broken were by trees and collapsed overhead services. Abandoned shelterbelts and group plantings — the greatest risk to and crashed vehicles could also disrupt roads. roads being blocked in rural and semi rural areas. Old roadside poplar shelterbelts and willow often suffering If snow freezes on road surfaces, traffic disruption from extensive decay are another high risk. On average would increase markedly as the ice could not be cleared deciduous trees are less likely to be windthrown than until the snow thawed. All vehicles would, in these evergreen trees as they are without foliage for approxi- conditions, require chains if the roads were passable. mately four to five months of the year when they The spreading of grit would help alleviate icy condi- present less “sail effect” to wind forces. Weeping tions. The effects would be magnified on the Port Hills willows are an exception as brittle timber and very where blockages could last several days or longer. weighty dense branches makes them particularly prone Alternative routes may be available if ongoing snow to wind damage. The trunks of most other open grown clearance during the storm allows emergency corridors deciduous broadleaved trees are usually stockier than to remain open. Power loss would result in loss of Pinus species and therefore less prone to stem break- traffic lights. age. Some communications difficulties would occur, par- Research following the catastrophic storm in England ticularly as access could be difficult. This would limit in 1987 found that trees that had been pruned to thin the amount of available information on conditions and crown density stood, while unthinned trees fell. Judi- problems. cious pruning obviously plays an important part in Flooding is likely from snowmelt, particularly when significantly reducing the degree of risk from combined with and tidal effects. windthrown tree hazards. Of concern in urban areas is the long term effect trenching/excavation operations Initial priorities would concentrate on establishing may have had on the root systems and stability of trees emergency corridors for services and the rescue and growing on or near public roads. movement of people.

Activities required include: Mitigation measures • opening up routes with priority decided by the Road System hierarchy of primary routes, sector distributors Mitigation measures for snow and wind hazards are followed by remaining arterial, distributor and lo- discussed in Section 6.7 “Major Snowstorm”. cal loads (this is the hierarchy identified in this report);

Airport • assessing limits, restrictions and long-term rein- The design of aerials and masts should be assessed in statement priorities; 186 • Risks and Realities

• removing snow, tree, etc. debris and re-establish- the ability to clear snow without specialist equipment ing maintenance regimes; and is limited.

• clearing landslides, slips and blockages on hillsides. Restoration of Services 1 Immediate action will be taken during the snow Snow clearing plant is located throughout the city in storm to attempt to keep the runway system clear. various contractors’ depots. The City Council’s in- house contractor has three yards, each able to respond 2 In freezing snow conditions: in some way. This will be of value as the immediate No action will be possible until snow thaws. A availability of external contractors’ plant cannot be further two to three days will be required to clear guaranteed. melting snow. Service Authority work in restoring all overhead services is likely to take several months, but with effective 3 If temperatures remain above freezing: co-ordination will not impinge markedly on key trans- Limited operations may be possible after one day of port routes. Initial road clearance will be followed by snow clearing with full services available in two to detailed clean-up programmes, retaining wall con- three days. struction and the like. Normal maintenance schedules will be phased in. Port of Lyttelton vulnerability Rail system vulnerability Road access to the port would be blocked thereby affecting operations. Shipping would be affected by Rail operations would be suspended with the build up land-based disruption. The closure of roads means the of snow on tracks. As with roads, trains could resume port cannot be serviced by personnel. The signals operations once snow clearance or snow melt had tower, built in the 1970s, has a flat roof which could be occurred. damaged under snow overload. Container crane opera- Damaged trees and overhead power lines adjacent to tions would be suspended until snow could be cleared. tracks would also disrupt operations, requiring clearance prior to the resumption of services. Trees The planting of large branched evergreen trees in Airport vulnerability positions where the branches are likely to overhang the The effect of snow on airport operations will depend on road carriageway should be avoided. Commonly planted what happens after the snow fall and to some extent trees such as Pinus radiata, Cuppressus macrocarpa how “wet” the snow is when it falls. During the and Acacia (wattle species) and cedar species are snowstorm the airport may be closed and efforts will particularly susceptible to major branch and stem break- begin immediately to clear snow from the runways, age under the weight of snow. In wet ground conditions taxiways and apron areas. However, the ability of the evergreen trees heavily laden with snow may lift at the airport maintenance staff to clear snow is limited by roots and fall over, particularly when accompanied by equipment. high winds. Pinus radiata suffered a high incidence of this in the August snow of 1992. If the snow freezes to runway surfaces the airport will be closed until the snow thaws. There are no acceptable If large exotic conifers, evergreens etc. are required for methods available for clearing frozen snow/ice from planting near roads they should be of a species indig- the runway surfaces. enous to high snowfall areas and therefore naturally resistant to snow damage. The main strip area may be cleaned reasonably quickly if freezing does not occur but clearing of snow from Generally it would be safer to plant deciduous trees around navigation aids will be slow and manpower (with the exception of willows) as the weight of snow intensive. This may mean that limited daytime visual on bare branches would not ordinarily be sufficient to operations only may be possible in the short term. cause major breakage or root lift.

The impact of snow falls of varying depths has been Another consideration with regard to deciduous trees is considered. The restoration and mitigation measures that bare branches allow sunlight to reach the road described are for snow falls of up to 250 mm. In carriageway causing the snow to melt faster. extreme events where the snow fall exceeds this level Transport • 187

Mitigation measures 6.8 Vulnerability — Road Network (Snow and Wind Hazards) Waimakariri River Flooding • Investigate the resilience of and the need for backup power facilities for traffic monitoring cameras and General effects display units to enable ongoing conditions moni- Breakout affecting the transport system could occur at toring of key city sites from the City Council’s any of three locations as detailed below. Tuam Street building. Upstream breakout from Halkett • Establish priorities for putting overhead services underground on key transport routes, including The most upstream breakout, from Halkett, would primary routes, sector distributors and other arte- result in flows crossing the Main South Road in several rial routes. locations in the Hornby/Wigram area, flowing into the upper reaches of the Heathcote and Halswell Rivers. • Clarify and document emergency event procedures River water would also flow across the airport, cross- (disaster plan), responsibilities and information ing Russley Road and flowing east of Memorial Av- flows across transport agencies and other lifelines enue to reach the Avon River. This event would have agencies including establishing response plans and the following effects: reconnaissance plans. • Road network — disruption to State Highway 73, • Provide more 4-wheel drive vehicles in the Coun- Main South Road and State Highway 75 to Akaroa, cil’s fleet. water depth approximately 0.5 metre. The road would be closed to traffic blocking access to the • Procedures should be developed for the spreading south and west. of grit. • Similar disruption to Russley Road would occur. As a general guideline, the following are recommended Flooding of the Avon River from Hagley Park to for the management of trees in the vicinity of all public the Estuary could block access from the south of the roads: city to the north. There could be some washouts. Clearance of roads following the flood should be • Plant fewer large growing conifers/evergreens on able to be effected quickly. the north side of roads than the south. • Access from the north of the city to the north should • Avoid planting damage-prone conifers/evergreens be available. Road access to the port could be such as Pinus radiata, Cuppressus macrocarpa, threatened by high flood levels in the Heathcote Acacia/Racosperma (Wattle species) and most River. Eucalyptus species within 35 metres of roads. • Rail network — washouts of ballast would close • Plant at wider spacings to ensure individual trees the south line. Temporary repairs could be effected grow and adapt to wind forces and have adequate quickly once floodwaters subsided. room for lateral root development. • Port of Lyttelton — no direct effects. • Carry out frequent pruning to thin crown and reduce excessive end weight on large branches. Breakout floodwaters from Crossbank • Progressively remove suspect pines, macrocarpas, Breakout floodwaters from Crossbank would flow evergreens, debilitated poplars and willow and across the airport, following both the Avon River path replant with more suitable species. above, and also along Johns Road into the Upper Styx River. Effects are as follows.

Airport • Road network — Russley/Johns Road would be A disaster plan should be written to highlight action to closed, depth up to 1 metre. Main North Road and be undertaken and contingency measures and a resto- Marshland Road would be restricted to heavy vehi- ration plan for scenarios should be written. Relation- cles and four-wheel drives, depth up to 0.5 metre. ships with business partners involved in contracting Access would be available from the south and west, should be formed to enable access to more equipment with limited access from the north. to speed snow clearing. • Rail network — ballast washouts on the north line would affect rail transport to the north, probably for 188 • Risks and Realities

several days. Access remains available from the ponding areas will depend on being able to dispose of south and the port. the water.

• Port of Lyttelton — no direct effects. Limited services operating during the hours of daylight under visual flight rules might be operational in two to Downstream breakout at Englebrechts three weeks. Full restoration of services is likely to take four to six months. The downstream breakout at Englebrechts would flow directly into the South Branch of the Waimakariri, flooding the Chaneys and northern motorway areas Mitigation measures south of the main river. Water would also enter Kaputone The most effective mitigation measure is the Canter- Stream to the Styx River. Effects are as follows. bury Regional Council’s proposed flood alleviation scheme, comprising of a stopbank system west of the • Road network — Main North Road, northern mo- airport. This would not reduce the likelihood of the torway and Marshland Road likely to be closed. breakouts described above from occurring, but would This would cut access to the north during the divert these flows safely back into the Waimakariri. duration of the flood event. Access would be avail- The possibility of reducing velocities in the old flood able to the south, the airport and the port. channels near the airport should be investigated.

• Rail network — ballast washouts would close the The possible development of a second parallel runway north line only. system in the longer term (20 to 25 years) may give an • Port of Lyttelton — no effects. opportunity to reduce the impact of flooding in the design of the earthworks associated with such a pro- Even without breakout, there could be some disruption posal. An airport restoration plan should be developed to the transport system. While the twin motorway to speed the process of returning to normality. bridges and rail bridge across the Waimakariri are expected to survive this flow, the older Main North Road bridge with its shallower piles could well be 6.9 Vulnerability — Local Flood severely damaged. In transport terms this would not be a major problem, although it would bring into question Hazard the issue of whether to rebuild the bridge at all. The flooding scenario is described in Section 2.4, “Local Flooding Hazard”, an event being close to the Airport vulnerability “Probable Maximum Flood”. Effects on the transport A major flood event breaking out from the Waimakariri system would arise from flooding of the Avon, upstream of the airport could result in damage to the Heathcote and Styx Rivers, together with the accumu- runway and taxiway system, probable flooding of the lation of stormwater throughout the city, unable to airport terminal building basement and other airport enter the drainage system. There would be widespread buildings (aircraft operators, freight forwarders etc.) local ponding throughout the city, with secondary and severe damage to power, water, sewerage and surface flow paths developing to tributary streams, stormwater reticulation. Depending on water veloci- open drains and river channels. These flow paths ties in the old flood channels the runway may be would very likely be concentrated on roads, primarily damaged by scouring and will definitely be left cov- because adjoining properties tend to be raised above ered with silt and debris. The flooding of the terminal road level, and because of the blocking effect of fences, building basement will result in a loss of power, heat- etc. In general, most roads would remain passable by ing, cooling and some communication (see Section vehicles, with some roads totally blocked and others 6.9). requiring four-wheel drive type vehicles. There are a number of likely “trouble spots” as highlighted below. The damage to site reticulation services such as power There could also be significant effects in hill catch- and water will be significant as the floodwaters drown ments, with saturated soils slipping and affecting roads pumping stations and electrical substations on site. on or immediately below the Port Hills. These are discussed in Section 6.11, “Slope Hazards”, in relation Restoration to the transport system. The main limiting factors in restoring services will be A further form of local flooding could occur in the the availability of water and power. It is also likely that event of extreme high tide levels in the Estuary. This flood waters will dump silt into the soak pits on site. could cause significant inundation effects in low lying The ability to pump out flood water from buildings and areas of eastern Christchurch. However, the attenua- Transport • 189

tion effects over wide areas, together with limits to the Four Avenues Primary route amount of sea water that can physically enter the No major river-based effects are anticipated. Estuary and the blocking effect of existing stopbanks, would restrict the maximum inundation depth (see Other effects Section 2.4, “Local Flooding Hazard”). A number of roads crossing the Heathcote River would be closed for several hours. These include Sparks Road network vulnerability Road, Cashmere Road, Ashgrove Terrace, Fairview Northern Primary route Street, Centaurus Road, Tennyson Street, Waltham Road and Opawa Road. All would require detours of up Blockages and flooding of the Styx River could cause to 3 km. backing up of floodwaters at Main North Road (culvert) and the alternative Marshland Road bridge, al- The flat gradients of the city’s river channels result in though it is unlikely that this would be sufficient to relatively low river velocities. This means that bridge affect traffic. Styx River flooding would cause disrup- foundation scour and/or road undermining effects have tion in the Brooklands area, probably blocking Lower not to date been significant. There are likely to be some Styx Road in several locations. Cranford Street would locations however (e.g. Heathcote River - Colombo be affected by local floodwaters in the Dudley Creek Street bridge) where the bridge/culvert waterway area and Innes Road areas. These could have significant is undersized and where a hydraulic “head” can be traffic disruption effects, however the road is expected developed. High local velocities in these situations to be passable. Papanui Road could be similarly af- could cause local scour effects and, potentially, under- fected by Dudley Creek flooding. mining of foundations.

Eastern Suburbs Primary route While the more localised flooding and secondary flow paths create a nuisance, affected areas could generally Major river-based disruption is expected near the Pages be expected to be accessible by four-wheel drive vehi- Road Bridge over the Avon River where floodwaters cles. This “storage” of stormwater also helps to mini- could be up to 900 mm deep for several hours coincid- mise peak river flooding and associated effects at road ing with very high tide levels. Alternative routes via crossings. However, the continuing operation of road- Bridge Street, New Brighton Road or Travis Road side sumps and stormwater collection systems is essen- would be similarly affected, although to a lesser extent. tial in allowing the local effects to be minimised.

Lyttelton Primary route Rail system vulnerability This route to the port is not expected to be affected by No major flooding effects are anticipated as the rail river-based flooding, although local flooding effects infrastructure and key facilities are above likely flood could be disruptive. Slips are discussed in relation to levels. Slip hazard effects are discussed in Section slope hazards. 6.11. Alternative routes via Linwood Avenue or Ferry Road would be affected by flooding associated with high Railway embankments themselves can create flooding estuary level and backing up of the Heathcote River. problems, if waterways become blocked. Ferry Road would probably be impassable for several hours. The Sumner route would also be affected by Airport vulnerability road flooding of Wakefield Avenue and hill slips. The airport is prone to local flooding in a prolonged heavy rainfall scenario. The airport site relies on soak Southern Primary route pits for the disposal of stormwater. None of the major No major effects are anticipated. Localised flooding on river systems (Avon, Heathcote or Styx) have signifi- the alternative Riccarton Road route in the Church cant effects on the airport. Many of the existing build- Corner vicinity would disrupt traffic flows on both ings on site are built with ground floor levels which are Riccarton and Yaldhurst Roads. at or very near ground level. The main terminal complex is built across the natural fall on the site which makes it vulnerable to flooding in basement areas. The Airport Primary route terminal basements contain service areas and flooding No major river-based effects are anticipated. Local would result in the loss of power, sewerage, heating, flooding in the vicinity of the Russley/Memorial inter- cooling and some communications. section could potentially disrupt access both to the airport and on the north-south State Highway 1 route. 190 • Risks and Realities

Restoration flood flow velocities and could cause scour of In the worst case scenario with significant local flood- riverbanks. It is expected that velocities would exceed ing affecting the terminal basement, the restoration of 1.5 m/s for less than half an hour. services would hinge on the availability of power, Effects will be confined to coastal areas and routes and heating and sewerage system. These services could be structures associated with the Waimakariri, Styx, Avon restored within two to three weeks. Limited operations and Heathcote Rivers. The primary routes and their could be available within one to two days. alternatives are detailed below.

Port vulnerability Northern Primary route No major flooding effects are anticipated. Information in Section 2.5, “Tsunami Hazard”, suggests that bore velocities in the Waimakariri River may Mitigation measures be high enough to damage river protection works. This could affect major bridge structures, not only on the Road network Waimakariri but also the Cam and Kaiapoi Rivers. • Ensure that good records of sump locations and These effects have not, however, been quantified. stormwater pipelines are available to key people in an emergency situation. Eastern Suburbs Primary route • Adequate sump maintenance and street cleaning This route will be significantly affected. The tsunami procedures are needed to ensure that available staff bore is expected to propagate up the Avon River to near are not overloaded in clearing blocked sumps and Fitzgerald Avenue, flooding riverside roads and poten- stormwater systems in an emergency situation. tially damaging bank protection works. The initial wave height at the Estuary mouth of some three • Develop self-cleansing sump inlet structures that mreduces as it travels upstream, but would disrupt this retain functionality in wet weather yet avoid the route in the Pages Road/Avon River vicinity, with intrusion of “rubbish” into the river system during inundation depths of around 0.5 m to 1 m. normal weather. Alternative routes would be similarly affected. Inun- • Disaster planning. dation is not expected to have any major lasting effects on the road infrastructure, however scouring of sandy Airport berms could undermine paths and kerbs and channels. • Future buildings on site should be designed with Access to areas east of the Avon River would be crucial services above ground floor level. Ground available to the north, e.g. Travis Road. floor levels should be established at a minimum height of 300 mm above surrounding levels. Lyttelton Primary route • Investigate feasibility of interim stormwater stor- Although the tsunami bore wave in the Heathcote age facilities. River is expected to travel as far as the Opawa Express- way, its height will not be sufficient to disrupt the road. A disaster plan should be written to cover this scenario Velocities are not expected to be sufficient to threaten aimed at minimising the impact of the flooding and a the Expressway bridge. restoration plan should be written highlighting the reinstatement of key services and systems. However, alternative routes to the port and the Sumner/ Redcliffs area will be significantly affected. Linwood Avenue and Ferry Road will both be affected by the surge in the Estuary, with inundation depths likely to be 6.10 Vulnerability — Tsunami about 0.5 m.

Road network vulnerability The causeway across McCormacks Bay would be Peak river velocities arising from the tsunami bore inundated by over 1 metre, with significant inundation were calculated using a model of wave propagation also in the Redcliffs (Moncks Bay) area. Water veloci- excluding stopbank overflow effects. The tsunami ties some six times greater than normal tidal velocities wave takes the form of a “rapid” three hour “tidal could undermine sea walls that protect the road in this cycle”. The velocities calculated are 2.3 m/second at vicinity. Disruption can also be expected in Sumner, Pages Road (Avon River) and at Tunnel Road with overtopping between Shag Rock and Cave Rock (Heathcote River). These are about double “normal” flooding the road and flowing into Sumner. Long term damage to roads is not expected to be significant in this Transport • 191

area. However, access to Sumner could be lost, al- Airport vulnerability though alternative access would be available via Evans No effects. Pass. Port of Lyttelton vulnerability Southern Primary route, Airport Primary The tsunami would inundate harbour facilities by up to route 1 m to 2 m. Given adequate warning of the event, No effects. precautions such as the turning off of electric power, evacuating the harbour of vessels, etc., would limit Effect on bridges loss. Experience at Lyttelton has shown that tsunami waves do not shoal into breaker type waves. Loss of Scour any structures is extremely unlikely, however some As peak velocities during the assumed event are greater damage to the Cashin Quay breakwater may occur. in magnitude than flood velocities but are of much Siltation of the main channel could occur, but this shorter duration, it is believed that there may be some could be dealt with through the maintenance dredging local scouring effect at bridges on the lower reaches of programme. Overseas experience has shown that fires the rivers. in electrical and fuel installations are frequently associated with tsunamis. It is probable that the port would Drag on bridge structure be operable immediately following the event. Electri- Although an exhaustive check on all the bridges has not cal services would take several days to restore. been carried out, it is clear that a number of bridge decks will obstruct the passage of the ‘worst case’ Mitigation measures surge. For example, the surge level at the Ferrymead • Identification of road routes that would be avail- Bridge is within 500 mm of deck level, and at Rutherford able for evacuations from eastern and low-lying Street, the deck is overtopped. However, calculations suburbs (more work required in comparing road indicate that the drag forces are not high, certainly no levels with wave profiles in rivers). worse than the earthquake forces that the bridges have been designed for. • Technical assessments of vulnerable road network structures (e.g. Moncks Bay sea wall, Ferrymead Buoyancy Bridge, Bridge Street and Pages Road Bridges) and the need for scour protection measures. A potentially serious problem is buoyancy created by the air trapped under deck slabs in the spaces between the bridge beams. Although it is unlikely that any bridge decks will be lifted off their bearings, it is 6.11 Vulnerability — Slope conceivable that the reduction in net downwards load Hazard on prestressed beams could result in excessive tensile stresses in the tops of the beams. The Rutherford Street Introduction bridge, with its low deck soffit level of RL 11.4, is This section reviews the overall likely performance of potentially vulnerable to this condition, with stress the main road transport routes given the slope hazard levels being close to acceptable limits. Other bridges, scenarios outlined in Section 2.8, “Slope Hazard”. with soffit levels higher their RL 12.0 are less vulnerable. However, Ferrymead Bridge is also vulnerable as Slope hazards considered here include inundation by set out in Chapter 12. soil falls and flows and rock falls, and the loss of foundations or formation by mass movement of either Overall, it is expected that most or all of the bridges will the subgrade or underlying natural soil and rock. be able to survive the buoyancy that may result from the ‘worst case’ surge. This assumes that the peak surge The triggering events are either extreme rainfall or level quoted in the “Barnett Report” (see page 34) is serious earthquake, particularly when the earthquake overly conservative. occurs when groundwater levels are high. The approximate return period for both events is 100 years.

Rail system vulnerability Given the limited time available and the aims of the The only potential effect is scouring of the bridge piles project only those locations with current evidence of at the Heathcote River crossing. Bore velocities are not significant hazard have been noted. In any major event expected to be sufficient to cause significant problems. there will be some unpredictable “surprises” and in 192 • Risks and Realities

addition there will be many minor slides, falls and Shag Rock extend over a distance of 500 mto the retaining wall collapses. Clifton Hill turnoff and could generate major rock falls, particularly at the Clifton Hill end. Fortunately, The Slope hazard — transport locality plan (Map 8, p there may be enough room to construct a temporary 291) shows the locations referred to in the following bypass around the beach and car park to give access to text. Sumner. There is no feasible mitigation option to prevent the rock falls. Primary transport routes Evans Pass: The Sumner side of Evans Pass is not as These are defined as routes serving primary destina- bad as the Lyttelton side and with one possible exceptions. The only primary route crossing the hill areas is tion (location 12, Table 6.5) the road formation is likely the port route of Port Hills Road-Tunnel Motorway- to stay intact. Debris and rocks will need to be cleared. Road Tunnel-Lyttelton. However it is highly likely the Lyttelton side will be Table 6.5 (page 195) locations 16 - 18, summarises the much more seriously affected in a number of locations details of three areas of this route potentially at risk. In where the road has been built on very high placed stone turn, these are as follows. “retaining walls”. In addition there are potentially Port Hills Road: Only a short length of road is at risk unstable high rock cliffs in many locations above the from significant debris flows or rock falls in this area. road with some very large blocks of rock which could While there is a chance of both lanes of the road being fall. This section of road may be closed for weeks to blocked by debris, fortunately there is a possible by- months. pass around this area until it is cleared, either through the Old Orchard or around the nearby industrial prop- Hill feeder roads erty. Although this not a category listed elsewhere in this Tunnel Motorway: There is a risk of debris flows and report, after consideration it seems justified to evaluate rock falls off the high slopes above the cut benches and at least one main route up to each of the main hill areas from the cut benches themselves. This extends over for both civil defence and access to the water reser- almost 2 km of road with varying degrees of risk. In voirs, etc. The most important of these, Dyers Pass most cases, the width of the motorway will ensure at Road, is also important for access to the various trans- least one lane will remain passable, but it is possible mitters and Governors Bay. that short stretches will be completely blocked. These should be able to be cleared within hours provided the Dyers Pass Road — Governors Bay appropriate machinery is available. Locations 1 to 8 (Table 6.5) cover this route. The first section of Dyers Pass benefits by having duplicate Road Tunnel, Lyttelton Portal: High cut benches in routes available if problems do develop. Hackthorne loess on the west side of the portal may block part of the Road up to Dyers Pass is free of significant potential roundabout with debris, but the east side should remain problems and could provide alternative access in the passable. lower section of Dyers Pass Road. Above the Sign of In summary, this route may be closed by debris imme- the Takahe the road up through Victoria Park is an diately following the triggering event but it is likely at alternative route for part of the way, although it is least one lane can be reopened within hours of the relatively narrow. event. The road up to the summit is likely to remain passable As explained in the subsection “Hill feeder roads”, in for at least one lane traffic, except for a matter of hours the event of the closure of the tunnel, Gebbies Pass is immediately following the event while debris is cleared. likely to be the best alternative route to the Port. The most significant area where both lanes of the road formation may slide (location 2, Table 6.5) can either be bypassed using the Victoria Park route, or the Major sector distributors carpark adjacent to the area widened slightly to allow Ferrymead - Sumner - Evans Pass has been designated more space around the fill area. as the only major sector distributor in the hill areas. Locations 12 to 15 and location 24 (Table 6.5) record The continuance of the route on over to Governors Bay details of the potentially affected areas. In turn, these is likely to require much longer to reopen because of are as follow. larger scale debris and rock falls. Also, lower down, there is the possibility of losing sections of the road Shag Rock: The very high rock cliffs commencing at formation itself. Transport • 193

Lyttelton — Governors Bay — Gebbies south of the Ferrymead entrance (location 21, Table Pass 6.5). This should be able to be reopened quickly and if This route was not initially considered as important required a temporary road formed in the adjacent until it became clear that Evans Pass, the alternative empty paddock. route to Lyttelton, was likely to suffer major damage. Access to the port over either Dyers Pass or Gebbies Major Hornbrook and Mt Pleasant roads Pass, and then around the edge of the harbour basin, is No major problems for the same reasons as Huntsbury the other alternative route. The section of road from Hill. Lyttelton to Governors Bay is common in this route to both the hill passes and is therefore separated out in the discussion below. Soleares Avenue A steep fill embankment just down from Valencia Place may affect the outer lane and Telecom cables. Lyttelton — Governors Bay Otherwise there are no major problem areas. Soil debris onto the road and loss of the outside road shoulder are likely in many places along this route (Table 6.5, locations 32 to 39). The entire road forma- Moncks Spur road tion was lost in a rainfall triggered slide near Cass Bay Similar to Huntsbury and Mt Pleasant. in 1986, but fortunately a small loop of the old road remained intact nearby so the road remained open over Clifton the months required to repair the problem. The fill at The approach to the climb up to Clifton Hill is subject this location is now much stronger and better drained to potential rock fall from the cliffs which extend round and is unlikely to slide again. Several similar gully from Shag Rock. This could be cleared relatively crossings are nearby and may result in loss of the outer quickly providing no large volume is involved. lane but hopefully a one lane bypass can be created reasonably soon after the event. Potentially more serious is the possibility of a foundation failure in the crib walling commencing approxi- Governors Bay — Gebbies Pass mately 30 m past the first hairpin and extending over 40 m to 50 m (location 29, Table 6.5). In this area the crib This route is reasonable all the way to the end of the wall which supports the road formation is built out over Teddington Flats. Soil debris falling onto the road is a high 50 to 60 degree loess bank which could fail in a likely in a number of places and may be extensive at the strong earthquake. The same section is vulnerable to box cutting near Allandale. At Gebbies Pass itself, both loess slides from the high cut batters above the road lanes of the current seal are at risk and have been (part of this failed in the snow storm in 1992). This slide progressively subsiding at location 45, Table 6.5. For- material could surcharge the road putting further strain tunately the inner road shoulder is quite wide at this on the crib foundations. point and a one lane bypass may be possible. Work will probably be needed soon at this location just to main- Further up the road (location 31, Table 6.5) there may tain the current road under normal conditions. be minor debris falls and slides and a section of road formation crossing a gully between Kinsey Terrace Apart from debris falling onto the road, the route and Tuawera Terrace could result in loss of one lane. generally should be able to be reopened reasonably quickly (i.e. within 24 hours) for at least one lane However, there has been recent reconstruction work of traffic. Clifton Terrace which has markedly improved the situation. Gebbies Pass is thus probably the best alternative route to the port and, compared to either Evans Pass or Dyers Pass, requires the least mitigation to give reasonable Scarborough confidence of future access. The greatest potential risk along this section of road is at the bottom near the lifeboat slipway and before the Huntsbury Hill first hairpin bend. The rock cliff at this point overhangs the road (location 26, Table 6.5). This could fail in a No major problems are likely due to the gentle gradient similar way to the rock fall 300 m south of this point and absence of major cuts. which collapsed the Edwin Mouldey track in 1986. This would probably block both lanes and bypass Bridle Path road options are very limited. However more serious is the One area may be closed by debris and rock fall just threat such a rock fall poses to the road above as it 194 • Risks and Realities

switches back beyond the hairpin (location 29, Table Airport 6.5). The road is actually built out on the overhang. Earthquakes There are likely to be rock falls and soil falls onto the Christchurch International Airport Limited (CIAL) road from the first hairpin up to No 127, after which has undertaken a survey of existing buildings and the point the batters are less steep. status of plant and equipment. A list of mitigation measures has been made. Where necessary designs Rail tunnel have been done and are in the process of being imple- There is a complex interrelationship at the Lyttelton mented. rail portal between the roads and the rail tunnel which passes underneath. This is essentially a structural analy- A study of the impact of seismic activity on the runway sis problem, however there are no signs of road defor- system has been commenced. mation to date and a brief inspection of the first 30 m CIAL has held discussions with Telecom and of tunnel has not revealed any sign of lining failure. Southpower on the security of supply and the status of In contrast, at the Heathcote Portal there are 10 to 15 m equipment. 1 high, /4 to 1, unsupported batters in loess and loess colluvium on both sides of tunnel mouth. These would Flooding (Waimakariri) be vulnerable in a serious earthquake and could block CIAL is undertaking a study of the implications of the tunnel at least for several hours until the debris is Waimakariri flooding looking particularly of the old cleared. flood channels on relation to the airport.

Conclusions Flooding (Local) Slope hazards associated with a major earthquake or CIAL has adopted a policy of designing new buildings rainstorm are likely to have only moderate impact on with local flooding setting design parameters for floor the one primary transport route which crosses the hills levels and siting of critical services. Discussions are (City - Road Tunnel - Lyttelton). Similarly only one taking place with the Christchurch City Council’s major sector distributor, the road to Sumner, is poten- Drainage Unit to look at strategies for stormwater tially affected by slope hazards however impacts along disposal. the Sumner end of this could be severe.

Locally severe impact is likely on some of the hill feeder roads, particularly Evans and Dyers Pass and the Lyttelton Harbour roads. With the exception of Scarborough, and possibly Clifton Hill, the nominated feeder roads for the main residential suburbs should all perform reasonably well.

6.12 Mitigation Measures — Status of Measures

Road network • Bridge strengthening requirements being taken into account in bridge maintenance programme of the Christchurch City Council.

• Rolling programme of low cost earthquake strengthening of key State Highway bridges by Transit New Zealand.

• Recognition of lifeline routes in undergrounding programme of Southpower.

• TNZ strategy for tree maintenance on State High- ways. Transport • 195

VULN'Y IMPACT

ROAD MITIGATION MEASURES COMMENT

RISK RATING

MAP REFERENCE MAP

IMPORTANCE 1 - 5 IMPORTANCE

PERIOD FOLLOWING

IMMEDIATELY AFTER IMMEDIATELY

SLOPE HAZARD TYPE SLOPE HAZARD

EARTHQUAKE HAZARD EARTHQUAKE RETURN TO RETURN TO NORMALITY DURING EARTHQUAKE 1 321111 Retainment Telecom cable affected Dyers Pass 2 43222 1 Retainment Victoria Park alternative route to 3 1,2 3 22 1 1 Planting above road 500 m length Sign of Kiwi 4 1,2 2 221 1 Planting above road 250 m length 5 23 11 1 1 Retainment One lane at risk only

Sign of Kiwi 6 1,2 2221 1 Plantings 820 m in two lengths to 7 1,2 3 3331 Plantings 1000m length Governors Bay 8 3 3 22 1 1 Retainment Major work involved

12 3 2111 1 Retainment One lane at risk only Evans 13 1,2 3 3311 Plantings Pass 14 1,2 3 33 2 1 Limited Plantings may help at Lyttelton end 15 433333 Retainment Major work involved

Port Hills Road 16 1,2 22211 Batter plantings Alternative route available Batter plantings through road tunnel 17 1,2 33311 Existing shrubs too small 18 2 11233 Very limited Partial blockage only

Rail tunnel Heathcote end 19 1 33311 Retainment or Bench 100 m on both sides

Bridle Path Road 20 3 21111 Retainment Telecom also affected 21 1,2 11223 Very limited Debris off private land

Mt Pleasant 22 311111 Retainment One lane at risk only 23 3 11222 Retainment Telecom also affected

Sumner Road 24 233321 Diversion onto beach

25 2 3332 1 Mesh rock bolts Scarborough Road 26 4 33333 Ground anchors Major engineering to improve 27 2 3 2211 Limited

28 22333 1 Limited Clifton Hill 29 4 33333 Ground anchors Crib foundation failure 30 1 3331 1 Retainment Private land? 31 3 222 11 Retainment

32 1 3 22 11 Limited 33 3 222 11 Extensive retainment Outer lane only Lyttelton 34 4 2 3333 Retain and drain to 35 1,3 22211 Limited Governors Bay 36 3 2 22 11 Retainment One lane only affected 37 3 3 22 11 Retainment One lane only affected 38 4 2 3333 Retainment Major engineering 39 1 322 11 Limited 2 - 3 km of unstable loess batters

40 3 3 1111 Bypass available 41 1 333 11 Plantings Governors Bay 42 1 22211 to 43 1 22 2 11 Gebbies Pass 44 1 33311 Limited 45 4 33 11 Retainment drains Bypass probably possible 46 3 22211 Retainment One lane only affected 47 1,2 3 22 11 Limited

Vulnerability Chart: Define components and elements of network at Regional and District level

For each component: Assess importance 1 to 3 — 3 most important Slope hazard type: 1 = soil debris onto road 2 = rock debris onto road 3 = slide of shoulder and one lane 4 = slide of both lanes Risk 1 to 3 — 3 most at risk Assess impact of damage 1 to 3 — 3 most impact

Table 6.5: Slope hazard vulnerability chart (to be read in conjunction with Maps 5 (p288) and 8 (p 291)) 196 • Risks and Realities Emergency Buildings • 197

Chapter 7 Emergency Buildings

7.1 Buildings Investigated Transmitter Masts • Gebbies Pass. Concrete column and wall building Introduction with timber roof trusses, built in 1932. Grossly Buildings associated with the necessary services fol- overdesigned for present style of equipment. lowing a major earthquake or natural disaster cover a • Ouruhia. Tidy reinforced concrete block construc- very wide range. Because of the nature of their use, tion. such buildings have, over the years, been generally well designed and constructed to state-of-the-art of the • Masts. Masts are generally very resilient structures day. and should perform satisfactorily providing there is no foundation failure. Attention is drawn to secur- It is particularly noticeable, however, that the performing equipment and considering face loads, etc. ance attained is very strongly related to the age of the building. The appreciation that the performance, design and construction of structures has advanced con- Lyttelton Container Terminal — Signal Tower siderably over the years, has to be acknowledged. Main building is of substantial construction with reinforced concrete with shear core. Tower adequate. At- It is noted that in most situations consultants are tention drawn to vulnerability of other ancillary build- involved in reporting and investigating on the possible ings, and to securing services and general office equip- vulnerability of the buildings. The state of buildings ment. tend to “speak” for themselves, but the contents are generally overlooked, and this could be the controlling factor at the time of a natural disaster. Police Hereford Street Police Station Buildings investigated Substantial reinforced concrete building with gener- The buildings investigated in relation to an earthquake ous columns and beams with basement. Attention hazard are reported in the following categories. (See drawn to fixing of water storage tanks and general also Tables 7.1 and 7.2.) office equipment.

Broadcasting Hornby Police Station CTV building Sound timber and reinforced concrete building built in 1990. Attention drawn to securing services. Early (1929) reinforced building with frames and walls. Well designed for its specific occupancy. Attention drawn to securing water storage tanks, electrical com- NZ Fire Service (seven stations, workshop ponents and office equipment. and training school) The NZ Fire Service has many fire stations and ancil- Television House lary buildings throughout the greater Christchurch area. They are fortunate that good simple design and Substantial reinforced concrete framed building with construction has been carried out over the years. They basement. Attention drawn to securing general office have full backup operations from other stations. Atten- equipment. tion drawn to securing of battery backup and general office equipment. Radio New Zealand — Kent House Well constructed reinforced concrete core and frame St John Ambulance building built in 1978. Attention drawn to securing St. John Ambulance have recently moved into new services and general office equipment. premises in the central city area. The building is part 198 • Risks and Realities

mid-1960s construction and part late-1980s. Renova- ing services was not undertaken by the Emergency tions and strengthening were being carried to suit the Buildings task group. new occupancy. They have access to two city streets and good visibility for their vehicles. Attention drawn The 1991 CAE project Lifelines in Earthquakes: Wel- to securing general office equipment. lington Case Study reports in detail on building services in commercial, industrial and public buildings.

Suburban bus companies It was felt that the findings for the Wellington report Buildings represent a small percentage of site area. were sufficiently general enough to apply to the Christ- Buses are stored in the open. Maintenance building church study. That being the case, this section draws built 1920s, offices 1960 and 1980. Attention drawn to heavily on the material presented in Section B9 of that fixing of equipment for office and operation use. report, but is specific to emergency or essential buildings, and modified where necessary to suit Christ- church conditions. Pumping Station — Scrottons Road Early unreinforced brick columns and walls building General with timber roof trusses. Attention drawn to checking wall stability to avoid possible damage to electrical Building services can be important lifelines within equipment. buildings and are necessary in most of the emergency or essential buildings to enable the occupants to function effectively. Buildings mitigation measures Service organisations should endeavour to obtain a They should be capable of continued use following structural report on their buildings. This may prove a (and during, for some) an earthquake. Otherwise they formality for the later buildings, but certainly a warn- should be readily repairable within the timeframe ing of hazards for the earlier ones. This report should imposed by the availability or restoration of the life- be endorsed to include restraint recommendations to lines external to the building to allow the essential all fittings, fixtures and other equipment. building services to be used where no in-house emergency services are available. The range of vital serv- Secondary power systems to stand-by equipment should ices can vary with the different emergency buildings. be adequately restrained. The following range of services can be vital for emer- Records should be stored in such a manner that they gency buildings: and their containers do not become dislodged in the event of an earthquake. • electrical;

Essential communication equipment should be ad- • water supply (and plumbing); equately restrained on desks etc., or where operating. • fire protection;

Building adequacy should be reviewed as demands and • heating, ventilating and air conditioning (HVAC); equipment change. This is particularly noticeable with electrical equipment where size diminishes as time • communications; advances. • vertical transportation;

• security; and 7.2 Building Services • special processes. Introduction The focus of this section is on building services for General vulnerability issues emergency buildings or essential buildings. A lot of the In any emergency building, provided it remains structur- comments will apply equally (or more so) to non- ally sound during and following an earthquake, contin- essential commercial, industrial and public buildings ued occupation and use will be dependent upon the outside the scope of this study. continuity of supply of water, electricity, communications services and other essential services. The task group involved in surveying the emergency buildings also viewed and commented on the building Electricity is required to maintain: services, where appropriate. A specific survey of build- • lighting; Emergency Buildings • 199

• ventilation — supply, exhausts and conditioned air • hospitals. for some processes and emergency equipment; For each main category, assessments were made for • water pumping; high, medium and low quality buildings. These assessments are generally quite applicable to the essential • communications — telephone and data; Christchurch buildings.

• boiler operation; For each building category, assessments of vulnerabil- • domestic hot water; ity and impact of loss of each service were made in tabular form. The vulnerability of each service within • freezers and coolrooms — drugs, goods etc.; the building category and the impact of the loss of the service were graded on a points system of l to 10, a • kitchens; score of l indicating low vulnerability or importance • medical services; and a score of 10 indicating high vulnerability or importance. A score of l for vulnerability means the • stormwater and sewage pumping; services would not suffer significant damage and would be available immediately after the earthquake. A score • lifts, particularly for patients or disabled people in of 10 for vulnerability means that the service would hospitals and medical centres; and suffer substantial damage and would take considerable • security. time to reinstated.

Buildings provided with emergency power generators The vulnerability charts were assessed on the basis of would normally be expected to be able to supply, as a an earthquake of 0.9g intensity for the Wellington minimum, the essential services for up to 24 hours. Region. The charts will generally be relevant to the This is sufficient for intermittent installation outages Christchurch study where both the design earthquake but could be insufficient to cope with the longer out- level is less as are the seismic design standards that ages due to major seismic events. apply to services installations.

Water supply would need to be maintained for the It should be acknowledged that the assessments were satisfactory operation of: made subjectively, but with the knowledge of both past and present design and installation standards. The • domestic cold water systems; tables presented can also apply to the various building service systems present in typical commercial, indus- • domestic hot water; trial and public buildings, likely to be found in the • fire hose reels; greater Christchurch area.

• fire sprinklers; In the following schedules, the classification of quality (high, medium and low) relate to the quality of installed • kitchens; and services. This assesses the ability of a public building to withstand a moderate earthquake and remain func- • laboratories. tional, which will enable it to provide a service imme- Buildings with water storage usually can provide 24 diately following the seismic disturbance. hours usage for domestic purposes but not for such In assessing the immediate ability to respond, the services as fire. Without mains water and fire water predominant factors were: storage most fire protection systems would be completely inoperable and ineffective. • the ability of the building to remain intact and structurally sound for use; Although electricity and water are absolutely vital for the continued operation of building services, even • the ability of building services to be able to function basic requirements for occupancy and functioning on in-house power; require restoration of other services. • the ability of furniture and essential portable first- aid equipment to remain intact; Public buildings The Wellington study split public buildings into two • the adequacy of housing of general medical sup- broad categories: plies and their availability for use; and

• government, transport and “emergency” buildings; • the availability of communication facilities either and immediately or at least within two hours. 200 • Risks and Realities

Building classification — public and will be subject to limited use and their HVAC will, emergency buildings most probably, not be functional. No generator for electrical emergency power. Battery essential lighting. High quality Fire protection limited to code with no wet systems. This would most probably only relate to buildings and services installed within the past five years and those Hospitals for which proper seismic engineering was applied to The various departments (in Wellington) have been the building services. During construction the instal- scheduled and reviewed in overall terms, mainly to lations would have been subject to good supervision as measure the impact of the loss. distinct from observation or no review at all. Hospitals by their very nature tend to be a series of Some minor damage will most likely occur. buildings linked by services. As such, some depart- Medical and essential services would be adequately ments are wholly dependent upon central services for housed and have suffered little or no seismic damage their continued operation. and be immediately available. Government HVAC systems will be operational and will be gener- In this area, central government, regional and local ally air types with no major weight items in the ceiling body buildings, Emergency services such as ambu- spaces. lance, police and Civil Defence, and transport centres Electrical systems will be fully functional and power were considered. available to ventilation fans and all essential services from in-house or on-site emergency generators. Fuel Assessment and vulnerability and supplies would be expected to provide at least five days impact of loss supply. Tables 7.1 (for hospitals) and 7.2 (for government, transport and emergency buildings) indicate the results Fire protection will generally be by sprinklers with of the Wellington Group’s analysis. Overall results dual water supplies or on-site storage and will be fully and trends will be very similar for the stock of public functional together with hose reels and alarm systems. and emergency buildings in the greater Christchurch Fire pumps will also most probably have a diesel region. As with previous assessments, reliance has driven standby. Water storage in this grade of building been placed on experience and judgement. should be adequate to provide at least 24 hours usage.

Expected common failures in Medium quality earthquakes This would most probably only relate to buildings The charts reflect the wide and critical vulnerability of completed over the past five to ten years or those for building services, and provide reminders of the com- which some upgrading of services has been completed. mon failures of such systems in earthquakes. These buildings would not have necessarily been sub- In many cases, a service fails due to the failure of an mitted to the same degree of detailed design as high initial item of equipment which comprises only a small quality buildings, have limited essential services and part of that particular service. The general vulnerabil- not have received detailed construction supervision. ity of services is demonstrated by the following review HVAC systems will probably be limited in operation. of failure examples. Electrical systems will be operational, emergency power most probably limited to say lifts, emergency lighting Heating ventilation and air conditioning and fire and sump pumps. systems Fire protection will be operational and where sprin- Inadequately restrained equipment: klers are installed there will be no supply other than city • In ceiling spaces relatively heavy plant such as fan mains. Some damage will have occurred and some coil units able to move sideways, may break away services may be limited or not available. from fixings and fall, damaging ceiling, components and possibly causing injury. Excessive move- Low quality ment may break water or electrical connections Buildings in this category will be those of an age with risk of flooding, fire and loss of function. outside the parameters above, have limited services, • Similar results may occur if piping systems are Emergency Buildings • 201

RISK 1 - 10 1 - 10 RISK 1 - 10 1 - 10 VULNERABILITY IMPACT OF VULNERABILITY IMPACT OF LOSS LOSS BUILDING TYPE H M L H M L BUILDING TYPE H M L H M L DEPARTMENTS MEDICAL Accident & Emergency 2 6 8 10 9 9 Bulk O2 2 2 - 9 9 9 Ward Blocks 1 5 7 9 8 7 Bottle O2 1 2 2 10 10 10 Theatre Suites 2 6 8 10 9 9 Med Air 2 2 3 8 8 8 X-Ray Department 2 6 8 9 8 7 Med Vacuum 1 2 3 8 88 Fracture Departments 2 5 7 9 7 6 N2O Bulk 2 2 - 7 7 7 Laboratories 2 5 7 9 6 5 N2O Bulk 1 2 2 8 8 8 CSSD 2 5 7 8 8 8 Drug Coolers 1 2 2 9 9 9 Kitchen 2 5 7 7 7 7 X-Ray Equipment 2 3 3 888 Laundry 2 6 7 666 Theatre Lights 1 3 2 10 10 10 Boiler House 2 7 8 7 7 7 Anaesth. Equipment 1 2 2 10 10 10 Bulk Medical Gases 2 7 8 9 8 8 Equipment Racks 1 2 2 9 8 8 Central Power 2 679 8 7 FIRE Mains Water SYSTEMS 2 2 2 3 7 7 HVAC 1 4 5 8 67 Water Storage 1 - - 5 - - Electrical 2 3 5 9 9 9 Mains Power 2 3 6 1 22 Medical 2 5 6 9 9 9 Sprinkler Pipework 2 2 - 8 7 7 Fire 2 4 5 9 9 9 Hose Reels 2 2 2 7 66 Plumbing 2 4 7 9 9 9 Extinguishers 2 2 2 6 6 6 Lifts 2 5 8 8 8 7 Alarms 2 2 2 7 7 7 Life Support 2 6 7 9 9 9 Fuel Supply 2 4 - 7 7 8 Communications 1 3 4 8 8 8 PLUMBING

ELEMENTS Water Storage 1 5 8 10 88 HVAC Tanks 2 3 4 9 99 Vent Fans 1 2 4 5 5 4 Mains Water 3 4 4 6 7 7 A/C Fans 1 2 4 5 4 3 Pipework 1 2 4 8 7 7 Pumps 1 2 3 8 8 7 Water Pumps 2 3 3 9 8 8 Boilers 1 2 4 7 7 7 Sewage Pumps 1 2 2 10 10 10 Chillers 1 2 4 7 7 6 S. Water Pumps 1 2 2 9 99 Piping 2 2 4 8 8 8 H.W. Storage 1 2 2 9 9 9 Ductwork 1 2 4 8 7 4 Waste and Vents 1 2 2 8 7 7 Fuel Systems 2 3 5 8 7 7 Cooling Towers 35 7 8 7 6 LIFTS Tanks 1 3 6 7 6 6 Passenger 1 3 5 6 6 6 Mains Power 2 2 2 4 7 8 Patient 1 3 5 8 8 8 Essential Power 1 2 3 2 3 4 Goods 1 3 5 4 4 4 MCCs 1 3 5 2 3 4 Mains Power 2 2 2 9 9 9 Cool Rooms 2 2 4 2 2 2 Essential Power 1 1 - 9 99 Freezers 2 2 4 222 Seismic Switch 1 5 8 - - - Current Code 1 5 8 - - - ELECTRICAL Mains 2 2 2 4 6 9 COMMUNICATIONS Transformers 2 2 2 4 6 9 PABX 2 3 5 8 8 8 MSB 2 2 2 10 8 7 Computer 2 3 5 6 6 6 Generator 1 2 8 10 10 10 Radio 1 2 2 5 5 5 Gen Fuel 1 2 8 10 10 10 TV 1 2 2 4 4 4 Batteries 1 2 8 10 10 10 Paging 1 2 2 7 7 7 DBs 1 2 8 9 9 8 Nurse Call 1 2 2 7 7 7 Light Fittings 1 2 2 6 6 6 Security 2 4 6 8 88 MSB Location 1 2 5 ---

Table 7.1: Hospital risk assessment (from the Wellington Lifelines Study) 202 • Risks and Realities

RISK 1 - 10 1 - 10 RISK 1 - 10 1 - 10 VULNERABILITY IMPACT OF VULNERABILITY IMPACT OF LOSS LOSS BUILDING TYPE H M L H M L BUILDING TYPE H M L H M L BUILDINGS FIRE Central Government 2 4 7 8 8 8 Mains Water 2 2 2 3 7 7 Local Government 2 4 7 8 88 Water storage 1 - --5 - Regional Councils 2 4 7 10 9 9 Mains Power 2 3 6 1 2 2 Civil Defence 1 3 5 10 9 9 Sprinkler Pipework 2 2 - 8 7 7 Airports 2 4 7 7 7 7 Hose Reels 2 2 2 7 6 6 Rail Stations 3 6 8 8 8 8 Extinguishers 2 2 2 6 6 6 Fire Stations 2 6 7 10 10 10 Alarms 2 2 2 7 7 7 Ambulance 2 5 7 10 10 10 Fuel Supply 2 4 - 7 7 - Police 2 4 7 9 9 9 Military 2 4 7 9 9 9 PLUMBING Water Storage 1 5 8 10 9 9 SYSTEMS Tanks 2 3 4 7 7 7 HVAC 2 3 4 5 6 6 Mains Water 3 4 4 6 6 6 Electrical 2 2 2 7 7 7 Pipework 1 2 4 9 8 7 Fire 2 3 3 6 6 6 Water Pumps 2 3 3 7 7 7 Plumbing 2 3 4 6 6 6 Sewage Pumps 1 2 2 10 10 10 Lifts 2 3 3 8 8 8 S. Water Pumps 1 2 2 9 9 9 Communications 1 1 2 2 2 2 H.W. Storage 1 2 2 6 6 6 Waste and Vents 1 2 2 8 8 8 ELEMENTS HVAC LIFTS Vent Fans 1 2 3 5 5 4 Passenger 1 3 5 6 6 6 A/C Fans 1 2 3 6 5 3 Goods 1 3 5 4 4 4 Pumps 1 2 2 8 6 7 Mains Power 2 2 2 6 6 6 Boilers 1 1 1 7 7 7 Essential Power 1 1 - 6 - - Chillers 2 2 - 7 7 7 Seismic Code 15- - - - Piping 1 2 3 6 6 6 Ductwork 2 3 3 7 8 8 COMMUNICATIONS Fuel System 1 2 3 8 7 6 PABX 2 3 5 9 9 9 Cooling Towers 2 3 4 7 7 7 Computer 2 3 5 9 9 9 Tanks 2 3 4 7 6 6 Radio 1 2 2 8 8 8 Mains Power 2 2 2 2 3 4 TV 1 2 2 4 4 4 Essential Power 1 2 - 2 2 2 Security 2 2 2 9 9 9 MCCs 1 1 1 2 2 2 FUEL SERVICES ELECTRICAL Fuel Oil 2 3 5 9 9 9 Mains 2 2 2 4 6 8 Petrol 2 3 5 9 9 9 Transformers 2 3 3 4 6 8 Aviation 2 3 5 9 9 9 MSB 1 2 2 9 9 9 Fuel Pumps 1 2 3 9 9 9 Generator 1 2 - 10 9 9 Gen Fuel 1 2 - 10 9 9 Batteries 1 2 3 10 8 8 DBs 1 2 8 9 9 8 Light Fittings 1 2 2 6 6 6 MSB Location 1 2 5 - - -

Table 7.2: Government, transport and emergency buildings risk assessment (from the Wellington Lifelines Study) Emergency Buildings • 203

poorly anchored to the building, allowing move- • Hot water cylinders toppling over. ment relative to secured equipment. Inadequate allowance for movement: • Movement of main plant items such as boilers, chillers, cooling towers and tanks may break water • Failure of water main at underground entry. and electrical connections. • Failure of sanitary plumbing or drains causing Inadequate allowance for movement: flooding, health risk.

• Failure of pipework connections at building entry Lifts points, between buildings or across seismic gaps. • Lift shaft damage, rail misalignment. • Failure of ductwork connections across seismic • Movement of hoist machines. gaps in the building. • Damage to controls. Electrical • Flooding of lift well and pit. Inadequately restrained equipment: • Jamming of landing or car doors. • Generator starting batteries moving and breaking terminals and connections. Refrigeration • Main switchboards moving sideways or toppling • Movement of unrestrained evaporators causing the over breaking connections. refrigeration pipework to shear at the point of connection to the unit. • Light fittings falling out of ceiling fixings and breaking connections with a risk of injury and fire. • Restrained and unrestrained evaporators, fixings pulling through insulating sandwich panels to which • Flooding of below-ground substations and genera- they are fastened. tor rooms.

Communications systems Fire protection • Loss of unrestrained items of equipment such as Inadequately restrained equipment: power supplies and electronic equipment not fixed • Movement of battery boxes and alarm panels break- in restrained racks, e.g. PABX’s, amplifiers. ing connections with loss of detector circuits and • Flooding of basement located PABX rooms. brigade connection • Microwave dishes and satellite dishes out of align- • Loss of power or fuel supply to sprinkler or wet ment. riser booster pumps.

• Failure of water storage tank or connections. Security systems Inadequate allowance for movement: • Jamming of locks due to doors and frames racking.

• Pipework rigidly restrained too close to sections of • Personal computers falling off desks. building structure designed for differential shear- • Power supplies falling over. ing movement, causing flooding and loss of function. Fittings and fixtures • Failure of water supply at building entry. • Free-standing, unrestrained shelving overturning or racking. Plumbing and drainage • Personal computers, facsimile machines, word- Inadequately restrained equipment: processors and other loose desk-top items falling to • Water storage tanks moving, impacting on adjacent the floor. structural members or collapsing. • Magnetic as well as weaker mechanical cupboard • Pipework connections shearing causing flooding door locks opening and spilling their contents. or loss of water. • Freestanding uplighters overturning. 204 • Risks and Realities

• Large pot plants overturning. for the substitution of failed lifelines such as:

— power generation by diesel generators; General conclusions It was found that significant services damage is un- — uninterruptible power supplies; likely in small- to medium-strength earthquakes. Al- — 24 hours water storage; and though some components may fail, the overall effect will not be serious. — emergency battery or battery inverters supplying lighting. Significant damage, however, may occur to services where they: • Provision of systems to augment weak lifelines such as: • pass between buildings; — power conditioning equipment to smooth • pass across seismic gaps, at junctions between out voltage dips and spikes; different structural systems; and — water pumps to boost weak water supply • are attached to unrestrained plant items where the lateral support has been inadvertently left off. pressures; and

It should be remembered that some services may be — cartridge water filters to clean dirty water disrupted temporarily because of the correct action of supplies and chemicals to manually/au- an in-built safety device. This applies particularly to all tomatically disinfect water supplies. forms of vertical transportation, for which it may take • Provisions of alternative lifelines such as: up to three days to reset safety devices. — Dual water supplies where twin independ- Even minor repairs are expected to take a significant ent supplies are available, e.g.: time due to the low availability of the appropriate skilled manpower and materials. — two independent town mains;

Significant services damage is unlikely in an earth- — town mains plus well; quake of moderate intensity provided the building and its services are designed in accordance with good — town main plus on-site water storage; current practice. Some components may fail, but the and effect is not expected to be serious. — dual fired heating hot water boilers (electricity and oil). Mitigation measures — Alternative communication links, e.g.: Detailed mitigation measures are as wide and varied as the systems and their elements. The following items — Telecom wire and fibre optic, together provide a general outline of important aspects which, with microwave and possibly UHF when consistently addressed, will significantly reduce earthquake risk. Many of the items do not involve large — Alternative supplies where possible, e.g.: capital expenditure. They should be attended to by suitably qualified persons on new projects in the design — 11 kV and 400 V, or phase, and on existing buildings at the earliest possible — two x 11 kV (feed from different reticu- time. lation systems or feed from a ringmain)

• Seismic restraint of plant and equipment including — Provision for the connection of mobile units, housekeeping items, storage shelves and appli- e.g.: ances in accordance with NZS4219. — generators; • Provision of flexible elements in services passing across structural and architectural discontinuities, — sewage/waste disposal tankers; and e.g. seismic joints and foundation beams. — water supply tankers/milk tankers. • Location of fluid-filled tanks and containers away from items of electrical and communications equip- • Ensure that all expendable items have replacement. ments available and storage tank levels are maintained. • Provision of on-site services connections to allow Emergency Buildings • 205

• Planned preventative maintenance contracts to * Because of the projected shortage of skilled tradesmen available maintain equipment, energy and water supplies, immediately after the event, building owners or operators of both incoming and storage. buildings that are essential to the recovery phase such as hospitals, energy suppliers, local authorities, etc., should look to • Contractual arrangements that are legally binding arranging contracts with local contractors and contractors out- with the appropriate service contractors for early side the areas so that they have first call on those contractor’s response to assist in rehabilitation after the event.* resources immediately after the event. These contracts, once • Provision and safe storage of essential spares for all established, should be reviewed annually to ensure the con- critical items of equipment, and ensuring these are tracted parties are readily equipped and can react in the required accessible. time. This enables the contracted parties to be forewarned so that the individuals who will respond are aware of their obligations • Compilation of up-to-date as-built documentation and can make alternative arrangements for their immediate on site for all services and operation and mainte- family during their expected absence. nance manuals.

Central city buildings, with Victoria Square in the middle left 206 • Risks and Realities New Zealand Fire Service • 207

Chapter 8 New Zealand Fire Service

Rather than investigate the effect on the capability of able for duty in the event of a major disaster hitting the the Fire Service on a hazard-by-hazard basis, as was the city are shown in Table 8.2. practice adopted for the other utilities, a different and more appropriate method has been adopted here. This is because of the unique nature of the Fire Service with On-duty In reserve its mobile provision of service and lack of dependence Permanent 40 125 on fixed buildings. Volunteers (ChCh) 45 75 85 200

8.1 Scenario Table 8.2: NZFS on-duty and reserve staff numbers Scenario The adopted scenario is shown in Table 8.1. The Task forces scenario quantifies the resources likely to be available after an earthquake. This compilation was done after There are three task forces (north, south and west) consideration of the reports of the other utilities’ likely available, which each contain 50 staff members plus performance as a result of the earthquake. additional support staff. This number usually indicates 16 2x2 wheeled units with 60 staff as each Task Force minimum. Task Forces can be broken down into 10 New Zealand Fire Service — staff single units operating under a team leader (officer) and numbers three staff, but they should not be further reduced. Staff members (both permanent and volunteer) avail-

Loss of stations 4 (8 remain) Loss of units 3 (30 remain) Injuries - treated on site Minor only Loss of power to substations 70% Loss of water - reticulated 40% (shutting down valves) Loss of communications Intermittent Staff immediately available 40 permanent; 45 volunteers Staff — first recall 40 permanent; 25 volunteers Staff — full recall 85 permanent; 50 volunteers Roading 10% impassable Bridges 15% impassable Collapsed buildings 30 identified Persons trapped in buildings (60) 20% of damaged buildings Hazardous substances 3 events (minor) Fires 40 fire calls; 60+ assistance in first hour Run out of fuel 4 hours; most units NB 1: See Table 8.4 “Scenario based on the work of the other task groups” NB 2: Should the event occur in darkness, increase losses of stations, units and injuries, as well as times to gather reconnaissance, intelligence and reporting times of additional staff.

Table 8.1: Scenario adopted for NZFS 208 • Risks and Realities

8.2 New Zealand Fire Service Six key outputs Initial Reaction Plan • extinguish; • rescue; Priorities — first life, then property • water; Priorities under the New Zealand Fire Service Initial Reaction Plan are to protecting life first, then property. • containment; Priorities are to: • co-ordination; and 1 restore 111 and turnout lines to New Zealand Fire Service; • consolidate.

2 gather status intelligence — local and area;

3 prioritise requests for responses to all emergencies; 8.3 Hazardous Substances Scenario Problems 4 match against resources (primary) and build-up (10 hours); As the majority of bulk chemicals are in industrial estates and these are, in the main, not labour intensive, 5 reinforce those in field, man relief and spare units; then evacuation of plant is the primary role, followed 6 restore stations for welfare, shelter, replenishment by containment by plant staff/or NZ Fire Service. No of operational and social needs; clean-up or removal is envisaged for some considerable time following an event unless reactions takes 7 power up all New Zealand Fire Service facilities place. (batteries, generators, LPG); As incidents of fire and collapse are serviced, some 8 co-ordinate with other agencies (e.g. police, ambu- elements of hazardous materials may be present. These lance, Civil Defence) on water, tankers, transport, will be dealt with depending on their severity, and their rescue units, etc; effects on the general public and services, systemati- cally. 9 ensure overview of all New Zealand Fire Service operations within 50 miles of affected area; Presently, the NZ Fire Service has one dedicated “Hazmat” unit capable of decontaminating NZ Fire 10 maintain Task Force welfare and changeover (regu- Service staff and a small number of civilians. The larly); Health authorities should bear in mind the possible 11 produce situation reports of all requirements and needs of larger groups for decontamination and should activities to Fire Command and Civil Defence; and gear up accordingly.

12 review priorities 1 to 11 on a regular basis.

Inputs and Outputs There are six key inputs and six key outputs in the New Zealand Fire Service Initial Reaction Plan as listed below. In all cases, it should be remembered that darkness doubles difficulties.

Six key inputs • communications;

• power;

• shelter;

• relief;

• supplies; and

• fuel. New Zealand Fire Service • 209

Functions and events Ramifications Mitigation disrupting NZ Fire Service operation directly after event 1 Family commitments - On duty - Want to return; check status Welfare plan - implemented June - Off duty - Delay in response while making 94 safe Local station assembly area 2 Tankers availability - Local authority tankers on Ensure TA aware of our station requirements (List of all tankers within 50 km of Christchurch attached as Table 8.6) 3 Overland mains - High demand Hold until prioritised. All reserve (1.75 km - 4 pumps) - Fires pumps and “A” and “B” type with spare hose - Portable supply 4 State of Fire Service - Doors jammed Procedure “Earthquake” drive buildings through or remove from framing - Loose fittings 5 Communication - Transmitter out Switching; simplex; cellular. - main channels - Realigned Priority restoration, Mt Pleasant 6 Rescue role - Overload of requests Split resources; unite with Red Cross and trained CD groups, - priority possibly allocate Officer or Senior Firefighter to semi-trained groups of volunteers 7 Hazardous materials - Mixed cocktails High risk plans, cordon, blocked - Fire drains Stabilise; leave signage. Only able to decontaminate own - Large numbers to decontaminate staff and small number of general public 8 LPG bulk tanks and - Leaks See list of locations of all units fixed installations - Fires more than 4,0001tr (Table 8.7); auto shut-off valves on most of - Pipelines excess flow valves. Portable cylinders on heating units would create numerous fires if operating during predicted earthquake due to pilot lights and lack of auto excess flow valves. Ruptures of fixed piping would result in entire contents of cylinders leaking into structures, which in turn may be ignited by naked flames or electrically, causing explosion and fire 9 Assembly - Task Force required Initial base HQ, if available or - staging areas - Activate Training and relief Woolston Training Centre pumps (Workshops) 10 Use of media for - Congestion of airways Pre-recorded tapes to radio recall of staff - priority CD stations; part recall or all staff

Table 8.3: Post-earthquake Task Group 210 • Risks and Realities

Functions and events Ramifications Mitigation disrupting NZ Fire Service operation directly after event 11 Feeding and shelter - NZ Fire Service self-sufficient Use of houses adjoining FS 24 hours but after that CD properties required - Harewood - Russley Hotel/ Commodore Hotel; - St Albans - Redwood Hotel - Wigram/Sockburn - Blenheim Road Hotel - Addington - Cashmere Club - Woolston - Ferrymead Tavern - Headquarters - Christchurch City, Travel Lodge, etc. Various volunteer stations on outer ring of City 12 Toilets Use fire stations after reco. Rubbish bins survey Plastic sheeting Screens 13 4WD vehicle - Marginal 3 units Have understanding with Defence - Only 2WD on all other units FS to send 4 vehicles for joint operations. - Police have number of 4WD in rural division . - Territorial authorities = 10 (4x4 tankers) 14 Reconnaissance - Require urgent overview of each Pre-plan route cards of each zone zone to key premises. These route cards are being researched prior to preparation 1994/95 year and will include primary service buildings. High occupancy an special identified risks premises, i.e. power substations, hospitals and chemical plants 15 111 system - Overload several hours Request Telecom to transfer all calls to Dunedin or Wellington Fire HQ's. If able to contact, or to Christchurch Police HQ's for Liaison Officer Fire to re- despatch 16 Resources (local) - Demand on these from several Priority on hire pumps, agencies tarpaulins, and heavy lifling gear. May need suppiy of empty sandbags 17 24 hour rostering - Overload on manpower after 6 Use buses, vans or taxis to tum hours around crews. Task forces may need shuttle type unit to change over 18 Aerial reconnaissance - With breakdown of normal Endeavour to have one unit communications, aerial exclusive Fire Service use (in observations imperative CD plan) Rotary wing preferred (fixed wing too fast)

Table 8.3: Post-earthquake Task Group (continued) New Zealand Fire Service • 211

Functions and events Ramifications Mitigation disrupting NZ Fire Service operation directly after event

19 Electricity - High demand for electricity Portable generators or - Computers down; dedicated substations; supply required - Fixed generator HQ's station; - Police have generators (Central, New Brighton, Hornby and Papanui) 20 Signage - Various indicators in use for Universal type of signage needed damaged, dangerous, trapped for all agencies. Need to develop etc this inter-agency wide (same format) 21 Inter-agencies - Loss of power All need urgency by Fire Service - Loss of communications Liaison Officer at CD HQ's; Monitoring, requesting and - Restricted roading updating Fire Control wherever - Loss of water it may be based

Table 8.3: Post-earthquake Task Group (continued)

Christchurch Headquarters Station 212 • Risks and Realities

Event/ Recovery Provision Provision Comment Function of basic of 50% of full Mitigation service or service service control

Loss of 30% 2 days 4 days 1 month Based on priority stations (3 stations allowances untenable)

Loss of 10% fire 1 hour 2 hours 1 day Use of reserve fleet and units Training units

Loss of water 3 hours 3 days Several weeks Use tankers from volunteer stations Power to stations 2 hours 24 hours 5 days Use generators to best advantage. Merge some stations and units

Loss of Comms

[VHF 4 hours 1 day 4 days ]Switching links (other [Phones 6 hours 1 day 6 days ]repeater sites). Use [Cell 4 hours 1 day 2 days ]simplex, unit to unit only. ]Request priority (111 ]exchange). Hire extras; ]high demand for shrinking ]resource Recall staff 2 hours 8 hours 1 day Depends on contact. Accountability or send around pick-up unit Loss of sewage 24 hours 3 - 4 days 1 month Use of equipment on stations Loss or overload 2 hours 1 day 2 days (Estimate 100 calls). Use 111 system Police/Wellington or Dunedin Roading 3 hours 1 day 3 weeks Re-route after reconnaissance Bridges 3 hours 2 days 3 months Re-route after reconnaissance Collapse of 24 hours 10 days Several Pre-planned route check to buildings months identify Fuel loss 10 hours 24/48 hours 2 weeks Bring in mobile transfers to service fleet from central depot

Table 8.4: Scenario based on the work of the other taskgroups New Zealand Fire Service • 213

1 Access Roads to Centre City 6 Inter-Agency Responses • Marginal • Need to use inter-agency responses co- ordinated through HQ's Station and Civil Defence HQ's 2 Bridge Access South 7 Dedicated Resources • Critical • Dedicated resources such as tankers, road dozers and 4WD units will require forrning and • Bridge Access North, West operating as a mobile response group to the larger incidents • Marginal 3 To Hospitals 8 Identification • Delays predicted • Identification of severely affected areas and the planning of resources to combat and stabilise situations by primary reconnaissance and tasking must have high priority 4 Water Supplies 9 Time • Severely restricted • Time and the advent of darkness add depth to • Need for use of river and land mains the above scenario and will need to be pre- • Urgent actuations of stop valves to conserve planned as an important consideration; water in affected areas 5 Communication 10 Training in Triage • Intermittent and simplex (line of site, may be • Training in triage for fireslcollapses/rescues. only initial system available) Water supplies need evaluation, showing team "skills" levels.

Table 8.5: Indicated problems For NZ Fire Service one hour into event 214 • Risks and Realities

Category B

Dedicated Rural Fire Engine Carter Holt Harvey Forest Ltd 5 units 4x4 or 4x2 1800 ltr+ 4x4 Selwyn Plantation Board 2 units 4x4 Department of Conservation 1 unit 4x4 Christchurch City Council 1 unit (Southpower Team) Canterbury Regional Council 1 unit Waimakariri District Council 1 unit 4x4 Total within 50 km 10x(4x4) + 1x(2x4) 11 Reserve Ashburton District Council 1 unit 4x4 Hurunui District Council 1 unit Total Reserve 02

Category C

Dedicated Rural Fire Engine Waimakariri District Council 1 unit 4x2 (ex NZ Fire Service) Selwyn District Council 2 unit up to 1800 ltr, 100 HP+ Canterbury Regional Council 1 unit HP/HV pump Christchurch City Council 1 unit (Southpower Team) Total 05 Reserve Ashburton District Council 04

Category E

Dedicated Water Carrier or Tanker Christchurch City Council 5 units Tanker Selwyn District Council 5 units 4x4 or 4x2, 3000 ltr+, 100 HP+ Waimakariri District Council 2 unit LP/HV pump Selwyn Plantation Board 1 unit Banks Peninsula District Council 3 units Total 16 Reserve Hurunui District Council 3 units Ashburton District Council 10 units Total Reserve 13

Category F

Dedicated Water Carrier or Trailer Christchurch City Council 1 unit 6x4, 8000 ltr+, 160 HP+ Selwyn District Council 2 units LP/HV pump Waimakariri District Council 1 unit Ashburton County Council 1 unit Total 05

Table 8.6: Fire units/water carriers (not part of NZFS Service Fleet) New Zealand Fire Service • 215

Category G

Slip on or Trailer Tanker Christchurch City Council 2 units 4500 ltr+ Waimakariri District Council 4 units HP/LV pump Carter Holt Harvey Forests 3 units Ashburton County Council 1 unit Total 10

Category H

Dedicated Smokechaser Waimakariri District Council 2 units 4x4, 280 ltr+ Ashburton 3 units HP/LV (diaphragm pump) Christchurch 3 units Banks Peninsula 1 unit Total 09

Category I

Slip on Smokechaser Tank & Pump Selwyn District Council 1 unit 200 ltr+ Waimakariri District Council 2 units LP/HV pump (diaphragm pump) Selwyn Plantation Board 1 unit Carter Holt Harvey Forests 4 units Total 08 Reserve Hurunui District Council 01 TOTALS Fire Units/Water Tankers in 64 Local Bodies and Territorial Authorities 20 Reserves

Table 8.6: Fire units/water carriers (not part of NZFS Service Fleet) (continued) 216 • Risks and Realities

ORGANISATION LOCATION TANK SIZE Adams Print - Div. of PRF NZ Ltd 234 Annex Road 4,300 Canterbury Carton Co Ltd 122 Antigua Street 4,315 Peterson Chemicals Ltd 122 Bamford Street 7,870 Mobil Bealey 268-270 BealeyAvenue 5,000 Bascands Ltd 30 Birmingham Drive 16,080 Firestone Retreading & Warehouse 51 Birmingham Drive 7,440 Europa Sales Yards Service Station 63 Blenheim Road 7,500 Caltex Blenheim Road 149 Blenheim Road 7,435 Wrightson Wool Centre Ltd 503 Blenheim Road 4,300 Blighs Road Service Station Ltd Blighs Road 4,500 Borden Filmpac (ChCh) NZ Ltd 74 Branston Street 15,000 Firth Industries Ltd 7 Broughs Road 9,300 Shell Brougham 495 Brougham Street 7,440 Toyota NZ (ChCh) Ltd 81 Burchanans Road 48,000 Burwood Hospital 300 Burwood Road 4,740 Christchurch Gas Ltd 95 Byron Street 40,000 Caledonian Hotel 101 Caledonian Road 5,000 Mainguard Packaging Ltd 66 Carmen Road 6,700 Princess Margaret Hospital Cashmere Road 4,740 Shell NZ Ltd Chapmans Road 82,500 Shell Oil NZ Ltd 69 Chapmans Road 15,000 Allan Autogas Ltd Cnr Clothier & Essex Streets 4,325 Shell Cashmere 25 Colombo Street 7,440 C E Boon Ltd 221 Colombo Street 15,000 Christchurch Women's Hospital 885 Colombo Street 4,740 Caltex Cranford Street 500 Cranford Street 7,000 Shell Curletts Road Cnr Curletts & Blenheim Roads 7,480 The Chateau on the Park 187-189 Deans Avenue 4,000 Air New Zealand - Technical HQ's Dury Road, Christchurch Airport 7,500 BP Edgeware 71 Edgeware Road 4,500 BOC Gases NZ Ltd 21-27 Epsom Road 150,000 Country Fare ChCh Ltd 38 Essex Street 48,500 Mobil Aldwins Road Ltd 375 Ferry Road 7,500 Shell Ferry Road 417 Ferry Road 7,300 Radley Motors 619 Ferry Road 5,000 Estuary Energy Centre Ltd 1105 Ferry Road 4,876 National Can NZ Ltd 45-49 Fitzgerald Avenue 48,500 Dallington Service Station Ltd 712 Gloucester Street 7,499 MM Cables NZ Ltd 650-652 Halswell Junction Road 30,000 BP Oaklands 246 Halswell Road 7,499 Nicholl Bros (Halswell Garage) 345 Halswell Road 4,876 Cardwell Motor Services Ltd 210 Harewood Road 7,000 New Brighton Service Station 40 Hawke Street 7,500 Central Police Station 48 Hereford Street 18,200 Ansett NZ Ltd - Maintenance Ivan Crescent, Christchurch Airport 7,500 Caltex Lincoln Road 55 Lincoln Road 4,490 Parkroyal (ChCh) Ltd Crn Kilmore & Durhams Streets 18,100

Table 8.7: LPG storage tanks (over 4000 litres) as of August 1994 New Zealand Fire Service • 217

ORGANISATION LOCATION TANK SIZE Shell Raceway 250 Lincoln Road 7,499 BP Gainsborough 457 Linwood Avenue 7,499 CWF Hamilton & Co Ltd 20 Lunns Road 28,600 Boral Gerrard Springs Ltd 25 Lunns Road 4,454 Metropolitan Service Station 355-359 Madras Street 7,499 Regents Park Main North Road 4,990 Belfast Service Centre 752 Main North Road 4,500 Fulton Hogan Canterbury Ltd Main South Road, Hornby 28,600 BP Sockburn 222 Main South Road 4,300 Mobil Wigram 243 Main South Road 7,500 Crowes Service Station Ltd 720 Main South Road 8,000 Mayell Foods Ltd 789 Main South Road 4,800 BP Marshlands 432 Marshlands Road 7,499 Fendalton Service Station 1 Memorial Avenue 7,499 Craddocks Service Centre Ltd 546 Memorial Avenue 4,800 Shell Moorhouse Avenue 40 Moorhouse Avenue 4,925 Big Fresh Food Co 347 Moorhouse Avenue 4,876 Christchurch Polytechnic 369 Moorhouse Avenue 4,995 Christchurch Polytechnic 369 MoorhouseAvenue 4,996 Lancaster Park Motor Co Ltd 511 Moorhouse Avenue 28,650 Opawa Garage Ltd 11 Opawa Road 4,315 Winstone Wallboards Ltd 219 Opawa Road 91,200 Stadium Auto Centre Ltd 165 Pages Road 4,876 Avon City Motors Ltd 461 Papanui Road 7,499 Arctic Coldstorage Ltd 58 Parkhouse Road 7,440 Tiffany Foods Ltd 10 Print Place 4,995 BP Riccarton Road 41 Riccarton Road 8,000 Caltex Riccarton Road 64 Riccarton Road 7,500 Bush Inn Hotel - Cobb & Co 340 Riccarton Road 4,876 Russley Hotel 75 Roydvale Avenue 6,700 Harewood Truck Stop Ltd 527 Sawyers Arms Road 48,500 NZ Fibre Glass Ltd Shands Road 43,165 Wattie Frozen Foods Ltd Shands Road 4,136 Branston Auto Services 25 Shands Road 4,876 BP Hoon Hay 69 Sparks Road 4,500 Canterbury Health Ltd 21-45 St Asaph Street 7,440 BP QEII 308 Travis Road 7,499 Flexipac - Borden NZ Ltd 66 Treffers Road 7,000 Medical Waste (Canterbury) Ltd Wairakei Road 4,995 US Naval Antarctic Support Unit Christchurch Airport: Public Works 7,435 Burnside Motors Ltd 449 Wairakei Road 7,499 Norths Bakery Ltd 584 Wairakei Road 28,650 Mobil Wainoni 175 Wainoni Road 7,375 Gasson Motors Ltd 196 Waltham Road 7,480 Waltham Road Services Ltd 229 Waltham Road 5,000 Yaldhurst Motors Ltd Main West Coast Road, RD6 4,740 Port-A-Gas Ltd 119 Wrights Road 30,000 Coro Trading Ltd 150 Yaldhurst Road 4,490 Table 8.7: LPG storage tanks (over 4000 litres) as of August 1994 (continued) 218 • Risks and Realities New Zealand/Los Angeles Workshop • 219

Chapter 9 New Zealand / Los Angeles Workshop

9.1 Lessons from the The following personnel from New Zealand attended the Northridge and Loma Prieta workshop: Earthquakes Christchurch Engineering Lifelines Group: — David Bell; Introduction In August 1994, 7 months following the Northridge — Brian Hasell; earthquake, a team from Wellington and Christchurch — John Lamb; Lifelines groups went to Los Angeles for a joint NZUS Lifeline Workshop. — John Lumsden; and

— Allan Watson. The Earthquake The magnitude 6.8 (surface wave magnitude) Northridge Wellington Earthquake Lifelines Group; earthquake struck at 4.31 am on 17 January, 1994. The — Nick Coad; epicentre was located in the community of Northridge in the San Fernando Valley, approximately thirty kilome- — David Hopkins; tres northwest of downtown Los Angeles. The earthquake had a focal depth of 20 kilometres. — Rachael Hughes; — Peter Leslie; The Joint NZ/US Lifeline Workshop — John Norton; and The Northridge earthquake was of considerable interest to New Zealand engineers and lifelines operators due to — Bill Smith. similarities in both seismicity and infrastructure, facilities and other buildings. Note: The 1994 Wellington Earthquake Lifelines group Report contained a large section dealing with Northridge The Centre for Advanced Engineering recognised the investigation to which some of the Christchurch mem- level of interest in the field of lifelines, and promoted the bers contributed. This chapter, which has a Christchurch idea of a joint New Zealand and United States lifelines perspective, is written by Christchurch members of the workshop involving representatives of the Wellington team, and includes comments on a visit to San Francisco Earthquake Lifelines Group and Christchurch Lifelines to investigate the effects of the Loma Prieta earthquake. Group. This workshop was held in the offices of the This information was presented at the 1994 Christchurch Southern California Gas Company on 15 and 16 August Workshop. 1994. The workshop was organised from New Zealand, with the assistance of the Southern California Gas Com- If there was a common theme to the Workshop it would pany. be the value of co-operation, recognition of interdependency, and the benefits of prior planning. The purpose of the workshop was as follows: The fact that 11 people from New Zealand Lifeline • to identify the current approach to lifeline earthquake Organisations and Consultancies attended this Work- preparedness in both countries; and shop is indicative of the co-operation that exists within New Zealand to reduce the impact of earthquakes on • to identify outcomes from the Northridge earthquake engineering lifelines. The representation from Los An- from utility operators. geles and Californian organisations was most gratifying Eleven representatives from New Zealand lifelines or- and the participation of a representative from the Califor- ganisations and consultancies attended the workshop, nia Utilities Emergency Association served to underline along with approximately twenty United States counter- the importance attached to co operation. parts. The meeting allowed the New Zealand representatives to 220 • Risks and Realities

hear first-hand accounts and experiences of the Loma 4 Obtain information on the time required to restore Prieta, Northridge and other recent California earth- adequate service. (The times achieved were most quakes. New Zealand representatives were mindful of impressive with basic service restored in nearly all the different context of the Northridge earthquake in cases within one to three days.) terms of its relative effects on the city as opposed to a similar size earthquake in central Wellington or Christ- 5 Obtain views on acceptable levels of risk and the link church. to investment — from Lifelines providers, politicians, community leaders etc. (Not much informa- The intention of this section is highlight the aspects of the tion on these matters was obtained, but there were presentations and visits relating to interdependence and different approaches from different organisations). co-operation amongst lifeline utilities, with particular emphasis on mitigation measures, preparedness, and 6 Make contact with appropriate staff with a view to recovery. It is intended to supplement reports by others “in depth” discussions and inspections. ( The inspec- focusing on technical matters within each utility. tions on the final three days in Los Angeles and particularly the San Francisco/Oakland visits were a highlight.) Workshop overview The focus was on the Northridge earthquake, but inevi- 7 Discuss the applicability of “Response Plans”. (There tably others were mentioned as were other emergencies was an emphasis on generalised response plans rather such as California wild fires and the Los Angeles riots of than for specific events.) 1992. 8 In San Francisco/Oakland, assess the effects of lique- A wide range of topics from research funding to techni- faction. (The three days there most valuable in the cal detail was covered. Los Angeles Lifeline Utility reassurance given about the relative resiliency of Organisations were represented at a high level through- pipelines, although it was realised that the Loma out, and this added greatly to the value of the seminar for Prieta earthquake was not as large as that in the the New Zealand group. Many of the presenters and scenario for Christchurch.) participants had been in key roles in the response to the Northridge earthquake and had clearly taken consider- Workshop presentations able trouble to gather information for presentation at the Workshop. Initiatives in lifeline earthquake preparation Presenter: Anshel Schiff, Stanford University, a noted There was a general preparedness to share information US lifeline expert. and the Workshop could have easily have had another day devoted to follow-up discussions and questions. This presentation traced the United States Federal Gov- ernment initiatives from the formation of a US-Japan Objectives of the Visit Earthquake Engineering Group in 1963, through meas- In the team’s visit to Los Angeles (and for San Francisco/ ures following the San Fernando earthquake of 1971, to Oakland) the desired outcomes in relation to Christ- the passage of the Earthquake Hazard Reduction Act church were largely achieved, as summarised below. 1990, which urged improvement of engineering lifelines. 1 Assess the comparability of the California events with Christchurch. (Some of them were, but ground Notably, this Act called on the Federal Emergency conditions were not equivalent and the sheer size of Management Association (FEMA) and the National Los Angeles means that not all of the city and nearby Institute of Standards and Technology (NIST) to submit jurisdictions were affected, which with “mutual aid”, a plan, including precise timetable and budget estimates meant many resources were available for restoration for developing and adopting, in consultation with appro- of services.) priate private sector organisations, design and construction standards for “lifelines”. 2 Identify Lifeline problems, both the expected and unexpected, including interdependence problems. Formation by the American Society of Civil Engineers (They were, but most were expected.) (ASCE) of a Technical Committee on Lifeline Earth- quake Engineering (TCLEE) in 1974 gave evidence of 3 Obtain expert views on mitigation measures, particu- professional co-operation and recognition of the impor- larly those previously adopted but now changed as a tance of the field of lifeline earthquake engineering. The result of previous earthquakes. (The higher than focus of this group is on mitigation measures and it has previously experienced ground accelerations are re- been instrumental in developing standards and improv- quiring a reconsideration of designs.) ing awareness since its inception. New Zealand/Los Angeles Workshop • 221

Private sector utility organisations are co-operating by • facilities for generator operation during the Northridge sharing information on damage and preparing guides to earthquake; good practice. They are also developing response plans and conducting training exercises. • assisting in prioritisation of electrical power restoration to vitally impacted locations in the Loma Prieta Overall, Professor Schiff’s presentation indicated the earthquake;and importance that had been attached by Federal, State and Professional Organisations to the subject of lifelines and • the airlifting of two 250 kV circuit breakers from the this was evident in the level of research funding being east coast to the west coast in the Loma Prieta made available. earthquake. CUEA’s current planning activities include participa- California Utilities Emergency Association tion in: (CUEA) • utility advisory committees; Presenter: Katherine Latipow, Executive Director. • government and industry task forces on potable wa- This presentation covered the role of the State of Califor- ter distribution and emergency cellular communica- nia Governor’s office of Emergency Services (OES). tions; This office has several divisions including one for utilities and another for telecommunications. • preparation of planning documents for critical facility security and for mutual aid agreements; and The OES seeks to establish common organisational structures for response to emergencies, with a regional • facilitation of access to information sources, work- focus. It aims to assist at the government and industry shops, conferences and training programmes. interface and is pushing for improved communications such as the Operational Area Satellite Information Sys- CUEA exists because of the recognition of the impor- tem (OASIS). OES also promotes the development and tance of the interdependence of engineering lifelines. It use of mutual aid agreements between organisations. was originally chartered in 1950 as part of California’s Civil Defence Plan, but it was not until 1990 that it was CUEA is an autonomous organisation of public and established as representing all utilities in an emergency. private utilities fostering co operation in emergency Significantly, transportation utilities are not included, planning, response, mitigation and recovery. CUEA though clearly close liaison exists. represents utilities in both emergency and non-emergency situations and provides a link to the State-run CUEA aims to encourage wider membership to provide OES. a larger base from which to promote its objectives.

In an emergency, CUEA establishes a Utilities Opera- Clearly, there are parallels for New Zealand. Utility tions Centre staffed by CUEA representatives from gas, organisations seeking to improve co-operation in emer- electric, water and telecommunication utilities. This gency and who wish to examine the effects of centre co-ordinates utility-related emergency response interdependencies, should maintain liaison with CUEA. and recovery issues during major emergencies. Principal concerns in California include: New Zealand perspectives • damage assessment; Presenters: John Lumsden (CAE), Peter Leslie (WRC), John Lamb (CRC), John Norton (THCC) and David • outage duration; Hopkins (Kingston Morrison).

• restoration priorities; The New Zealand presentations covered key aspects of the Wellington Lifelines Project, the Wellington Earth- • distribution of potable water; quake Lifelines Group, the Christchurch Engineering • co-ordination in government agencies in affected Lifeline Project, preparedness, response plans and inter- areas; and dependence.

• media releases. The interdependence of utilities was identified as important in response planning and in prioritising mitigation Examples of CUEA contributions to disaster response measures. Techniques for analysing relative depend- include: ence, importance and priority were described.

• co-ordination of emergency diesel fuel delivery to It was gratifying to note that many of the issues raised in telecommunications; the New Zealand presentation, particularly on interde- 222 • Risks and Realities

pendence and response planning, proved to be of particu- Significant damage to monitors and work-stations re- lar relevance to many of the subsequent presentations. sulted in reduced surveillance in the early stages. This prompted a complimentary comment from one US participant on our achievement, especially in the absence In the Northridge earthquake, 187,000 lines were af- of a major earthquake in a New Zealand city. fected for up to 22 hours.

One particular point covered in the New Zealand presen- Key findings and lessons from an interdependence and tation which emerged as important in almost all other response planning perspective included: presentations was the need to handle media releases • Three hour battery reserve is not long enough. responsibly and pro-actively. • Power to alarm/monitoring systems is vital (back-up Effects of a large earthquake on Los Angeles power for this is critical). Presenter: Ron Eguchi (Vice-President, EQE Interna- • Testing of stand-by generators is recommended. tional). • A Network Management Centre is critical to dial tone This presentation reminded participants that the delivery. Northridge earthquake was not the worst that LA could face. It consisted of the results of a analysis of movement • Back-up systems for communications are most worth- of the Elysian Fault with a magnitude 7.0 earthquake while. near downtown Los Angeles. • Integrated utility response plans are valuable. Impressive use of GIS systems and the resulting col- • Mutual aid agreements between utilities are vital. oured graphs was made in presenting the results of predictions. The County of Los Angeles has US$500 An established network of communication between utili- billion at risk, while the city has a US$200 billion ties is critical. An example was quoted where a telecom exposure. This exposure includes 4,000 steel frame field worker would not venture into an area of telecom- buildings, 300 of which did not perform well in the munications damage for fear of a dam bursting. Those in Northridge earthquake. control of the telecommunications recovery were able to obtain authoritative clearances as to the safety of the dam Plots of MM intensity, damage, injury, deaths, and from their counterparts in the Department of Water and uninhabitable residences were presented. The overall Power. Telecommunications repairs could thus proceed losses were computed to be US$75 billion, split US$50 without delay. billion for general damage and US$25 billion for lifelines. The total of US$75 billion compares with an Exercises are the best preparation. estimate done by similar means for the Northridge earthquake of between US$13 and US$20 billion in direct A clear impression was given that good planning and damage. A feature of the damage ratios used for houses mitigation had reduced damage and impact, but there was that they were significantly higher than those com- was room for improvement. monly accepted for values of MM less than eight.

Overall, this presentation served as a reminder that Water supply in Northridge earthquake damage and disruption can rise exponentially as the size Two organisations gave presentations, the Los Angeles of an earthquake increases. Survival in Northridge was Department of Water and Power and the Metropolitan not necessarily the ultimate test. Water District of Southern California.

Telecommunications in the Northridge Los Angeles Department of Water and Power earthquake (DWP) Presenters: Mike Caren, Director, Emergency Prepared- Presenter: Bob Giles. ness, Pacific Bell, and Bill Sevido, Pacific Bell. This was an excellent blow-by-blow description of DWP’s A detailed account of outages in Loma Prieta and response. It is of interest to note that it took 12 days to lift Northridge was given resulting in some interesting com- the “boil water” notice. parisons. Telecommunciations damage in Loma Prieta Other key points to emerge were: was $70 million compared with $26 million for Northridge, both figures being for repairs only. • There were many offers of mutual aid, including several from out of State. Because of the existence of Calls in both earthquakes were around twice normal. these aid agreements, phone orders for materials and New Zealand/Los Angeles Workshop • 223

labour were acceptable. Assistance from surround- • MWD plan to purchase a helicopter, with pilot, to ing authorities came in the form of equipment and assist with emergencies in future. personnel. Improvement projects scheduled are: • Mutual aid crews from outside Los Angeles were used during the day time, reserving the night-time • updating of emergency response plans; work for locals who were more familiar with the • installation of GIS systems to give better and quicker system. Provision for accommodation and meals was information; difficult. Care had to be taken to avoid fatigue amongst workers. • developing a business resumption programme; and

• Communications had been difficult because there • formalising mutual aid contracts. was no common radio frequency.

• Cellular phones had proved beneficial, especially the Electrical systems in Northridge earthquake use of ones personally owned by staff. This was in Two organisations made presentations. These were South- spite of the fact that a repeater station failed. ern California Edison and the Los Angeles Department of Water and Power. The comparative size of these two • A 24-hour information centre was set up for custom- organisations which, between them, serve the greater ers. Los Angeles area can be seen by the fact that Southern • Media response was via a single source. California Edison has 4.2 million customers and DWP 1.36 million.

Metropolitan Water District of Southern California (MWD) Southern California Edison Presenter: Bill Cooper. Presenter: Dennis Ostrom, Seismic Consultant.

This presentation provided an interesting overview of There were US$50 million in damage and related costs, the damage to the Jensen Filtration Plant which suffered and 1.1 million customers lost power initially, although US$4.8 million damage. A US$11 million programme half a million of these were restored to power almost for repair and mitigation is underway. immediately. After 12 hours all but 180,000 customers had power, after 16 hours this had reduced to 40,000 and From an interdependency and response planning per- after 30 hours to 3,000. All service was restored within spective the following comments were noted: 56 hours of the earthquake.

• A viable emergency preparedness programme is es- Edison’s distribution facilities experienced 471 sus- sential. tained and 200 momentary circuit interruptions. Thir- teen poles and 61 transformers were damaged and re- • Mitigation work carried out since 1971 reduced im- quired replacement. Over 200 spans of conductor re- pact considerably. quired replacement or repair. Over 40 pad-mounted • Better recognition of emergency personnel ID is transformers shifted on their pads and required resetting. required. An interesting point to emerge in discussions was that the • Provision of lodging and meals for staff was difficult, peak daily demand in winter (January) is some 40% of especially with no water or power in hotels and the peak daily demand in summer. This fact allowed motels. repairs and re-energising to be effected using spare equipment. • MWD is contemplating the free storage of staff recreational vehicles (campervans) at key sites on the No significant problems were mentioned regarding ac- basis that they would be available in emergencies. cess or coping with the response. Dennis Ostrom, a keen motorbike user, found that a mountain bike was the best • MWD proposes to establish prior contracts with local means of transport immediately following the earth- equipment suppliers and hirers for use in emergen- quake. He was able to by-pass lines of stationary traffic cies. and gain access to otherwise inaccessible areas. • Communications were difficult in the first 48 hours. Los Angeles Department of Water and Power • A special Emergency Operation Centre was estab- (DWP) lished at the treatment plant. This was separate from Presenter: Ron Tognazzini, Seismic Manager. the area control centre. 224 • Risks and Realities

DWP has 1.36 million customers serving 3.5 million system flexibility. The considerable redundancy helped people and has 11,000 employees. At the time of the as did the low load in January. earthquake it was generating approximately 1,900 MW out of a total capacity of around 5,000 MW. Rigid An interesting observation from the Manager of the ECC conductors, transformers, cantilever supports, discon- was that in a previous emergency, senior management nect switches and structural systems were damaged. A was allowed on the dispatch room floor and this proved graph of the restoration is shown in Figure 9.1. counter-productive. In the Northridge earthquake, special efforts were made to keep senior management (Emer- The presentation provided a valuable and close insight gency Command Centre) separate from the Energy Con- into the restoration process, proving the value of redun- trol Centre activities. This worked well. dancy in the systems. The presentation included a sequence of colour slides showing the step-by-step resto- The Manager of the ECC, Marcie Edwards, made some ration of the power grid. telling points at the Workshop:

DWP have been active in seismic mitigation since 1971 • don’t tie up key low level people on other tasks; and this has paid off well. Seismic qualification of • prepare access methods in advance, e.g. have maps of equipment also paid dividends. patrol roads ready; The state of DWP in seismic mitigation is indicated by the proposed future design changes: • keep the media well informed — explain what, when and why things are being done; • more rigorous design procedures; • review station layouts for by-pass options; • higher design loads; • technical aspects are largely known — management • more full scale testing; is the key to recovery; and

• use of alternative materials (to porcelain); and • decide in advance about logistic support of personnel, e.g. payroll, accommodation, shifts. • system configuration changes.

Central to the recovery of DWP’s system was the Energy Gas supply Control Centre (ECC), a robust facility located in the Presenters: John May, Martin Remmen, Randy Ragos. mountains to the north-west of the city. The day-to day experience of dispatchers in balancing load and genera- The presentation covered the more spectacular aspects tion was helpful in giving the necessary knowledge of of the mobile home fires and the house fires in Balboa

99.5% 97% 100 93% 90 80

% of customers 30

20 10

0 0244872 Hours after earthquake

Figure 9.1: Power service restoration in the Northridge earthquake New Zealand/Los Angeles Workshop • 225

Street. A lack of water to fight fires was highlighted and Monica freeway west of Los Angeles city. This is one of it was noted that the toppling of water heaters accounted the busiest freeways in the world carrying some 340,000 for 39% of fires. vehicles per day.

Martin Remmen was in the emergency operation centre All freeways were scheduled to be restored by December for four days. The gas system took 12 days to restore 1994. The key Santa Monica freeway was substantially fully. The bulk of the damage occurred to residential and replaced by April 1994 which is equivalent to building light industrial installations. the Thorndon Overpass two or three times over in 65 days. The controversy over automatic shut-off valves also emerged. The gas company now has instructions for Funding reimbursement is available from FEMA which people not to turn-off their gas unless there is evidence of pays 100% of costs in the first 180 days and between 80% leakage. This represents a reversal of previous policy. and 92% after that. Even so, Los Angeles City Fire Department requires seismic shut-off valves for new construction. This is Special contract procedures and conditions were put in being resisted by Southern California Gas since it causes force to expedite reconstruction. These included bonus/ more problems than it solves. penalty clauses, 24-hour, seven day working with no allowance for rain days. A mixture of bidding and force Little was raised regarding interdependency and re- account methods were used, depending on scale of the sponse planning, but the gas company consider them- work and its urgency. selves to have been lucky in this earthquake. Design of freeway replacements complied with post- The Workshop was held at the offices of Southern Loma Prieta standards, and it can be assumed that the California Gas and this afforded the opportunity of extensive research and development work initiated after visiting their spacious, well-equipped, high-tech Emer- Loma Prieta enabled a quicker response to the Northridge gency Control Centre. This impressive facility had damage than would otherwise have been possible. Some special screens for monitoring television news, a valu- discontinuities in the bridges were removed, but gener- able source of information. Communication terminals, ally geometry was kept the same. Mitigation measures computers, and special overhead TV cameras were in- had been carried out on a number of bridges in the Los stalled. There was also provision for a media interview Angeles area, notably steel jacketing of columns. These area as well as supplies of food. Interestingly, there were performed well. full height glass partitions, suspended ceilings, and unsecured equipment in this facility. Nevertheless, it is clear Detours were set up quickly, especially at the Santa that emergency response is being taken seriously by this Monica freeway, aided by the Los Angeles city’s earth- organisation. quake traffic management centre. This is used every day to control traffic within the city, and in particular to cope Transportation Systems in Northridge with unforeseen blockages and accidents. The ATSAC Earthquake system (Automatic Traffic Surveillance and Control) of Responsibility for transportation systems within the LADOT provides graphical and video signals which greater Los Angeles area is split between California enable computer control of traffic lights. The system was Department of Transportation (Caltrans), which is simi- reprogrammed to give priority to designated detours. lar to Transit New Zealand and looks after the freeway The urgency of replacement of freeways is indicative of network, and the Los Angeles City and other cities which the reliance on the motor vehicle in Los Angeles gener- are equivalent to our territorial authorities. Presentations ally. were made by both Caltrans and the Los Angeles City Department of Transportation (LADOT). Los Angeles City Department of Transportation

Caltrans Presenter: Tom Conner, Assistant General Manager. Presenter: Jack Hallin, Deputy District Director. Tom Conner gave a comprehensive overview of the earthquake damage to buildings and infrastructure. From Although repair costs will total US$300 to US$350 the perspective of the Los Angeles Department of Trans- million, only a small proportion of freeway structures portation, it emphasised the value of co-operation be- were significantly affected by the earthquake. This meant tween Caltrans, LADOT, Santa Clarita City, Los Ange- that alternative routes were available in most cases, les County, FEMA, OES, and the Federal Highways although at some inconvenience. Critical areas were the Administration. Through this co-operation, approvals intersection of the Interstate 5 and Freeway 14 which is for work were streamlined. A small example was that a large complex of elevated interchanges, and the Santa 226 • Risks and Realities

new traffic lights could be put in within a matter of hours • Redundant communications and emergency systems when approval would normally take up to one month. paid off.

The metro systems, consisting of bus and rail were • Electricity is vital and back-up should be provided relatively unaffected by the earthquake and this allowed where possible. commuters an alternative to the collapsed freeways, particularly from Santa Clarita. Ridership was many • Stand-by generators should be serviced by those who times normal at the peak, and remains about double the depend on them, not by an outside agency or depart- pre-earthquake level. Generally, traffic signals survived ment. and performed satisfactorily once energised. • Structural and non-structural mitigation work car- Key points from an interdependence and response plan- ried out by the Bureau of Sanitation cut losses and the ning perspective to emerge were: impact of the earthquake overall. It allowed greater focus on recovery of the sewerage system. • Two-way radios worked well. Rafael Solorzano focused on damage to sewer lines and • The phone system was satisfactory but was over- the problems of inspection, collation of damage records, loaded. and procedures and documentation for funding of re- • Cellular phones were handy, but there was no provi- pairs. The collection system totals 7,000 miles of pipe sion for priority for emergency management use. with an average age of 50 years. GIS systems were used to log damage. Interestingly, plots of building damage • Preparedness plans and training were invaluable. and its severity were used to prioritise sewer inspections by closed circuit television. In all US$200 million dam- • Daily press briefings were given and it was found that age was done to sewer lines. the people respond well to accurate and helpful information. Because of the difficulties in identifying damage, the • Agency co-ordination was essential and works well Bureau of Sanitation asked FEMA for an extension of in the emergency context. As Tom Conner said later time in lodging its claims. Mr Solorzano emphasised the “People want to be part of the solution, not part of the need to liaise with funding agencies in the early stages. problem”. Efforts in repair were categorised as emergency and non- emergency. Progressive discovery of leaks occurred as • In spite of the value of pre-planning of response, water was restored to the city. Broken water mains much was achieved on the “just-do-it” principle. caused wash-outs in some places compounding damage This applied particularly to the construction of some to the Bureau’s sewers. detours which were designed and constructed from the bulldozer seat. On the funding side, there was concern that the city must bear the cost of inspection if no damage was found, and For transportation systems, there are close parallels in that FEMA would pay for the CCTV inspections but not the LA experience for New Zealand, especially in the for the interpretation of the video film. need for good communications, co-operation amongst agencies and preparedness planning. Recommendations were made as follows:

• Employ three teams in recovery, one for inspection Sewage system in the Northridge and damage assessment, one for project definition earthquake and co-ordination with funding agencies and one for Presenters: Del Biagi, Director, Bureau of Sanitation, design and construction. Rafael Solorzano and others. • Liaise early with funding agencies on arrangements Del Biagi explained the system and noted that the Bu- for inspection and repairs. reau’s emphasis has shifted from response in the late 1970s to preparedness and mitigation. • Know the “hot-spots” of your system, i.e. those likely to give trouble in earthquake. Not much damage occurred at treatment plants, but considerable damage was done to lines. The following points were made: Summary Of Key Issues In reviewing the Workshop, the presentations and visits, • Employee preparedness paid off. the following points stand out as providing worthwhile pointers for New Zealand’s Lifeline Organisations. • Response plans must cover all times of day and night and cater for holidays. New Zealand/Los Angeles Workshop • 227

Effect of a larger event that prior contact is made and the likely scope of A bigger event would have increased problems in required support defined. non-linear fashion. Extrapolation of the Northridge experience to more critical facilities should be done Cellular phone network with this in mind. This is clearly growing in popularity and had an important role to play in the Northridge earthquake. Damage assessment There is a need to establish a dedicated part of this This has an important role to play in a number of network for use by emergency operators. There is ways. also a need for the various organisations involved in cellular networks to co-operate in an emergency to Firstly, prior to the earthquake, it allows the identifi- ensure full coverage of the affected area. cation of “hot-spots” by examining the possible impact on the system of failure of particular ele- Planning and training ments. The importance of this was emphasised by many of Secondly, it raises staff awareness of the facilities those involved. We all know that this is the case, but and their vulnerability generally. it is more difficult to do something about it. Califor- nia has had 25 federal disasters in the last 10 years, so Thirdly, and increasingly, GIS systems are helpful in that the need to plan and carry out training and providing a tool to prioritise inspections in the imme- exercises is widely accepted. People will only know diate aftermath of an earthquake. This has been what to do if they have the opportunity to practise. assisted by the CUBE (Caltec USGS Broadcast of Earthquakes Facility) which provides on-screen iso- seismal maps. The use by the Bureau of Sanitation of Timing of earthquake plots of building damage to prioritise their inspec- Some reminders of this were given in the time of day tions of sewers is a classic example of the use of GIS that the earthquake occurred and the fact that it was and inter-utility co-operation. a holiday. A further factor was that the electrical load was some 40% of what it would have been in the summer. Options for repair and recovery would have Communications been limited had higher loads been on the system. These hold the key to effective response.

This applies to automatic information systems on the Special contractual measures status of plant and to person-to-person communica- These have immense potential benefit and organisations both within a lifeline organisation and between tions should give them prior consideration. The very one utility and another. Prior personal contact amongst rapid completion of the Santa Monica freeway re- key managers is helpful. pairs and reconstruction are testimony to the value of innovative and flexible arrangements. Organisational preparedness There is a need to separate Emergency Command Handling of the media Centres (senior management) from Emergency Con- A consistent message came through from all in- trol Centres (system operators). volved — the media demand, and deserve, attention. They are also a source of valuable information. It is Professional associations and inter-utility co- vital that communication with the media is controlled operation through a single source for each organisation. Infor- These are clearly valuable in establishing networks mation must be accurate and explain what, why, so that key engineering lifeline managers are known where and when things will be done to restore nor- to one another. The California Utilities Emergency mality. Association could be a valuable source of information and guidance for New Zealand engineering Seismic mitigation lifeline organisations. It was clear from many of the presentations and discussions held that mitigation, both structural and Mutual aid contracts non-structural, had been seen to be extremely worth- The value of these was emphasised by a number of while. Testimony to this is the ongoing programmes presenters. These can be anything from an informal of mitigation by all organisations. arrangement to a formal contract. The key factor is 228 • Risks and Realities

Traffic management Francisco/Oakland where they have experienced liq- One advantage that Los Angeles had was a sophisti- uefaction problems. They were able to relate the cated system for the automatic surveillance and con- damage that had occurred to ground conditions. trol of traffic. This allowed detours to be put in place through city suburbs with the lights automatically Engineering services are resilient controlled to give necessary priorities. Equivalent One major impression was the way in which the New Zealand response may not require such sophis- engineering services are remarkably resilient, par- tication, but at least the value of pre-planning for ticularly when compared with the amount of building these events is evident. damage which received so much publicity.

Each earthquake is unique Damage is still visible, but not all effects It is important to remember that each earthquake is The drive through the various affected areas in Los unique in its location, its size, the length of time Angeles most worthwhile and the team was amazed involved in the earthquake, the timing of the earth- at the extent of damage still visible in so many areas quake (day or night, weekday or weekend/holiday). seven months after the earthquake. Although all The effects of the earthquake are also unique with engineering services were again operable, that did factors such as the location of services, their age, not mean that the effects on the community had their design, their construction, their materials, the passed. soils, the relationship to buildings and to other services, to roadways, to rivers, etc., all affecting the It is important to remember that utilities are there to damage sustained. serve the public who may be very badly affected emotionally and financially and therefore even more Lifeline engineering is not a precise science demanding than usual. Earthquake engineering, and in particular lifelines engineering, is not a precise science and it is therefore Mutual aid impossible to predict accurately what will happen in In relation to the restoration of services the concept of Christchurch, or elsewhere, as the result of an earth- “mutual aid”, where adjoining authorities came to quake. Notwithstanding this, in California most of the assistance of the affected ones, was most interest- the mitigation measures undertaken on services as a ing in the California area, but the sheer size of Los result of lessons learned from previous earthquakes, Angeles meant that although many had been affected were largely successful (although they may not have by the earthquake, there was still many more nearby been had the earthquakes been of longer duration) ready and able to assist. and it is therefore well worthwhile mitigating in accordance with present knowledge. FEMA funding The resources provided by FEMA (Federal Emer- The Christchurch approach is appropriate gency Management Agency) was very impressive The above means that the imprecise way in which the and the way in which restoration of services was Christchurch Project has had to be approached does undertaken was often affected by the way in which not detract from the benefits likely to be obtained as funding was available. a result of the Project. In fact, the approach taken trying to relate the services not only to their design, People response methods of construction etc., but also the soil condi- Provision of accommodation and meals for those tions, is entirely appropriate. The somewhat crude involved in making repairs and organising recovery classification of vulnerability on a scale of 1 to 3 is emerged as a concern. So too did the tendency of perfectly appropriate in assisting in the determina- those involved to work extended periods and so tion of the priority for mitigation measures, and, suffer fatigue. This had to be watched. notwithstanding the tremendous amount of money “thrown at the problems” by the well-resourced Although the value of planning prior to the event is American authorities, the Christchurch approach on well recognised, considerable scope must be given to a shoe-string is still very worthwhile. those in the field to solve the problems on the ground. This is the “just-do-it” approach. Clearly, with proper The team was surprised that the Americans did not communication and pre-planning combined with this seem to take into account the soil conditions in spirit, much can be achieved. explaining, or predicting, damage as much as expected, although that situation was different in San In relation to people response, it is worth concluding New Zealand/Los Angeles Workshop • 229

with Tom Conner’s observation which sums up why • Don’t have a detailed emergency response plan be- cities cope much better than expected with major cause it never occurs like it is predicted! disasters: • In setting up mitigation measures, test rather than “People want to be part of the solution not analyse their likely performance if at all possible. part of the problem” • Think carefully about the number of spare parts you have because if you simply replace those that have Miscellaneous Quotes been damaged with the same then they may well be The following miscellaneous quotes noted during the damaged in an aftershock. Workshop and inspections are of interest. • Don’t take the operators away from the work just for • A visit to an earthquake makes you a believer. the information services — managers who don’t do the physical work can be very good as liaison people. • Mutual aid agreements determined in advance are a help. • Develop media response in advance to explain the reasons why services are being restored in a certain • A GIS helps in evaluating the effect of an earthquake. order.

• A “disaster day” practising a co-ordinated response • A clean/clear change is needed to the emergency to a simulated disaster across the whole of the com- organisation rather than overlap with the normal munity is worthwhile. management structure.

• Before Northridge the geologists thought they had • Waterheaters that moved were a major problem with sorted out the faults — they are now rethinking. gas supplies.

• The Northridge repair costs were of the order of • All trenches throughout Los Angeles are backfilled $US13 to 20 billion. with one, two or three sacks of cement slurry mix (the • Don’t publicise that it will be three days before the team saw virtually no settlement of trenches.) power is restored — it becomes the expected per- • LPG gas bottles and gas heaters could be a real source formance regardless of the effects of the earthquake. of fire in Christchurch.

• Telecom equipment survived but not the structures. • In Los Angeles dependence on vehicles is a way of life — is it any different in Christchurch? • A lot more outside plant damage was expected (with the exception of fibreoptic cables which were thought • Make sure the public know what is involved in would perform well) than actually occurred. restoration. If they know, they will put up with it. It is essential that the media has good information or • Test/use the emergency power plants regularly for an they will make up their own! extended period on full load. • The refuse collection programme to the Los Angeles • Audit to see that all procedures of mitigation are Landfill was lost for one day only but they needed a actually continuing in place — regular testing is special household hazardous waste roundup because essential. of the dangerous/hazardous materials dumped on the • The team did not hear of any damage and we believe floor from shelves. there was no damage to cable ducts in the Loma • Bolt down or secure small items. Prieta or Northridge earthquakes. Nearly all the problems were at the structures. • Have an employee emergency preparedness programme. • Vehicles taken home by staff aid tremendously in the inspections necessary immediately after an earth- • The person who has to use the emergency equipment quake and before planning restoration. in an emergency should maintain it.

• Having a core staff of your own is essential in being • Have a separate emergency preparedness budget. able to cope with damage quickly. • In sanitary sewers, it was mainly the clay pipes that • Offices performed very badly, particularly with com- were damaged and the concrete undamaged. puters, drawers and plan rooms and material off shelves. • In looking for sewer breaks authorities used the GIS 230 • Risks and Realities

to overlay the water breaks (which were visible) in emergency response decision-makers can make a the first instance. sensible decision. Los Angeles has a CUBE system which produces within minutes of an earthquake • Damage to sewer pipes was not so much related to the being monitored a map on screen showing the loca- materials but more to the shallow depths (shallow is tion of the earthquake, its time and its magnitude. five metres). Lesson 4: Even though it may be thought that a consid- • There seemed to be no connection with the soil type erable amount is known about the seismic geology of and damage in Los Angeles unless liquefaction was an area, each earthquake is unique and it is very involved. difficult to forecast accurately what will happen. The • One often has to proceed on very limited data but that more is known of the present conditions the better is not an excuse for not doing anything. will be our prediction of the future response to earthquakes. • The demand on the system (the time of the day, the Lesson 5: Damage can occur at a considerable distance week, month, etc.,) makes a big difference in the from the epicentre of an earthquake and reflections ability to cope with outages. from the regional geologic structure may play a • The impact of an earthquake increases in a non-linear significant part. This needs to be borne in mind in way in relation to magnitude/intensity. view of the close proximity of the Port Hills. It is important also to emphasise that damage may occur • Modified Mercalli is a measure of the damage that from smaller magnitude earthquakes close to Christ- occurred to other than engineering structures. It is not church, not only from the “Big One” on the main a measure of ground acceleration and the length of alpine fault. time an earthquake shakes is most significant. Lesson 6: There are surprising variations in the intensity • The source of funds available for restoration of the and nature of strong ground motion resulting from an damage often determines the way the restoration is earthquake and engineers must not assume that the carried out and its priority. Earthquake Code is adequate for all sites. The nature and location of the substrata is particularly important Practical Lessons from the Loma Prieta knowing that about half of Christchurch could be Earthquake subject to liquefaction. Lesson 7: As has been observed in many earthquakes, the Acknowledgement intensity of seismic shaking is critically dependent The following lessons were adapted from those listed in upon the nature of the local soils and shallow geo- Practical Lessons from the Loma Prieta Earthquake, logical structures. It is important that the impact of Report from a Symposium sponsored by the Geotechnical possible hazards be incorporated into landuse plan- Board and the Board on Natural Disasters of the National ning or building codes. Research Council. Ron Eguchi, the Visiting Fellow at the Christchurch Engineering Lifelines Workshop, was Lesson 8: In the Loma Prieta earthquake the geological a joint author on the lifeline perspective of this report. maps prepared for the purpose of identifying potential areas of liquefaction proved adequate in defining Many of the general observations made in the report will the locations of major occurrences of liquefaction apply in Christchurch and the “lesson” approach used in and lateral spreading. the National Research Council publication has been used. In Christchurch there must be continued refining of the accuracy of our maps and where facilities must be Lesson 1: Investments made in earthquake preparedness built in areas identified as being susceptible to lique- and hazard and risk mitigation do pay off. faction, measures must be taken to ensure that the ground improvement techniques are used to mini- Lesson 2: Those who are aware of the risks involved mise earthquake damage. from earthquakes have an obligation to ensure that the policy-makers are aware of the risks, the costs and Lesson 9: A significant effect of earthquakes is land- benefits of various strategies and must make strong slides, and it is important that maps be developed for recommendations for earthquake hazard and risk landslide potential. Some work was done on slope mitigation. hazards on the Port Hills but it needs to be borne in mind that detailed geological studies were not under- Lesson 3: Utility operators need to know quickly follow- taken. The time of the year at which an earthquake ing an earthquake what are the affected areas so that New Zealand/Los Angeles Workshop • 231

occurs, the duration of the shaking and the water Lesson 19: When considering possible damage to instal- content of the material very much affects the extent lations note must be taken of adjacent hazards which of landslides. may impact on the particular lifeline.

Although there was no mention in San Francisco Lesson 20: Old cast iron pipes are most prone to damage regarding rockfalls, these could be a significant prob- and an earthquake will find the weakest (most cor- lem on the Port Hills. roded) pipes.

Lessons 10 and 11: Properly engineered fills perform Lesson 21: Well-designed water and waste water treat- well and it is important to ensure that proper placement facilities will probably experience little dam- ment and compaction of material is undertaken. In age but the non-structural components are also im- Christchurch the stability of some of the older fills portant. has not been examined and there must be some concern regarding the likely performance of these, Lesson 22: Structures founded on soft soil are very particularly the bridge approaches and some of the vulnerable even to earthquakes of short duration, but cut and fill on the hill areas. are much worse so in the event of longer duration earthquakes. Lesson 12: Skilled earthquake engineers can considerably reduce the loss of life and damage resulting from Lesson 23: The structures that are necessary in sub- the loss of structures. This applies in lifelines as well stations need to be seismically resistant as well as as buildings. remote from the human hand. Many electrical installations could be designed to resist earthquake motion Lesson 13: Although in some cases bridges and build- if they were laid horizontally rather than vertically, ings that had been retrofitted to resist earthquakes but space requirements often mean there are vulner- sustained structural damage, many bridges that had able vertical components in substations. not been retrofitted sustained little or no damage, but this is not a reason for not undertaking mitigation Lesson 24: It is important to convey the results of measures. Not enough is known about the way in earthquake research to the potential users of these which structures perform, but any mitigation meas- findings and it must be ensured that the engineering ure is obviously better than none at all. community within Christchurch is aware of the knowledge that is actually available. Lesson 14: “Hinge restrainers” generally performed well on highway bridge structures and on the Los Angeles Lesson 25: A disruption of water pipelines, particularly Freeways the relative simplicity and fragility of some from elevated reservoirs, can deplete the water avail- of these is of interest. able for firefighting. However, although automatic cut-off valves are considered, it has to be borne in Lesson 15: It is important that both building owners and mind that an earthquake-operated valve which will lifeline operators realise the fact that although work retain the water in reservoirs may still leave a com- may be done to mitigate the effects of an earthquake munity without firefighting supplies. (The valves there is no absolute surety that damage will not ensue. proposed in Christchurch are activated by flow rather Lesson 16: Inspections during the building of engineer- than by earthquake motion.) ing lifelines and their buildings are important to see Lesson 26: Good emergency response in recovery plans that the designer’s intentions are actually carried are essential to facilitate co-ordination and quick through into the construction. response to an earthquake, but the response plans Lesson 17: Unreinforced masonry buildings and fences need to be general rather than specific. The results of can be dangerous, not only to occupants but also to the earthquake are not always predictable and there the control structures in Lifelines. In many cases the may well be a different situation from that originally lifelines themselves performed satisfactorily but it predicted. was the buildings that enclosed their control structures that caused problems. Sometimes the buildings Lesson 27: Although Christchurch does not have a were satisfactory but the contents had not been re- reticulated gas supply, there is a potential for ignition strained. of bottled gas by the reactivation of power supply.

Lesson 18: Simple measures are available for greatly Lesson 28: Following an earthquake there could be a reducing earthquake losses and the anchoring of tendency to simply restore what was already there as equipment is essential. is the basis of many insurance policies, but in restoration it needs to be borne in mind that there may be 232 • Risks and Realities

a need to upgrade the damaged elements to mitigate Lesson 39: In the event of an earthquake authorities need against similar damage from later earthquakes (or to capitalise on the effect on the community through perhaps even aftershocks?). their increased awareness of earthquakes and it may be that we should be referring to recent earthquakes Lesson 29: The earthquake often will damage old instal- as a “wake-up call”. lations which should have been replaced but have not been because of lack of finances. The earthquake Lesson 40: Recovery from a destructive earthquake is does not help in this. expensive for everybody and anything to reduce the cost is worthwhile. Lesson 30: Although a community is very resilient it needs to be aware of the potential for earthquake Although the report on the Loma Prieta earthquake is damage so that the community is better able to acknowledged as above, the 40 lessons derived for function in response to the earthquake. Christchurch do not, in many ways, relate directly to the Lesson 31: Rushed post-earthquake inspections of in- lessons included in the National Research Council Book. stallations are not necessarily accurate and inspec- Having seen the damage first hand that occurred in the tions often have to be redone. It is important that Northridge earthquake, the team was more than ever adequate records of inspections are kept and staff convinced of the obligation to do everything possible to trained in advance, if possible. mitigate the effects on Christchurch of an earthquake. Lesson 32: Fibre-optic cable is remarkably resilient but The team was very reassured as a result of the visits that it is important to have slack in the cable to allow for they are on the right lines in their approach to investigat- movement. ing Christchurch engineering lifelines. What has been Lesson 33: Different sections of the community will achieved so far will be a very significant step in enabling respond in different ways to the stress imposed on it Christchurch to better withstand an earthquake. by an earthquake and these issues may affect restoration priorities. 9.2 Geotechnical Aspects Lesson 34: Earthquakes seek out the weakest buildings and if these are historic buildings work needs to be undertaken in advance to ensure their retention. This Introduction will apply for engineering lifelines as well. At 4:31 am on 17 January 1994, a magnitude 6.7 earthquake struck the Los Angeles metropolitan area, causing Lesson 35: The public need to be aware of the risks direct damages estimated at US$15 to US$20 billion and related to gas leakage and hazardous substances making this the single most costly natural disaster to following an earthquake. It is important that those affect the United States of America. The earthquake, involved in the restoration of services are not ex- centred beneath the city of Northridge, resulted in 61 posed to even further hazards. deaths, caused serious injury to a further 1,500 people, and left 22,000 homeless. Widespread damage to infra- Lesson 36: Disagreement over damage estimates, the structure and lifelines occurred, extensive landslides and cost of repairs and the level of expected performance will occur following an earthquake and it is important ground failures were triggered, and significant structural that these are resolved quickly so that restoration of effects on residential and commercial buildings resulted services is not held up. in more than 3,000 being declared unsafe for re-entry. The following summary of geotechnical aspects of the Lesson 37: Although there are many successful mitiga- Northridge earthquake is based on two specific publication efforts, many people elect not to undertake them tions, Goltz (1994) and Stewart et al (1994), as well as and simply blame the earthquake. With the increase personal observations during site visits in mid-August of knowledge in Lifeline Engineering this approach 1994. may not be so acceptable for services in the future.

Lesson 38: We need to bear in mind the factors that Geology and seismology motivate service operators and building owners re- The Northridge earthquake occurred on a south-dipping lating to the return on investments, lowering ex- fault adjacent to the north-dipping structures involved in penses, curtailing losses and avoiding liability, and the 1971 San Fernando Earthquake (M = 6.6). The point out that although the cost benefit of measures rupture initiated about 18 km below the surface, propa- may be difficult to quantify, there are very definite gated up-dip to a depth of about 5 km (Figure 9.2), and benefits in undertaking mitigation measures. had a source duration of approximately 6 seconds. As the rupture did not reach the ground surface, it is generally New Zealand/Los Angeles Workshop • 233

considered to have occurred on a “blind thrust” that had elongate zones that experienced VIII and, very locally, not previously been identified as an active fault capable IX. of causing widespread damage. Six aftershocks of magnitude M > 5 were recorded, the largest having M = 6.0, Ground motions and site effects and these all had a similar mechanism to the main shock. The maximum horizontal acceleration recorded was Vertical uplifts to a maximum of about 60 cm in the 1.82 g on thin alluvium over bedrock at Tarzana some 17 epicentral region were recorded from GPS resurveys km from the fault rupture surface, and at the Jensen following the earthquake. Filtration Plant a value of 0.98 g was measured on soil The Los Angeles area is underlain by a variety of only about 10 km distant (Figure 9.5). At Pacoima Dam consolidated sedimentary, volcanic and crystalline base- left abutment an MHA value of 1.58 g was recorded on ment rocks (Figure 9.3), and active north-south com- rock, and the next largest value on rock was 0.49 g at a pression is occurring across the region. Extensive site 36 km southeast of the epicentre. Contours of maxi- Pleistocene alluvial and marine terrace deposits are also mum horizontal acceleration, ignoring those at Tarzana present, whilst basin areas such as the San Fernando and Pacoima Dam because of possible topographic in- Valley are infilled by Holocene age (i.e. postglacial) fluences, reflect both the south-to-north rupture mecha- sediments (Figure 9.3). Damage patterns from the 1994 nism and the effects of local site geology (Figure 9.6). In Northridge earthquake show strong correlations with general, the vertical ground accelerations were about geology, the majority of red-tagged (= unsafe) buildings two-thirds of the horizontal values obtained. being located on softer surficial Holocene deposits in the The distribution of red-tagged (= unsafe) buildings can San Fernando and Santa Clarita Valleys (Figure 9.3). be assumed to reflect variations in ground conditions and The Modified Mercalli intensity plot (Figure 9.4) simi- shaking intensities, but may also be influenced by the age larly reflects the distribution of softer Holocene sediments, and type of construction employed. It is significant that, with intensities of VII throughout most of the epicentral in addition to the Northridge epicentral area, a number of area (about 250 km2) on which are superimposed local other concentrations of structural damage are apparent at

Elysian Park Fault

Hollywood Newport Fault Inglewood Fault San 110 Fernando Los Angeles San Fault Gabriel 2 Fault V 10 e rdugo M

134 Santa Monica Mtns Pacific Ocean ountains 5

San 170 210 Fernando 405

S Valley an G 101 ab ri S el Northridge anta M S C o a unt n * la ains ta ra S R u 5 iv sa 118 er na F a Santa ult Susana Mtns 14 N ew hall

Northridge Rupture Bridge failures *Focus

Figure 9.2: Geological model of the Northridge earthquake (view to the south east) 234 • Risks and Realities

Sherman Oaks, Hollywood, central Los Angeles, The soils in many of these areas are also mapped as Newhall, Santa Monica, and the Santa Clarita Valley potentially liquefiable, and whilst liquefaction is not (Figure 9.3). These areas are all underlain by deep generally evident, the soft soils may have yielded under Holocene deposits in obvious alluvial basins, and in the relatively high ground accelerations to produce the some areas, such as Sherman Oaks and Santa Monica (to observed damage. It has been suggested, however, that in the north and south of the Santa Monica Mountains, the Santa Monica area site amplification may have respectively), energy focusing due to basin and edge occurred because of the presence of deeper, older and effects is considered to have been a significant factor. stiffer soils.

ANGELES

LOS PADRES NATIONAL 14

NATIONAL FOREST

FOREST

Piru

Filmore 126 S AN 150 14 G Santa AB RI Clarita EL SANT A S US AN 5 MO M UNT 23 IG OU A M AINS B NT OU AIN NTAINS San Fernando

118 118 118 Simi Valley Northridge 34 S ILL 5 H ★ Burbank 23 MI SI 105 101 2 Sherman Glendale Pasadena Thousand Oaks EPICENTER Oaks 134 S 101 AN Universal TA City

TAINS MOUN Hollywood MONICA 27 104 10 23 Pacific 1 Malibu Palisades

1 10 10 Culver City 5 Santa Monica Los Angeles 110

Sa 42 nta Marina del Rey Watts M Inglewood 605 Scale (mi) o 710 ni 2 024 68 ca B a 1 y Compton P 91 A Redonda Beach C

IFIC 1 1

103 110 O Long Beach C E A LEGEND N

Holocene sediments

Pleistocene alluvial and marine terrace deposits

Rock - undifferentiated consolidated sedimentary, volcanic and crystalline basement rocks

Figure 9.3: Damage patterns from the 1994 Northridge earthquake and generalised geologic conditions in the Los Angeles area New Zealand/Los Angeles Workshop • 235

119.0 118.5 118.0

5 V Palmdale 5

F 5 + + 7 + 34.5 5 5 Fillmore 7 Sta. Clarita 7 8 5 7 Sta. 7 V Susana 5 9 VII Mtns 9 8 IX San Fernando 8 VIII 8 7 9 7 5 6 8 9 6 Chatsworth 9 9 8 5 6 8 7 7 5 6 Oxnard 8 7 8 8 IX 6 8 5 9 8 Glendale 6 VI 7 7 6 Tarzana 5 6 7 Santa Monica Mtns 7 5 4 6 7 VIII 8 6 9 5 6 9 7 8 8 6 VII 5 8 6 6 5 5 34.0 + Sta. Monica + 6 7 7 Montebello+ 5 5 5 5 VI 5 6 5 6 5 5 Epicenter 5 5 5 5 9 Felt at Intensity 9 5 020 6 Torrance 5 5 F Felt 5 5 5 km 6 5 5

Figure 9.4: Contours of Modified Mercalli Intensity (MMI) observed during the Northridge earthquake

Ground failure and landslides resulting in intact block movements at some localities, Ground failure by soil liquefaction and dynamic ground but the evidence for such a failure mechanism has yet to compaction is considered to have occurred up to 50 km be demonstrated. In central Los Angeles partial liquefac- north and south of the epicentre, and clear evidence tion of Holocene alluvial and dune deposits was also exists for lateral spreading due to liquefaction in sandy suggested as a cause of distress in buried utility pipes, but fills at Redondo Beach (Figure 9.3). In the San Fernando other infrastructure components (such as pavements) did and Simi Valleys widespread “deep” liquefaction has not behave similarly and the observed failures may have been postulated with water table depths of 6 m to 8 m, more to do with age and the type of pipe welds. Of

0.1 Alluvium Free field Structure 2nd % 16th % 50th %

Peak horizontal acceleration (g) 84th % 98th % 0.01 1 10 100 Distance to surface projection of rupture (km)

Figure 9.5: Comparison of selected recordings of mean peak horizontal acceleration on alluvium with predictions based on the attenuation relationships developed by Boore et al 236 • Risks and Realities

0.4

0.2 14

0.2 126 0.4 126 150 14

0.4 0.6

23 0.8 5

EPICENTER 210

118 118 118 0.8

0.8 0.2 5 23 ★ 405 101 0.4 SURFACE PROJECTION OF 27 2 APPROXIMATE FAULT RUPTURE PLANE 0.4

134 0.6 101 0.2 0.2

110 0.2

101 10 0.3 0.2 23 0.4 0.4 0.4 0.2 0.6 10 10 1 0.8 0.2 N 110

710 Scale 605 2 0 2 4 6 8 miles 0.2

1 P 91 A

C 0.1 I 710 F 1 I 1 C 110 103

O 0.2 Legend C Rock Station E A Soil Station N

Note: Site classification was provided by the owner/agency except for USGS stations. Sites were typically classified by surficial geology; however, USC stations were classified based on Seed et. al. (1976) in which surficial materials with Vs > 800 m/s are designated as "rock" sites.

Figure 9.6: Contours of maximum horizontal acceleration based on recordings at rock and soil sites

particular significance in terms of the observed “partial” Palisades (Figure 9.2). The predominant modes of fail- liquefaction is the potentially much more severe effects ure were shattered ridges, shallow surficial slides, that would have resulted had shaking continued for a rockfalls, and rockslides. Their impacts included road longer duration. closure, direct property damage, and increased susceptibility to subsequent storm-induced debris flows. The Northridge earthquake caused hundreds of scattered rockfalls and landslides throughout Los Angeles and Ventura Counties up to a distance of approximately 25 Performance of geotechnical structures km from the epicentre. The principal failure sites were in Many geotechnical structures such as dams, hillside the Simi Hills, Big Mountain, the Santa Monica, San structural fills, earth retaining structures, and solid waste Gabriel and Santa Susana Mountains, and along the landfills were strongly shaken by the Northridge earth- marine terrace bluffs in Santa Monica and the Pacific quake, and although most performed very well there New Zealand/Los Angeles Workshop • 237

were several notable failures. There are approximately active faults exist beneath the Canterbury Plains near 65 dams, both concrete and earth- and rock-fill, within 40 Christchurch which could produce similarly high km of the epicentre, and many act as debris dams and had ground accelerations. The present seismicity model little active reservoir storage at the time of the earth- for Christchurch relies heavily on work completed on quake. With the exception of the concrete arch Pacoima the Porters Pass Tectonic Zone, and on inferences Dam, all dams performed well and observed damage was regarding other major active faults to the west and confined to surface cracking and some shallow slides, northwest. and no significant damage was reported to any of the earth structures. At Pacoima, however, which was within • Severe shaking of the thinning sediment wedge against 10 km to 12 km of the rupture, directional shaking effects the Port Hills is considered possible given the ob- produced extensive rockfalls and the left abutment ap- served effects of the Northridge earthquake, and the pears to have moved downstream by some 10 mm to 15 potential for significant ground amplification effects mm relative to the arch. Pacoima Dam was subjected to should be further evaluated. Extensive liquefaction a peak horizontal acceleration of 1.25 g on the upper left in the saturated sandy and silty sediments of the abutment during the 1971 San Fernando Earthquake, and Avon-Heathcote estuary is still regarded as probable, one instrument on the crest recorded about 2.3 g during especially given the high water table there. the 1994 Northridge earthquake. Investigations are pres- • Local rockfalls and shallow landsliding must be ently being carried out to determine what (if any) reme- anticipated on the Port Hills, with consequential dial measures are required for the Pacoima Dam, which potential for damage to houses and injury to occu- is still operational. pants. Some protection measures can be implemented The only significant flow slide to occur during the 1994 for house sites in hill or footslope locations, but Northridge earthquake was a small tailings dam at a sand prediction of specific problems or failure sites is not and gravel aggregate mining operation in Tapo Canyon considered generally feasible. near Simi Valley. The dam was approximately 25 m • Well engineered geotechnical structures can be ex- high, was located only about 12 km from the rupture pected to perform satisfactorily in a similar earth- source, and experienced MHA values in the range 0.4g quake, and the generally sound performance of the to 0.6 g. Portions of the dam slid up to 30 m downstream, Los Angeles dams, fills, etc., is most encouraging. releasing the saturated tailings which then flowed down Localised ground cracking, especially in older fills or existing watercourses, and the failure is thought to have stiff liners, must still be anticipated, however. been caused by liquefaction. Some hillside structural fills and earth retaining structures experienced failures, and significant damage was reported to concrete and References concrete crib retaining walls subjected to horizontal Goltz, J D (ed.) (1994). The Northridge, California accelerations of about 0.6 g. The overall performance of Earthquake of January 17, 1994, General Recon- landfills during the Northridge earthquake is regarded as naissance Report Technical Report NCEER-94- encouraging, with no signs of major instability and 0005 National Center for Earthquake Engineer- damage limited to cracking of the soil covers due to a ing Research State University of New York at variety of possible causes. Buffalo. Stewart, J P, J D Bray, R B Seed, N Sitar (eds.) (1994). Some implications for Christchurch Preliminary Report on the Principal Geotechnical There are a number of conclusions that can be drawn Aspects of the January 17,1994 Northridge Earth- from the Northridge earthquake in relation to the possi- quake, Report No UCB/EERC-94/08 Earthquake bility of seismic damage in Christchurch, as follows: Engineering Research Center College of Engi- neering, University of California at Berkeley. • Although the Northridge earthquake was located beneath Los Angeles, rather than 50 km to 100 km distant, it produced similar shaking intensities (MMVII and locally MMVIII/IX) to those antici- 9.3 Transportation pated in Christchurch. The high horizontal accelerations experienced were to a significant ex- Introduction tent offset by the relatively low duration (6 to 10s) of Most engineering services in the greater Los Angeles shaking, and this (together with the timing of the area were essentially fully restored within three days of main shock) was critical in limiting the potential for the 17 January 1994 Northridge earthquake. As far as damage and loss of life. they impacted directly on the public, many were restored within 24 hours. The exception to this was the transpor- • Further seismic studies are needed to determine if tation network. 238 • Risks and Realities

In an area where goods and people movement is domi- five new stations were opened and bus and shuttle nated by the freeway system, the earthquake produced services provided. Ridership leapt from 1,000 to 22,000 severe and long-lasting transport disruption. A total of passengers per day, stabilising later at 5,000. By Febru- eight overpasses collapsed blocking some of the busiest ary, 13 of the 22 highway lanes were usable, with one freeways. More than a month later there was still designated carpool lane in each direction. Other recov- massive traffic disruption with sections of (Interstate) I- ery measures included detours onto old surface roadways. 5 and I-10 still closed. Seven months later all routes were open with some temporary sections and work was still Simi Valley (R-118) corridor proceeding at a number of sites. A completion date of December 1994 has been set, almost one year after the The R-118 freeway was closed due to severe damage to earthquake. bridges (see 4,5,6 on Figure 9.7) requiring 187,000 vehicles per day to find alternatives. Fortunately, there Although the damage to commercial and residential was a good grid system of surface arterial highways in buildings was severe close to the epicentre and the the area with excess capacity available to accept diverted impact of this type of damage on the local area continued traffic. The Ventura County Metrolink rail line also for months, within days the event became a “transporta- assisted after a three-day delay needed to clear a derail- tion earthquake”. There were dramatic pictures of col- ment. The line was extended 25 km and three new lapsed sections of freeways in the media and a wider stations built. Ridership increased from 2,100 to 3,000 effect on commuters living well beyond the area of the boarding per day. damage. This emphasises yet again the vital role played by the media in such events. In the days and months At the Balboa Boulevard crossing of R-118 (site 6) the following the earthquake extensive traffic detours, bus overbridge carried power, water, gas and telephone and rail system expansion and demand management services. The water lines suspended from the structure were required to restore mobility to the affected corri- ruptured and washed out the soil beneath an abutment dors. Information services and the media generally were which then settled. The bridge satisfactorily resisted the vitally important in implementing these measures. seismic forces, its damage and closure being a direct result of the damage to another lifeline. The information in this section has been summarised from material obtained during a visit to Los Angeles as I-10 Freeway corridor part of the Joint New Zealand/Los Angeles Lifelines The Santa Monica Freeway (I-10) is considered to be the Workshop on 15 to 18 August 1994. In particular, busiest freeway in the world, carrying over 340,000 gratefull acknowledgement is made for the assistance of vehicles per day. The earthquake heavily damaged two Mr Jack Hallin of Caltrans and Mr Thomas Conner of the long bridges spanning four major surface streets (see 7,8 City of Los Angeles. A full list of references is attached. on Figure 9.7) at a location where the average daily flow was 271,000 vehicles per day. Impact on the freeway system The impact on the freeway system occurred in three areas The freeway traffic was diverted to the parallel arterial (see Figure 9.7): Golden Gate Freeway I-5 and SR-14 at street system, including preferential treatment for carpool Gavin Canyon, Simi Valley Freeway SR-118 and Santa lanes. An example of the routing used, taken from a Monica Freeway I-10. publicity leaflet, is shown in Figure 9.8. Carpool or high occupancy vehicles (HOV) were those with two or more occupants. Assisted by the computer controlled traffic 1-5/Route 14 corridor signal system (ATSAC Smart Corridor) delays on carpool The Golden Gate Freeway (I-5) is Southern California’s routes averaged 3 to 5 minutes, with 10 to 15 minutes on main link with Central and Northern California and is the longer mixed flow detours. Existing transit (bus) especially important for produce and goods movement. services were expanded but experienced only modest The Antelope Valley Freeway (Route 14) provides resi- increases in ridership. dents from the Santa Clarita and Antelope Valleys with commuter access to employment centres in Los Angeles Retrofitted structures and surrounding cities. Prior to the earthquake, there were 22 freeway and surface highway lanes carrying Following the 1971 San Fernando earthquake, it was more than 216,000 vehicles per day through the moun- found that the unseating of bridge decks at abutments and tain pass where I-5 and R-14 intersect. After the earth- expansion joints was the principal cause of collapse. A quake only six lanes were usable. programme of retrofitting existing state-highway bridges with hinge restrainers was started. This Phase I pro- Fortunately, the Santa Clarita Metrolink rail line parallel gramme was extended in 1988 with a Phase II pro- to I-5 was undamaged. This line was extended by 86 km, gramme of strengthening single-column piers by enclos- New Zealand/Los Angeles Workshop • 239

KEY S an 1 Gavin Canyon Undercrossing (I-5) 1 3 G a 2 SR14/I-5 Separation and Overhead (southbound) 2 br 3 SR14/I-5 North Connector ie l 4 Bull Creek Canyon Channel Bridge (SR118) Santa Susana M 5 Mission-Gothic Overcrossing (SR118) Mts San ts Fernando 6 Balboa Boulevard Overcrossing (SR118) 7 Fairfax-Washington Undercrossing (I-10) 8 La Cienega-Venice Undercrossing (I-10) 6 Chatsworth 5 4 Sunland Tujunga San FernadoValley S an G Northridge Ve a Epicenter rd br u ie goMt l Canoga Sun Valley M Park Reseda s ts S Altadena Van Nuys a Burbank n Tarzana Ra H fa North il el Hollywood ls Woodland Hills Encino Sherman Pasadena Oaks Glendale

onica Mts ta M San South Pasadena Hollywood Alhambra West Hollywood

Beverly Hills Pacific Palisades Monterey 8 LOS ANGELES

Topanga Beach East Los Angeles Santa Monica Culver City 7 Commerce Ladera Huntingdon Heights Venice Park Maywood Florence- Graham Bell Bell Gardens Inglewood Walnut Cudahy Marina Park Santa del Ray Los Angeles Watts Downey Intl Lennox Lynwood

Monica Willowbrook Hawthorne El Segundo Norwalk Bay Lawndale Compton Manhattan Beach Paramount Gardena Bellflower

North Long Beach Lakewood Redondo Beach Carson Torrance

Figure 9.7: Location map of bridges with major damage ing them in steel jackets and some associated strengthen- quakes measured in terms of year designed, column ing of abutments and footings. A further programme type, abutment type, skewness, bent redundancy, and (Phase III) was initiated following the Loma Prieta potential for drop-type failure. earthquake and involves retrofitting multi-column and • The seismic hazard at the bridge site measured in complex structures. terms of maximum expected peak rock acceleration, Caltrans staff evaluated state highway and local bridges maximum expected duration of ground shaking, and and selected these for possible retrofitting based on the local soil conditions. bridges vulnerability, the seismic hazard at the site and • The impact of a possible bridge failure on the local the impact of failure. The evaluation considered: community measured in terms of average daily traf- • The bridge’s vulnerability to strong-motion earth- fic on the structure, average daily traffic under or 240 • Risks and Realities

FAIRFAX PICO BL NORTH VENICE BL

OVERLAND

WESTWOOD BL LA CIENEGA 405 CADILLAC LA BREA PICO BL WASHI NGTO N BL Single Occupant SAWTELLE FAIRFAX Eastbound HOV Vehicles must exit APPLE ST. at La Brea lane begins at ROBERTSON BL 2

Overland LA 2 AVE 2 2 10 2 NATIONAL BL WASHINGTON ADAMS BL 2 Westbound HOV CIENEGA lane begins at JEFFERSON Western

Single Occupant LA BREA Vehicles must exit at Robertson VENICE BL HOV Only Alternate Route WESTERN NATIONAL BL WASHINGTON Mixed flow Alternate Route

Figure 9.8: Re-routing of freeway traffic with preferential carpool lanes

over the structure, type of leased air space (residen- freeway structures that collapsed. A free-field motion tial, office, parking, storage), type of facility crossed, recorder fixed to this bridge showed a peak acceleration route. at deck level of 1.83 g. The single-column piers of this structure had been retrofitted with steel jackets in 1991. Those bridges identified in this process were then re- No significant damage occurred to this structure. viewed in detail by experienced Caltrans design engineers. This process was only partially complete at the It is apparent that the bridge retrofit programmes were time of the Northridge earthquake. Also, while the work effective and Caltrans are seeking to have funds pro- required in Phase I (hinge restraint) and Phase II (single vided so that the programme can be speeded up. It should column jacketing) was nearly complete there had been be noted, however, that the Northridge event had a very little progress with the more extensive and costly duration of only 9 seconds, and a longer duration event task of retrofitting multi-column pier bridges. State- would be a more severe test. wide this task was only 2% complete with a further 7% Other initiatives to reduce risk in future designs and underway. either adopted as policy by Caltrans or under active Of the seven concrete bridge structures that collapsed consideration include: during the Northridge event, none had been retrofitted in • avoidance of large skews in geometry; the Phase II programme. The eighth bridge (site 6), Balboa Boulevard Overcrossing of SR-118, was dam- • consideration of alternative routing to reduce the aged by subsidence caused by a waterpipe rupture and so number of multi-level crossings; and is not considered in this discussion. • use of freeway over-crossings rather than under- The two adjacent bridges on SR-118 (sites 4,5) had been crossings. evaluated by Caltrans as not requiring retrofit but failed due to shear in short columns. The other five bridges had Impact on other transportation systems been scheduled for Phase II retrofit but the work had not yet been started. These bridges either failed when deck Very little damage occurred to the local roading system sections came off their supports or by shear failure in remote from the freeway system. Any problems were short columns when the hinge restrainers were sufficient due to landslides blocking roads, rather than structural to keep the deck on the supports. failures.

In contrast to these collapses there was no significant Airports and railway tunnels were closed briefly for damage to structures that had been retrofitted under the inspection as a precautionary measure but no serious Phase II programme. Some of these were in locations damage was found. where they were exposed to shaking at least as severe as At Los Angeles International Airport (LAX) minor flood- that impacting on the collapsed structures. Some 24 ing occurred after the sprinkler system was activated. retrofitted bridges experienced peak ground accelerations The airport closest to the epicentre was Van Nuys Air- over 0.5 g and all performed satisfactorily. An example port, some 6 km to the southeast. The glass panels of the is a 320 m long curved box girder structure at the I-10/I- control tower broke during the earthquake, interrupting 405 interchange located some 6 km west of the I-10 operations of the tower. New Zealand/Los Angeles Workshop • 241

A 64-car freight train passing through Northridge de- Secondly, there is the recovery from that damage — railed at the time of the earthquake. There were 16 tank involving human and material resources. The implemen- cars carrying sulphuric acid causing a 8,000 gallon acid tation of emergency plans, co-ordination with other spill. There was also a 2,000 gallon diesel fuel spill from related utilities, interaction with funding agencies and the locomotive. This incident was not immediately decision makers, public relations and information gath- responded to by the Los Angeles Fire Department due to ering and analysis. the number of priority calls at the time. Train service was restored two days later. Thirdly, among the three areas of interest is preparation and contingency planning including design and investigation adequacy, emergency planning, mutual aid, miti- References gation of damage, retrofitting, material and plant avail- Yates, Robert R (1994). Northridge Earthquake Trans- ability, investigation, analysis and reporting, damage portation Response After-action Report, City of prediction and risk analysis. Los Angeles Department of Transportation, March 1994. In this chapter, the experience of visiting the sites of two major recent earthquakes in southern California, focus- Yates, Robert R (1994). Northridge Earthquake — Long ing on water and wastewater pipelines and facilities, is Term Transportation Recovery Task Force, City presented under the above three headings. Throughout of Los Angeles Department of Transportation, the text a particular statement of lessons gleaned from the June 1994. visit is followed by referenced examples illustrating the point. In this way it is hoped that readers with an interest Buckle, I G (1994). “The Northbridge, California earth- in utility management will be able to find parallels to quake of January 17, 1994: Performance of high- their own concerns. way bridges”, Technical Report NCEER-94-0008, National Centre of Earthquake Engineering Re- search, State University of New York at Buffalo, The earthquake and accompanying March 1994. damage

EQE International (1994). The January 17, 1994 Pipelines Northridge, California Earthquake. Basic rules for reducing buried pipeline damage during earthquake are now well known, for example as devel- Seismic Advisory Board (1994). Background Report oped and set out by the Wellington Earthquake Lifelines DBL1 — Bridges, February 1994. Group. Lund, Le Val (1994). Northridge Earthquake, January Applicability of these findings was confirmed during the 17, 1994, Lifeline Performance (Draft), ASCE Northridge earthquake where more recently laid services Technical Committee on Lifeline Earthquake generally featuring more ductility and/or flexibility in Engineering, February 1994. their material and joint type showed remarkable resistance. Close analysis attempting to detail correlations between pipe damage and shaking intensity, soil type, 9.4 Water and Wastewater material jointing and age was lacking, but the impression Utilities gained was that New Zealand lifeline studies have probably overestimated the number of breaks likely for the Introduction various combinations of pipe size, soil type, material, age and earthquake intensity. Discussions about the performance of water and wastewater pipelines and facilities during seismic activ- Distribution water main breaks in the City of Los Ange- ity inevitably centre on one of three main areas of les numbered over 1100, almost half of which (47%) interest. There is the earthquake itself and its accompa- occurred in the West Valley District which includes the nying damage — its size, location, depth, intensity, time San Fernando Valley epicentre area. The other half were of occurrence, its primary manifestation, the explana- to the east and south but in broad bands corresponding tions of its origins and propagation and the scientific and with surface damage. Breaks were concentrated in cast subjective measurement of its effects. There is the dam- iron piping with rigid joints or in steel pipes with some age measured in loss of life, injury, homelessness, social level of corrosion. disruption, disorientation, delays, damage to structures (domestic and commercial) to infrastructure and to the ‘Boil water’ notices were posted and the requirement natural environment as well as consequential damage, was gradually lifted as connection was restored with the aftershock effects, and risk of fire. final notice removal occurring 12 days after the event. 242 • Risks and Realities

Damage to major aqueducts supplying Los Angeles City mm to 375 mm diameter), the number of breaks likely in Water and Power occurred at 20 locations, the most an MM VIII earthquake, taking into account pipe ductilily significant damage being to steel pipelines with lap and soil stiffness. The results need to be checked against welded joints where the vulnerability occurs at the curved breakages in actual events, but that type of analysis was portion of the bell, resulting in circumferential cracking. not available for Northridge. Pullouts at mechanical joints also occurred as did buck- ling in steel pipes where large strains were imposed. In the San Francisco/Oakland area a higher awareness of regional geology was evident amongst Utility managers Discussion with officers from the Pacific Gas and Elec- and for the Loma Prieta earthquake plots relating water tricity Utility based in San Francisco confirmed the main breaks to liquefiable soils had been made for the vulnerability of screw joint steel and cast iron pipe and marina district showing the relationship between vertical noted their company’s programme to replace with settlement caused by liquefaction and pipe breaks. polyethylene. Statistics provided by the Southern Cali- fornia Gas Company showed that in Los Angeles, of the One interesting plot of above ground damage caused by 181 distribution main failures, 154 were in metal pipe the Northridge quake superimposed on a regional relief and 27 in plastic. While their breakage rate seemed map showed concentrations of damage remote from the surprisingly low, the Gas Company had to deal with epicentre. These were either very close to the intersec- some other very large numbers: tion of low ranges of hills with the sediment filled basins or appeared to follow old river channels across the flats. The possibility that this shows magnification for effects Number of outages restored by company 122,886 close to hilly areas raises questions for Christchurch Customers who turned gas off unnecessarily 107,865 relating to vulnerability of the city’s southern boundary Total number of leak investigations 61,172 defined by the Port Hills. A GIS-based analysis of the likely sewer system damage identified 54.4 km of mains in the high risk category (see Major facility damage “Recovery” below). Of these, 16% needed emergency Recovery to full operation of the wastewater and water repair, 49% sustained damage that may require repair treatment facilities located relatively close to the epicen- and 35% were undamaged. Most breaks occurred in tre (i.e. within 15 km) was remarkably fast despite some sewers smaller than 400 mm diameter, which make up significant damage (see Table 9.1). 90% of the collection system. Close to the epicentre damage occurred regardless of soil type, but with more These fast recovery times probably contributed to the distant areas the liquefaction potential of the soil became somewhat cursory reporting that the damage has re- a factor in the degree of damage. Los Angeles City ceived. Managers were justifiably proud of the recovery reported that their large diameter reinforced concrete performance and outside investigators tend to lose inter- mains were undamaged, that breaks occurred most fre- est when the plant is reported fully operational. In fact it quently at joints and in locations with a high water table. was a salutary experience to check these installations more closely and receive detailed descriptions of the damage, (much of it non-structural) and the costs to Correlation of breaks with geology remedy it. There was disappointment at Northridge for those members of the team from New Zealand hoping to examine At the Jensen Filtration Plant for example, much was work aimed at correlating damage (to underground pipes made of one major break in an 84 inch diameter influent for example) with soil structure and geology or with pipe pipe. This plant has instituted major mitigation work material and age. It was clear from the Utility Managers following the 1971 Sylmar earthquake and operators that their focus was on assessing and repairing damage, were satisfied with the performance and the fast restora- while at the same time maintaining the documentation tion to full service. Nevertheless a detailed listing and records that would secure the Federal funding that showed that damage though minor was extensive and the was available. Even six months after the event these were estimated cost to repair amounted to US$4.8 million for some pressing, overriding concerns. But, particularly (NZ$8.0 million). in the case of the private utilities, once the supply to the Similarly, at the new Don Tillman sewage treatment customer was restored, there was little evidence of a plant located 7 km from the epicentre, which was back in desire to establish the correlations sought by New Zea- operation within hours, the listing of damage runs to land team members. seventeen pages with repairs estimated at US$1.5 mil- An exercise carried out by the Wellington Earthquake lion (NZ$2.5 million). It was clear in these examples that Lifelines Group calculated, for the 375 km of water while overall facility performance may be excellent in supply pipe in the Hutt City network (ranging from 100 terms of returning to full operational capacity, miscella- New Zealand/Los Angeles Workshop • 243

neous damage can be both extensive and expensive. Recovery Furthermore, much of this minor damage can be avoided by good design or by subsequent careful mitigation Use of GIS work. The use of computerised maps and GIS to assist decision making during the recovery period can provide a power- A damage prone area perhaps not well identified by ful tool hitherto unavailable to Utility Managers. How- treatment facility managers occurs at free water surfaces ever its usefulness is enhanced only in proportion to the in structures such as ponds, clarifiers, tanks and reser- amount of preparation of the essential underlying map- voirs. Forces developed by sloshing water can cause ping layers that have occurred prior to the event. Manag- damage well beyond that predicted by normal horizontal ers where facilities are or can be plotted in a CAD or GIS seismic loads. Roof structures, flight systems, test equip- environment need to consider the demands that will be ment and floating or poorly secured apparatus close to placed on the mapping system during the emergency and the surface are prone to damage. The debris can then ensure the system is ready to meet them. Acquiring and cause secondary problems at gates, pipe outlets or other digitising base data is time consuming and not possible discharge routes. during the recovery period.

The clearest example was provided by the City of Los Stand-by plant Angeles Bureau of Engineering which had the problem The failure of stand-by power generation plant either to of locating sewer damage. A real spur to the exercise was fire from a “black start” mode or to remain in service for provided by the Federal funding which was available for an extended period was a common complaint amongst 90% of the repair work cost, providing the City Council both utility managers and custodians of large buildings. located the damage and convinced FEMA (Federal Emergency Management Administration) it was earth- Problems were varied and highlighted the need to ana- quake related. Once water was available some severe lyse carefully the many circumstances which could go damage evidenced itself, but closed circuit television wrong. In the 1989 Loma Prieta earthquake, the East Bay (CCTV) was necessary to find structural damage that Municipal District sewage treatment plant generators was not causing any discernible interference to flow. A which normally supply surplus power to the city grid, means of prioritising the CCTV work was required to came on line after the city power failed and immediately reduce the 11,400 km of sewer to a manageable level, tripped out on overload. A switching error had allowed while retaining confidence that the sewers likely to be the plant to attempt to supply the whole city with power. damaged were televised. To do this the Bureau hypoth- More common were complaints that stand-by plant failed esised that known surface and subsurface damage were to fire automatically from a black start on high load, and the best indicators of likely sewer distress and accord- there was concern that regular testing of this plant was ingly plotted in a GIS environment the location and not thorough and did not test start-up on full load. Plant condition of all buildings inspected for damage, water frequently showed an inability to remain operational leak positions and street, kerb and sidewalk damage. under load for an extended period. Total grid scores were counted and the totals were used Examination of stand-by plant needs to go beyond the to set inspection priorities. This lead to 54.4 km of sewer plant itself. If cooling systems depend on an external being identified as high risk and requiring CCTV inspec- water supply the plant will obviously fail if the water tion. supply has been damaged. Security of such things as fuel storage, fuel lines, switchboard and control cabinets Earthquake information must all be checked and upgraded if necessary. Early accurate information describing the nature of the

Plant Type Capacity Damage Recovery to Full Operation Significant, but Jenson Full water treatment 400 million gpd localised 1 week Van Norman Full water treatment 600 million gpd Significant 6 days Donald Tillman Sewage treatment 150 million gpd Minor 10 - 12 hours Glendale Sewage treatment 20 million gpd Minor 10 - 12 hours

Note: Los Angeles’ major sewage treatment plant (the Hyperion, located on the coast) suffered no significant damage and continued operation under its own power.

Table 9.1: Time for recovery to full operation for Los Angeles sewage and water treatment plants 244 • Risks and Realities

earthquake is essential for utility managers whose re- Preparedness sponsibilities are over a large area since it can influence the early decisions made and save wasted effort. The Recovery plans size, location, depth to focus, duration and type of fault Being prepared for an earthquake requires commitment movement must be immediately available so that, cou- to both the things that can be done now (design of new pled with an understanding of the area geology intelli- work, mitigation by way of replacement, duplication, gent guesses can be made at an early stage, of where the retrofitting, securing, strengthening, relocating) and plan- vulnerable and therefore damaged areas are likely to be. ning the emergency procedures that operate when the event occurs. This is disaster recovery planning and for For a utility with wide spread facilities this knowledge Utility Managers is quite distinct from civil defence allows efficient direction of scarce inspection resources. involvement. In Christchurch the Civil Defence emergency caused by a large earthquake is likely to be over in In Southern California each major utility is linked to a a matter of a week or two, by which time the threat to life network called CUBE (Caltrac USGS Broadcast of Earth- will have been controlled. For the Utility Manager, quakes facility) which records and maps all quakes disruption to the utility’s normal operation could last a within minutes of their occurrence. Users of this system year or more as reticulation damage is found and repaired have a range of options for the sorting and viewing of the and structural damage to the treatment and pumping growing database and have more or less immediate facilities is addressed. access to information about an earthquake event. Emer- gency control centres will thus have more or less imme- In California, State law in the form of SB 1841 requires diate access to earthquake information. local government and public agencies to adopt the Stand- ardised Emergency Management System (SEMS) which Aftershock records requires that, in times of disaster, local jurisdictions co- ordinate their response activities with the Counties, The role of aftershocks, in adding flesh to the bare bones which in turn must co-ordinate with the State. This of information provided by the initial quake, may not be requirement as it applies to water utilities originated as a widely appreciated. As secondary ruptures occur and result of problems in multi-agency co-ordination during their epicentre located, the pattern defines a subsurface the 1989 Loma Prieta quake and the Oakland fires of plane which in turn identifies both the fault surface and 1991. the extent of the rupture. Of more importance to the Utility Manager is the need to keep in mind the likelihood A clear message from utility ‘earthquakees’ is that the of aftershocks and ensure field crews understand the response plan should deal with organisation, procedures, ensuing dangers and problems. Aftershocks can cause lines of command and resources in terms of sustaining failure of weakened structures and endanger those in the relief effort rather than attempt to anticipate specific inspection or rescue roles. They can also disrupt repairs events. It can be assumed that if the utility’s skilled staff that have been carried out on a temporary basis. will be available to direct and undertake the recovery work, attention in response planning should be directed Mutual aid instead at ensuring that decision making personnel are in place and well informed, that communication is estab- Assistance between adjoining local authorities or utility lished and maintained, that relief and recovery effort is operators can be formalised to a much greater degree resourced and sustained and that publicity and public than occurs in New Zealand. In the United States this relations are properly addressed. takes the form of mutual aid agreements which set out the basis for assistance. For example, both the San Francisco Thus the Director of the water supply side of Los Angeles Water Department and the Eastern Bay Municipal Util- Department or Water and Power reported deficiencies in ity District (Oakland) sent fully-equipped teams to the emergency planning in such simple areas as providing Northridge area to assist the Water and Power depart- for relief workers who slept at the yard between shifts — ment of Los Angeles City in water main repair. To make it was impractical for them to get home; accommodating this work, compatibility of materials and skills is essen- the vehicles and staff that came from other areas; com- tial which clearly requires previous planning and co- municating with vehicles on non-city radio frequencies; ordination. One recent initiative by the OES (California providing meals when many food outlets were closed Office of Emergency Service) is to establish the Califor- (holiday) or damaged. nia Utilities Emergency Association (CUEA), a joint public/private autonomous organisation set up to achieve Code design requirements inter utility co-ordination. The association has 125 Of profound significance for all Utility Managers is the member companies as of September 1994. evidence in recent years that recommended design New Zealand/Los Angeles Workshop • 245

accelerations have been too low for both horizontal and ways, oil pipelines, major water aqueducts, emergency vertical directions. The Northridge event added to this services, high voltage transmission lines, the railway discrepancy by recording surprisingly high accelerations network, etc.). Relationships are included that, for exam- for a quake of this size (6.7), and reports are talking about ple, link each utility to likely damage and restoration further code revisions to accommodate them. For utility time for various sizes of earthquake. The program can operators the implications are clear. For all existing then simulate a particular earthquake and predict for the important facilities the design assumptions used to cal- authorities the likely effects, costs and restoration peri- culate and detail seismic resistance should be deter- ods. Calibration of the model is possible given the mined, compared with today’s code requirements, and information that comes out of events like Northridge. mitigation measures considered. Presentation of these Provided the data is available, the program can work at measures to the decision makers should include an any level of detail and an example was viewed where a assessment of the risks and costs of not proceeding as city’s water supply network was analysed and various described elsewhere in this report. levels of earthquake mitigation developed and costed for investment decision. East Bay Municipal Utility District were anxious about reservoirs designed in 1973 for a maximum horizontal East Bay Municipal Utility District based in Oakland has ground acceleration or 0.1 g. Code requirements today undertaken a similar simulation exercise on their water were set at 0.5 g and they were anticipating a new code system which conveys water from the headwaters of the level of 0.7 g. Mokelume River via three adjacent pipelines 90 miles in length, that traverse major faults associated with the San Those engaged in “walk-down” surveys to assess seis- Andreas group. Their analysis has lead to a recom- mic risk need to be aware that the high vertical mended programme of mitigation amounting to US $162 accelerations being recorded, coupled with the horizon- to 202 million. tal component means that unsecured installations will “walk off’ their foundations. Large tanks, for example in An interesting point made in the analysis is that a utility addition to being susceptible to side plate compression must achieve another balance — that between mitigation failure, can move off foundations and damage inlet and and preparedness. It was found more cost effective to outlet piping. Even more alarming is the magnification reduce customer outage times following an earthquake of acceleration that can occur at above-ground levels in by deliberately planning for extra maintenance crews to facility buildings. Forces on elevated plant such as over- deal with breaks than to undertake wholesale pipe im- head cranes, storage hoppers, etc. can be higher than provement as a mitigation measure. those occurring at ground level and their securing needs special attention. The Donald C Tilman Reclamation Plant (sewage treatment) which was located only 7 km from the epicentre, reported damage to three overhead cranes all of which were subsequently locked out of service pending structural evaluation.

Use of GIS in mitigation scenarios Establishing the appropriate level of investment in mitigation measures is clearly a vexed issue for a utility’s directors. There is a balance which establishes a sensible relationship between the likely cost of an element’s failure during an earthquake and the cost of mitigating works designed to reduce or eliminate the failure. The analysis assesses the probability of the design earthquake, assigns to the event the appropriate future date and then compares discounted damage costs with present day mitigation.

The use of interactive intelligent GIS environments to assess and cost these scenarios is well advanced in the Parking building, Northridge States. As an example, a firm of San Francisco based consultants has developed a program that covers the USA on a 25 km square grid on which is transposed soil data (average of first 30 m depth), major lifelines (free- 246 • Risks and Realities Interdependence of Lifelines • 247

Chapter 10 Interdependence of Lifelines

Introduction risk management of, say, a nuclear power facility where the expectations for reliability would justify the Task groups considered hazards to discreet networks cost of analysis. It is not appropriate for this project, and devised mitigation measures on that basis. But however, particularly given the uncertainties in hazard there is a strong interdependency between lifelines. and system knowledge. The uncertainty in defining Water supply systems need electricity. Communica- failure mechanisms within a particular service network tions systems are of critical importance to the operation would make establishing connections to other net- of, and disaster response to, all lifelines. Restoration of works highly problematic. water, power, sewer and telecommunications services is often dependent on the roading network. So, for the moment at least, we are left with a top down approach. Interdependencies arise in relation to network performance (if A fails then B fails), and in relation to recovery (we can’t fix A until B is fixed). Where to start? An assessment of interdependency in recovery is im- The key consideration for interdependency analysis is plicit in much of the individual lifelines assessments as the length of down time of the failed system. This is the consideration must be given to demands on other case whether in relation to operational or recovery services in deciding impact factors. But, as with the interdependencies. So, looking at a particular lifelines- Wellington project, a specific focus on inter-depend- dependent service, mitigation effort should be focused encies is still considered worthwhile, notwithstanding on reducing the downtime through failure to accept- the extreme difficulty in doing so. able levels. Acceptable levels of downtime will be based upon community expectations of the service concerned. How can interdependencies be provided for in analysis? Public expectations can be managed through preparedness programmes and thereby enable costs of mitiga- From an analytical point of view, there are two broad tion to be minimised, but in any case it is necessary that approaches available: top down, and bottom up. these expectations be known.

A top down approach would look at some valued The approach taken by the Wellington project team is service which is highly dependent on lifelines, and helpful here as a beginning. simply ask the question: what lifelines failures could cause loss of that service or an unacceptable delay in Table 10.1, based on the Wellington project, usefully restoring it? Those causes may then be addressed by illustrates indicative general interdependencies mitigation measures. in recovery between lifelines services. The interdependencies will, however, vary site-to-site, as The service in question could be a particularly impor- well as over time. Table 10.2 is an indication of service tant lifelines facility, for example a key substation, or recovery times from a general point of view, which, could be some other service, for example a hospital. again, is useful as a starting point, but there is no substitute for looking at the particular cases. A bottom up approach would treat the combined lifelines services as one system, and, component by com- In the real world, an exhaustive analysis would only be ponent, work through the full implications of any applied to an actual disaster situation where the prob- particular component failure. lems are known and clear objectives and priorities are able to be set. Project management techniques could The approach used for the individual lifelines analyses then be used to optimise interdependent recoveries. was, broadly speaking, a bottom up approach. Under- taken properly, this approach is exhaustive, and there- It would be of value in the meantime that essential fore very time consuming. It would be appropriate to community services highly dependent on lifelines be 248 • Risks and Realities

identified so that suitable arrangements any required viders and in the sharing of information about each for collaboration can be made. other’s networks. If a disaster were to occur today, there would be a high level appreciation of the need for priorities to be established across the services, and a high level of cooperation and expertise would be THESE ARE immediately available to address both priorities. DEPENDENT ON THESE

This team building effect of the lifelines project fosters

Water Supply Water Mains Electricity Storm Drainage Sea Transport Roading Railways VHF Radio Air Transport Broadcasting Fuel Supply Sanitary Drainage Telephone Systems Telephone Fire Fighting Standby Electricity an appreciation of lifelines as one system and will Water Supply 2 3 facilitate cross-utility mitigation measures where they Sanitary Drainage are identified.

Storm Drainage 2 When particular network owners are considering their Mains Electricity 2 32 33 2 31 budget allocations it is most important that the value to Standby Electricity 3 23 33 3 23 other lifelines of any proposed mitigation be taken into VHF Radio 1 132 3 22222 3 account. The success of this project can be judged, in Telephone Systems 2 111 1132 no small way, to the extent that they do this. Roading 2 223 222233 233

Railways 1 At the workshop, three groups looked at particular

Sea Transport 1 critical services as an exercise using the following

Air Transport 1 approach:

Broadcasting 1 2 11 1 Fuel Supply 3 213 2132 11 3 Tasks Fire Fighting 1 2 1 1 The seismic hazard scenario applies. Equipment 3 3233233333322 2 Select an element of service or network component which is dependent on other networks. Note: 3 = High Dependence 2 = Moderate Dependence 3 Agree on an acceptable outage time for 2 above. 1 = Low Dependence = No Dependence 4 Identify dependencies specific to 2. (Table 10.1 may help). Table 10.1: Interdependence of lifelines (first week after earthquake) 5 Identify (likely) damage to supply networks.

6 Establish recovery time for interdependent ele- Observations about ments. (Table 10.2 may help). Christchurch lifelines 7 Identify mitigation opportunities. interdependencies — an 8 Prepare an action plan. intuitive view 9 Report back 2 to 8 to full forum. Has/would this While an analytical approach poses almost insurmount- exercise cause a review of mitigation measures able difficulties, there are some things we can take presently proposed? comfort from. Firstly, there is usually a high level of redundancy in Christchurch networks. Even when fail- (NB 6, 7, 8 require co-operation with other opera- ures occur, re-routing will often be an option. tors/task groups.)

However, utility and emergency services operations The next step would be to consider the action plan as a whole and review as necessary. have to be arranged so that emergency provision of vital services such as power, water, toilet facilities etc., The results of the 12 hour exercise are not reported in will have to be made to provide for the time between the this book in that more time was required to produce event (say an earthquake) and the provision of services useful results, although an understanding was achieved by other utilities. of the processes involved. The participants in the exercise all agreed that a detailed look at the concepts Secondly, the lifelines project has in itself provided for involved in interdependencies heightened their aware- interdependencies in a management sense through ness of the fact that in a large emergency event no opening lines of communication through network pro- utility can stand alone. Interdependence of Lifelines • 249

UTILITY RECOVERY OF PROVISION OF PROVISION OF COMMENT BASIC SERVICE 50% SERVICE FULL SERVICE OR CONTROL Water Up to 2 days 2 weeks 12 months Access and equipment. Pipe replacements. Radio communications. Drainage: As for water, 6 • Sewage Up to 2 weeks 12 weeks minimum 24 months minimum months min for for control inspection - CCTV.

• Stormwater 2 - 3 days 12 weeks 12 months Important for early sewage control. Electricity: Radio • Transpower Up to 3 days 2 weeks 9 - 12 months communication. Access and equipment. Transmission towers.

• Southpower Up to 3 days for 4 weeks 12 months Sub stations. control. 1 week for Cable structures. service Telecom: Access and • VHF Mobile & 002 weeks equipment. Standby Fleetlink fuel. • Local Telephone Up to 2 days 3 - 6 weeks 12 months Buildings. Line structures.

• Tolls Up to 4 days 2 weeks 1 month Transmission towers • Cellular Up to 6 hours 4 days 2 weeks Switching units. Transmission towers. Broadcasting 2 - 3 days 2 - 3 weeks 2 - 6 months Access and tower antennae realignments. Standby power. Telephone link. Roading 1 - 2 days. Some 3 - 4 weeks 18 - 24 months vehicular access to most areas.

Rail 1 - 2 weeks 1 - 2 months 6 months Track alignment. Buildings. Signalling. Slips. Port 1 - 2 days 6 - 8 weeks 12 - 24 months Access. Power. Airport 1 - 2 days 1 - 2 weeks 4 - 6 weeks Controls and navigation aids.

Definitions Ability to provide a Provision of general basic manageable service to most service for priority areas. Some use. queuing or overload. Temporary fixes in place.

Table 10.2: Recovery of service — a preliminary assessment of the time to recover following a major earthquake

The participants in the Workshop were aware of the Summary of Findings Wellington Earthquake Lifelines Group (WELG) Interdependence of lifelines is a key concept to be Project on “The value of interdependence analysis on integrated into the response plan of utilities. The response planning” fully reported in Sections 4.1 - 4.6 analyses in this report are illustrative only and are of the 1993 report of the WELG. Section 4.5, the based on particular scenario assumptions and ex- Summary of the findings of the WELG Report, is pectations. The principal findings of this Group are reproduced below. as follows: 250 • Risks and Realities

• Interdependence analysis is fundamental as a re- There is considerable benefit in each utility prepar- sponse planning tool for lifeline utilities. ing its response plan individually and collectively reviewing the plans against a variety of events and • Response planning for utilities provides a means of assumptions both to heighten their awareness of minimising the impacts of damage from earth- each others needs and to enhance those plans. quakes and providing for the earliest possible res- Doing so would identify critical interdependencies. toration of services. It is important for meeting both Where practicable, these interdependencies can be social and commercial objectives. modified by mitigation or allowed for in the re- • An outline for a Utility Response Plan is presented sponse plan. based on: WELG has done further work on response plans since — identifying a hierarchy of key components to be then. inspected (and if necessary, responded to) fol- In Christchurch, work has been undertaken since the lowing an event. October 1994 workshop by the individual utilities and — identifying contingency preparations which are this is reported in Chapter 11. No further work has been needed to activate key responses. done on interdependencies as yet (but see possible further work in Chapter 13). — determining a basis for making decisions on allocating resources following an event. Notwithstanding the above, in the event of a major emergency all utilities are now aware of the impor- — providing realistic assessments of times to re- tance of interdependencies as is the Regional Control- cover. ler of Civil Defence who may have the ultimate responsibility to determine priorities and allocate resources. • In the Response Plans, the impact of damage to The Chief Engineer in Civil Defence (at present the other services needs to be addressed and planned Director of Operations of the Christchurch City Coun- for. cil) has accepted the task of co-ordinating the responses of the various utilities if required for Civil • From the three analyses reported, the indications are that assessed times for recovery are increased Defence. significantly when the impacts of interdependence It may well be that future formal work on are included. Implications from this follow in the interdependencies is not warranted and the current area of overall Civil Defence planning. focus on response plans (which take into account interdependencies) is the best approach. In due course • Interdependence analysis has application in the areas of risk assessment and in identifying mitiga- an exercise which will highlight interdependencies tion priorities. will be held.

Multiple services in one area Summary of Benefits • 251

Chapter 11 Summary of Benefits and Work Undertaken or Proposed

11.1 Christchurch City Council • Removing spare (loose) electric motors and other equipment (from pumping stations) that could move Water Services and damage installed equipment. A comprehensive report was presented to the Christ- • Increased instrumentation, monitoring equipment church City Council City Services Committee and this and battery capacity in pumping stations so that, in followed a seminar organised for elected members in the event of a power outage, a longer period can September 1996. The report dealt with projected miti- elapse before operations staff lose contact and no gation measures and financial implications, emergency longer are aware of the situation. response and business continuance planning and insurance cover relating to infrastructural assets managed • Obtained budget provision and arranged for major by the Water Services, Waste Management and City ($270,000) seismic strengthening of the two 11,000 Streets Units. The following sections have been ex- cubic metre Worsley Spur reservoirs that have been tracted from that report. identified as likely to collapse in a severe earthquake. Of the hazards examined (earthquake, tsunami, flood, snow, wind, slope hazard) earthquake has the most • Installed remotely operated (from Colombo Street impact and, as it is a “reasonably probable” event, water supply control room) motorised valves at justifies significant attention in terms of mitigation most major bulk storage reservoirs. Following a measures. staff investigation, it was decided that in the event of an earthquake the motorised valves on the The following sections of the report summarise the Worsleys, Cashmere, Hackthorne, Sutherlands and mitigation programmes currently being developed by Huntsbury No 1 reservoirs will fully close. All the Water Services Unit. others will stay open and be closed at 50% capacity. This minimises the risk of water damage, or wast- Water Supply age, due to ruptured pipelines in the vicinity. The attached schedule and tables (Tables 11.1 and • Undertaken pipework alterations at some loca- 11.2) sets out Lifelines work expanded for the coming tions, that allow water to bypass (by operating ten year period, including 1996/97. Budget require- valves) normal supply routes, this allowing limited ments are not particularly high and have been included continuity of supply if a pump, pipe, or reservoir in current programmes. fails on the usual arrangement. Considerable work has already been done in the water • Investigated the feasibility of “water proofing” supply area. A survey has been completed, followed by pumping stations sufficiently to minimise damage remedial work, of all water supply facilities to identify, if a tsunami occurs. This feature has been incorpo- then secure or remove any unsecured equipment that rated in the new Aston Drive pump station. could move in an earthquake and damage itself or other essential equipment. Examples include: • Commenced a programme of installing isolating valves on key trunk mains that pass over at risk • Bolting down standby generator sets, batteries etc. bridges. • Installing brackets to prevent overhead cranes jump- • Consciously paying more attention to design detail ing their travelling rails and falling down. of new works to ensure mitigation measures are • Securing computers and other control equipment included in design. (that were sitting on desk tops) to ensure they will • Initiated, and managing a research project, funded not fall. by the NZ Water and Waste Association to inves- 252 • Risks and Realities

1,000

101,000

1,000

101,000

1,000

101,000

1,000

101,000

1,000

100,000 100,000 100,000 100,000 100,000

101,000

1,000

101,000

100,000

?????? ?

50,000

151,000

100,000

1,000 1,000

50,000

151,000

100,000

5,000

8,000

3,000

5,000

2,000

10,000

25,000

58,000

8,000

3,000

5,000

3,000

2,000

5,000

96,000

10,000

132,000

1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06

CCTV inspection of brick barrel CCTV stormwater pipelines Brick barrel stormwater pipe repair/replacement Review of Design Manual and publication of new version Study mitigation measures at five stormwater pumping stations Inspect and secure pump station components and stored equipment Complete study of stopbank liquefaction potential and define works Stopbank relocation or strengthening Complete study of mitigation for pump station inundation Pump station protection against inundation Investigate effects of power supply Investigate effects loss Prepare emergency response plans Prepare emergency and train staff

Figures in normal type are already budgeted. already in normal type are Figures Figures in italics are not budgeted. in italics are Figures

Totals

WATER SERVICES UNIT SERVICES WATER

ENGINEERING LIFELINES MITIGATION PROGRAMME ENGINEERING LIFELINES MITIGATION

LAND DRAINAGE SYSTEM

Table 11.1: Engineering lifelines mitigation programme — water supply system Summary of Benefits • 253

1,000

21,000

1,000

10,000 10,000

21,000

1,000

21,000

36,000

1,000 1,000

10,000 10,000 10,000

15,000 15,000

36,000

15,000

36,000

17,000

10,000 10,000 10,000 10,000 10,000 10,000 10,000

1,000 1,000 1,000

53,000 38,000

10,000 10,000 10,000

17,000

10,000

15,000

5,000

90,000

15,000

17,000

10,000

147,000

5,000

10,000

180,000

200,000

1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06

NC = no direct cost NC = no direct

Worsley Reservoirs - strengthening Worsley Complete reservoir isolation and automatic operation project Install isolation valves at all bridges Complete project for nationwide pooling of water supply materials by 30 December 1997 Fix flexible joints at critical pump stations (above and below ground) Investigate and strengthen reservoir pipework Review design practice and specifications Prepare emergency response plans Prepare emergency and train staff Waterproof vulnerable pump stations Waterproof

Totals

WATER SERVICES UNIT SERVICES WATER

3 4

ENGINEERING LIFELINES MITIGATION PROGRAMME ENGINEERING LIFELINES MITIGATION

WATER SUPPLY SYSTEM SUPPLY WATER

Table 11.2: Engineering lifelines mitigation programme — land drainage system 254 • Risks and Realities

tigate the feasibility of setting up a mutual assist- 11.2 Christchurch City Council ance agreement (materials, plant, labour, and technical support) between Water Supply Authorities in Waste Management (Sewerage) times of emergency.

• Commenced a programme of installing controlled 1. Reticulation standpipes on artesian well heads so that water can (a) Work is proposed to construct emergency overflow be drawn from the well, even if the pump station is installations to allow emergency discharge to the inoperable. stormwater system at 19 pumping stations (numbers 12, 13, 14, 22, 28, 33, 34, 38, 39,41, 44, 45, 50, • Become involved with the Council study on the 51, 52, 54, 56, 63 and 67). vulnerability, and mitigation measures for the Ferrymead bridge site. (b) Work is proposed to replace the rising main on Pages Road bridge. No. 57 rising main on Ferrymead The scheduled programme shows the continuation and bridge is programmed for 1997/98 as a capital work completion of the above projects but also allows for: as a result of the poor condition of the pipeline. It will also act as a lifeline improvement. • The fitting of flexible joints both above and below ground at selected pump stations. (c) Brick sewer renewal and/or replacement is proposed for 3 km out of 6 km of sewer in Kilmore, • Strengthening reservoir pipework. Madras, Fitzgerald, Tuam and Moorhouse. • A review of design practice to ensure lifeline considerations are part of the design phase. 2. Pumping Stations (a) It is proposed to provide flexible couplings for • Preparation of Emergency Response and Business those pumping stations which are at critical points Continuance Plans. in the network, either at terminals where several pipelines meet, at inline locations with other pump- Land Drainage ing stations or in the eastern earthquake liquifac- Land Drainage Lifelines work is not as advanced as tion zone. There are 18 pumping stations (numbers Water Supply and the ten year programme contains a 7, 9, 10, 12, 13, 14, 15, 28, 30, 31, 35, 36, 37, 38, 45, number of investigations with future spending very 46, 54 and 57). dependent on the outcome. (b) It is proposed to waterproof buildings and electrical • CCTV inspection of the 15 km of stormwater brick control cabinets at risk of flooding at pumping barrel culverts is being completed this year, pre- stations 15, 18, 19, 21, 22, 23, 39, 43, 45, 46, 63, 77. liminary to decisions on action that may be taken to (c) Strengthening of the only two pumping stations strengthen them. They are generally still in very assessed as being at risk in an earthquake (pumping good condition and current thinking is that it may stations 2 and 5). be wiser to be properly prepared to repair damage than to spend large amounts on their renewal. There (d) It is proposed to install standby power plants at may be a much cheaper non-destructive means of critical points in the network similar to 2(a) above. securing and strengthening the brick lining and this There are 14 pumping stations (numbers 20, 33, 34, will be considered. In the meantime a nominal 36, 38, 40, 41, 42, 60, 67, 75, 76, 77 and 78). $100,000 per year is shown, but this is not included (e) Purchase of portable pumping equipment is pro- in current programmes. posed for service in emergencies at other pumping • Investigation of stopbank stability on the lower stations assessed as either at less risk or at less Avon is proceeding, but it is simply not possible to critical locations than 2(d) above. predict what works may be recommended. 3. Treatment Plant • Emergency Response Plan preparation is commencing now as a single exercise involving City (a) Investigation of locations where flexible couplings Streets, Waste Management and Water Services. at buildings are needed is proposed.

• Other initiatives are as described in the 10 year (b) These will then be installed in as yet unidentified schedule. but critical locations. Summary of Benefits • 255

(c) It is proposed to upgrade the basement sump pumps • Undergrounding overhead services, particularly where the current pumps are totally inadequate. on significant lifelines routes.

(d) Tying down of control equipment against displace- • Service authorities to consider lifelines issues when ment under earthquake attack has been done at the planning service installations in key roads. treatment plants. • Selection of trees which are more likely to survive The work proposed is shown in Table 11.3. an earthquake, particularly on key routes. Plant only known rootfirm species in areas where lique- City Streets faction may occur. General earthquake and other hazards mitigation meas- • Earthquake proofing traffic signal control and sur- ures recommended in the Lifelines Project Report are veillance systems. proposed as follows. Some of these involve Transit NZ and other authorities. • Investigation of the need for railway and road bridges when they become due for reconstruction or rehabilitation, and safeguarding land for ground Earthquake level replacements. The elimination of some of (a) Bridges those bridges would decrease route vulnerabilities for road and rail. • Strengthen connections particularly between superstructures and substructures. • Identification of alternative river crossings (such as fords) including investigation of rail bridges (and • Increase column strength and ductility. the tunnel) for road vehicles, and safeguarding the • Strengthen or renew retaining and approach struc- land for this. tures. • Review route diversion procedures and informa- • Strengthen lateral/longitudinal restraint mecha- tion systems, these could be used for minor emer- nisms. gencies such as weather induced route severances.

• Construct landing slabs. • Investigation and recording of temporary emergency bridging materials and equipment (including • Addition of information plaques to bridges over bailey bridges) and their locations. which key lifelines services pass, describing their location and nature (services authorities). • Securing road tunnel mechanical/electrical services. • Further technical assessment of the Sutton Quay rail overbridge/road retaining structure and clarifi- • Review standards of construction and maintenance cation of responsibilities (TNZ, NZ Rail, BPDC). of all transport facilities to reduce the vulnerability of the road network, for example trench backfilling • Detailed geotechnical and structural investigation and reinstatement requirements. of vulnerable bridges located on Primary Routes, or significant alternative routes, for which simple Snow and Wind mitigation measures are not immediately apparent. • Investigate the resilience of and the need for backup (e.g. Pages Road, Marshland Road-Styx River). power facilities for traffic monitoring cameras and display units to enable ongoing conditions moni- (b) Roads toring of key city sites from the City Council’s • Clarify ventilation requirements to allow contin- Tuam Street building. ued operation of Road Tunnel to Lyttelton, provide • Establish priorities for undergrounding overhead backup power facilities for emergency use. services on key transport routes, including primary • Development of integrated response plans for the routes, sector distributors and other arterial routes. road and bridges network, recognising the accessi- • Tree management procedures, particularly on arte- bility needs for evacuations, supplies, fuel sup- rial roads (e.g. type, pruning, spacing, removal of plies, and services reinstatement. “suspect” trees). • Early replacement of older, vulnerable underground • Clarify and document emergency event procedures pipes (e.g. stormwater) on key lifelines roads. (disaster plan), responsibilities and information 256 • Risks and Realities

1,000,000

10,000

(Pages Road)

100,000

(Investigation)

30,000

100,000

10,000

130,000

100,000

20,000

10,000

40,000

30,000 30,000

20,000

40,000

4,000

1,000

20,000

95,000 100,000 110,000 140,000 140,000 130,000 110,000 1,000,000 1,000,000 1,000,000

10,000 30,000

20,000 10,000 10,000 10,000

40,000

1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06

(b) Rising Mains on bridges replacement (c) Brick sewer renewal and/or strengthening

(a) Provide flexible couplings at Pumping Stations (b) Continue waterproofing programme at buildings (c) Strengthen PS2 and PS5 (d) Install standby power plants (e) Purchase of portable standby pumping equipment

(a) Investigate locations where flexible couplings at buildings are needed (b) Install flexible couplings (c) Upgrade basement sump pumps (d) Tie down control equipment Tie (d)

Totals

(a) Emergency Overflow installations (a) Emergency

WASTE MANAGEMENT UNIT MANAGEMENT WASTE

ENGINEERING LIFELINES MITIGATION PROGRAMME ENGINEERING LIFELINES MITIGATION

WASTE WATER SYSTEM WATER WASTE

1 RETICULATION

2 PUMPING STATIONS

3 TREATMENT PLANTS TREATMENT 3

Table 11.3: Engineering lifelines mitigation programme — waste water system Summary of Benefits • 257

flows across transport agencies and other lifelines deteriorating crib walling on the Blenheim Road agencies including establishing response plans and Overbridge and this will include seismic strengthen- reconnaissance plans. ing.

• Provide more 4-wheel drive vehicles in the Coun- cil’s fleet. Year Amount ($) 1996/97 400,000 • Procedures for the spreading of grit. 1997/98 281,300 1998/99 357,629 Flooding 1999/2000 357,629 • Ensure that good records of sump locations and 2000/01 192,666 stormwater pipelines are available to key people in 2001/02 570,360 an emergency situation. 2002/03 418,337 • Adequate sump maintenance and street cleaning 2003/04 462,591 procedures are needed to ensure that available staff 2004/05 363,462 are not overloaded in clearing blocked sumps and 2005/06 366,000 stormwater systems in an emergency situation. Table 11.4: Bridge strengthening programme • Develop self-cleansing sump inlet structures that retain functionality in wet weather Current Disaster/Emergency Management Planning Initiatives • Disaster planning. During 1996, a project for the development of ‘Busi- ness Continuity’ Plans was initiated by the Managers Tsunami of three Business Units, namely City Streets, Waste • Identification of road routes that would be avail- Management and Water Services, as a lead task within able for evacuations from eastern and low lying Council. suburbs (more work required in comparing road levels with wave profiles in rivers). The managers concerned quickly identified that a planning need existed far beyond the limits suggested • Technical assessments of vulnerable road network by the term ‘business continuity’ and in particular structures (e.g. Moncks Bay sea wall, Ferrymead recognised the need for a level of planning consistent Bridge, Bridge Street and Pages Road Bridges) and with and complimentary to that which currently exists the need for scour protection measures. in relation to Council’s roles and responsibilities in a declared Civil Emergency and over and above those Slope Hazard plans establishing response procedures for ‘minor • Retainment, plantings, drainage improvements, re- emergencies’. moval of unsafe rocks, ground anchors at a range of To that end, a planning project was drafted and agreed sites on the Port Hills. in late 1996, with the project to be completed in three stages (as illustrated in Figures 11.1 and 11.2). Programmed Work Most emphasis since the Project Report was completed has been on bridges and a scoping report “Bridge Disaster Management Planning Strengthening Programme” dated February 1996 was 'Time' and Terminologies circulated to Council Committee members at a seminar in September 1996. IMMEDIATE RESPONSE PLANNING Table 2 of that report is set out in Tables 11.4 and 11.5. 3 STAGE PROCESS BUSINESS/SERVICES CONTINUITY PLANNING It shows a possible programme of bridge strengthening and/or renewals. DISASTER RECOVERY PLANNING The total cost of the programme is $3,769,974 over 11 years. DISASTER MANAGEMENT PLAN These compare with a current annual budget provision of $110,000 per annum in the five year plan. In addition, work is being programmed in 1997/98 to renew Figure 11.1: Disaster management planning 1

258 • Risks and Realities

Prepared 4/03/96

Programme Year Programme

95/96 2002/03 2001/02 97/98 2002/03 98/99/2000 96/97 Retrofit cost allows for full replacement 2000/01 2004/05 2003/04 2003/04 97/98 2006/07/08 2003/04 2004/05 2005/06 2003/04 2003/04

• • • •

$31,000 $40,000 $97,000

$50,589

$209,482 $125,394 $218,337 $570,380 $154,000 $209,482 $107,000 $200,000 $715,259 $149,608 $400,000 $192,668 $174,300

$4,801,975 $1,000,000

1 4 8 7 2 5 6 3

16 12 18 15 17 13 14

9 10

Order Retrofit Cost

Proposed

Programme

Large structure and traffic Vital link to Sumner for services Vital replaced should it collapse Important lifelines Structure is unlikely to be replace expenditure required Expensive and disruptive to Lower Avon Lower Avon Bridge Important traffic route. Modest Important traffic Lower Avon Lower Avon Bridge Analysis includes both bridges. be determined Priority for widening yet to Lower Avon Lower Avon Bridge Retrofit design underway unavailable Very large detour if Bailey bridge Very 65 years old Brighton's commercial centre. It is This bridge is a vital link to

6 3 5 7 2 8 4 1 9

13 10 18 15 16 17 12 14

Relative Network

Importance Comments

Planned Traffic Mods Traffic Planned Widen downstream structure Add cycle and traffic lanes Add cycle and traffic Nil Nil Widen bridge (longterm) Nil Widen to 6 lanes Add cycle lanes Widen to 4 lanes Nil Clipon cycle lanes Nil Nil change vertical alignment Nil Add corner roundings and Add cycle lanes Nil Nil

ENGINEERING LIFELINES STUDY

BRIDGE STRENGTHENING PROGRAMME

Lifelines Present

= Telecom =

T 3 = Major 2 = Medium 1 = Minor S = Sewer W = Water P = Power P

T3, W2 T2, W3 P2, W1, S1 T3, P1, W3, S3 T2 T1 T3, P3, W3, S7, Police, Fire Station T2, W3 T1 T2, W3, Fire Station T2 Key: T1 T2 T2, W3 T2, W3 T2, W3, S1 T2, W3 T2, W3

4 8 3 5 7 2 6 9 1

14 18 10 16 13 17 12 15

(Average)

1.7 3.2 1.6 5.3 2.3 0.3 4.1 1.3 3.6 5.3 3.6 7.6 1.0 4.0 1.3 8.0 2,7 2.6

24.9

Relative BC BC Priority

Avondale Road Avondale Colombo O'B Marshland Road Moorhouse O'B Pages Road Straven Road Durham O'B Ferrymead Gayhurst Road Kilmore Street Swanns Road Old Main North Road Fitzgerald Twin Barrington Street Causeway Blenheim O'B Wainoi Road Wainoi Opawa Road

R104

Bridge No. Bridge Name R702 R704 R201 R131 R111 R801 R107 R501 R703 R705 R405 R214 R109 R103 R102 R106 R206

8 7 2 4 9 3 6 5 1

15 10 13 17 16 18 12 14

Table 11.5: Engineering lifelines mitigation programme — bridge strenghtening programme Summary of Benefits • 259

Disaster Management Planning water) to at least an acceptable minimum level will be 'Time' and 'Effort' a prime focus for this stage of the project.

Completion of this stage is targeted for immediately following stage one, although those continuity issues which will be required to support the immediate re-

EFFORT sponses are being addressed, as are possible service contingencies identified during the progress of stage Response Business Continuity Disaster Recovery Disaster one. These contingencies are being logged as they are identified, for later detailed review and inclusion in the

TIME plan.

Figure 11.2: Disaster management planning 2 Stage Three — Disaster Recovery Planning Stage One — Immediate Response To date, review and discussion of longer term recovery Planning and reinstatement strategies has not progressed beyond This stage of the project was commenced in early 1997 the initial concepts and what might be considered to and uses as it’s principal tool, the findings of the some of the overall objectives. However clearly this Christchurch Lifelines Study Group with particular stage will focus on delivery of key (albeit impaired and emphasis on the exposure presented by the maximum hence reduced) services over an extended period, strat- credible earthquake (mce). The likely planning time- egies which might hasten the restoration of full serv- frame envisaged for this stage is from zero to 24-36 ices (new technology) and the possible introduction of hours. replacement technology with greater resiliency (to The threat and vulnerability assessments established damage) characteristics. by the Lifelines Study are being expanded and refined by each of the business units in their specialist areas of Insurance Issues interest. The results of this work will establish a Re- In 1991 Central Government established a Disaster sponse Hierarchy which will in turn be used for two Recovery Plan which identified the need for substan- primary purposes: tial financial contributions from local authorities following natural disasters and emergencies. The Central 1 To identify and establish the resources required for Government Disaster Plan places obligations on local an immediate and acceptable emergency response. authorities with respect to reinstatement costs involved 2 To ensure the organisational structure and operat- following damage to infrastructural assets of natural ing procedures required to maximise the use of disasters and emergencies. The plan requires that above those resources are in place. the initial threshold level Central Government will provide 60% of the repair or recovery cost and the A secondary outcome of this work may be the identi- responsibility for the remaining 40% falls on the local fication of further hazard mitigation measures. authority. The plan also specifies that unless the local authority can demonstrate that it has adequately pro- The business units will be working closely together tected itself through proper maintenance, the provision (and in liaison with Regional Civil Defence) to ensure of reserve funds, effective insurance or participation in a composite and cohesive plan is prepared. a mutual assistance scheme with other local authorities Stage One planning is due for completion and testing to a level that it can meet its 40% obligation, then well prior to the end of 1997. Central Government’s 60% share will not be made available.

Stage Two — Business/Services Continuity Following the issue of the Central Government’s Dis- Planning aster Recovery Plan, the Local Government Insurance This next stage will be required to review and address Corporation (LGIC) established the Local Authority the various systems, support and procedural issues Protection Programme (LAPP). This is a mutual assist- necessary to identify, provide and maintain minimum ance scheme with the contributions of members held in acceptable levels of business and services to the com- trust and administered by LGIC. The fund is designed munity, for a period (some weeks/months) following a to cover infrastructure which is generally uninsurable. major event. Annual contributions to the fund are based on a series of factors and the contributions are fixed each year. Contingencies for the delivery of critical services (e.g.; 260 • Risks and Realities

In June 1993 the Council resolved to join LAPP. There Roads and road structures are not covered under the is no guaranteed distribution from the fund in the event LAPP scheme as they attract a subsidy from Transit of a natural disaster or emergency and the distributions New Zealand. The Council’s contribution to the fund are at the discretion of the trustees who may take into for 1995/96 was $229,450 (GST exclusive). account a variety factors including the amount of the annual contributions, the total fund of contributions In 1995 the Council commissioned Sedgwick Limited paid, the record of payment, the state of repair of the to identify and quantify (by frequency/severity) the infrastructure, progress with risk management, finan- risk of a major disaster affecting any area within the cial circumstances and other subsidies or compensa- city’s regional boundaries. The purpose of the study tion available to the member. However, during the was principally twofold: 1995/96 year a focus group was established and this • To provide evidence of and to ensure that the group has recommended limiting the discretion that the Council achieves a credible level of compliance Fund Trustees currently have in determining the quan- with the ‘responsibilities’ assigned to local authori- tum of any distribution. It is expected that this will give ties under terms of the Natural Disaster Recovery more certainty to members in terms of both coverage Plan. and distributions. The group also made recommendations on options for the use of the fund in the future. • In identifying the sum of any loss resulting from disasters, provide the Council with the tool with The fund reinsurance programme was renewed during which to measure the adequacy and cost effective- 1995/96 and additional support means LAPP now ness of existing funding strategies and consider provides $25,000,000 for its members. Assuming possible alternatives. Central Government’s 60% contribution, protection of $62,500,000 is now in place. Reinsurance support has Conclusions and recommendations of the Sedgwick increased as a result of the understanding the insurance report were as follows: market appears to now have of the aims and objectives of the fund. • That all planning be based on an expected loss of $50,000,000. This sum is a reasonable compromise The infrastructural assets of the Council covered under between the maximum credible loss ($36,931,300) the scheme relating to 1995/96 are shown in Table 11.6. Replacement Covered by Value as @ April 1995 $ LAPP Water water reticulation 207,941,128 yes pumping stations 6,066.100 part less pumping stations insured (59 of 69 insured) -5,968,400 insured 208,038,828 Stormwater stormwater reticulation 357,107,900 yes Heathcote river floodgates 1,500,000 yes 358,607,900 Waste sewer reticulation 489,506,903 yes pumping stations 4,898,500 part less pumping stations insured (7 of 87 insured) -1,965,700 insured Bromley oxidation ponds 2,000,000 yes Templeton oxidation ponds 240,456 yes Belfast oxidation ponds 271,768 yes 494,951,927

Total replacement value covered by LAPP 1,061,598,655

Table 11.6: City council infrastructural assets covered by LAPP Summary of Benefits • 261

and the maximum possible ($69,011,500) as iden- A number of more minor vulnerable areas were also tified within the report. detected and mitigated.

• That the minimum reserve of $5,000,000 earmarked The over all benefit of the Christchurch Engineering by the Council for disaster recovery purposes be Lifelines Project as far as Transpower is concerned has maintained to cater for the lower layer of funding meant that a number of “weak areas” were exposed and required and specifically to cover the $1,600,000 mitigated. This has resulted in the likelihood of outages excess or threshold above which Central Govern- being reduced and if they did occur, would be of a ment emergency assistance may apply. shorter duration, as a result of a major earthquake, flooding, tsunami, windstorm or snowstorm. • That capital and/or maintenance budget appropria- tion should form a significant part of catastrophe risk/loss funding, particularly through the middle layer. 11.4 Southpower A continual review and assessment has taken place The value of road and road structure infrastructural since the original report. assets were included in the Sedgwick report in order to identify the total cost of any loss. Often this review has been carried out in more detail than the original report in specific areas particularly The recommendations of the Sedgwick report are where additional drivers have been identified and lift being considered by the Director of Finance. the priority as a mitigation measure.

Southpower has demonstrated a high commitment to 11.3 Transpower (New Zealand) improving the electrical infrastructure assets to mini- Limited* mise the impact of a natural disaster or hazard. Transpower (NZ) Ltd, the owner of the national elec- It has recognised the fact that it is important for any tricity grid, took an active role in the Christchurch business to manage its risk and the possible impact this Engineering Lifelines Project. In addition to contribut- may have on its customers. It is also considered ex- ing financially to the project, Transpower carried out a tremely important that action be taken where appropri- review of its assets in the Christchurch region to ate as soon as possible to mitigate identified hazards in determine the reliability of its network during and after a structured manner. a moderate to severe earthquake, flooding, tsunami, The following work is a summary of the major mitiga- windstorm, snowstorm and slope instability. tion measures that have been actioned to date:

The survey was a broad brush review of the entire • The Control Centre has now been moved to a more system with the aim of pin-pointing ‘weak links’ in the seismically secure part of the new building and system. Consideration was given to both the effect on retained at first floor level because of the possibility the system and the time/costs of mitigation. For exam- of flooding. ple, the lead time to replace a large transformer could be up to twelve months. • The standby generator has now been replaced above ground level. As a result of the review of the Transpower system using a systematic methodology developed for a number • Demolition and removal of that part of the of lifelines participants, a number of vulnerable areas Southpower building considered seismically inse- were identified. Urgent action was taken to mitigate the cure is under way. areas of concern. • Emergency spares have been relocated after de- Major transformers at Transpower’s Islington and tailed assessment of the buildings identified a far Bromley substations were significantly strengthened more secure area of storage. so as to be able to resist current code earthquake loadings. The loss of any of the above mentioned • A seismic restraint project is underway to minimise transformers could have caused significant problems the effect of damage on emergency spares during for Transpower and Southpower, the local electricity an earthquake. distribution company. • Two Armagh bridges that carry major feeders supplying the central city have had their cable supports

* Reference: J Mackenzie (1995). Lifelines Surveys for Mitigation, substantially strengthened. A Christchurch Experience, NZNSEE Technical Conference 1995. 262 • Risks and Realities

• Transformer hold down review and bunding at undertaken although, fortunately, this has not involved district substation sites is almost complete. a lot of expenditure. A structural consultant John MacKenzie was employed and he developed and used • The sound attenuation wall at Sockburn substation a standard method of survey of all the key elements has been lowered. involved in the control and linking of the telecommu- • Spare circuit breakers, voltage transformers and nications in Christchurch for Telecom. The work un- auxiliary transformers within substations have been dertaken was mainly bracing and anchoring of equip- secured against movement. ment.

• Network substations have been assessed to deter- It is important to note that the Telecom network is mine their structural ability to withstand a moder- protected by the diversity so that the failure of an ate to severe earthquake. Of the 528 substation individual element simply results in a re-routing of a structures 253 will require further evaluation to call through undamaged elements. determine strengthening needs. Budget provision for work on Telecom engineering • Generic strengthening systems have been designed lifelines in Christchurch City is approximately 4 mil- and work has started improving the strength of lion dollars over 10 years. This will cover seismic network substations. bracing of telecommunication equipment, structural strengthening of selected buildings, additional fire • A suitable method of securely fixing transformers protection and physical security measures. in kiosk substations has been developed.

• An engineering review of pole mounted substa- 11.6 Christchurch International tions has been commissioned and new standards set. Airport Limited

• Battery banks and chargers in district substations Christchurch International Airport Limited (CIAL), and network substations have now been strapped the owner of Christchurch International Airport, is down. responsible for providing landing facilities for all air transportation in and out of Christchurch. Managing risk of a natural disaster is now an integral part of Southpower’s Asset Management plan. This is CIAL recognised that not only is air transportation updated annually and provides detailed information on essential following a major disaster, such as a severe asset planning for the next ten year period. earthquake, loss of key components could also have a detrimental effect on the airport’s ability to function The most significant cost to date has been bunding and properly. strengthening work at major transformer sites. The expenditure to date on this aspect of work has exceeded In order to assess and mitigate potentially vulnerable $700,000 and has included all major transformer sites areas, CIAL commissioned an independent lifelines at the 45 district substations which are part of survey of all the key elements of Christchurch airport Southpower’s distribution system many of which are and acted on the findings with urgency. outside the area of the original lifelines study. The key components covered by the survey included: The estimated cost of strengthening work at 253 net- • emergency power generation; work substation sites is expected to exceed $4,000,000, however this is only a fraction of the cost of replace- • electrical reticulation; ment of these assets. • water supply and reticulation;

• sewage disposal; 11.5 Telecom • telecommunications; Telecom New Zealand has an extensive risk management mitigation programme to protect its assets and • flight management systems; and engineering lifelines is part of this programme. Much of this work would have been undertaken as part of the • emergency services. normal risk management programme. As a result of the survey, a mitigation programme was As a result of the investigation in Christchurch a initiated. A significant number of areas have been considerable amount of mitigation work has been either attended to directly or are now superseded by the Summary of Benefits • 263

International Terminal rebuilding project. This in- Significant seismic upgrading has been started on No.7 cludes: wharf which is used by Pacifica Shipping. This wharf would be one of the most important wharves after an • all international mechanical and electrical services; earthquake to provide outside access for emergency • emergency generators, air compressors and refrig- service supplies to Christchurch. The vessels which eration compressors (Domestic Terminal); use this wharf are totally independent of wharf power.

• mechanical services (domestic boiler room); It is interesting to note that since November 1994 the Port Company has had a Risk Management Commit- • some heat pumps; tee.

• electrical switchboards and control cabinets;

• hot water supply tanks (airport fire service); and 11.8 Tranz Rail

• foam storage tank (airport fire service). The train control monitoring equipment has been secured. The stability of embankments and cuttings have Some areas have yet to be attended to, but will be as part been reviewed. Investigations were undertaken of the of the ongoing mitigation programme. However, just Lyttelton Tunnel portal and the bridges’ stability and as importantly, an awareness has been raised of the the Heathcote underpass bridge is to be replaced. correct installation of the new building services as part of the substantial International Terminal rebuilding As a result of the Christchurch engineering lifelines project. study, work may be undertaken investigating the performance of the Picton-Christchurch line. As a result of the Christchurch Engineering Lifelines Project, the Christchurch International Airport is becoming a more dependable part of the transportation 11.9 Transit New Zealand system, especially during and after a major disaster such as an earthquake, flooding. etc. The study highlighted the importance of seismically rating bridges and initiating the remedial works on those bridges to improve their seismic resistance. The 11.7 Lyttelton Port Company lifelines work has given some impetus to the national rating of bridges on the State Highway network and The Lyttelton Port Company has ensured that all of the work is still proceeding. A particular methodology for issues raised in the vulnerability assessment of the port use by Transit New Zealand has been developed and is have been investigated. The principle concern is still now underway. It should be noted that the Thorden the possibility of the loss of the use of Cashin Quay overbridge strengthening arose directly from the Wel- because of damage at the face of the reclamation as a lington lifelines investigation and is top priority for result of an earthquake. Although most of the port New Zealand. structures are unlikely to suffer severe damage, the interfaces between them and the land are expected to drop, necessitating remedial work with temporary bridg- 11.10 Petroleum Products ing which is readily constructed from material already at the port. The tank farm at Lyttelton is to be redeveloped and planning is currently under way. The installations will The provision of electrical power to the port has been be built to the latest European safety specifications, but reviewed and a new alternative line is to be laid under the direction of New Zealand engineers who will between the port and Woolston giving much more ensure that the local standards (particularly earth- security. The two overhead lines over the Port Hills quake) will be met. have been modified so that the configuration now allows the servicing of the lines independently and some repoling has been undertaken. Major modifica- 11.11 Conclusion tions were also made at the substation. There is a possibility that an alternative emergency generator It is gratifying to see such a large amount of work has may be installed. been done or is proposed as a result of the engineering lifelines work. It shows that the considerable effort of The equipment at the command at control centres have so many people at surprisingly little cost has resulted in been checked and secured. budget provision for mitigation and planning work that 264 • Risks and Realities

has and will continue to make Christchurch much better able to withstand the effects of natural hazardous events.

However, there is an acknowledged need to continue improving the infrastructure and this is addressed in Chapter 13.

Christchurch Cathedral earthquake damage, 1 September 1888.

(RW Meers & Co. photo, Canterbury Museum Ref: 7068. Reproduced with permission of Canterbury Museum)

Flood in Gloucester Street, Christchurch, 4 February 1868 with the old Gloucester Street footbridge on left.

(DL Mundy photo, Canterbury Museum Ref: 1401/8. Reproduced with permission of Canterbury Museum) Ferry Road/Ferrymead Liquefaction Investigations • 265

Chapter 12 Ferrymead Bridge /Ferry Road Liquefaction Investigations

12.1Ferrymead Bridge rating. That rating system, contained in the February 1994 Bridge Strengthening Programme Report, placed The following material is summarised from a docu- the Ferrymead Bridge second behind the Durham ment entitled “Ferrymead Bridge Lifelines Investiga- Street Overbridge. tion - Initial Report”, prepared by City Design, Christ- church City Council in December 1996. Structural investigation work carried out thus far on the performance of the Ferrymead Bridge during a seismic event has shown that the structure and its services are Background extremely vulnerable. The best mitigation option is to Ferrymead Bridge (see Figure 12.1) was first flagged replace the existing structure with a new one. as a lifelines hazard by Works Consultancy Services in their hazard report prepared for the October 1994 Traffic increases as a result of growth in the Sumner Lifelines Workshop. Work then began on assessing the area will require an additional two lanes to the bridge impact of a major lifelines event on this structure and within ten years and significant traffic alterations to the its associated services as the bridge provides the only intersection on the Sumner side of the bridge. link to Sumner for most of them. A likely scenario may be a parallel widening scheme Concurrent with this work on the Ferrymead Bridge all whereby a new structure that can withstand all lifelines Christchurch City Council controlled structures desig- events is built upstream from the existing one. The nated as a hazard in the October 1994 report were impetus for spending a considerable sum of money on

prioritised for mitigation in terms of a Cost/Benefit strengthening the existing structure is therefore lost as

o R

s t r e e e r y t

D S h t r o w E s t u a r y s le r a h C

H u m T p id h a r l V Ferrymead e i y e s w F D Bridge ER r RY iv d a e o R in a ROAD M

T u n n e

N h t a

R P

d le H id eath r r cote Rive B

Figure 12.1: Ferry Road/Ferrymead Bridge location 266 • Risks and Realities

is the need for an immediate continuation of an inves- and there are rubber bearing pads between the deck and tigation into strengthening options. the columns. At the abutments there are hold down bolts and rubber bearings. The deck, which spans Nonetheless it is considered worthwhile to at least between the post-tensioned arched beams and interme- record the work that has been done to date and then diate transverse diaphragms, gives a honeycombed resurrect this work when traffic options are more appearance as shown in Figure 12.3. There are service advanced. ducts on either side of the bridge under the footpaths. On the downstream side the duct contains Telecom Introduction services and the upstream duct contains two 300 mm The Ferrymead Bridge provides a vital link to Sumner water mains. Other services run under the deck be- for transport (including a cycleway) and services. tween the bearings and through the diaphragms. Twenty-five thousand vehicles per day use the bridge at present, with traffic flows increasing at faster than average rates for the rest of the city due to the effect of high growth in the Sumner area. The bridge carries all the water, sewerage and telephone lines for Sumner. The bridge is also a vital link in the overweight/ overwidth vehicle route for vehicles to and from the port of Lyttelton. The bridge is shown in Figure 12.2.

Figure 12.3: Honey comb structure under the Ferrymead Bridge

The deck was recently widened by 800 mm on the downstream side to make room for a cycleway. The approaches consist of filled embankments approximately three metres high.

Lifelines Hazards Following the lifelines approach three events have Figure 12.2: The Ferrymead Bridge been considered, namely earthquake shaking, tsunami and flood, initially all with a 150 year return period. The existing Ferrymead Bridge was built in 1967 to (The modern design code return period earthquake for replace an existing narrow structure. Based on Transit a structure of this type would be 1,000 years.) Follow- New Zealand’s design criteria the life of a bridge is 100 ing a description of the events concerned, the potential years, which means for the Ferrymead Bridge a re- effects of these hazards are described. maining life of at least 70 years. An investigation of possible strengthening/mitigation options for the bridge is therefore seen as worthwhile. Considerable effort Earthquake has therefore been put into analysing the possible Transverse and longitudinal shaking from a 150-year response of the existing structure to various hazard return period earthquake is expected to produce shak- events including earthquake shaking, tsunami and flood. ing intensities of between VIII and IX on the Modified Mercalli Scale. This is likely to leave alternative routes Structural Description to and from Sumner via Evans Pass and Bridle Path Road closed by slips. The bridge, should it survive, The Ferrymead Bridge is a continuous, three-span, would therefore provide the only available access to post tensioned arched beam structure. The L-shaped Sumner. Leaving the community isolated. abutments and the piers are piled. The piles are at least 10 metres long founded on rock at the Sumner end and in sand at the city end. The piers have four tapered Tsunami columns, a pile cap and 16 prestressed concrete piles. Ferrymead Bridge is the only bridge affected by the The columns are connected to the deck by steel dowels 150 year return period tsunami event. A water level Ferry Road/Ferrymead Liquefaction Investigations • 267

three times above normal high water level is expected, Calculation of joint behaviour led to the following which equates to road level. Water velocity is given as conclusions: 7 m/s or seven times the normal maximum flow rate. This imparts an extremely large load onto the bridge • The steel pin connection at the top of the pier together with huge scour potential so a further search column is sufficient to fail the column in bending. was made to confirm whether or not the bore height and • There is no pile cap steel within the area of the joint water velocity were realistic. within the pier column. Behaviour of this joint is Derek Todd of the Canterbury Regional Council was therefore difficult to predict. One way to restore able to confirm that the bore height and velocity used capacity is by post tensioning the pile cap. in this investigation were realistic, but felt that the • The transfer of horizontal transverse load through return period was less certain and could be much the hold down bolts and bearings at the abutments greater. is a complicated combination of shear, bending and friction. The joint is not strong enough to transfer Flood the full code load earthquake force from the deck. The maximum flood velocity is predicted as only Sharing of the deck load between piers and abut- 1 m/s compared to the 7 m/s for a tsunami generated ments requires a computer analysis and soil stiff- bore. Flooding is therefore not a critical hazard for this ness properties to determine. structure and is not considered further. • Horizontal transverse load transfers through the abutment wall to the abutment piles and also to the Structural Analysis wingwalls. The interaction between the wingwalls The structural analysis was carried out in two stages. and the backfill and the piles and soil cannot be Stage 1 involved calculating member strengths, trying determined without definitive soil properties and a to determine a basic failure hierarchy, getting “a feel full computer analysis. for the structure” and assessing the need for a site investigation to determine soil properties. Structural Performance Under Earthquake Longitudinal Shaking Stage 1 Structural Analysis To sustain longitudinal horizontal load passive resist- The effects of longitudinal and transverse earthquake ance of the soil behind the abutment wall is required shaking were considered on the superstructure and together with sufficient ductility in the pier columns to then on the piles. Calculations developed for trans- withstand the displacement required to generate the verse earthquake load were then used to determine the passive resistance. Flexibility in the abutment wall and performance under tsunami loads. its supporting piles is also required.

Calculations of member strengths and relative stiffness Structural Performance Under Earthquake gave the following results: Transverse Shaking Calculation of member strengths led to the following • Passive pressure behind the wall is sufficient to conclusions. resist full code earthquake forces.

• Pier columns will yield before the pile cap. • The piles are flexible enough in bending to allow sufficient rotation to enable generation of passive • Pier columns will yield before the piles assuming pressure behind the abutment wall though whether sufficient lateral soil strength. bending can take place or not may depend on the soil stiffness, which can only be determined by soil • The column shear strength is adequate to resist the testing. shear associated with column bending overstrength. • Shear capacity of the piles is not known as the • There is sufficient longitudinal column steel for amount of transverse steel present is not shown on strength, though the bar spacing is greater than the drawings. allowable under NZS 3101:1995. Ductility is therefore a problem. • Soil strength around the abutment piles should be sufficient to sustain the seismic loads. CPT testing • There is sufficient area of transverse confining is required to confirm this. steel but the spacing is greater than allowable, especially at the critical column/pile cap interface Capacity calculations for the pier column/pile joint causing a ductility problem. were inconclusive. The longitudinal bending capacity 268 • Risks and Realities

of the pier columns and the piles is similar. The joint, Mitigation possibilities are to strengthen the abutment/ i.e. the pile cap, should be able to transfer the moments deck connection and to strengthen the piers by infilling and shears. Laboratory testing would help to confirm between the columns with concrete or cross bracing, joint behaviour and to provide a more accurate assess- assuming that the piers are the weak link and not the ment of column/pile capacities. Soil stiffnesses are piles. Soil testing is required to determine this. required to determine actual pile behaviour, together with computer modelling. Conclusion From Stage 1 Structural Analysis The abutment/deck connection can be improved and Structural Performance Under Tsunami Loading pier column ductility can be improved, though there is A tsunami has three effects: flotation, scour and trans- little point in doing this if the soil strength and stiffness verse force on the bridge. is such that failure occurs in or is transformed to, the piles. Liquefaction potential is also a concern. A site investigation was therefore required. Stage 2 of the Flotation structural analysis could then proceed on the basis of Flotation reduces the dead weight of the bridge to the results from the site investigation and decisions could point where it can more easily be pushed sideways and be made as to what mitigation methods, if any, to use. also reduces the effect of dead load on the post- tensioned beams to the point where explosive failure is possible. Mitigation possibilities are as follows: Stage 2 Structural Analysis

(a) Strengthening the connection between the deck Site Investigation Results and the abutment. The results of the site investigation are somewhat alarming as the site is the most susceptible to seismic- (b) Drilling holes in the honeycomb deck to allow air generated liquefaction so far found in Christchurch. to escape. The reason for this is that the Heathcote River and its (c) Putting in place a contingency plan whereby trucks predecessors have probably always entered the sea or can be placed on the deck to reduce post tensioning estuary at this site, as the rock headland of St Andrews compression stresses. (As the most likely origin of Hill will have held the outlet. The site has therefore a tsunami is South America there would be suffi- been subject to a long period of aggradation in shel- cient time available to arrange this, but there may tered tidal conditions. Such low-energy environments be difficulty in persuading truck owners to put their are often identified with loose sediments and a lique- trucks at risk.) faction hazard.

The implications of liquefaction for the structure are Scour severe. There are three problems with liquefaction; Based on the effects of a tsunami which originated in loss of lateral support to the piles, approach embank- South America and reached New Zealand at the end of ment settlement and lateral spreading. These problems last century, the river channel dimensions could be are discussed in the following sections. expected to increase substantially though quantifying this effect is complicated. Loss of Lateral Support To Piles The bridge and its approaches form an impediment to Liquefaction is predicted to a depth of up to ten metres the flow. This further complicates any attempt at a for the 150 year return period lifelines event. The actual theoretical prediction. To properly quantify this prob- extent of liquefaction predicted depends on the method lem a university model study is required. used, but even the most conservative method still predicts liquefaction sufficient to cause failure of the Transverse Force piles in bending from transverse loading for the 150 The expected force on the structure is related to the year return period earthquake. Collapse is certain for square of the bore velocity, assumed equal to its maxi- the design code 1 in 1,000 year return period earth- mum value of 7 m/s. This imparts a large force of quake. The question is, then, what minimum return similar magnitude to a full code earthquake. The struc- period earthquake would cause failure under trans- ture as it stands could not withstand this with collapse verse loading? Further investigation of the site inves- at the piers a likely mode of failure. The exact nature of tigation results revealed that a 100 year event is still the failure would depend on soil properties. A flexible likely to cause collapse. soil would result in pile failure. A stiff soil would result Even a 30 year return period earthquake is predicted to in pier column failure. The abutment/deck connection is also vulnerable. Ferry Road/Ferrymead Liquefaction Investigations • 269

cause liquefaction of certain layers to over seven • 150 year event, or greater in the next 10 years; p = metres depth at the Sumner end of the bridge; however, 7%. with a much reduced seismic-generated horizontal force accompanying this, the piles should remain capa- • 30 year event, or greater in the next 10 years; p = ble of supporting the super structure. 29%. • It has been 68 years since the last earthquake within Settlement a 30 year return period or greater, p (30 year event The bridge piles are not expected to settle as they are or greater not occurring for 68 years) = 10%. founded on rock at the Sumner end and in firm sand at • The TNZ Bridge Design Code could stipulate a one the other, however the approaches are expected to in 1,000 year earthquake to design for, for a new settle under 150 and 100 year return period earth- structure at this site. quakes. 1,000 year event or greater in the next 10 years; p= Approach settlement of one metre relative to the bridge 1%. deck during the 150 year return period event is conceivable. This is clearly an unsustainable situation for 1,000 year event over bridge design life of 100 services crossing the bridge. Settlement is also theo- years; p = 9.5%. retically possible under a 30 year return period earthquake though this is considered unlikely due to the Conclusion from Stage 2 Structural Analysis mitigating influence of the approach overburden material and the silty nature of the surface layer. The Ferrymead Bridge is extremely vulnerable to seismic generated liquefaction. The critical element in Accurately assessing the relationship between return this case is the piles, though a definite mode of failure period and settlement is not possible. could not be determined without a 3D computer analysis. Column ductility and the deck/abutment connec- Lateral Spreading tion could be improved, but with the probable mode of failure being lateral spreading induced pile failure, Lateral spreading is a phenomenon whereby liquefac- there seems little point in investigating strengthening tion of sand at or below bed level allows the embank- further. ment to move over the liquefiable layer towards the centre of the waterway. This can apply passive pres- However, one aspect that did require further investiga- sures to abutment walls, piles and pier pile caps and is tion is the effect of liquefaction on the surrounding area also the cause of the very high approach embankment and the rest of Ferry Road in particular. With this in settlement predicted. The site investigation showed mind a site investigation for Ferry Road was carried out that this is the most likely cause of collapse of the and the results and the implications for lifelines are structure during seismic shaking. summarised in Section 12.2.

Calculations carried out subsequent to the site investi- The behaviour under a tsunami load could be investigation show that the abutment piles would fail under gated further. However, it was felt that as this event lateral spreading induced pressures for a 150 and a 100 carried with it so many uncertainties determining miti- year return period earthquake, but while a 30 year gation measures, it could await the outcome of propos- return period earthquake may induce high stresses, it als for traffic generated upgrades to the bridge and should not cause failure. There is also a mitigating proposals for services. effect of overburden pressure. As the amount of overburden increases the potential for liquefaction de- Services creases. This is not enough to eliminate liquefaction The Ferrymead Bridge is packed with vital services for under the approach fills for 150 and 100 year events the Sumner/Mt Pleasant area including water, sewer, and even theoretically for a 30 year return period event, telecommunications, power and streetlighting. The however Soils and Foundations feel that predicting bridge provides the only link to the area for water, liquefaction for a 30 year event is overly conservative. sewerage and telecommunications (the main power transmission lines run across the Port Hills). Liquefaction Probabilities Given the dramatic predictions of damage as a result of Services Vulnerability Assessment liquefaction from relatively low-return period earth- As previously mentioned, settlement of the approaches quakes, it is interesting to consider some of the relevant during a seismic event with respect to the bridge deck probabilities: 270 • Risks and Realities

would cause rupture of the services at the abutment. A Southpower’s response plan is simply to replace any tsunami may also disrupt services either by damaging damaged cables with overhead lines at this location. the bridge or by washing out the approaches. A major Southpower are currently directing their lifelines ef- flood is not expected to cause any damage to services. forts at their substations, particularly those constructed before 1964. Prior to strengthening they are capable of Response Plans withstanding only a 40 year return period earthquake. It is hoped to improve their performance to enable Existing Services and their respective authorities pro- survival of at least a 150 year event. jected response plans to date are as follows:

Possible Mitigation Measures Water Several mitigation measures have been proposed for Two 300 mm diameter water mains cross the bridge on services in order to get around the problem at the the upstream side in a service duct. Water Services abutment/backfill interface, including finding an alter- response plan is to have isolating valves on either side native route to Sumner and thrusting under the river of the bridge with 50 mm take-offs that hoses can be away from the bridge. connected and then strung across the river.

Thrusting Sewers In August of 1996, Waste Management successfully Two pressurised sewer mains run under the centre of thrust a new 325 mm diameter sewer under the the bridge, a 375 mm diameter AC pipe and a 150 mm Heathcote about 400 mm upstream from the Ferrymead diameter AC pipe. West of the bridge the sewers are Bridge to service part of Mt Pleasant The cost was gravity fed. The 375 mm diameter sewer is in need of approximately $1,000/m for the 200 m line. This does replacement as it is choking from the swelling effects not provide an alternative route for the lines across the of hydrogen sulphide. bridge as they connect into different systems, however Waste Management’s response plan is to allow the should the lines across the bridge be lost then a 500 m sewage to discharge into the estuary (as would be the overground line could be used as a temporary connec- case at any other structure in the city). tion between the two systems. There is also the possibility of a Telecom or power Telecommunication service being laid inside a sewer line thrust under the The Telecom service consists of a fibre optic trunk line river. Watermains can be thrust directly. that serves an exchange on the Mt Pleasant side of the bridge, and a 4,000 pair cable that runs back across the The system uses a welded plastic pipe which is pulled bridge to serve Woolston. A cellphone site on Mt through a bentonite stabilised hole. The hole is made by Pleasant is also accessed by lines across the bridge. a reamer that is pulled through the hole as the thrusting This is the only cellphone site located on the hills and device is removed. Thrusting deep enough to avoid is therefore the site with the best coverage in Christ- tsunami-related scour is not a problem. church and a site that could be important after a major lifelines event should other services on the flat be Automatic Shut - Off Valves disrupted. There is also a VHF site on Mt Pleasant but Another suggestion from Kobe for protecting water this does not require a link across the bridge to operate. stored in reservoirs from reticulation leaks/failures is to install automatic shut-off valves, which are acti- Telecom’s response plan for the fibre optic trunk is to vated by seismic shaking. on some reservoirs. The replace it with an overhead line. This can be done fairly remaining reservoirs are left connected to the system to quickly. The 4,000 pair cable however, could take provide water for fire fighting (work is already underway months to replace depending on priorities in other parts on this). of the city. Telecom expect half their response effort after a lifelines event to be repairing damage caused by other authorities in the course of their repair opera- Alternative Route tions. One of the keys to any service authority’s response plans to any problem is diversity, an option not cur- Power rently available in the case of Ferrymead Bridge. All the service authorities are therefore very keen to ad- An 11 kV service crosses the bridge along with dress this and, as a result, favour investigation of an streetlighting and other minor cables. This is not the alternative route to Sumner from the Brighton Spit. main supply and is not so important. Ferry Road/Ferrymead Liquefaction Investigations • 271

Transportation Issues Bridge is the vehicle access required following an Sumner is a rapidly growing suburb and this together event for the repair of services damaged by a tsunami with average projected traffic growth of 2% to 3% will or an earthquake. In the case of a tsunami damaging the mean the bridge will require at least an extra two lanes bridge, alternative access is available along Bridle Path by the year 2003 to cope with an expected traffic flow Road or the hills. However, after a major seismic event of 35,000 vpd. It may also be opportune to provide a slips will have closed these two routes leaving the area cycle crossing on the south side of the bridge as the isolated, perhaps with some water left in reservoirs cycleway on the north side is difficult to access from should they have been fitted with shut-off valves but the south. This could have extremely important rami- with no access for equipment to repair reticulation fications for retrofitting plans for the bridge. problems. The water services across the Causeway are also vulnerable. Breaks are likely on other reclaimed or The bridge currently carries 25,000 vpd. If this link was filled areas around McCormacks Bay, which is unfor- lost, the traffic network model predicts that 18,000 vpd tunately very close to the two big reservoirs on Glenstrae would use Bridle Path Road and that the rest would use Road. Evans Pass. The additional travel costs of losing the link are calculated as $69,000 per day. This figure is An earthquake significantly smaller than that required important when considering benefits derived from to collapse the bridge could cause sufficient settlement mitigation measures. to sever all of the services. Maintaining traffic flow while water, sewer, Telecom and power are simultane- For example, should the bridge be only slightly dam- ously repaired will not be possible. Significant delay aged and be able to be reopened within only two weeks, costs are inevitable. the traffic related cost to the community is still a million dollars. Summary This adds impetus to the argument for either replacing The Ferrymead Bridge is extremely vulnerable to the existing structure or building a new two-lane struc- seismic generated liquefaction. The most likely cause ture adjacent to the existing one, which will survive an of failure is from lateral spreading causing failure of event intact and therefore maintain at least two open the abutment and piles. lanes while the remains of the existing structure are removed and a replacement structure built. Some improvement to performance during transverse and longitudinal shaking and tsunami effects may be Traffic growth predictions also require consideration possible though this will require a computer analysis. of the Bridle Path/Ferrymead/Main Road/St Andrews Hill intersection. This needs to be worked through. Planned traffic modifications may have a significant impact on the future of the bridge. There is another argument for providing a new bridge from Sumner to the Brighton Spit. At a cost of five or Recommendations six million dollars, the idea seems initially to be folly, (a) Await study of traffic options for the bridge, the however this route would provide an alternative emer- intersection and a Spit Crossing. gency access to Sumner and would also provide a beneficial link in the network. Simply considering (b) Then investigate options for a new structure, wid- traffic travel time benefits over 25 years, discounted ening the existing, strengthening the existing or a back to present values, the benefit is around 25 million combination. dollars. Including other costs and benefits compared to a Do Minimum option, the Net Present Value of benefits is nearly 40 million dollars. 12.2 Ferry Road Initially service authorities felt a bridge across the Spit The following material is summarised from a Christ- would provide a useful link for services, however it church City Council document entitled “Ferry Road was felt that services are best kept away from struc- Liquefaction - Sketch Report”, prepared in October tures. The Spit Bridge idea will therefore be given 1996. consideration from a traffic network perspective only in the meantime and as provision of an emergency The report covers the effect of liquefaction from a 150 access route following a lifelines disaster/event. year return period event on Ferry Road/Ferrymead and its lifelines, from the Ferrymead Bridge to the Tunnel Interdependencies Road Roundabout. The main interdependency regarding the Ferrymead From site investigation soil test results, an assessment 272 • Risks and Realities

was made of the impact on the various lifelines, those tanks at the service station by the bridge are prime of primary concern being the buried services (which candidates for flotation with consequent environmen- may float), the riverbank (which may suffer from tal damage and fire risk. lateral spreading) and the general disintegration of the road surface. Sewers The investigation indicated a similar soil profile along Manholes and pipework are at risk from the effects of the length of Ferry Road in question with approxi- flotation during liquefaction. Calculations show that mately two metres of interbedded silts and sands over- for the solid thick walled sewer manholes, many are in laying sands (see Figure 12.4). fact more likely to sink than float. However, the pipe itself is subjected to a considerable buoyancy load if Lifelines in the Ferry Road/Ferrymead area are as empty at the time, though mitigating this is the un- follows: known effect of the pipes concrete haunching. Some displacement in plan can be expected adjacent to the • Transport — Ferry Road carries 15,500 vehicles river due to lateral spreading. per day and provides access for the local fire station. The rubber ring joints will reduce the impact this has, though experience from Northridge suggests that if • Sewers — A 675 mm diameter foul sewer runs nothing else this area would be a good starting point under the centre of the carriageway, a small 150 when looking for damage to pipework from a major mm diameter main runs part way along the northern shake. property boundaries and various lengths of stormwater piping traverse the street. Lateral spreading and major ground rupture aside, stormwater sewers are less likely to be affected by • Water — A 300 mm diameter AC Main runs under liquefaction as they are relatively shallow and there- the southern side of the carriageway. fore above the liquefaction zone. • Telecommunications — Twelve ducts run under the northern footpath. There are 16 access man- Water holes from the bridge to the roundabout. Steel or AC water pipes when full of water have neutral • Power — Overhead powerlines run the length of buoyancy and are not greatly affected by liquefaction Ferry Road on the north side. On the southern side but they are affected by lateral spreading, particularly underground cabling services the streetlights. Two as the watermain is the nearest underground service to high voltage transmission lines supported by lattice the river. The sidelines are quite likely to be damaged towers cross Ferry Road about halfway along the at the joints to where there is a variation in the bare section of road in question. material. Water Services response plan should take into account probable loss of this section of the main. • Streetlighting — Streetlights run along the southern side of Ferry Road and are supported by con- Telecommunications crete poles and single outreach mastarms except at Manholes are the susceptible part of the Telecom the Tunnel Road end where the streetlights are system. However, the liquefaction zone does not supported by the power poles on the north side. extend to the surface and the buoyancy load can therefore be resisted by the dead weight of the manhole, Effects on Lifelines overburden and the shear strength of the ducts them- Details of the expected effects of seismic-generated selves. liquefaction on services in Ferry Road/Ferrymead are as follows. Calculations in fact show that if anything the manhole may tend to sink, with this tendency increased should the assumed wall thickness of 150 mm be exceeded. Transport Resisting this tendency is the shear strength of the Widespread liquefaction along the carriageway with ducts. Shear strength varies depending on the material surface cracking and sand boils is expected. Close to but with 12 ducts entering either side of each manhole the Ferrymead Bridge and along the riverbank at the the Telecom manholes should in theory remain stable. Tunnel Road end severe damage is likely from lateral spreading to the extent where the carriageway may Power become as impassable. A response plan is therefore required for restoring access to the fire station (which, Power poles are only considered susceptible should itself, may also be subject to damage). The fuel storage they be subject to a lateral load from cables changing

Ferry Road/Ferrymead Liquefaction Investigations • 273

Pole

Concrete

Street Light

Pole

Power

Concrete

Timber or Timber

Telecom

t h

o a

e l

i r

Invert

1.94m

to 3.95 Bridge

Ferrymead

HA1

(2 tests) Sewer

E s t u a r y w r

D ie e

V v l i

a O

id e R

T riv R

D e s t ey hr

p co

Hum

th HA6

HA2

ea Probable

H High

Liquefaction

Very High

High HA5

Y Probable

R Liquefaction

CPT2

HA3 t

r t

Test Sites Test t

CPT2

r High

o N

Probable s

Liquefaction a

HA4

W.T.

. .

x x

x .

. .. .

Silt

xx x

Log

x x .

. Bore

Sand x Road CPT2

HA5 nel CPT1 Tun

o R

r e

D Silty

Sand

HA6

Depth

2 1 3 (metres)

Comments

is indicative only. This analysis is from scala results and Only with unusually high water table. is indicative only. This analysis is from scala results and is indicative only. This analysis is from scala results and is indicative only. This analysis is from scala results and indicates liquefaction at depth. Test probably in fill, CPT from bridge probably in fill, CPT Test bank will result in lateral displacement. if water table high. Proximity to river above 2.2m have high susceptibility Water table approximately 2m. Soils Water is indicative only. This analysis is from scala results and is indicative only. This analysis is from scala results and

Susceptibility

(m below ground) below (m

Liquefiable Layers Liquefiable

2.0 - very low

5.2 - 5.8 mod 2.2 - 4.0 mod 5.6 - 6.6 mod 1.8 - 2.2 mod-low 3.8 - 4.4 mod 2.4 - 3.8 low 1.5 - 2.0 low 1.5 - 3.0 low 2.5 - 3.0 low 2.0 - 2.5 mod 1.8 - 2.5 low 2.0 - 3.0 high

Location

HA1 1.0 - 1.5 mod HA2 1.0 - 1.5 mod HA3 1.5 - 2.0 high HA4 1.5 - 1.8 high HA5 1.0 - 2.0 very high HA6 2.0 - 3.0 high

CPT1 1.2 - 1.8 low CPT2 1.2 - 2.4 high

Figure 12.4: Summary of test results, location of tests and services locations for Ferry Road site investigations 274 • Risks and Realities

direction. Each power pole in this category needs to be Conclusions and Recommendations assessed on an individual basis. The site investigation showed that for the 150 year return period event the liquefaction risk though not as The high voltage transmission lines which traverse great as the Ferrymead Bridge, is still significant. Ferry Road are supported by lattice towers. The critical loading case for these structures is wind and they are Disruption to services, particularly in the area affected therefore unlikely to overturn as a result of seismic by lateral spreading will be severe. loading. They are, however, quite heavy. Minor settlement is therefore probable (assuming the foundations Conclusions/recommendations are as follows: are not piled). • A response plan is required for Fire Station access There are also some cables underground though these and for flotation of fuel storage tanks. are relatively shallow and therefore less likely to be • The sewer manholes and pipework are not at great directly affected by liquefaction. high risk except adjacent to the riverbank, where a response plan is required. Streetlighting Concrete streetlighting poles extend further into the • Water pipes (assuming full of water) are only at risk ground than power poles and therefore further into the by the riverbank also, where total loss from lateral liquefaction zone, to account for the overturning effect spreading can be expected. of the streetlamp outreach arm and the weight of the • Telecom manholes being close to the surface and concrete pole. From previous earthquakes it is likely on the north side of the street away from the river that many of the streetlight poles will be left at odd are expected to perform satisfactorily. angles and require straightening. • Power poles are vulnerable if subject to lateral load. This should be mitigated against where possible.

• Streetlights are likely to get out of plumb. Continuing Work • 275

Chapter 13 Continuing Work

13.1 Introduction actually necessary beforehand by way of mitigation. The engineering lifelines work will continue in Christ- church, and there are many activities that could be • However, each lifeline does not exist in isolation undertaken. The following is a list of possibilities that and what has to be considered also is the interde- will be considered from time to time by the Steering pendence with other lifelines, i.e. the effect that a Committee with a view to having some projects under- lifeline or other lifelines may have one on another taken each year. has to be considered.

It is important to acknowledge that one of the continu- • It is necessary also to look at the benefits of what is ing benefits of an engineering lifelines project is the proposed then to be sure that the work is cost inter-utility contact between persons who would be effective. required to work together in the coordination of their • Initially this work is done on a very broad scale work in the event of a major emergency. At least an which is what the project has done so far, but by far annual project/function will be required. most of the work that is undertaken as a result of the lifelines project is not done in conjunction with In the final session of the October Workshop in 1994 other services, but arises from an individual de- Dr David Hopkins, the External Reviewer, presented a tailed investigation of a service by that utility chart depicting the Lifelines process as set out below. owner. It is the follow-up of the project which results in the mitigation measures.

• Once an initial detailed assessment has been done, unfortunately that is not the end of the process. More information will be available in the future regarding the hazards, the risks, the various new materials and methods dealing with problems, so the process continues until the diminishing returns are such that it is no longer worthwhile doing anything.

The chart is based on an earthquake hazard but is appropriate for the other hazards.

All of the service authorities in Christchurch have now • David Hopkins pointed out that the process is not a been once round that cycle, many in an informal way, “lifeline” it is a “lifecycle”. but the opportunity is now for the full process to be revisited. The following discussion addresses the is- • Once the hazard has been determined and the sues that arise in the future as possible work based on vulnerabilities assessed, the two together form the the lifelines cycle. risk. • It is then possible to postulate some mitigation to 13.2 Hazards deal with that risk. In all areas of natural process, there is always knowledge • The cost that may be required to deal with the risk to be discovered. The hazards in Christchurch are no can be estimated but that is not the end of the exception. Research in some areas is ongoing as part of process. other programmes, particularly for the seismic hazard. • It may also be necessary to analyse what would be the response if nothing had been done. The method Earthquake hazard of response may be so simple that no work is The University of Canterbury and the Institute of 276 • Risks and Realities

Geological and Nuclear Sciences (IGNS) have ongo- To achieve this aim the following tasks need to be ing research programmes investigating North Canter- completed: bury faulting, the magnitude and recurrence intervals of movement on the alpine fault, and the nature of the (1) Summarise the tectonic setting of the Canterbury recent Arthurs Pass earthquakes. In fact a paper dealing region with respect to the principal causes of earth- with the earthquake hazard in Christchurch produced quakes. by D J Dowrick, K R Berryman, G H McVerry and J X (2) Compile existing records of historical and instru- Zaho has been received in late 1997 and is yet to be mental seismicity in the Canterbury region. considered by the Hazards Group. It is obviously worthwhile to attempt to resolve any ongoing differ- (3) Compile existing information on active or poten- ences in approach in relation to the extent and effects tially active faults and other tectonic structures in of the hazard. the Canterbury region and nearby (including off- shore) that may impact on the region. This should As part of their lifelines mitigation investigations, the include information on: Christchurch City Council will soon have completed in-situ testing for liquefaction potential at over 30 sites (a) The location and known accuracy of location of in eastern Christchurch. faults; The Canterbury Regional Council is currently investi- (b) The length of individual fault segments, the size gating coordinating research in the region, collating of past individual displacements, the frequency recent and current work and some details of this is set and magnitude of past earthquakes associated out below. There is a need to revisit the seismicity with these surface displacements, and the tim- model for the region and Christchurch and this will also ing of the last event; and assist to resolve current divergent opinions on the probability of damaging seismic shaking. (c) The location of tectonic structures (other than faults) and if appropriate, the frequency and In recognition of the earthquake threat to Canterbury, magnitude of any past earthquakes associated and the Regional Council’s statutory responsibilities, with these features, and the timing of the last the Canterbury Regional Council will complete the event. first stage of a comprehensive Earthquake Hazard and Risk Assessment study in 1997/1998. The long-term (4) Undertake aerial photograph analysis in areas programme will address the earthquake hazard itself, where no existing data are available to determine the risks posed, possible mitigation options and mitiga- the location of active faults. Priority areas include tion implementation methods. central and south Canterbury, and the area surrounding the Culverden basin. The general aim of the first stage of the study is to determine the location, average return time and associ- (5) Undertake a probabilistic hazard analysis using ated earthquake magnitudes caused by geological struc- the information from (2), (3) and (4) above to tures (such as active faults) capable of generating determine the MM Intensity and peak ground moderate to large magnitude earthquakes. accelerations (PGA) for the main urban areas (refer Section 3) for return periods of 50, 150, 500 Future stages of the programme will look at ground and 1000 years. shaking hazard, liquefaction and associated ground damage, other related earthquake hazards such as slope (6) Define appropriate earthquake scenarios for Can- instability and tsunami. The final stage of the pro- terbury, or parts of, to be used in assessing future gramme will be a risk assessment and will include ground shaking, liquefaction, slope failure and (if estimates of potential damage and casualty estimates appropriate) tsunami inundation studies. This for the main urban areas of the region. should involve obtaining a consensus among appropriate experts. Details of stage one of the study is set out below. (It is important to note that the investigation is not limited to (7) Identify future work that could be undertaken to the Christchurch Urban Area as is the current Christ- better understand the hazard from major large, church Engineering Lifelines Project.) damaging earthquake sources in Canterbury region.

Aim of Study Other Hazards The aim of Stage I of the study is to locate and characterise the active geological structures in the Flood plain of the Avon, Heathcote and Canterbury region capable of generating moderate to Styx Rivers. large earthquakes. Since the initial Project, computer-based modelling Continuing Work • 277

has included the simulation of the extent, depth and amount of work was undertaken with the aid of a duration of flooding within the flood plains of the three computer. small coastal rivers — Avon, Heathcote and Styx. Various scenarios have been modelled including an Many of the utilities are involved in recording details “extreme” event and the impact of the Waimakariri of their service electronically using Geographic Infor- River overflows. mation Systems (GIS). It would seem that when sufficient records are available, and probably with the new It is intended that floodplain management strategies earthquake hazard information, it will be worthwhile will be adopted jointly by the Canterbury Regional investigating vulnerabilities with the aid of GIS. Council and the Christchurch City Council for implementation in their Annual Plans and Budget and City/ At the October 1994 Workshop Mr Ron Eguchi dem- Regional Plans. As at July 1997 a strategy for the Styx onstrated the then-current usage of a computer in the River had been adopted and adoption of strategies for cost assessment and prediction of damage in California the Avon and Heathcote Rivers is intended by end of and sophisticated software is now available in New 1997. Zealand. Depending on the extent of GIS usage in the various services, it may be worthwhile to model/ana- The general thrust of the strategies is to recognise (and lyse the damaging effects on infrastructure using the protect if appropriate) existing ponding and floodplain now available software. Set out below is the main areas, have appropriate development rules, consider features of one suite of programmes which is available the effects of rising sea levels and identify any neces- and consideration should be given as to whether or not sary special flood warning or emergency procedures. each services wishes to undertake further investiga- These strategies apply generally not only to the engi- tions at this stage. neering lifelines. It is not anticipated that the more detailed work since 1994 will be likely to require any Opus International Consultants have provided the fol- modification of the work on the lifelines. lowing information on “Earthquake Loss Assessment Analysis Developments”

Tsunami and Wind “There are two main developments: The tsunami and wind scenarios have been reviewed • The use of Geographic Information Systems (GIS) since 1994 and no further work is proposed at this stage to process, store, analyse, retrieve and present on them. The wind hazard was extended to include spacial data. GIS use is eminently suited to hazard, comments on a cyclone hazard. asset, and risk data applications whereby hazards (earthquake sources and characteristics, ground Biological, hazardous substances, conditions — amplified shaking, liquefaction, land- Vandalism/Terrorist slip and fault movement); inventory (buildings, The Auckland Engineering Lifelines Project is consid- lifelines, emergency services facilities); and de- ering a different range of hazards from those consid- mography etc. are established as “data layers” ered in the Christchurch area. Although it is obviously from which composite models can be established. not appropriate to consider a volcanic hazard for Christ- • Hazard, loss, risk, analysis models and systems church, there is some merit in considering some of the based on data inputs from GIS layers and common other hazards under investigation in Auckland (i.e. database systems (Lotus, Excel, Access, dBASE biological hazards, hazardous substances spill, vandal- etc.). ism/terrorism) and it may be that these should also be investigated for Christchurch. Thus there are now modelling/analysis capabilities to specify a wide range of earthquake scenarios and to forecast the damaging effects on infrastructure, esti- 13.3 Use of GIS mate damage repair costs, and to assess numbers of displaced people, casualties etc. At the time of the original project it was considered that for nearly all services the assessment of vulnerabilities Such analyses producing quantified damage estimates by an experienced person reviewing a map of the are useful as a basis for risk management practices and service under consideration overlaid over the particu- emergency preparedness. Models can be established lar hazard was by far the best way of assessing for specific intersets, e.g. roading, or telephone com- vulnerabilities. There are so many variables that have munications, upon which to base business risk deci- to be taken into account, most of which are not readily sions (mitigation by addition, replacement and retrofit quantifiable, which means a computer analysis is not to appropriate design standards), preparedness plan- only difficult but may not be worthwhile. However, in ning (post event business continuity, emergency re- the telecommunications assessment a considerable sponse and recovery measures: Insurance provisions 278 • Risks and Realities

etc.), or where multiple infrastructure facilities are to make loss assessments for particular or general modelled then to then determine damage effects inter- infrastructure facilities, producing maps of damage/ actions and interdependencies. For example damage cost consequences and tabular output in database form to water supply systems can be matched to residential for subsequent analysis according to use.” building property (non or low) damage areas and emergency accommodation situations as a basis for planning water provisions in emergency response. 13.4 Risk Balancing The easier application of scenario based loss assess- As at August 1997 it is most likely that a Ph.D. student ments enables economical and effective studies of will be working on a thesis based on Christchurch various magnitude earthquakes from whatever loca- engineering lifelines and this may well involve GIS tion to be used to suit the intended purpose. For capability. The following is an extract from the pro- example emergency response planning may be better posed thesis abstract. set up upon a more frequent strong earthquake rather than the very much less probable maximum probable “The importance of lifelines to human civilisation is or extreme event. Such an approach presents response significant such that any loss in lifelines can conse- and planning difficulties that are realistic, more within quently result in severe damage to the performance of the appreciation of the personnel involved, and are the community and its environment. The application of more manageable, resulting in better and more serious proper risk management techniques for different natu- consideration. ral or man-made hazards can provide good control of risks that resolved many performance and economic The quantified loss assessment process is useful for problems. The researcher on her review of literature risk management decisions and preplanning and train- found out that there are no available performed studies ing exercises before an earthquake event, and is also that employs easy-to-use and formal risk management very useful immediately following an event where approaches that are applied to lifeline engineering and either actual earthquake can be modelled to provide its safety. The proposed study addresses risk manage- the first estimates of levels and locations of damage to ment for lifelines in easy to-use manner to identify the be used as a basis for managing response activities. probable hazards, vulnerability of lifelines, risk evalu- It is clear that once hazard and infrastructure data ation and assessment; that eventually provide the have been collected then the task is to undertake means to mitigate and control risks to lifelines. The analysis to determine the effects of events. The more study will start by the case study of Christchurch as a ‘sophisticated’ the analysis, taking into account the pilot study and later a generalised methodology is to be resolution and quality of data available, then the developed. higher the confidence in the analysis results. Where there is doubt about the quality of any particular data The objective of the study is to provide a systematic then opportunity should be taken to conduct sensitivity approach that supplies a step-by-step risk manage- analyses to show the outcome variations arising from ment process that can be applied to existing and new use of the data value range. Such multiple analyses are lifelines and incorporates risk balancing as guideline. more cost effective with good computer based method- This intended guided risk management scheme shall ologies. suffice for providing optimum mitigation measures that can be applied to various risk situations, in order Funded by the United States Federal Government to be utilised by the Government, Departments of through the Federal Emergency Management Agency Conservation and Environment, regional or city coun- and the National Institute of Building Science, such an cils and utilities in lifelines planning to minimise the earthquake loss modelling system has been developed adverse effects of different risks. The expected time with intended application to all US state, regional and span of this research is about three years.” local governments use, initially for estimating regional earthquake losses for emergency planning. Risk Management Solutions (RMS) of California have de- 13.5 Use of TCLEE Guide for veloped the necessary computer systems and data linkages for this Regional Loss Model (RLM). Review of Vulnerabilities

Through a joint venture relationship between RMS and It is possible now to undertake a review of utilities Opus International Consultants (Opus) the RLM sys- vulnerability to an earthquake hazard in light of check- tem is to be made available for New Zealand applica- list lists produced by the Technical Council on Lifeline tions over the next few months (by the end of 1997). Engineering (TCLEE). This could be easily achieved Application of the RLM system takes advantage of the and could be worthwhile. These lists reprinted from current establishment of the respective GIS data layers “Guide to Post-earthquake Investigations of Lifelines” Continuing Work • 279

edited by Anshel J Schiff and published by TCLEE in • Were there loads applied through interconnections August 1991 may be worthwhile. to adjacent equipment?

This guide to “Post-earthquake Investigation of Life- • Has the base of the equipment moved on its footing? lines” has included within it reference to a checklist for • Is there a gap around the footing or equipment possible damage to engineering services as a result of pedestal indicating differential movement? an earthquake. A possible project could be the review of the vulnerabilities assessed as part of the original There are other checklists i.e. project using the checklists provided in the TCLEE Report. • equipment checklist; e.g. Anchorage checklist • power systems/power plant checklist;

• Are the anchor bolts cast in place or expansion • power systems/substation checklist; anchors? • power equipment check list;

• Can you identify the type or manufacturer of the • water systems checklist — transmission and distri- anchor? bution systems;

• What is their length of embedment? • water systems checklist — pumping facilities;

• What is their diameter? • water systems checklist — reservoirs;

• How did they fail? • water systems checklist — ground water basins;

• Did they pull out of concrete? • water systems checklist — pressure reducing and • Is the concrete cracked? Is it in a tension zone? relief facilities;

• Did fracture cones develop in the concrete? • water systems checklist — treatment facilities;

• Did the bolts stretch or pull out slightly so that they • water systems checklist — treatment plant; are not tight? • sewage systems checklist — manholes;

• Did the bolts break? • sewage systems checklist — sewers;

• Is there any indication that they were installed • sewage system checklist — treatment plant and incorrectly? pump stations equipment;

• What were the standards, if they existed, when the • sewage systems checklist — treatment plant or equipment was installed? pump station structures;

• How many bolts were there and how were they laid • airport checklist; out? • port checklist; • Did the bolt pass through a structural member in the equipment framing? • highway checklist — roadway, bridge abutments, bridge column/pier supports, bridge deck • Are there signs of distress in the equipment in the region around the anchor bolt: cracked or chipped • communications checklist — power room; paint, deformation of metal? • communications checklist — rectifiers and distri- • Does the equipment introduce a prying action to bution panels; the bolt? • communications checklist — busbar; • Is the bolt hole appropriate to the bolt diameter? • communications checklist — engine generators; • Does the load path from the equipment frame to the bolt or weld introduce flexibility in the anchorage • communications checklist — distribution frame; system? • communications checklist — switch room; • What are the sources of loading on the anchorage: equipment weight, height of centre of gravity, di- • communications checklist — transmission room; mensions of the base of the equipment? • communications checklist — building facilities; 280 • Risks and Realities

• communications checklist — data room; search has been carried out on responsive pipelines. More work is required in this area. • gas and liquid fuel lifeline issues checklists; From a review of the literature it is quite apparent that • tank checklist; the identification of potential ground failure areas • emergency power- backup batteries; represents a critical step in assessing the earthquake vulnerability of underground pipes. There are factors, • emergency power — engine generators; such as pipe material and joints that also affect the performance of underground pipes, but to a lesser • emergency power — uninterruptable power extent. supply; From observation of the earthquake effects in Kobe it • emergency power — impact or lack of emergency is quite apparent that considerable damage can be power. expected in Christchurch in the areas subject to lique- Note: These checklists were prepared for post-earth- faction and very much more information is required quake investigations and as such are not strictly appro- regarding the potential for this to assist in evaluating priate for a vulnerability analysis. However, they do the economics of replacing underground pipes. At this highlight the types of failures that have occurred in stage it seems that an awareness of the factors affecting previous earthquakes and should therefore be worth- underground pipe failures is important in design con- while to use in assessing installations prior to earth- siderations and for new piping work the vulnerability quakes. to damage must be considered.

The above summary of the checklists has been in- Further work on appropriate design codes should be cluded to indicate a possible method of reconsideration one of the outcomes of a lifelines investigation in of the vulnerability to the earthquake of the various Christchurch and perhaps New Zealand-wide and un- facilities, bearing in mind that it will be nearly four dertaking the initial work on such a code would be an years since the original investigation was undertaken. excellent project for Christchurch.

It would be of value to revisit this area to ensure the The Christchurch City Council Design Manual for issues originally raised (or perhaps overlooked) are drainage, pipework and waterways is currently being considered by current staff. updated and will include deliberate reference to lifeline considerations. It should be also important to consider whether in the design of new facilities the vulnerabilities and mitigation measures identified are being built into new de- 13.7 Critical Areas and Critical signs and the currently used design codes take into account the earthquake hazard, particularly to the equip- Buildings ment rather than the traditional design where the build- Although Christchurch, because of its topography, ing codes often were the only considerations in consid- does not have the same number of “critical areas” as ering vulnerability to earthquake. Wellington, or some of the more hilly cities, one such area is already under investigation — the Ferrymead bridge area. There are others and a multi-utility inves- 13.6 Underground Services tigation should be undertaken on these. Damage Research Likewise, there are some “critical buildings”, some of It is relatively easy to assess vulnerability of above- which have already been identified. However, consid- ground structures when sufficient is known regarding eration should not only be given to ways of maintaining the ground conditions, but much more difficult to services to those in an emergency, but to others such as assess the damage to underground services. The Wel- banks and industries that will be very vital in the lington Project acknowledged this and established a recovery after an emergency. subgroup of its Civil Services and Gas Task Group to investigate pipelines. The Christchurch project is indebted to that group for their investigations. It is noted 13.8 Interdependencies however, that although considerable effort and resources have been expended in New Zealand on re- Elsewhere in this section is reference to the search into the earthquake resistance of buildings and interdependencies between the various utilities in- bridges, and particularly components of reinforced volved in restoration of services following an emer- concrete, structural steel and timber, little or no re- gency. Continuing Work • 281

This subject has proved to be extremely difficult to deal — assessment of the status of the service in with in detail although the project and the work of the relation to key customers Task Groups has highlighted the fact that the services • Recovery of basic service are most often dependent on others for their satisfactory operation. — prioritising of a repair programme

One of the main factors in interdependencies is the time — alternatives for service taken to restore a particular service with it being accepted that underground pipe services will prove to — resource allocation be the most time-consuming in restoration. Notwith- • Restoration programme standing the difficulties however, more work should be done in attempting to develop mitigation opportunities across the various utilities. 13.10 Mutual Aid Agreements The 1993 Wellington Project presented a Case Study Elsewhere reference is made to mutual aid agreements which may be a suitable model for a similar investiga- in the water and waste water fields which are being tion in Christchurch. However, it probably has to be developed nationally and it would be worthwhile at- accepted that the co-location of the various representa- tempting to develop these in the other utilities. (Some tives from the utilities in the event of an emergency will utilities are already nationwide but others are not — probably achieve, on the day, the best coordination. some organisations have already made arrangements What has not been determined is what would be the in advance.) best preparation for this coordination. It is therefore suggested that a combined exercise for the utilities would be most worthwhile. 13.11 Internet With the increasing use of the Internet, consideration 13.9 Combined Exercise/ should be given to putting lifelines information onto the Internet, provided it can be kept up-to date. Consid- Response Plans eration should be given to establishing a small Task The advent of the new Emergency Management Groups Group to investigate this proposal. proposed in the Emergency Services Review currently (1997) being undertaken may assist in coordination but in the meantime the ‘Controllers Advisory Group’ for 13.12 Risk Management the Regional Controller in the event of a declared Civil Seminar Defence Emergency undertakes much of the same role. Since the Christchurch Engineering Lifelines Project The ‘Engineer’ for a Civil Defence Emergency (at first began there has been an increased understanding present the Director of Operations Christchurch City of the fact that the lifelines investigation is really only Council) has the coordination responsibility to advise part of responsible risk management. Consideration the Regional Controller. should be given to putting together a Risk Management Previous Civil Defence exercises have tended to focus Seminar focusing particularly on lifelines in the Christ- on the operation of the various headquarters dealing church area. We are particularly fortunate to have with welfare response needs but it would be appropri- Professor David Elms of the University of Canterbury ate to arrange a lifelines emergency exercise. Arising and Janet D Gough of Lincoln University resident in from this exercise would be a better understanding of the Christchurch area. Both are acknowledged experts interdependencies and it may be that mitigation plans on risk management and they would be able to contrib- will arise. ute most usefully. Professor Elms will be the supervi- sor for the Ph.D. thesis mentioned above and Janet A possible method of approach in producing response Gough has particular expertise in relation to the com- plans is for each utility to address response planning in munity perception of risk which may well be advanced three phases: if the suggestion of involvement with a social scientist is progressed as is suggested in the next paragraph. • Initial actions Any such seminar should address the processes as set — activation out in the Australian/New Zealand Standard “Risk Management” AS/NZS 4360:1995 which was pro- — identification of damage to the hierarchy of duced after the 1994 workshop. The project generally the service followed the methods set out in the Standard. 282 • Risks and Realities

13.13 Review Project with management of these services and to raise the aware- Social Scientist ness of the public to their importance”. Bearing in mind that the decision makers of the utilities Mr Ron Eguchi, the invited reviewer from California have to be aware of the issues involved, it will be suggested in the final session of the October 1994 appropriate to visit the various Committees and /or Workshop that it would be worthwhile attempting to Boards of Directors of the various utilities to present have the Project reviewed using a social scientist. This the findings of the Project and obtain a continuing is an acknowledgment that the engineering decision- commitment to the ongoing work. makers involved in the Lifelines work are in effect making decisions on behalf of the community without reference to the community. In the Workshop it was submitted that the community should be formally 13.16 National Forum consulted but the magnitude of this task is formidable. On 22 October 1997 it is intended that a National The suggestion then for a review of the various deci- Forum for representatives of lifelines groups be held sions required and made by the Engineers in the project and it is hoped that this will be the first of many such may result in a better response to meet the needs of the meetings. community. Dr David Hopkins the external reviewer in his final summing up of the Workshop made the It is hoped that representatives from current or pro- point that the participants in the Project as Engineers posed lifeline groups from Auckland, Waikato, Hawkes were obligated as professionals to have undertaken the Bay, Wairarapa, Wellington, Christchurch, Timaru, work in a thoroughly professional way. This will mean Dunedin, Southland and the Ministry of Civil Defence that the inevitable post-emergency review of the deci- will participate. As one of the prime movers in this sions taken both in preparation and response will be group it is intended that Christchurch will continue a defensible. national involvement. Already through the association with the Wellington group it was possible for Christ- church people to take part in investigations following 13.14 Task Group Meetings earthquakes in Northridge and Kobe.

One of the very significant benefits in the project work Some of the items for discussion at the first Forum are was the grouping together of various utility representatives with dissimilar working interests to produce the • Guidelines for the process — what is required? Task Group reports and act as peer reviewers in the • Knowledge sharing. work of others. As a result of the project, very useful contacts were made in preparation, as it were, for a • List of contacts ( both in New Zealand and over- major emergency requiring the various utilities to work seas). together. It is important that these contacts are not lost and as staff move into different positions the current • Funding approaches and strategies. staff still know their counterparts in other organisations. One method of maintaining these contacts is by • Implications of the new Ministry of Civil Defence the group as a whole attempting a project across all of for lifelines groups. the utilities at least annually (since the completion of The Christchurch Group should, if requested, provide the project various utility representatives have been support to other groups, in the same way as it has working together on the Ferrymead problem). How- worked with Wellington and Auckland. ever, if a major project is not adopted then at least annual task group meetings to review progress in the previous year would be worthwhile. 13.17 Summary It is intended that with the publication of this book the Task Group Chairmen will call together their various Dr. David Hopkins in his final summary of the 1994 Task Groups to consider possible future work. Workshop drew attention to the need to remember that the whole engineering lifelines process was aimed at “Value for money through engineering”. This means making the best possible use of available information 13.15 Promulgation and technology so that no matter when the earthquake With the publishing of this book, it is timely to refer to occurs, we can look back at money well spent on one of the original objectives of the project being: “To reducing earthquake risk to lifelines. This will apply communicate the issues to people involved in the for all hazards. Maps • 283

Maps

Reference list for colour maps

Map 1: Seismic Hazard — Location of Emergency Services and Major Contractors Yards 284 Map 2: Waimakariri River Floodplain — Location of Emergency Services and Major 285 Contractors Yards Map 3: Local Rivers Floodplain — Location of Emergency Services and Major 286 Contractors Yards Map 4: Tsunami Hazard 287

Map 5: Slope Hazard Zones — Hill Areas 288 Map 6: Seismic Hazard — Transport System Network 289 Map 7: Bridge Vulnerability 290

Map 8: Slope Hazard — Transport System Locality Plan 291 Map 9: Seismic Hazard — Telecom Cables 292 Map 10: Local Flooding Hazard — Telecom Cables 293 Map 11: Tsunami Hazard — Telecom Cables 294 Map 12: Waimakariri Flooding Hazard — Telecom Cables 295 Map 13: Seismic Hazard — Fire Services 296 Map 14: Seismic Hazard — Police Radio Services 297 Map 15: Seismic Hazard — Electricity Network 298 Map 16: Electricity Network 299 Map 17: Seismic Hazard — Broadcasting Services 300 Map 18: Seismic Hazard — Stormwater System 301 Map 19: Seismic Hazard — Foulwater Sewer and Pressure Mains 302 Map 20: Seismic Hazard —Water Services (Major Pipelines in Metropolitan Area) 303 Map 21: Seismic Hazard — Petroleum Products 304

Map Limitations

For the siesmic hazard maps, zone boundaries have been determined from geological maps of the area. These in turn are derived largely from borelog information from about 15,000 sites in the Christchurch area. Of these, the majority are shallow (less than 3 m deep), and are often concentrated in particular areas (e.g. the central city). The boundaries between soil types is therefore frequently ill-defined, particularly with increasing depth. In places the boundaries can be moved two or three hundred metres to either side, and still fit the available data.

The accuracy of the hazard zone boundaries is even less than for the soil type boundaries as these relate to the complex 3-dimensional changes in soil types beneath the city. For any critical structure, a specific site study is needed to determine the actual degree of hazard. 284 • Risks and Realities

CONTRACTORS YARDS CONTRACTORS

PRIMARY ROUTE PRIMARY

MAJOR ARTERIAL MAJOR ARTERIAL ROAD/ MINOR ARTERIAL MINOR ARTERIAL ROAD/ RAILWAY LOCAL DISTRIBUTOR ROAD DISTRIBUTOR LOCAL FIRE STATION

POLICE

HOSPITAL

CIVIL DEFENCE SECTOR H.Q. DEFENCE SECTOR CIVIL

CITY COUNCIL SERVICE CENTRE SERVICE COUNCIL CITY

EMERGENCY KEY EMERGENCY

Extent of Detailed

Earthquake Hazard Study Area Earthquake Hazard Study

ESTUARY

O R

L OF LYTTLETON PORT

S NE

E UN

G D T

A A P RO

D BAY

A CORSAIR O

T AVE STYX D O O W LIN BAY

GOVERNORS

R WINTERS ROAD

I A M AVON

E JOHNS ROAD

U HEATHCOTE

A H

RIAL AVE D

L EMO

S S

U M R

WIGRAM AIRPORT

I A

W D A

R T S R U H D L A Y AIRPORT (HAREWOOD) AIRPORT CHRISTCHURCH INTERNATIONAL

3km

A O

A M 1

N 0 1 Predominantly sands 2-10m depth 20-30% of Zone 2A could liquefy Predominantly silts and sandy silts 2-5m depth 10-15% of Zone 2B could liquefy Predominantly sands 2-10m depth could liquefy 2--30% fo Zone 3A Predominantly silts and sandy 2-5m depth 10-15% of Zone 3B could liquefy ZONE 2A ZONE 3B ZONE 2B ZONE 3A

LIQUEFACTION ZONES LIQUEFACTION

± 300m in areas

± 20m in a few areas Bedrock at shallow depth Shaking intensity MM VII-Vii Sediments 50-800m depth intensity MM VIII-IX, high amplification of ground motion. Within Zone 3 there will be areas of pronounced amplification and intensity IX Sediments less than 50m deep Shaking intensity MM VII ZONE 1 ZONE 2 ZONE 3

Please note the limitations of this map, as outlined on page 283

SHAKING ZONES

SEISMIC KEY

IMPORTANT NOTE IMPORTANT

1. The Zone boundaries on this map are approximate 1. only. Accuracy can vary from only. with closed spaced borelogs to with few widely spaced borelogs.

2. For important structures a site specific study is needed to accurately determine the soil profile, the degree of likely shaking, and to assess the liquefaction hazard.

SOURCES

Detailed Earthquake Study Area:- Detailed Earthquake Study The Earthquake Hazard in Christchurch Elder et al (EQC) 1991

Area North of detailed Study Area:- Area North of detailed Study Water Well Data Sheet S76/7-8 Belfast - Style Brown NZGS 1975

Water Well Data Sheet S76/4-5 Water Kaiapoi-Kairaki Brown NZGS 1976

Map 1: Seismic Hazard - Location of Emergency Services and Major Contractors Yards Maps • 285

CITY COUNCIL SERVICE CENTRE SERVICE COUNCIL CITY

CONTRACTORS YARDS CONTRACTORS

PRIMARY ROUTE PRIMARY MAJOR ARTERIAL ROAD/ MINOR ARTERIAL MINOR ARTERIAL ROAD LOCAL DISTRIBUTOR ROAD DISTRIBUTOR LOCAL RAILWAY FIRE STATION

POLICE

HOSPITAL

CIVIL DEFENCE SECTOR H.Q. DEFENCE SECTOR CIVIL

EMERGENCY KEY EMERGENCY

ESTUARY

PORT OF LYTTLETON PORT S

A P ROAD TUNNEL BAY

A CORSAIR

R E

V T

X A

T D S O O W IN L BAY GOVERNORS

WINTERS ROAD MAIN NORTH ROAD NORTH MAIN AVON JOHNS ROAD E T O C H T A E H

DUDLEY CREEK

A H

E V

A L IA R O M

M RUSSLEY ROAD RUSSLEY

WIGRAM AIRPORT

A W

RST RO U

YALDH AIRPORT (HAREWOOD) AIRPORT CHRISTCHURCH INTERNATIONAL

3km

R 2

A M 1

N 0 1

KEY

FLOODPLAIN

Map 2: Waimakariri River Floodplain — Location of Emergency Services and Major Contractors Yards 286 • Risks and Realities

CONTRACTORS

PRIMARY ROUTE PRIMARY MAJOR ARTERIAL ROAD/ MINOR ARTERIAL MINOR ARTERIAL ROAD LOCAL DISTRIBUTOR ROAD DISTRIBUTOR LOCAL RAILWAY FIRE STATION

POLICE

HOSPITAL

CIVIL COUNCIL SERVICE CENTRE SERVICE COUNCIL CIVIL

CITY COUNCIL SERVICE CENTRE SERVICE COUNCIL CITY

EMERGENCY KEY EMERGENCY

ESTUARY

PORT OF LYTTLETON PORT S

A P ROAD TUNNEL BAY

A CORSAIR

R E V T A

D STYX O O W IN L BAY GOVERNORS

WINTERS ROAD MAIN NORTH ROAD NORTH MAIN AVON

E JOHNS ROAD

U HEATHCOTE

A H

E V A

R D

O O

Y M

LE E

S S

U M R

WIGRAM AIRPORT

I A

W D A

O R

T S R U H D L A Y AIRPORT (HAREWOOD) AIRPORT CHRISTCHURCH INTERNATIONAL

3km

R 2

A M 1

N 0 1

KEY

FLOODPLAIN

Map 3: Local Rivers Floodplain — Location of Emergency Services and Major Contractors Yards Maps • 287 ESTUARY

PORT OF LYTTLETON PORT

T S AVON

R E T

C O Y C

E H L T A D E U H D

WIGRAM AIRPORT

A W AIRPORT (HAREWOOD) AIRPORT CHRISTCHURCH INTERNATIONAL 3km 2 1

N 0 1

INUNDATION FROM BEACH OVERTOPPING INUNDATION

INUNDATION FROM ESTUARY INUNDATION

COMBINED INUNDATION FROM BEACH & ESTUARY COMBINED INUNDATION

INUNDATION OF LYTTELTON PORT OF LYTTELTON INUNDATION

TSUNAMI HAZARD

Map 4: Tsunami Hazard 288 • Risks and Realities 3km 2

Slope Hazard Zone 1

Slope Hazard Zone 2

Slope Hazard Zone 3 1

Slope Hazard Key 0 1

ESTUARY

PORT OF LYTTLETON PORT

A ROAD O

R TUNNEL S

G PA BAY

CORSAIR

E V

A D O O W IN L

BAY

GOVERNERS

AVON

HEATHCOTE

A H

Low risk. Likely damages to services negligible.

area,normally where Zone 2 is closest to Zone 3. A similar A area,normally where Zone 2 is closest to Zone 3. proportion (10% to 20%) of roads and overhead services could be affected. For buried services damage is likely to could be affected. occur only from undermining or deep-seated failure reducing the extent of damage to 5% to 10% of service length.

similar proportion of the length of roads and overhead services could be affected. Once again, buried services are likely to could be affected. suffer less damage, say 10% to 20% of service length. suffer

WIGRAM AIRPORT

Slope Hazard Zones with respect to services Zone 1: Zone 2: 10% to 20% of Moderate risk. Likely damage may affect

Zone 3: A up to 40% of area. High risk. Likely damage may affect

Map 5: Slope Hazard Zones — Hill Areas Maps • 289

BRIDGE NUMBER

MAJOR ARTERIAL MAJOR ARTERIAL ROADS

MINOR ARTERIAL MINOR ARTERIAL ROAD/ LOCAL DISTRIBUTOR ROAD DISTRIBUTOR LOCAL RAILWAY

RAIL BUILDING RAIL

ROAD BRIDGES RAIL BRIDGES RAIL

TUNNEL PORTAL TUNNEL

Extent of Detailed

Earthquake Hazard Study Area Earthquake Hazard Study

TRANSPORT KEY TRANSPORT

R806

R801 ESTUARY

R102

A O

R101 R

PORT OF LYTTLETON PORT S NE

E UN

G D T

R201

A OA P R

R103

T713

D BAY

T710

T201 CORSAIR O

R413

T712

R104

R E

V T711 T

X A

F201

R202 Y

T S

R806

R203

R105 LINWOOD

R106

R205

R206

R207 BAY

R108 GOVERNORS

R303

R208

T701

R209

R501

R210

T501

R WINTERS ROAD

R502 N

I A M

R213

R214

T702 AVON

R217

R218 T401 E

R125

E JOHNS ROAD R E

C T

O Y C

R219

T703

E H R220

R705 L T A D E

T704 U H D R161

R131

R159

T705

R221

R804

T702

T706

R146 HALSWELL ROAD HALSWELL

R411

R149

R412

R150

E R142 V A T707

R D

O O

Y M E

L E

U M R T709

T708 WIGRAM AIRPORT

WAIMAKARIRI D A

O R

T S R U H D L A Y AIRPORT (HAREWOOD) AIRPORT CHRISTCHURCH INTERNATIONAL

3km

R 2

A M 1

LIQUEFACTION ZONES LIQUEFACTION

± 300m in areas

Please note the limitations of this map, as outlined on page 283 ZONE 1 ZONE 2 ZONE 3

SHAKING ZONES

SEISMIC KEY

IMPORTANT NOTE IMPORTANT

1. The Zone boundaries on this map are approximate 1. only. Accuracy can vary from only. with closed spaced borelogs to with few widely spaced borelogs.

2. For important structures a site specific study is needed to accurately determine the soil profile, the degree of likely shaking, and to assess the liquefaction hazard.

SOURCES

Detailed Earthquake Study Area:- Detailed Earthquake Study The Earthquake Hazard in Christchurch Elder et al (EQC) 1991

Area North of detailed Study Area:- Area North of detailed Study Water Well Data Sheet S76/7-8 Belfast - Style Brown NZGS 1975

Water Well Data Sheet S76/4-5 Water Kaiapoi-Kairaki Brown NZGS 1976

Map 6: Seismic Hazard — Transport System Network 290 • Risks and Realities

PRIMARY ROUTES PRIMARY MAJOR ARTERIAL MAJOR ARTERIAL ROADS MINOR ARTERIAL MINOR ARTERIAL ROAD/ LOCAL DISTRIBUTOR ROAD DISTRIBUTOR LOCAL RAILWAY USABLE (MINOR DAMAGE) USABILITY RESTRICTED USABILITY (MODERATE DAMAGE) (MODERATE SEVERE DAMAGE, LIKELY TO BE UNUSABLE TO

BRIDGE NUMBER

R805

BRIDGE VULNERABILITY KEY BRIDGE VULNERABILITY

R801 ESTUARY

R102

A O

R101 R

L OF LYTTLETON PORT

S NE

E UN G T

R201 D

A A P RO

R103 BAY

T713

T710 CORSAIR

T201

R413 R

T712

R104

R E

V T711 T

X A

F201 Y R202

T S D O O W

R806

IN R203

R105 L

R106

R205

R206

R207 BAY

R108 GOVERNORS

R303

R208

T701

R209

R501

R210

T501

R WINTERS ROAD

R502 N

I A M

R213

R214

T702 AVON

R217

R218

T401 E

R125

E JOHNS ROAD

R219

T703

R220

R705 LEY

T704 U HEATHCOTE D R161

R131

R159

T705

R221

R804

L E

T702

T706 W

R146 S

A H

R411

R149

R412

R150

E R142 V

T707

L A

R D

O O

Y M E

L E

U M R T709

T708

WIGRAM AIRPORT

I A

W D A

O R

T S R U H D L A Y AIRPORT (HAREWOOD) AIRPORT CHRISTCHURCH INTERNATIONAL

3km

R 2

A M 1

N 0 1

Map 7: Bridge Vulnerability Maps • 291

12 km 12

PRIMARY ROUTES PRIMARY MAJOR ARTERIAL MAJOR ARTERIAL ROADS MINOR ARTERIAL MINOR ARTERIAL ROADS LOCAL DISTRIBUTOR ROADS DISTRIBUTOR LOCAL RAILWAY SITE NUMBER EXTENT OF HAZARD AREA OF HAZARD EXTENT

KEY

S A

P 0

A V E

28-31

(Clifton Tce)

Lyttelton Reclamation Lyttelton

FERRY ROAD

PORT HILLS ROAD

SUMMIT ROAD SUMMIT

COLOMBO STREET 4 DYERS PASS ROAD

T E E TR S N TO G IN R R A B

43-47 Gebbies Pass

42 south of Allandale 42 south of

41 at Allandale Governers Bay 41 at

40 at Ohinetahi Governors Bay CASHMERE ROAD CASHMERE

Map 8: Slope Hazard — Transport System Locality Plan 292 • Risks and Realities

E RAD E PA

RIN D D

MA OA A D

Y R O A

AR O

STU R R

T E N I R

S A U

M P

S D

D S

E R T

G S

W O D

N E I E I

O R V

E M

B D A A W

AR IN E P A

NE A

I S AR H M M ROAD HO

D R

D NBR

A O RD A OK

BOWE O

R AV R E R

S E RD

E TH

Z D PA G

E A E

D DL O

A E I

A R

R B P R

D O B S

BEACH

A R R D

Y O A

D FROSTS D

O R O

O ROAD N I

O O E W RD

R T I V A A

U H D B W A

D G

I O Y A R

R R

B R TU E NNEL RO R AD S D F W E S

D I O E Z

R V E O N

A E W

R D R N

A A I

T R

B O L

A E R

D R N

O V I A

OI R

OD RO A

O A T

MA W D W S

UR D B O

O IL G

R H

A D T H R

P A RO

YHURST RD A GA BUCKLEYS O D R A E O

V Y R

R A

D A D W

R R A

A P

O E O

R F

N I

D W

O L WOODHAM ROAD O A

N D AD I R

SHLANDS RO L

MAR S

SHLANDS ROAD R

MAR O

D E

N I AD T

STANMOR RO NS

E R O A S T IL

SHIRLEY E M W RD HILLS ROAD AVE FITZGERALD E

T T S S

AM ROAD

AVE WALTH

E E

Seismic Hazard and Telecom Cables Telecom Seismic Hazard and

R LICHFIELD ST M

D ST CASHEL A WORCHESTER ST GLOUCESTER ST

R HEREFORD ST ST ASAPH ST ST D

BEALEY AVE BEALEY H

EDGEWARE ROAD EDGEWARE A

O WORDSWORTH O

F ARMARGH ST COLOMBO ST COLOMBO G STREET COLOMBO R AN

U S D R S

C O

D A A

KILMORE ST

Christchurch Engineering Lifelines Study 1994 R STREET ST DURHAM P A

B S O R PETERBOROUGH ST

R DYE

R S T ST MONTREAL

S E

N E N AVE

R R TO O

S S T T E ST ANTIGUA

E L

T S

E L T

O S

T D S

N R E R N E

O IN R

F O T R

W N E S

P ILT

A V

R A

C M 2 km

R MOORHOUSE E

P N

R O

T T

E T G H

N N

E L S I

R A R D

T S R T

A S

H A A

S O O

T R B R E R

N BROUGHAM STREET

IN C

AD A

A O

T R

M O O T AY

E H

R N

C D N

E R

A A O E

R A H

O O

T H

R IG E

L D L

B R

N D

E A N D S A

N S A

O O Y C T

ROAD L R

A W S K

D R

0 D R

N O A

E T OA P

R O S

D M

D R

S D

K A N R

AD E

A A O E R

O R R I O D G E

R A R

R A

W O

SAWYERS ARMS I T N

U E

O L

H L O T

1 R

HAREWOOD E

S L

B ROAD SPARKS A

IC H

R AD MEMORIAL AVE D

A S RO S O R S T

D RAHAM E

A G L

O R D A

R U D

C O

I O

E R L

L L

K L W

A S

D L A W

R A H I S

D L

D ROA O A EA A ONH H Predominantly sands 2-10m depth could liquefy 20% to 30% of Zone 2A Predominantly silts and sandy 2-5m depth. 10% to 15% of Zone 2B could liquefy Predominantly sands 2-10m depth could liquefy 20% to 30% 0f Zone 3A Predominantly silts and sandy 2-5m depth. 10% to 15% of Zone 3B could liquefy AV R W S R A D B A N O D R BLENHEIM RD U A D D MEMORIAL AVEA O

ZONE 2A: ZONE 2B: ZONE 3A: ZONE 3B: E R ROAD SOUTH MAIN H T N S O R V U A H ROAD D L A Y

D D

A O A O

R O

Y R D E

L R

S S A N

S A

U N O R O E I A T T R A N C W A H N A U H T J

C U U

O MAIN SO R

D M R

A A O C R ELL T

S LSW R D

N A

U HA O

Bedrock at shallow depth Shaking intensity MM VII-VIII Sediments less than 50m deep Shaking intensity MM VIII Sediments 50-800m depth Intensity MM VIII-IX; high amplification of ground motion. Within Zone 3 there will be areas of pronounced amplification and Intensity IX+ H R S

D G

L N I

A R

Y P S

map, as outlined on page 283 ZONE 1: ZONE 2: ZONE 3: BUCHANANS ROAD

Please note the limitations of this SHANDS ROAD SHANDS

Map 9: Seismic Hazard — Telecom Cables Maps • 293

E RAD E PA

RIN D D

MA OA A D

Y R O A

AR O

STU R R

T E N I R

S A U

M P

E R

N E I E

O R V

E M

B D A

ARA W E

NE P A

I S AR H M MAIN ROAD ROAD HO

D R

D NBR

A O RD A OK

BOWE O

R AV R E R

S E RD

E TH

Z D PA G

E A E

D DL O

A E I

A R

R B P R

D O B S

BEACH

A R R D

Y O A

D FROSTS D

O R O R O N I

ROAD D

O O E W R

R T I V A A

U D B W A

D GH

I O

A R R

O B TUNNEL RO R AD S D FERRY W E S

D I O E Z

R V E O N

A E W

R D R N

A A I

T R

B O L

A E R

D R N

O V I A

OI R

OD RO A

O A T

M W D W S

UR D B O

O IL G

R H

A T H R

P ROAD

YHURST RD A CKLEYS GA BU O D R OA E Y R

V R A D

A D W

R R A

A P

O E O

R F N

D D L O A

WOODHAM ROAD O

W D

IN R

MARSHLANDS ROAD L S

MARSHLANDS ROAD O

D E

AD T NS

STANMORE RO R O A S T IL

SHIRLEY E M W RD E HILLS ROAD AVE FITZGERALD

T T S S

AM ROAD

AVE WALTH

R LICHFIELD ST M

CASHEL ST CASHEL A WORCHESTER ST GLOUCESTER ST HEREFORD ST

ST ASAPH ST ST D

BEALEY AVE BEALEY H A

WORDSWORTH O

EDGEWARE ROAD EDGEWARE ARMARGH ST COLOMBO ST COLOMBO G STREET COLOMBO R ET U S E S

CRANFORDTR O A KILMORE ST R S ST DURHAM P

B S R

PETERBOROUGH ST DYE T ST MONTREAL E VE

RE TON A

S S T T

S LE ST ANTIGUA E L S

D O

N R

R N

O I

F O

WINTERS ROAD N E STREET L

PRESTONS ROAD I

A V

R A

C M

R MOORHOUSE

2 km E

P N

R O

T T

E T G H

N N

E L S I O

R A R

T S R

T S

H A A

S O

T R B ROAD E R

N BROUGHAM STREET

IN C

AD A

A O T R

M O T AY O O

E H

N R

C D N

A A O E

R A H

O O

Local Flooding Hazard and Telecom Cables Telecom Local Flooding Hazard and O

T H

G R

R I E

L D

O Christchurch Engineering Lifelines Study 1994 B

R K

H N AVE DEANS

N S

D A

A O O Y C O T

R L R

A W S K

D R S R

O 0 M AD A

E T R O P

F A R O S I M

S E D R D

D K A

E A N

A A O

Y O D R R

I O

O E R

W R RD GREERS A

H O R

A W

S IM T

W N

U E

E A O

H O

R E O T

V ROAD ELL

S A N R S

A K

H E

1 L A R

L A

IA C P

R B S

O IC HALSW

M R D

A E O M D

A R

S O

M R

A S

H T

A T

D R E

A G L

O R D A

R U D

C O

I O

R R

E L

L L

K L W

A S

D L A W

R A H I S

D L

D ROA O A EA A H H AVON R W E V S A R L A IA D B A R N O O D R BLENHEIM RD U M A E D D M A O

E R ROAD SOUTH MAIN H T N S O R V U A H ROAD D L A Y

D D

A O A O

R R O

Y D

E R

L S S A

S A

U N R O E A T R A N W A H A H T JUNCTION

C U U

O MAIN SO R

D M R A A

O C L R L E T W

S S

L R D

A A

U H O

H R S

D G

L N I

A R

Y P S

BUCHANANS ROAD SHANDS ROAD SHANDS

Map 10: Local Flooding Hazard — Telecom Cables 294 • Risks and Realities

E RAD E PA

RIN D D

MA OA A D

Y R O A

AR O

STU R R

T E N I R

S A U

M P

E R

N E I E

O R V

E M

B D A

ARA W E

NE P A

I S AR H M MAIN ROAD ROAD HO

D R

D NBR

A O RD A OK

BOWE O

R AV R E R

S E RD

E TH

Z D PA G

E A E

D DL O

A E I

A R

R B P R

D O B S

BEACH

A R R D

Y O A

D FROSTS D

O R O R O N I

ROAD D

O O E W R

R T I V A A

U D B W A

D GH

I O

A R R

O B TUNNEL RO R AD S D FERRY W E S

D I O E Z

R V E O N

A E W

R D R N

A A I

T R

B O L

A E R

D R N

O V I A

OI R

OD RO A

O A T

M W D W S

UR D B O

O IL G

R H

A T H R

P ROAD

YHURST RD A CKLEYS GA BU O D R OA E Y R

V R A D

A D W

R R A

A P

O E O

R F N

D D L O A

WOODHAM ROAD O

W D

IN R

MARSHLANDS ROAD L S

MARSHLANDS ROAD O

D E

AD T NS

STANMORE RO R O A S T IL

SHIRLEY E M W RD E HILLS ROAD AVE FITZGERALD

T T S S

AM ROAD

AVE WALTH

R LICHFIELD ST M

CASHEL ST CASHEL A WORCHESTER ST GLOUCESTER ST HEREFORD ST

ST ASAPH ST ST D

BEALEY AVE BEALEY H A

WORDSWORTH O

EDGEWARE ROAD EDGEWARE ARMARGH ST COLOMBO ST COLOMBO G STREET COLOMBO R ET U S E S

CRANFORDTR O A KILMORE ST R S ST DURHAM P

B S R

PETERBOROUGH ST DYE T ST MONTREAL E VE

RE TON A

S S T T

S LE ST ANTIGUA E L S

D O

N R

R N

O I

F O

WINTERS ROAD N E STREET L

PRESTONS ROAD I

A V

R A

C M

R MOORHOUSE

2 km E

P N

R O

T T

E T G H

N N

E L S I O

R A R

T S R

T S

H A A

S O

T R B ROAD E R

Tsunami Hazard and Telecom Cables Telecom Hazard and Tsunami D

N BROUGHAM STREET

IN C

AD A

A O T R

M O T AY O O

E H

N R

C D N

A A O E

R A H

O O

T H

G R

R I E

L D

O Christchurch Engineering Lifelines Study 1994 B

R K

H N AVE DEANS

N S

D A

A O O Y C O T

R L R

A W S K

D R S R

O 0 M AD A

E T R O P

F A R O S I M

S E D R D

D K A

E A N

A A O

Y O D R R

I O

O E R

W R RD GREERS A

H O R

A W

S IM T

W N

U E

E A O

H O

R E O T R

S A N R S

A K

H E

1 L A R

L A

IA C P

R B S

O IC ROAD HALSWELL

M R D

A E O M D

A R

S O

M R

A S

H T

A T

D R E

A G L

O R D A

R U D

C O

I O

R R

E L

L L

K L W

A S

D L A W

R A H I S

D L

D ROA O A EA A H H AVON R W E V S A R L A IA D B A R N O O D R BLENHEIM RD U M A E D D M A O

E R ROAD SOUTH MAIN H T N S O R V U A H ROAD D L A Y

D D

A O A O

R R O

Y D

E R

L S S A

S A

U N R O E A T R A N W A H A H T JUNCTION

C U U

O MAIN SO R

D M R A A

O C L R L E T W

S S

L R D

A A

U H O

H R S

D G

L N I

A R

Y P S

BUCHANANS ROAD SHANDS ROAD SHANDS

Map 11: Tsunami Hazard — Telecom Cables Maps • 295

E RAD E PA

RIN D D

MA OA A D

Y R O A

AR O

STU R R

T E N I R

S A U

M P

E R

N E I E

O R V

E M

B D A

ARA W E

NE P A

I S AR H M MAIN ROAD

ROAD HO

D R

D NBR

A O RD A OK

BOWE O

R AV R E R

S E RD

E TH

Z D PA G

E A E

D DL O

A E I

A R

R B P R

D O B S

BEACH

A R R D

Y O A

D FROSTS D

O R O

O ROAD N I

O O E W RD

R T I V A A

U H D B W A

IG O Y A R

R R

B R TU E NNEL RO R AD S D D F W E S

R I O E Z V E O

U A E W

A D

R D R N

A A I

H T

B O L

D R N E O

V A I I

OD RO A

O A T

M W D W S

UR D B O

O IL G

R H

A T

H R

YS O

P ROAD

C D

YHURST RD U A GA B O D R OA E Y R

V R A

A D W

R R PA

E O

S F N

D D L O A

WOODHAM ROAD O

W D AD R

SHLANDS RO IN

MAR L S

SHLANDS ROAD R

MAR O

D E

I T S

SHIRLEY AD N

STANMORE RO R O A S T IL

E M W RD E HILLS ROAD AVE FITZGERALD R

T S ST

AM ROAD

AVE WALTH

ROAD LICHFIELD ST M

D ST CASHEL A WORCHESTER ST GLOUCESTER ST

R HEREFORD ST ST ASAPH ST ST

BEALEY AVE BEALEY H

O WORDSWORTH

F ROAD EDGEWARE ARMARGH ST

COLOMBO ST COLOMBO AN ET G STREET COLOMBO R E U AD

C TR O O KILMORE ST R S ST DURHAM R

B S

PETERBOROUGH ST A MONTREAL ST MONTREAL P ET S

E ON AVE R

TR ST DYE E ST ANTIGUA S LL T D O E

R R E O INNES R

F T WINTERS ROAD N E S

PRESTONS ROAD

A V

R A

C ST MILTON R MOORHOUSE

2 km

P R

T A S H AL

SS H

STREET O

T R BARRINGTON ROAD R

HEATON

N BROUGHAM STREET

IN C

AD A

A O T R

M O T AY O

E H N R

C N E

A R O E

H R

A O

T H ROAD

R E

R S

BLIGHS R

N D

E A N D S

N S

D A

A O O Y C O T

R L R

A W S K

D R S R

N O AD A

E T O P

R O S

I D M

E D

S K A N

A E A

O R

YERS ARM D E R

R I O E D

O G

ROAD A R

O R

W O

SAW IM T N

D R

U E

O L

H L

E O T

V E

S N

R S

A W

HAREW E

L A R

S 1

L A

IA C P

B S

R A

O IC H M R E M D A O R S T T

D E

A ROAD GRAHAMS L

O R D

R U A C O

E R L

K L

A D E W

R A ROAD HALSWELL

I S

L D D

D ROA O A A A E A H H

AVON R W E O

V R

R A H

L A T IA D U B

R A

O O N

O D S

BLENHEIM RD U

M A

N D

E D I

O M A A

E R M H T N S O R V U A H ROAD D L A Y

D D

A O A O

R O

Y R D E

L R

S S A N

S A

U N R O IO A T R ATE N C W A H N A U H T J

C U U B

IN SO

O MA R

D M R A A

O C L L R E T W

S S

L R D D

A A A

U H O

O R

R S

D G

L S N I

A N R

Y A P

N S

A D

H A

C O R

U S

B D

H S

Map 12: Waimakariri Flooding Hazard — Telecom Cables 296 • Risks and Realities

E RAD E PA

RIN D D

MA OA A D

Y R O A

AR O

STU R R

T E N I R

S A U

M P

S D

D S

R E T

S G

O C

E N I

K R

O R V

E M W

AD B

R IN

A PA E

E A H

ARIN S M M ROAD HO

D R

D NBR

A O RD A OK

BOWE O

R AV R E R

S E RD

E TH

Z D PA G

E A E

D DL O

A E I

A R

R B P R

D O B S

BEACH

A R R D

Y O A

D FROSTS D

O R O

O ROAD N I

O O E W RD

R T I V A A

U H D B W A

D G

I O Y A R

R R

B R TU E NNEL RO R AD S D F W E S

D I O E Z

R V E O N

A E W

R D R N

A A I

T R

B O L

A E R

D R N

O V I A I

O R

OD RO A

O A T

MA W D W S

UR D B O

O IL G

R H

A D T H R

P A RO

YHURST RD A GA BUCKLEYS O D R OA E Y R

V R A D

A D W

R R A

A P

O E O

R F N

D D L O A

WOODHAM ROAD O

W D AD R

SHLANDS RO IN

MAR L S

SHLANDS ROAD R

MAR O

D E

AD T NS

STANMORE RO R O A S T IL

SHIRLEY E M W RD E HILLS ROAD AVE FITZGERALD

T T S S

AM ROAD AVE WALTH

Seismic Hazard and Fire Services

R LICHFIELD ST M

D ST CASHEL A WORCHESTER ST GLOUCESTER ST

R HEREFORD ST ST ASAPH ST ST D

BEALEY AVE BEALEY H A

O WORDSWORTH O

F ROAD EDGEWARE ARMARGH ST COLOMBO ST COLOMBO G STREET COLOMBO R AN

U S D R S

C O

D A A

KILMORE ST

Christchurch Engineering Lifelines Study 1994 R STREET ST DURHAM P A

B S O R PETERBOROUGH ST

R DYE

R S T ST MONTREAL

S E

N E N AVE

R R TO O

S S T T E ST ANTIGUA

E L

T S

E L T

O S

T D S

N R E R N E

O IN R

F O T

W N E S

P IL

A V

R A

C M 2 km

R MOORHOUSE E

P N

R O

T T

E T G H

N N

E L S I

R A R D

T S R T

A S

H A A

S O O

T R B R E R

N BROUGHAM STREET

IN C

AD A

A O

T R

M O O T AY

E H

R N

C D N

E R

A A O E

R A H

O O

T H

R IG E

L D L

B R

N D

E A N D S A

N S A

O O Y C T

ROAD L R

A W S K

D R

0 D R

N O A

E T OA P

R O S

D M

D R

S D

K A N R

AD E

A A O E R

R R

I O D G E

RO A R

R A

W O

SAWYERS ARMS I T N

U E

O L

H L O T

1 R

HAREWOOD E

S L

B ROAD SPARKS A

IC H

R AD MEMORIAL AVE D

A S RO S O D R A S

O T R AHAM T

I E

E GR L

K R D A

A U D A

C O

R R

L R L

L L

A E L

W D L A

A H S

D L

D ROA O EA A ONH H Predominantly sands 2-10m depth could liquefy 20% to 30% of Zone 2A Predominantly silts and sandy 2-5m depth. 10% to 15% of Zone 2B could liquefy Predominantly sands 2-10m depth could liquefy 20% to 30% 0f Zone 3A Predominantly silts and sandy 2-5m depth. 10% to 15% of Zone 3B could liquefy AV R S R A D B A RD O N R D U

A BLENHEIM D D MEMORIAL AVEA O

ZONE 2A: ZONE 2B: ZONE 3A: ZONE 3B: E R ROAD SOUTH MAIN H T N S O R V U A H ROAD D L A Y

D D

A O A O

Y R D E

L R

S S A N

S A

U N O R O E I A T T R A N C W A H N A U H T J

C U U

O MAIN SO R

D M R

A A O C LL R E T

S LSW R D

A A

U H O

D G

L N I

A R

Y P S ZONE 1: ZONE 2: ZONE 3: BUCHANANS ROAD map, as outlined on page 283

Please note the limitations of this SHANDS ROAD SHANDS

Map 13: Seismic Hazard — Fire Services Maps • 297

E RAD E PA

RIN D D

MA OA A D

Y R O A

AR O

STU R R

T E N I R

S A U

M P

D D S

R A K

O C

D W

N I E

E R K I O R

E M V

D B

A W IN

AR A P A

INE E

R H A S M M ROAD HO

D R

D NBR

A O RD A OK

BOWE O

R AV R E R

S E RD

E TH

Z D PA G

E A E

D DL O

A E I

A R

R B P R

D O B S

BEACH

A R R D

Y O A

D FROSTS D

O R O

O ROAD N I

O O E W RD

R T I V A A

U H D B W A

IG O Y A R

R R

B R TU E NNEL RO R AD S D F W E S

D I O E Z

R V E O N

A E W

R D R N

A A I

T R

B O L

A E R

D R N

O V I A I

O R

OD RO A

O A T

M W D W S

UR D B O

O IL G

R H

A D T H R

OA P R

KLEYS C D

YHURST RD A GA BU O D R OA E Y R

V R A D

A D W

R R A

A P

O E O

R F N

D D L O A

WOODHAM ROAD O

W D AD R

SHLANDS RO IN

MAR L S

SHLANDS ROAD R

MAR O

D E

AD T NS

STANMORE RO R O A S T IL

SHIRLEY E M W RD E HILLS ROAD AVE FITZGERALD

T T S S

AM ROAD

AVE WALTH

R LICHFIELD ST M

D ST CASHEL A WORCHESTER ST GLOUCESTER ST

R HEREFORD ST ST ASAPH ST ST D

BEALEY AVE BEALEY H A

O WORDSWORTH

Seismic Hazard and Police Radio Services O

F ROAD EDGEWARE ARMARGH ST COLOMBO ST COLOMBO G STREET COLOMBO R AN ET U S D R E S

C R O

A T A KILMORE ST

Christchurch Engineering Lifelines Study 1994 R S ST DURHAM P A

B S O R PETERBOROUGH ST

R DYE

R S T ST MONTREAL

S E

N E N AVE

R R TO O

T S S LE ST ANTIGUA

T S E OL

TE D S

R N R

E N

IN O I

F R W N E STREET

A V

R A

C ST MILTON 2 km

R MOORHOUSE E

P N

R O

T T

E T G H

E L S IN

R A R

T S R T

A S A A S

O RO B R E

K D BROUGHAM STREET

E C

OAD A

T R

1 MAIN NORTH MAIN O T AY O

E H

N R

C D N

A A O E

R A H

O O

T H

G R

R I E

L D

B R

N D

E A N D S A

N S A

O O Y C T

ROAD L R

A W S K

D R

0 D R

N O A

E T OA P

R O S

D M

D R

S D

D K A N R

A E

A A O

YERS ARMS D E R

O R R I O D G E

R A R

OO R

W O

SAW IM T N

D R

U E

O L

H L

S N

S 1 R

HAREW E

A R

L A

C P

B S A

IC H

R D

O MEMORIAL AVE D

R A

S O

M R

A S

H T

A T

D R E

A G L

O R D A

R U D

C O

I O

R R

E L

L L

K L W

A S

D L A W

R A H

I S

D D

D ROA O A A A E A ONH H

Predominantly sands 2-10m depth could liquefy 20% to 30% of Zone 2A Predominantly silts and sandy 2-5m depth. 10% to 15% of Zone 2B could liquefy Predominantly sands 2-10m depth could liquefy 20% to 30% 0f Zone 3A Predominantly silts and sandy 2-5m depth. 10% to 15% of Zone 3B could liquefy AV R

W O S R

H A T D U B

O O N

D S

BLENHEIM RD U

N D D I

O MEMORIAL AVEA A

ZONE 2A: ZONE 2B: ZONE 3A: ZONE 3B: E R M H T N S O R V U A H ROAD D L A Y

D D

A O A O

R O

Y R D E

L R

S S A N

S A

U N R O E IO A T T R A N C W A H N A U H T J

C U U B MAIN SO

D A

O ROAD CARMENS L L R E T W

S S

L R D D

A A

N A

U H O

O R

R S

D G

L S N I

A N R

Y A P N S

H AD

C O

ZONE 1: ZONE 2: ZONE 3: U R S

B D map, as outlined on page 283 N A

Please note the limitations of this SH

Map 14: Seismic Hazard — Police Radio Services 298 • Risks and Realities

SP 11 kV Feeder O/H 11 SP SP 11 kV Feeder U/G 11 SP SP 33 kV Feeder O/H SP SP 33 kV Feeder U/G SP SP 66 kV Feeder O/H SP SP 66 kV Feeder U/G SP ECNZ 66 kV Feeder O/H ECNZ 220 kV Feeder O/H

SP 11 kV District Substation 11 SP

SP 33 kV District Substation SP

SP 66 kV District Substation SP Transpower 66 kV Substation Transpower

Transpower 220 kV Substation Transpower

KEY

Extent of Detailed

Earthquake Hazard Study Area Earthquake Hazard Study

BARNET PARK BARNET

ESTUARY

PAGES

SIMEON PORT OF LYTTELTON PORT

BRIGHTON

BROMLEY BAY

CORSAIR STYX

HEATHCOTE

PORTMAN

PAGES-KEARNEYS

DALLINGTON BAY

LINWOOD GOVERNORS

MILTON

ARMAGH

MONTREAL

McFADDENS

GRIMSEYS-WINTERS

KNOX

AVON

OXFORD-TUAM

SPREYDON

HOON HAY

U HEATHCOTE D

HARRIS

FENDALTON

ADDINGTON

PAPANUI

MIDDLETON

HALSWELL

BISHOPDALE

HAWTHORNDEN E L M O R

I IR

SOCKBURN

W L

HAREWOOD Z I

W T

HORNBY

ISLINGTON

MOFFET ST MOFFET

ROAD

SHANDS

WAIPARA N

B N T

A O

T O S T

A S

A P G G

R N I

W N

I A

K O I R

I P K R V K S

O E I L H T

LIQUEFACTION ZONES LIQUEFACTION

Predominantly sands 2-10m depth 20-30% of Zone 2A could liquefy

Predominantly silts and sandy silts 2-5m depth 10-15% of Zone 2B could liquefy Predominantly sands 2-10m depth 2--30% fo Zone 3A could liquefy 2--30% fo Zone 3A

Predominantly silts and sandy 2-5m depth 10-15% of Zone 3B could liquefy

ZONE 2A

ZONE 3B

ZONE 2B

ZONE 3A

± 300m in areas

± 20m in a few areas

Bedrock at shallow depth Shaking intensity MM VII-Vii

Sediments 50-800m depth intensity MM VIII-IX, high amplification of ground motion. Within Zone 3 there will be areas of pronounced amplification and intensity IX

Sediments less than 50m deep Shaking intensity MM VII

SHAKING ZONES

ZONE 1

ZONE 2

ZONE 3

SEISMIC KEY

IMPORTANT NOTE IMPORTANT

1. The Zone boundaries on this map are approximate 1. only. Accuracy can vary from only. with closed spaced borelogs to with few widely spaced borelogs.

2. For important structures a site specific study is needed to accurately determine the soil profile, the degree of likely shaking, and to assess the liquefaction hazard.

SOURCES

Detailed Earthquake Study Area:- Detailed Earthquake Study The Earthquake Hazard in Christchurch Elder et al (EQC) 1991

Area North of detailed Study Area:- Area North of detailed Study Water Well Data Sheet S76/7-8 Belfast - Style Brown NZGS 1975

Water Well Data Sheet S76/4-5 Water Kaiapoi-Kairaki Brown NZGS 1976

Please note the limitations of this map, as outlined on page 283

Map 15: Seismic Hazard — Electricity Network Maps • 299

SP 11 kV Feeder O/H 11 SP SP 11 kV Feeder U/G 11 SP SP 33 kV Feeder O/H SP SP 33 kV Feeder U/G SP SP 66 kV Feeder O/H SP SP 66 kV Feeder U/G SP ECNZ 66 kV Feeder O/H ECNZ 220 kV Feeder O/H

SP 11 kV District Substation 11 SP

SP 33 kV District Substation SP

SP 66 kV District Substation SP Transpower 66 kV Substation Transpower

Transpower 220 kV Substation Transpower

BARNET PARK BARNET

KEY

PAGES

SIMEON

BRIGHTON

BROMLEY

HEATHCOTE

PORTMAN

PAGES-KEARNEYS

DALLINGTON

LINWOOD

MILTON

ARMAGH

MONTREAL

McFADDENS

GRIMSEYS-WINTERS

KNOX

OXFORD-TUAM

SPREYDON

HOON HAY

HARRIS

FENDALTON

ADDINGTON

PAPANUI

MIDDLETON

HALSWELL

BISHOPDALE

Y E HAWTHORNDEN L M O R B

SOCKBURN

HAREWOOD Z I

W T

HORNBY

ISLINGTON

MOFFET ST MOFFET

ROAD

SHANDS

WAIPARA N O

B N T

A O

T O S T S

G A G

A P N

R I W N

A I R

K O I

I P K

K R S I

O E L H T

Map 16: Electricity Network 300 • Risks and Realities

E RAD E PA

RIN D D

MA OA A D

Y R O A

AR O

STU R R

T E N I R

S A U

M P

E R

N E I E

O R V

E M

B D A

ARA W E P A

NE AIN ROAD

I S AR H M M ROAD HO

D R

D NBR

A O RD A OK

BOWE O

R AV R E R

S E RD

E TH

Z D PA G

E A E

D DL O

A E I

A R

R B P R

D O B S

BEACH

A R R D

Y O A

D FROSTS D

O R O R O N I

ROAD D

O O E W R

R T I V A A

U H D B W A

D G

I O Y A R

R R

B R TU E NNEL RO R AD S D F W E S

D I O E Z

R V E O N

A E W

R D R N

A A I

T R

B O L

A E R

D R N

O V I A

OI R

OD RO A

O A T

M W D W S

UR D B O

O IL G

R H

A T

H R S

Y O

P E ROAD

L K D

YHURST RD A GA BUC O D R OA E Y R

V R A D

A D W

R R A

A P

O E O

R F N

D D L O A

WOODHAM ROAD O

W D AD R

SHLANDS RO IN

MAR L S

SHLANDS ROAD R

MAR O

D E

AD T NS

STANMORE RO R O A S T IL

SHIRLEY E M W RD E HILLS ROAD AVE FITZGERALD

T T S S

AM ROAD

AVE WALTH

T E

R LICHFIELD ST M

D ST CASHEL A WORCHESTER ST GLOUCESTER ST

R HEREFORD ST ST ASAPH ST ST D

BEALEY AVE BEALEY H A

O WORDSWORTH O

F ROAD EDGEWARE

Seismic Hazard and Broadcasting Services COLOMBO ST COLOMBO G STREET COLOMBO R AN

U S D R S

C O ARMARGH

A A

KILMORE ST

Christchurch Engineering Lifelines Study 1994 R STREET ST DURHAM P

B S R PETERBOROUGH ST

R DYE S ET ST MONTREAL N VE

RE TON A O

S S T T

LE ST ANTIGUA

T S E L

S T

D O S

N R E

R N E

O IN R

F O T

WINTERS ROAD N E S L

P I

A V

R A

C M 2 km

R MOORHOUSE E

P N

R O

T T

E T G H

E L S IN

R A R D

T S R

A S

H A A

S O O

T R B R E R

O H

N D

BROUGHAM STREET

E N

I AD A A O

T R 1 M AY O

O H

C D N

A A O E

O O

O R

T H R

R E

STREET

LIG D

B R

KILMARNOCK

N D

E A N D S A

N S A

O O Y C

A R

ROAD L

A W S K

D R 0 R

N O AD A

E T O P

F R O S I M

E D

D K A

A A N

O R

O IR R GREERS RD GREERS E

R A H

W M

SAWYERS ARMS I T

U E

H OA E O

S N

A 1 R KS

HAREWOOD L E

A L

I PAR

R B S

O ROAD HALSWELL

M ROAD RICCARTON D

A E

O M D

R A

S O

M R

A S

H T

A T

D R E

A G L

O R D

R U A C O

E R L

K L

A D E W

R A ROAD HALSWELL

I S

OAD D O A A A NHEAD R H

V S

A R R H

L A

A T I D U B

R A O N O O

BLENHEIM RD U

M R S

E D IN M A A

ZONE 2A: ZONE 2B: ZONE 3A: ZONE 3B: E

H M N O D V A A O R

YALDHURST ROAD

D D

A A O

R O

Y D

L R

S A N

S A

U O R O E I T T R A C W H N A U T J

BUCHANANS ROAD MAIN SO

CARMENS ROAD CARMENS L L E W

L D D

A A

N A

H O

O R Bedrock at shallow depth Shaking intensity MM VII-VIII Sediments less than 50m deep Shaking intensity MM VIII Sediments 50-800m depth Intensity MM VIII-IX; high amplification of ground motion. Within Zone 3 there will be areas of pronounced amplification and Intensity IX+

R S G

S N I

N R

YALDHURST ROAD A P N S

A H AD C

ZONE 1: ZONE 2: ZONE 3: U

B map, as outlined on page 283

Please note the limitations of this SHANDS RO SHANDS

Map 17: Seismic Hazard — Broadcasting Services Maps • 301

∅ 450

area

Extent of detailed

earthquake hazard study

STORMWATER PUMPING STATIONS STORMWATER TRIBUTARY WATERWAY, OPEN DRAINS WATERWAY, TRIBUTARY

PIPED WATERWAY, GREATER THAN GREATER PIPED WATERWAY, PIPED WATERWAY, BRICK BARRELS

STORMWATER KEY STORMWATER

ESTUARY

D LYTTELTON

A O

R LL

S NE

E UN

G D T

A A P RO BAY

A CORSAIR

R E V

T A STYX D O O W IN L BAY

GOVERNORS

R WINTERS ROAD

I A M AVON

EE JOHNS ROAD R E T

C O Y C H LE T A D E U H

A H

E V A L

IA AD R O M

M A

SSLEY RO SSLEY E

R RU M

R A

K T A R

I P A IR

W D A

R T S R U H D L A Y

3km

IN A

1 M

LIQUEFACTION ZONES LIQUEFACTION

± 300m in areas

± 20m in a few areas Bedrock a5 shallow depth Shaking intensity MM VII-Vii Sediments 50-800m depth intensity MM VIII-IX, high amplification of ground motion. Within Zone 3 there will be areas of pronounced amplification and intensity IX Sediments less than 50m deep Shaking intensity MM VII ZONE 1 ZONE 2 ZONE 3

SHAKING ZONES

SEISMIC KEY

IMPORTANT NOTE IMPORTANT

1. The Zone boundaries on this map are approximate 1. only. Accuracy can vary from only. with closed spaced borelogs to with few widely spaced borelogs.

2. For important structures a site specific study is needed to accurately determine the soil profile, the degree of likely shaking, and to assess the liquefaction hazard.

SOURCES

Detailed Earthquake Study Area:- Detailed Earthquake Study The Earthquake Hazard in Christchurch Elder et al (EQC) 1991

Area North of detailed Study Area:- Area North of detailed Study Water Wkk Data Sheet S76/7-8 Water Kaiapoi-Kairaki Brown NZGS 1976

Please note the limitations of this map, as outlined on page 283

Map 18: Seismic Hazard — Stormwater System 302 • Risks and Realities

LYTTELTON WATERMAINS LYTTELTON

BRICK BARREL SEWER BRICK BARREL

SEWER OVERFLOWS

SEWAGE TREATMENT AREAS TREATMENT SEWAGE

PUMPING STATION NUMBER PUMPING STATION GRAVITY MAINS GRAVITY

SEWAGE PUMPING STATIONS SEWAGE

Extent of Detailed

FOULWATER KEY FOULWATER

Earthquake Hazard Study Area Earthquake Hazard Study

37 ESTUARY

57 PORT OF LYTTELTON PORT

D 59

A O 27

R ROAD TUNNEL

78 E

G 48

63 A

77 P BAY

A CORSAIR

R E

IS V

54 A

T 15 X

T LINWOOD S

120

19 BAY

7 GOVERNORS

41 PRIVATE

23 WINTERS ROAD

53 MAIN NORTH ROAD NORTH MAIN

58 AVON

E JOHNS ROAD R E

C T

O Y C

E H L T A D E U 68 H 42 D

50 HALSWELL ROAD HALSWELL

E 61 V A L IA R T

O R

M A

E O

R RUSSLEY ROAD RUSSLEY M

I R

W A

WAIMAKARIRI D A

O R

T S R U H D L A Y

79 D

A M PRIVATE

80 Predominantly sands 2-10m depth 20-30% of Zone 2A could liquefy Predominantly silts and sandy silts 2-5m depth 10-15% of Zone 2B could liquefy Predominantly sands 2-10m depth could liquefy 2--30% fo Zone 3A Predominantly silts and sandy 2-5m depth 10-15% of Zone 3B could liquefy ZONE 2A ZONE 3B ZONE 2B ZONE 3A

LIQUEFACTION ZONES LIQUEFACTION

± 300m in areas

± 20m in a few areas Sediments 50-800m depth intensity MM VIII-IX, high amplification of ground motion. Within Zone 3 there will be areas of pronounced amplification and intensity IX Bedrock at shallow depth Shaking intensity MM VII-Vii Sediments less than 50m deep Shaking intensity MM VII ZONE 1 ZONE 2 ZONE 3

SHAKING ZONES

SEISMIC KEY

IMPORTANT NOTE IMPORTANT

1. The Zone boundaries on this map are approximate 1. only. Accuracy can vary from only. with closed spaced borelogs to with few widely spaced borelogs.

2. For important structures a site specific study is needed to accurately determine the soil profile, the degree of likely shaking, and to assess the liquefaction hazard.

SOURCES

Detailed Earthquake Study Area:- Detailed Earthquake Study The Earthquake Hazard in Christchurch Elder et al (EQC) 1991

Area North of detailed Study Area:- Area North of detailed Study Water Well Data Sheet S76/7-8 Belfast - Style Brown NZGS 1975

Water Well Data Sheet S76/4-5 Water Kaiapoi-Kairaki Brown NZGS 1976

Please note the limitations of this map, as outlined on page 283

Map 19: Seismic Hazard — Foulwater Sewer and Pressure Mains Maps • 303

Ø200

WATERMAIN GREATER THAN GREATER WATERMAIN

PUMPING STATION RESERVOIR

LYTTELTON WATERMAINS LYTTELTON

WATER SERVICES KEY SERVICES WATER

Extent of Detailed

Earthquake Hazard Study Area Earthquake Hazard Study

ESTUARY

D LYTTELTON

S NE

E UN

G D T

A A P RO BAY

A CORSAIR

R E

V T

X A

T S

LINWOOD BAY

GOVERNORS

R WINTERS ROAD

I A M AVON

E JOHNS ROAD R E

C T

O Y C

E H L T A D E U H

D HALSWELL ROAD HALSWELL

E V A L

A R

E A

L M

S E

U R R M

AIRPORT

WAIMAKARIRI D A

O R

T S R U H D L A Y

3km

R 2

A M 1

N 0 Predominantly sands 2-10m depth 20-30% of Zone 2A could liquefy Predominantly silts and sandy silts 2-5m depth 10-15% of Zone 2B could liquefy Predominantly sands 2-10m depth could liquefy 2--30% fo Zone 3A Predominantly silts and sandy 2-5m depth 10-15% of Zone 3B could liquefy 1 ZONE 2A ZONE 3B ZONE 2B ZONE 3A

LIQUEFACTION ZONES LIQUEFACTION

± 300m in areas

SHAKING ZONES

SEISMIC KEY

IMPORTANT NOTE IMPORTANT

1. The Zone boundaries on this map are approximate 1. only. Accuracy can vary from only. with closed spaced borelogs to with few widely spaced borelogs.

2. For important structures a site specific study is needed to accurately determine the soil profile, the degree of likely shaking, and to assess the liquefaction hazard.

SOURCES

Detailed Earthquake Study Area:- Detailed Earthquake Study The Earthquake Hazard in Christchurch Elder et al (EQC) 1991

Area North of detailed Study Area:- Area North of detailed Study Water Well Data Sheet S76/7-8 Belfast - Style Brown NZGS 1975

Water Well Data Sheet S76/4-5 Water Kaiapoi-Kairaki Brown NZGS 1976

Please note the limitations of this map, as outlined on page 283

Map 20: Seismic Hazard — Water Services (Major Pipelines in Metropolitan Area) 304 • Risks and Realities

MOBIL OIL PIPELINE OIL MOBIL

LIQUIGAS PIPELINE

PETROLEUM KEY

ESTUARY

O R

PORT OF LYTTLETON PORT S

A P ROAD TUNNEL

D BAY

A CORSAIR O

R E

V T

X A

T S

LINWOOD BAY GOVERNORS

WINTERS ROAD MAIN NORTH ROAD NORTH MAIN AVON

E JOHNS ROAD R E

C T

O Y C

E H L T A D E U H

D HALSWELL ROAD HALSWELL

E V A

L IA R O M

M RUSSLEY ROAD RUSSLEY WIGRAM AIRPORT

WAIMAKARIRI D A

O R

T S R U H D L A Y AIRPORT (HAREWOOD) AIRPORT CHRISTCHURCH INTERNATIONAL

3km

R 2

A M 1

LIQUEFACTION ZONES LIQUEFACTION

± 300m in areas

Please note the limitations of this map, as outlined on page 283

SHAKING ZONES

SEISMIC KEY

IMPORTANT NOTE IMPORTANT

1. The Zone boundaries on this map are approximate 1. only. Accuracy can vary from only. with closed spaced borelogs to with few widely spaced borelogs.

2. For important structures a site specific study is needed to accurately determine the soil profile, the degree of likely shaking, and to assess the liquefaction hazard.

SOURCES

Detailed Earthquake Study Area:- Detailed Earthquake Study The Earthquake Hazard in Christchurch Elder et al (EQC) 1991

Area North of detailed Study Area:- Area North of detailed Study Water Well Data Sheet S76/7-8 Belfast - Style Brown NZGS 1975

Water Well Data Sheet S76/4-5 Water Kaiapoi-Kairaki Brown NZGS 1976

Map 21: Seismic Hazard — Petroleum Products Bibliography • 305

Bibliography

Advisory Notes on Lifeline Earthquake Engineering.. quakes. Wellington Case Study Project Report, Au- A Report prepared by the Technical Committees of the gust 1991. ASCE Technical Council on Lifeline Earthquake En- gineering. Centre for Advanced Engineering. Risk Assessment of Industrial and Natural Hazards. Proceedings of a Benuska, Lee. “Earthquake Spectra”. The Profes- Workshop 21-22 August 1991. sional Journal of the Earthquake Engineering Re- search Institute, Supplement to Volume 6, Loma Prieta Christchurch Swamp to City. A Short History of the Earthquake Reconnaissance Report. Published in Christchurch Drainage Board 1875-1989. California in 1990. Cotton, Elden A. Board of Water and Power Commis- Berrill J.B., Mulqueen P.C., Pautre J.-L.,. Ooi E.T.C, sioners’ Report on Increased Seismic Risk to Power March 1994.. Liquefaction at Kaiapoi in the 1901 System Facilities. Published in Los Angeles, Califor- Cheviot, N.Z., Earthquake Department of Civil Engi- nia in July 1 1994. neering, University of Canterbury, Christchurch, New Dames & Moore. The Northridge Earthquake January Zealand. Research Report. 17 1994. California, 1994.

Brown, L.J, .and Weeber. J.B. Geology of the Christ- Earthquake and Geological Hazard. Mitigation Strat- church Urban Area, Institute of Geological and Nu- egy. December 1995. Policy and Planning Depart- clear Sciences 1992. ment. Wellington Regional Council.

Buckle,I.G. (Editor). The Northridge California Earth- Earthquake Business Plan. A Planning Guide for quake of January 17 1994: Performance of Highway Commercial Organisations. Ministry of Civil De- Bridges, National Centre for Earthquake Engineering fence. Research, State University of New York at Buffalo. Published Buffalo, New York, March 24, 1994. Earthquake Business Plan. Planning Guide for Cor- porate Organisations. Ministry of Civil Defence. Bulletin of New Zealand National Society for Earth- quake Engineering, Volume 27, No.4, December 1994. Earthquake Business Plan. Saving Your Goods and Northridge Earthquake Reconnaissance Report. Pub- Chattels. Ministry of Civil Defence. lished in Waikanae, New Zealand in 1994. Earthquake Damage Evaluation Data for California. California Office of Emergency Services. Folder con- ATC Applied Technology Council 1985. taining miscellaneous pamphlets mainly relating to the Northridge Earthquake and methods of obtaining as- Earthquake Engineering Research Institute. Northridge sistance after the earthquake. Earthquake January 17, 1994. Published in California in March 1994. California Office of Emergency Services. Model Or- dinances for Post Disaster Recovery and Reconstruc- Earthquake Engineering Research Institute. Northridge tion. Published in California. Earthquake January 17, 1994. Preliminary Recon- naissance Report. Caltrans. Accelerat Caltrans’ Action Plan to get all our Freeways moving again. Spring 1994. Published Earthquake Engineering Research Institute. Slides of in Los Angeles, California. the January 17, 1994 Northridge Earthquake. Pro- duced in California, 1994. Caren, M.D. Pacific Bell California Joint New Zea- land/Los Angeles Workshop (paper only). Published East Bay Municipal Utility Department. Personal and in California in 1994. Family Emergency Programme Package. Published in California in 1991. Centre for Advanced Engineering. Integrated Risk Management Conference Abstracts, Wellington 12-13 East Bay Municipal Utility District. Emergency Op- March 1997. erations Plan. Published in Oakland, California in 1994. Centre for Advanced Engineering. Lifelines in Earth- 306 • Risks and Realities

East Bay Municipal Utility District. G. & E. Systems Giles, Robert. Joint New Zealand/Los Angeles Life- Inc. East Bay Municipal Utility District Seismic Evalu- lines Workshop, August 15/16 1994, Los Angeles ation Program Final Report. Published in California, Department of Water and Power. Published in Los January 21 1994. Angeles, California in 1994.

East Bay Municipal Utility District. The Mokelumne Gough Janet D. Risk and Uncertainty. July 1988. Aqueduct Seismic Upgrade Project. Published in Information Paper No.10. Centre for Resource Man- California, 1994. agement.

Engineering Lifelines Study. Bridge Strengthening Gough, Janet D A Review of the Literature Pertaining Programme. Design Services Unit, Christchurch City to ‘Perceived’ Risk and ‘Acceptable’ Risk and the Council, February 1996. Methods used to estimate them. May 1990. Informa- tion Paper No. 14. Centre for Resource Management. EQE Engineering. January 17, 1994 Northridge, California Earthquake. Published in California in Gough, Janet D. Risk Communication: The Implica- 1994. tions for Risk Management. December 1991. Informa- tion Paper No. 33. Centre for Resource Management. EQE Engineering. Summary of the 1987 Bay of Plenty, New Zealand Earthquake. Published in California in Governor’s Office of American Emergency Services. 1987. Earthquake Recovery: A Survival Manual for Local Government. Published in California in September EQE Engineering. The April 22, 1991, Valle de la 1993. Estrella Costa Rica Earthquake. Published in Califor- nia in 1991. Governors Office of Emergency Services. Northridge Earthquake, January 17 1994. Interim Report. Cali- EQE Engineering. The December 7, 1988 Armenia, fornia, April 1994. USSR Earthquake . Published in California, 1989. Greater Los Angeles Press Club. Earthquake Manual EQE Engineering. The July 16, 1990, Philippines - a Manual for the News Media on the first 72 Hours. Earthquake. Published in California in 1990. Published in California in March 1990.

EQE Engineering. The October 17, 1989 Loma Prieta Guilhem, Olivier and Berrill, John. Canterprise. Project Earthquake, October 1989. Published in California, No. CP/2245. Cone Penetrometer Results and Esti- 1989. mates of Liquefaction Potential at Some Key Lifeline EQE International. Review for 1993 Hokkaido Nansei Sites. August 1993. - Oki Earthquake of July 12 1993. Hopkins, David C. Extending the Lifespan of Struc- EQE International. Risk Safety Engineering World- tures. San Francisco. August 1995. Report by King- wide. ston Morrison.

EQE International. The January 17, 1995 Kobe Earth- Hopkins, David C. Fourth US Conference on Lifeline quake and EQE Summary Report. Earthquake Engineering. San Francisco. August 1995. Report by. Kingston Morrison. Estimating Losses from Earthquakes in China in the Forthcoming 50 Years. Department of Seismic Hazard Hutt City Council. Emergency Response Plan. Part 2. Prevention. State Seismological Bureau, China, 1992. Earthquake Emergency Water Supply.

Federal Emergency Management Agency. When Dis- Institution of Engineers, Australia. Are You At Risk? aster Strikes. A Handbook for Media. Published in Imagine Expectations. November 1990. Washington, D.C. December 1990. Jones, B.S. Los Angeles Earthquake Report, New Field Manual Post Earthquake Safety Evaluation of Zealand Fire Service, Wellington, March 1994. Buildings. ATC 20 - 1, April 1994. Lamb, John. Christchurch Engineering Lifelines Gas Company. Quick Reference Guide. Published in Project - Short Paper presented at the Joint New Los Angeles, California in 1994/1993. Zealand/Los Angeles Lifelines Workshop. Published in Christchurch, New Zealand on 11 August 1994. Gilbert, Jerome B. Bay Area Water Utilities Response to Earthquake (Paper only). Published in California in Latipow, Katherine Boxer. Paper presented to the 1989. Joint New Zealand/Los Angeles Lifelines Workshop Bibliography • 307

regarding state of California Governors Office of Owens F, Kirk R.M., .Bell D.H, Cowan H., Pettinga J. Emergency Services (paper only). Published in Los and Todd D. Hazards in Canterbury. Canterbury Angeles, California in 1994. Regional Council, June 1994.

Le Val Lund. Technical Council on Lifeline Earth- Pacific Gas and Electric Company. Programme for quake Engineering. A.S.C.E. Northridge Earthquake, Reducing Earthquake Vulnerability of Gas and Elec- January 17, 1994. Lifeline Performance. tric Systems by the Year 2000. 1993 Report. Published in California, June 1994. Lifelines Workshop Programme. Joint New Zealand/ Los Angeles Workshop 15/16 August 1994. Published Pacific Gas and Electric Company. Programme for in Los Angeles, California in 1994. Reducing Earthquake Vulnerability of Gas and Elec- tric Systems by the Year 2000. 1992 Report. Published Local Authority Engineering. Various articles on the in California, June 1993. Edgecumbe earthquake, March 2, 1987. Pacific Gas and Electric Company. Programme for Lopez, Marciano and Tognazzini, Ron and Minor, Reducing Earthquake Vulnerability of Gas and Elec- Lawren. Impact of the 1994 Northridge Earthquake on tric Systems by the Year 2000. 1991 Report. Published the Los Angeles Electrical System (paper only). Pub- in California, June 1992. lished in Los Angeles, California, 29 April 1994. Pacific Gas and Electric Company. Programme for Lund, Le Val. Northridge Earthquake, January 17, Reducing Earthquake Vulnerability of Gas and Elec- 1994. Lifeline Introduction, Water and Waste Water tric Systems by the Year 2000. Published in California, Lifeline Performance, ASCE - TCLEE Draft Report. December 1990. Published in California, August 15, 1994. Park, Professor Robert Standard “Risk Management” May, John et al. The Southern California Gas Com- AS/NZS 4360:1995 N.Z. Engineering March 1995. pany Response to the Northridge Earthquake (paper The Great Hanshin Earthquake. only). Published in Los Angeles, California, August 15/16, 1994. Procedures for Earthquake Safety Evaluation of Build- ings. Governors Office of Emergency Services. State Metropolitan Water District of Southern California. of California. ATC 20 1994. The Northridge Earthquake, January 17, 1994. Jensen Filtration Plant Report. Published in Los Angeles, Proceedings of the 6th US Japan Workshop on Earth- California in 1994. quake Disaster Prevention for Lifeline Systems, July 18-19, 1995. Public Works Research Institute, Minis- N.Z. National Society for Earthquake Engineering, try of Construction. Volume 27, No.3, September 1994. Pyle, Eric and Gough, Janet D. Environmental Risk N.Z. National Society for Earthquake Engineering. Assessment for New Zealand. July 1991. Information Draft Guidelines for Assessing and Strengthening Paper No. 29. Centre for Resource Management. Earthquake Risk Buildings. 10 February 1995. Review of the Seismological Work in China. State National Research Council. Practical Lessons from Seismological Bureau, 1992. the Loma Prieta Earthquake. Published by National Academy Press, Washington D.C. in 1994. Schiff, Anshel J. Joint New Zealand/Los Angeles Lifelines Workshop, Federation, Professional and Pri- Norton, Bruce. Background Reports on Northridge vate Initiatives in Lifelines Earthquake Preparedness Earthquake. California by California Seismic Safety (paper only). Published in Los Angeles, California in Commission, August 1994. 1994.

Norton, John and Hopkins, David. US/NZ Lifelines Schiff, Anshel J. Earthquake Spectra. The Profes- Workshop - Short Paper. Published in Wellington, sional Journal of the Earthquake Engineering Research New Zealand in 1994. Institute, Supplement A to Volume 7, Philippines Official Commuter Action Guide by Southern Califor- Earthquake Reconnaissance Report. Published in Cali- nia Transportation Providers (pamphlet only). Pub- fornia in 1991. lished in Los Angeles, California in February 1994. Schiff, Anshel J. Guide to Post Earthquake Investiga- Ostrom, Dennis. A Summary of the January 17 tion of Lifelines, Technical Council on Lifeline Earth- Northridge Earthquake. Impact on Edison facility (paper only). Published in Los Angeles, California, 1994. 308 • Risks and Realities

quake Engineering, Monograph No.3, August 1991. Published in New York in 1991.

Schiff, Anshel J. (Editor). Guide to Post Earthquake Investigation of Lifelines, Technical Council on Life- line Earthquake Engineering. August 1991.

Seismic Vulnerability and Impact of Disruption of Lifelines in the Conterminous United States. ATC Applied Technology Council 1985.

Solorzano, Rafael et al. Sewer Damage from the Northridge Earthquake (paper only). Published in Los Angeles, California, in 1994.

State of California Office of Emergency Services, Division of Utilities. Several Mutual Aid Agreements. Published in California in 1994.

State/Federal Hazard Mitigation Survey Team. Report for the October 17 1989 Loma Prieta Earthquake California. Published in California in January 1990.

Stewart, Jonathan P. et. al. Preliminary Report on the Principal Geotechnical Aspects of the January 17, 1994, Northridge Earthquake (index only). Earth- quake Engineering Research Centre, College of Engi- neering, University of California at Berkeley. Pub- lished in California in June 1994.

Taber, John and Cowan, Hugh A. Measurement of Earthquake Ground Shaking at Characteristic Sites in Christchurch, Report prepared for the Earthquake Com- mission and Canterbury Regional Council.

TEPHRA June 1996. Articles on Risk Management and Natural Disasters. Ministry of Civil Defence.

Trifunac, M.D. et al. A Note on Distribution of Uncor- rected Peak Ground Accelerations During the Northridge, California Earthquake of 17 January 1994. Published at the University of South California, Los Angeles in 1994.

Wellington Earthquake Lifelines Group, 1993 Report.

Wellington Earthquake Lifelines Group, 1994 Report.

Wellington Earthquake Lifelines Group, 1995 Report.

Wellington Regional Bulk Water Department. Emer- gency Response Plan. 1994.

Works Development Services Corporation (NZ) Ltd. Report for New Zealand Earthquake and War Dam- ages Commission on the Effects of Tsunamis in New Zealand, August 1988.

Yates, R. Northridge Earthquake. Transportation Response After Action Report. Published in Los Ange- les, California, in March 1994. Participant Contact Details • 309

Project Team Participants — Contact Details

Steering Mr John Lumsden Mr Tony Boyle Committee Projects Director Structures & Hazards Planning Manager Centre for Advanced Engineering Canterbury Regional Council Mr John Lamb University of Canterbury PO Box 345 CHRISTCHURCH Project Manager Private Bag 4800 Phone: (03) 365 3828 Christchurch Engineering Lifelines CHRISTCHURCH Fax: (03) 365 3194 42 Lynfield Avenue Phone: (03) 364 2219 Email: [email protected] CHRISTCHURCH 8004 Fax: (03) 364 2069 Email: [email protected] Phone: (03) 358 5884 Dr Ian Owens Senior Lecturer in Geography Mr Richard Keys Mr Kevin O’Kane Department of Geography Otago Regional Council Manager, Mitigation Research & University of Canterbury Private Bag Development Private Bag 4800 Ministry of Civil Defence DUNEDIN CHRISTCHURCH PO Box 5010 (formerly Development Engineer Phone: (03) 364 2927 WELLINGTON Canterbury Regional Council) Fax: (03) 354 2907 Phone: (04) 495 6836 Phone: (03) 474 0827 Email: [email protected] Fax: (04) 473 7369 Fax: (03) 479 0015 Email: [email protected] Email: [email protected] Mr Derek Todd Tonkin & Taylor Mr Brian Hasell Mr Ian McCahon PO Box 13055 Chief Executive Manager CHRISTCHURCH Ashburton District Council Soils & Foundations Limited (formerly Coastal Investigations Officer PO Box 94 PO Box 13 052 Canterbury Regional Council) ASHBURTON CHRISTCHURCH Phone: (03) 365 3703 (formerly Group Manager Corporate Policy, Phone: (03) 379 8432 Fax: (03) 365 3742 Email: [email protected] Canterbury Regional Council) Fax: (03) 366 7780 Phone: (03) 308 5139 Mr Jim Williamson Mr John Weeber Fax: (03) 308 1820 Director of Operations Hydrogeologist Email: [email protected] Christchurch City Council Canterbury Regional Council PO Box 345 PO Box 237 Mr Allan Watson CHRISTCHURCH CHRISTCHURCH Manager Water Services Phone: (03) 365 3828 Phone: (03) 371 1607 Christchurch City Council Fax: (03) 365 3194 Fax: (03) 371 1786 PO Box 237 Email: [email protected] CHRISTCHURCH Email: [email protected] Phone: (03) 371 1303 Fax: (03) 371 1387 Hazards Task Group — Task Group 1 — Email: [email protected] Other Advisors Hazards Mr Stephen Franklin Dr John Berrill Engineering Consultant Mr Ian McCahon Reader in Civil Engineering Telecom New Zealand Limited The Manager Department of Civil Engineering PO Box 1473 Soils & Foundations Limited University of Canterbury CHRISTCHURCH PO Box 13 052 Private Bag 4800 Phone: (03) 353 3035 CHRISTCHURCH CHRISTCHURCH Fax: (03) 353 3041 Phone: (03) 379 8432 Phone: (03) 366 7001 Email:[email protected] Fax: (03) 366 7780 Fax: (03) 364 2758 Email: [email protected] Mr Mark Gordon Mr David Bell City Streets Manager Senior Lecturer in Engineering Geology Dr Neil Cherry Christchurch City Council Department of Geology Department of Natural Resources PO Box 237 University of Canterbury Engineering CHRISTCHURCH Private Bag 4800 Lincoln University Phone: 025 220 7250 CHRISTCHURCH PO Box 94 Fax: (09) 575 0919 Phone: (03) 364 2717 LINCOLN UNIVERSITY Email: [email protected] Fax: (03) 364 2769 Phone: (03) 325 3828 Email: [email protected] Fax: (03) 325 3845 Email: [email protected] 310 • Risks and Realities

Mr Ken Couling Mr Mike Berry Mr Ray Basher Land Drainage Manager Liquid Wastes Engineer Operations Support Manager Water Services Unit Waste Management Unit Transpower New Zealand Limited Christchurch City Council Christchurch City Council PO Box 21 154 PO Box 237 PO Box 237 CHRISTCHURCH CHRISTCHURCH CHRISTCHURCH Phone: (03) 365 6948 Phone: (03) 371 1365 Phone: (03) 371 1388 Fax: (03) 379 1525 Fax: (03) 371 1384 Fax: (03) 371 1387 Email: [email protected] Email: [email protected] Email: [email protected] Mr Peter Brash Dr Derek Goring Mr Ken Couling (formerly Administration Manager National Institute of Water & Atmospheric Land Drainage Manager Television NZ Limited Research Water Services Unit PO Box 1945) PO Box 8602 Christchurch City Council PO Box 237 CHRISTCHURCH CHRISTCHURCH CHRISTCHURCH Phone: (03) 343 7868 Phone: (03) 371 1388 Mr Roger Smithies Fax: (03) 348 5548 Fax: (03) 371 1387 Engineer (Planning) Email: [email protected] Email: [email protected] Telecom NZ Limited Professor Bob Kirk PO Box 1473 Mr Allan Dowie H.O.D. CHRISTCHURCH Installation Manager Phone: (03) 353 3634 Department of Geography Shell NZ Limited University of Canterbury PO Box 24 Fax: (03) 366 3818 Private Bag 4800 LYTTELTON Email: [email protected] CHRISTCHURCH (Rep. of Petrol Industry Phone: (03) 364 2893 Emergency Action Committee) Mr Russ Botting Fax: (03) 364 2907 Phone: (03) 328 7395 Senior Engineering Consultant Email: [email protected] Fax: (03) 328 7452 Fire Protection Telecom New Zealand Limited Mr Mark Yetton Mr Dave May PO Box 1473 Geotech Consulting Limited Civil Engineer CHRISTCHURCH Charteris Bay Water Services Phone: (03) 353 3569 R D 1 Christchurch City Council Fax: (03) 366 9127 CHRISTCHURCH PO Box 237 Phone: (03) 329 4044 CHRISTCHURCH Mr Bud Chapman Fax: (03) 329 4044 Phone: (03) 371 1329 Specialist Fax: (03) 371 1387 Communications Centres Email: [email protected] PO Box 6147 Task Group 2 — Civil Mr Alan Marshall Marion Square Services Engineer (Protection Systems) WELLINGTON Telecom New Zealand Limited Phone: (04) 801 8939 Mr Allan Watson PO Box 1473 Mobile: 021 377 874 Manager Water Services CHRISTCHURCH Christchurch City Council Phone: (03) 363 8448 Mr John Coleman PO Box 237 Fax: (03) 363 8015 Supervising Technician CHRISTCHURCH Email: [email protected] NZ Police Engineering Workshop Phone: (03) 371 1303 PO Box 2109 Fax: (03) 371 1387 Mr Bob Watts CHRISTCHURCH Email: [email protected] Planning Manager Phone: (03) 363 7640 Water Services Unit Mr Neil Bennett Christchurch City Council Fax: (03) 365 3515 Principal Consultant PO Box 237 Email: [email protected] SERCO CHRISTCHURCH PO Box 19 683 Phone: (03) 371 1393 Mr Noel Maginnity CHRISTCHURCH Fax: (03) 371 1387 Technical Centre (formerly Asset Manager Radio Network Banks Peninsula District Council) PO Box 1484 Phone: (03) 384 5974 Task Group 3 — Electrical CHRISTCHURCH Fax: (03) 384 1985 & Communications Phone: (03) 379 9600 Email: [email protected] Fax: (03) 365 5635 Networks Mr Neville Stewart Mr John MacKenzie Senior Engineering Officer Mr Stephen Franklin Senior Engineer Waste Management Unit Engineering Consultant Montgomery Watson Limited Telecom New Zealand Limited Christchurch City Council PO Box 13052 PO Box 1473 PO Box 237 CHRISTCHURCH CHRISTCHURCH CHRISTCHURCH Phone: (03) 343 8715 Phone: (03) 353 3035 Phone: (03) 371 1366 Fax: (03) 366 7780 Fax: (03) 353 3041 Fax: (03) 371 1384 Email: [email protected] Email: [email protected] Email: [email protected] Participant Contact Details • 311

Mr Grant Roberts Mr John Robb Mr John Reynolds Assistant Manager (formerly Transportation Engineer Design Engineer Civil Lincrad Aerials Canterbury Regional Council) Opus International Consultants Limited 17 Washbournes Road PO Box 345 PO Box 1482 CHRISTCHURCH CHRISTCHURCH CHRISTCHURCH (formerly Works Consultancy Services (formerly Director of Resources Limited) Canterbury Television Limited) Mr Russell Herbert Phone: (03) 365 1530 Operations Group Phone: (03) 348 0659 Fax: (03) 365 7858 Tranz Rail Limited Fax: (03) 348 4043 Email: [email protected] Private Bag 4723 Mr John O’Donnell CHRISTCHURCH Mr Graeme Wilson Network Asset Manager Phone: (03) 372 8425 Programme Manager Southpower Fax: (03) 372 8282 City Streets Unit Private Bag 4999 (formerly Road Programming & Pavements CHRISTCHURCH Mr Ken McAnergney Engineer) Phone: (03) 363 9781 Airport Planner Christchurch City Council Fax: (03) 363 9707 Christchurch International Airport Limited PO Box 237 Email: [email protected] PO Box 14001 CHRISTCHURCH CHRISTCHURCH Phone: (03) 371 1656 Phone: (03) 353 7716 Fax: (03) 371 1864 Task Group 4 — Fax: (03) 353 7730 Email: [email protected] Transportation Email: [email protected] Mr Mike McGlinchey Task Group 5 — Fire Mr Mark Gordon Maintenance Manager Services City Streets Manager Lyttelton Port Company Limited Christchurch City Council Private Bag 501 Mr Barry G J Shields PO Box 237 LYTTELTON Assistant Area Chief Fire Officer CHRISTCHURCH (formerly Engineering Services Manager, NZ Fire Services Phone: 025 2207 250 Christchurch International Airport Limited) PO Box 13 747 Fax: (09) 575 0919 CHRISTCHURCH Phone: (03) 328 8198 Email: [email protected] Phone: (03) 371 3603 Fax: (03) 328 7828 Fax: (03) 371 3622 Mr David Bates Email: [email protected] Regional Highways Engineer Mr Neil McLennan Transit NZ Engineering Services Manager Task Group 6 — Building PO Box 1479 Lyttelton Port Company Limited Services CHRISTCHURCH Private Bag 501 Phone: (03) 366 4455 LYTTELTON Mr Iain Drewett Fax: (03) 365 6576 Consulting Engineer Phone: (03) 328 8198 Email: [email protected] 24C Bowenvale Avenue Fax: (03) 328 7828 CHRISTCHURCH 2 Mr Tony Barnett Email: [email protected] Phone: (03) 332 0133 (formerly Transport Planning Engineer Christchurch City Council) Mr Grant Wilkinson PO Box 237 Director CHRISTCHURCH Holmes Consulting Group Limited PO Box 701 CHRISTCHURCH Phone: (03) 366 3366 Fax: (03) 379 2169 Email: [email protected] 312 • Risks and Realities