Knock For Knock Indemnities and The Law Contractual Limitation and

KNOC K-F OR-KNOC K IN D EMN ITIES
AND T HE LAW
CONTEMPORARY COMMERCIAL LAW
Maritime Law in China: Emerging Issues and Blockchain Technology and the Law:
Future Developments Opportunities and Risks
Edited by Johanna Hjalmarsson and Muharem Kianief
Jenny Jingbo Zhang
Third Party Protection in Shipping
Illegality in Marine Insurance Law Carlo Corcione
Feng Wang
Multi-sided Music Platforms and the Law:
Insurance Law Implications of Copyright, Law and Policy in Africa
Delay in Maritime Transport Chijioke Ifeoma Okorie
Ayşegül Buğra
Reinsurance and the Law of Aggregation: Event,
Online Arbitration Occurrence, Catastrophe, Cause
Faye Fangfei Wang Oliver D. William
Double Insurance and Contribution Comparative Analysis of Interim Measures:

Nisha Mohamed Interim Remedies (England & Wales) v
Preservation Measures (China)
The Law and Autonomous Vehicles Vivek Jain, Thomas Macey-Dare QC and
Matthew Channon, Lucy McCormick and Shengnan Jia
Kyriaki Noussia
Knock-for-Knock Indemnities and the Law:
FIDIC Yellow Book: A Commentary Contractual Limitation and Delictual Liability
Ben Beaumont Edited by Kristofer Svendsen, Endre Stavang
and Greg Gordon
The Contract of Carriage: Multimodal Transport
and Unimodal Regulation Managing International Trade Risk: Customs,
Paula Bäckdén Revenue and VAT Compliance
Mark Rowbotham
FIDIC Red Book: A Commentary
Ben Beaumont
For more information about this series, please visit:

www.routledge.com/Contemporary-Commercial-Law/book-series/CCL
K N OCK - F OR -KNO C K
INDE M N IT IE S AN D T HE L AW
C ON T RAC T UAL LI MI TATI O N

A N D D E L I C T UAL LI A BI LI TY
EDITED BY
K R I S T O F F E R S V E N D S E N,
E N D R E S TAVA N G A N D G R E G G O R D O N
Cover credit: © Panuwat Dangsungnoen / Getty Images
First published by Informa Law from 2023

by Routledge
4 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Informa Law from Routledge

605 Third Avenue, New York, NY 10158
Informa Law from Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2023 selection and editorial matter, Kristofer Svendsen, Endre Stavang and Greg Gordon; individual
chapters, the contributors
The right of Kristofer Svendsen, Endre Stavang and Greg Gordon to be identifed as the authors of
the editorial material, and of the authors for their individual chapters, has been asserted in accordance
with sections 77 and 78 of the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by
any electronic, mechanical, or other means, now known or hereafter invented, including photocopying
and recording, or in any information storage or retrieval system, without permission in writing from
the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identifcation and explanation without intent to infringe.
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library
ISBN: 978-1-032-07408-5 (hbk)

ISBN: 978-1-032-07411-5 (pbk)
ISBN: 978-1-003-20679-8 (ebk)
DOI: 10.4324/9781003206798
Typeset in Times New Roman

by Apex CoVantage, LLC
C ON T E N TS
List of contributors xix

Editors’ preface xxi
Table of cases xxv
Table of statutes and Model Form Contracts xxxiii
PART I SELECTED TOPICS

CHAPTER 1 THE DEVELOPMENT OF KNOCK-FOR-KNOCK
CLAUSES IN THE LAST 15 YEARS 1
Kyriaki Noussia and Hanieh Bolourian
CHAPTER 2 AN INTRODUCTION TO RISK ALLOCATION IN

OIL AND GAS CONTRACTS FROM AN ENGLISH
LAW PERSPECTIVE 20
Greg Gordon
CHAPTER 3 CONTRACTING AROUND TORT DEFAULTS:

THE KNOCK-FOR-KNOCK PRINCIPLE AND
ACCIDENT COSTS 65
Gideon Parchomovsky and Endre Stavang
CHAPTER 4 ON KNOCK-FOR-KNOCK CLAUSES AND

THEIR OPTIMAL REGULATION 87
Henrik Lando
CHAPTER 5 KNOCK-FOR-KNOCK INDEMNITY PROVISIONS

AND LIABILITY INSURANCE: POTENTIALLY
STRANGE BUT ALWAYS COMPLICATED
BEDFELLOWS 98
Jay R. Sever and Lauren E. Burk
CHAPTER 6 THE EFFECT OF CHOICE OF LAW ON

KNOCK-FOR-KNOCK CLAUSES 106
Uisdean Vass
v
contents
PART II KNOCK FOR KNOCK IN SPECIFIC JURISDICTIONS
CHAPTER 7 INDEMNITY CLAUSES IN FABRICATION AND
CONSTRUCTION CONTRACTS IN NORWAY 117
CHAPTER 7A LIMITING AND CHANNELING LIABILITY UNDER

OFFSHORE CONSTRUCTION CONTRACTS
IN NORWAY 118
Knut Kaasen
CHAPTER 7B THE OBLIGATION UNDER NORWEGIAN

OFFSHORE CONTRACTS TO INSURE THE
CONTRACT OBJECT AS PART OF A
KNOCK-FOR-KNOCK AGREEMENT 153
Erik Brannsten
CHAPTER 8 LIABILITY AND INSURANCE CLAUSES IN

CONTRACTS FOR VESSEL SERVICES IN
THE NORWEGIAN OFFSHORE SECTOR:
THE KNOCK-FOR-KNOCK PRINCIPLE 164
Trine-Lise Wilhelmsen
CHAPTER 9 STATUTORY LIABILITY REGULATION

VERSUS CONTRACTUAL RISK ALLOCATION
IN UPSTREAM OIL AND GAS: THE NORWEGIAN
CASE 182
Kristofer Svendsen
CHAPTER 10 APPLYING KNOCK-FOR-KNOCK IN GERMANY 193

Eckehard Volz and Anna-Sophie Waldmann
CHAPTER 11 KNOCK FOR KNOCK UNDER BRAZILIAN LAW 204

Felipe T. Boechem and Márcio Opromolla
CHAPTER 12 THE IMPLICATIONS OF LITIGATION POST

DEEPWATER HORIZON ON KNOCK-FOR-KNOCK
CLAUSES IN US LAW 214
Cindy Matherne Muller, Jeanne Amy, and Jennifer David
CHAPTER 13 MUTUAL HOLD HARMLESS CLAUSES IN

FRANCE AND FRANCOPHONE CIVIL LAW
SYSTEMS 238
Bertrand Montembault and Paul Morton
vi
contents
CHAPTER 14 LESSONS FROM THE APPLICATION OF
KNOCK-FOR-KNOCK CLAUSES UNDER
MALAYSIAN LAW 251
Wan Mohd Zulhafz Wan Zahari
Index 291
vii
D E TAI L E D C O N TEN TS
List of contributors xix

Editors’ preface xxi
Table of cases xxv
Table of statutes and Model Form Contracts xxxiii
PART 1 SELECTED TOPICS

CHAPTER 1 THE DEVELOPMENT OF KNOCK-FOR-KNOCK
CLAUSES IN THE LAST 15 YEARS 1
1 Introduction 1
1.1 Defnition 1
1.2 Essential meaning of a K4K clause 1
1.3 Historical genesis of the K4K clause 2
2 Issues in using, implementing and enforcing K4K clauses 3
3 Various uses 3
3.1 SUPPLYTIME 2005 3
3.2 Insurance and P&I clubs 4
3.3 Use in ofshore contracts 4
4 Modern use 5
4.1 The BIMCO SUPPLYTIME 2017 5
5 A historical encounter of the K4K clauses in the shipping and
in the ofshore oil and gas sectors 6
6 The evolution of K4K clauses in standard contracts and in
English case law 7
6.1 The signifcance of the Piper Alpha litigation 9
7 The contribution of other jurisprudence to the historical evolution
of K4K clauses 10
7.1 US jurisprudence 10
7.2 Deepwater Horizon 11
7.3 Brazil 11
7.4 Germany 12
7.5 Australia 13
7.6 Nordic countries – the history and validity of K4K clauses from
a comparative perspective 13
ix
detailed contents
7.7 Later judicial trends 15
7.8 The way forward 16
8 Conclusive critique 17
CHAPTER 2 AN INTRODUCTION TO RISK ALLOCATION

IN OIL AND GAS CONTRACTS FROM AN
ENGLISH LAW PERSPECTIVE 20
Greg Gordon
1 Introduction 20
2 Indemnifcation and related concepts 21
2.1 Introduction to the concept of indemnifcation 21
2.2 Indemnity and hold harmless clauses 22
2.3 Mutual indemnity and mutual indemnity and hold
harmless clauses 24
3 Indemnity and hold harmless provisions in the oil and gas context 25
3.1 Introduction to simple indemnity and hold harmless clauses
in oil and gas contracts 25
3.2 Introduction to mutual indemnity and hold harmless clauses
3.3 The rationale for a mutual indemnity and hold harmless regime
in the oil and gas context 28
3.4 Back-to-back indemnity and hold harmless provisions
3.5 Qualifed indemnity and hold harmless provisions 31
4 Selected further issues in indemnifcation law and practice in the UKCS 34
4.1 Statutory control of indemnity and hold harmless clauses 34
4.2 The law’s default settings: the normal presumptions
about the distribution of risk 35
4.3 The position of third parties 36
5 Interpreting indemnity and hold harmless clauses 37
5.1 The traditional approach to interpretation 37
The general rules 37
Interpretation contra proferentem 38
Towards contextualism (and back again) 39
6 Some known problems of drafting and interpretation 40
6.1 Contra proferentem and the problem of negligence and
breach of statutory duty 40
6.2 Words delimiting the circumstances in which the indemnity
and hold harmless provision will take efect 46
6.3 Multi-party issues 47
6.4 “Full and primary” 48
6.5 Defnitional issues 49
Company groups 49
Employees and personnel 50
Property 50
x
detailed contents
7 The problem of multiple parties 50
7.1 Introduction to the problem 50
7.2 The Industry Mutual Hold Harmless Agreement (IMHH) 52
The Deed’s core indemnity provisions 53
Other important provisions 54
Entry into force 54
Geographical extent 55
Order of precedence 55
Extension of benefts to groups 55
Waiver of rights of subrogation 55
Right to defend 56
Exceptions 56
Transportation by air excluded 56
Landward areas 57
Operators 57
Commentary and conclusions on the IMHH scheme 57
The contractual nature of the scheme 57
Potential dangers of the IMHH 58
8 Liability for “consequential” loss 59
8.1 What is “consequential loss”? 59
8.2 Commentary and conclusions on consequential loss 61
9 Overall limitation of liability 62
CHAPTER 3 CONTRACTING AROUND TORT DEFAULTS:

THE KNOCK-FOR-KNOCK PRINCIPLE AND
ACCIDENT COSTS 65
1 Introduction 65
2 The knock-for-knock principle 68
2.1 The industrial setting 69
2.2 Illustrating the operation of knock for knock 69
3 Knock for knock, social norms, and private ordering 71
4 The efect of knock for knock on social welfare 73
4.1 The perfect separating equilibrium 74
4.2 The perfect pooling equilibrium 74
4.2.1 Self-regulation 75
4.2.2 Repeat interactions 76
4.2.3 Health and safety regulation 77
4.2.4 Insurance 78
5 The hidden perils of knock for knock 79
5.1 Risk interdependencies 80
5.2 Litigation externalities 80
5.3 Harm to environmental interests 82
5.4 Deterrence versus compensation 84
6 Conclusion 85
xi
detailed contents
CHAPTER 4 ON KNOCK-FOR-KNOCK CLAUSES AND
THEIR OPTIMAL REGULATION 87
Henrik Lando
1 Introduction 87
2 On the literature 88
3 When are knock-for-knock clauses useful? 89
3.1 A model of the use of knock-for-knock contracts 90
4 Should courts impose carve-outs for gross negligence? 92
4.1 A framework 94
5 Conclusion 97
CHAPTER 5 KNOCK-FOR-KNOCK INDEMNITY PROVISIONS

AND LIABILITY INSURANCE: POTENTIALLY
STRANGE BUT ALWAYS COMPLICATED
BEDFELLOWS 98
1 Introduction 98
2 K4K provisions and general liability insurance 99
3 Impact of blanket AI endorsements 100
4 “Other insurance” clauses 102
5 Incorporation of indemnity limitations in insurance policies 103
6 Waiver of subrogation 104
7 Conclusion 105
CHAPTER 6 THE EFFECT OF CHOICE OF LAW ON

KNOCK-FOR-KNOCK CLAUSES 106
Uisdean Vass
1 Introduction 106
2 Knock for knock and the full English legal treatment 107
2.1 The knock-for-knock clause itself 107
2.2 Simple knock for knock in contractual context 108
2.3 Knock for knock in its English legal context 109
3 Compare the United States 111
3.1 The jurisdictional issue 111
3.2 Federal maritime law – Deepwater Horizon 111
3.3 The anti-indemnity statutes 113
3.3.1 General observation 113
3.3.2 Louisiana Oilfeld Indemnity Act and
other state provisions 113
3.3.3 Texas Oilfeld Anti-Indemnity Act and
other state provisions 114
4 Compare Brazil 114
4.1 General observations 114
4.2 Knock-for-knock and Brazilian law 115
5 Conclusion 116
xii
detailed contents
PART II KNOCK FOR KNOCK IN SPECIFIC JURISDICTIONS
CHAPTER 7 INDEMNITY CLAUSES IN FABRICATION AND
CONSTRUCTION CONTRACTS IN NORWAY 117
CHAPTER 7A LIMITING AND CHANNELING LIABILITY

UNDER OFFSHORE CONSTRUCTION
CONTRACTS IN NORWAY 118
Knut Kaasen
1 The problem 118
2 The contractor’s liability for breach under ofshore
construction contracts 120
2.1 The liability 120
2.1.1 The need for a balanced regime 120
2.1.2 Types of sanctions against breach 121
2.1.3 Specifc features 122
2.1.4 Breach of secondary obligations 123
2.1.5 Obligation to notify – and the role of the
variation mechanism 123
2.2 The limitation of liability 124
3 The knock-for-knock system 125
3.1 The call for regulation 125
3.1.1 The risks of damage 125
3.1.2 Consequences of no regulation 126
3.1.3 Potential regulation 126
3.1.4 The insurance aspects 127
3.1.5 A global view of risk, liability and insurance 128
3.2 Provisions on liability 128
3.2.1 Overview 128
3.2.2 The family zones 130
3.2.3 The third-party zone 131
3.3 Insurance 131
3.3.1 Considerations 131
3.3.2 Main components 132
3.3.3 Insurance arrangements 133
4 Setting aside provisions limiting exposure resulting from default or damage? 134
4.1 Authority for censoring provisions limiting exposure? 134
4.2 Diferences and similarities between the two types of
liability situations 135
4.2.1 Liability situations 135
4.2.2 Liability regulations 135
5 Are the agreed limitations of liability for breach of contract valid? 137
5.1 Points of departure 137
5.2 The contractor’s own gross negligence 138
5.3 Who is “the contractor”? 141
5.4 The role of the concept “willful misconduct” 143
xiii
detailed contents
6 Censoring the knock-for-knock regime? 145
6.1 The question 145
6.2 “Organized fnancing of loss” 147
6.3 Can the waiver of subrogation be maintained? 148
6.3.1 Waiver of subrogation 148
6.3.2 The co-insured 148
6.3.3 Are the arrangements “reasonable”? 149
6.3.4 The conclusion 150
7 Do the NTK 15 provisions on exclusion and limitation of
liability hold good? 150
CHAPTER 7B THE OBLIGATION UNDER NORWEGIAN

OFFSHORE CONTRACTS TO INSURE THE
CONTRACT OBJECT AS PART OF A
KNOCK-FOR-KNOCK AGREEMENT 153
Erik Brannsten
1 Introduction and background 153
1.1 Indemnifcation, liability and insurance 154
1.2 Project all risk insurances 155
2 The knock-for-knock arrangement 156
2.1 Risk zones 156
2.2 Waivers of liability, indemnifcation and waiver
of subrogation 156
2.3 Damage to the contract object 156
2.4 Waivers and indemnities shall apply regardless
of cause 158
3 The ofshore project insurances 158
3.1 Introduction 158
3.2 Insured perils 160
3.3 Basis for recovery 162
4 Summary 163
CHAPTER 8 LIABILITY AND INSURANCE CLAUSES IN

CONTRACTS FOR VESSEL SERVICES IN
THE NORWEGIAN OFFSHORE SECTOR:
THE KNOCK-FOR-KNOCK PRINCIPLE 164
1 Introduction 164
2 Overview of the relevant tort law and insurance legislation 165
2.1 Tort law 165
2.2 Insurance law 166
3 The content and structure of the knock-for-knock principle 166
3.1 Type of loss and basis for liability 166
3.2 Who is included in the liability provisions – “the group concept” 167
3.3 Freedom of liability, indemnity and subrogation 169
3.4 The insurance regulation 170
xiv
detailed contents
4 The rationale for the knock-for-knock principle 171
4.1 The need for contractual control of the liability risk 171
4.2 Efcient insurance coverage 172
4.3 Loss prevention? 173
4.3.1 Loss prevention and efcient liability rules 173
4.3.2 The efcient basis for liability in contractual relationships 174
5 The validity of the regulation 175
5.1 Some starting points 175
5.2 NL 5–1–2 176
5.3 The Contract Act § 36 178
5.3.1 Overview and some starting points 178
5.3.2 The content of the agreement 179
5.3.3 The insurance clauses 181
CHAPTER 9 STATUTORY LIABILITY REGULATION

VERSUS CONTRACTUAL RISK ALLOCATION
IN UPSTREAM OIL AND GAS: THE NORWEGIAN
CASE 182
Kristofer Svendsen
1 Introduction 182
2 Statutory regulation of risk under Chapter 7 of the Norwegian
Petroleum Act 182
2.2 Liability for petroleum pollution damage 184
2.2.1 Unlimited no-fault liability 184
2.2.2 Pollution damage 185
2.2.3 Channelling liability to the licensee and licensee’s recourse 186
3 Contractual allocation of risk on oil and gas contracts in Norway 187
3.2 Industry-negotiated model contracts 188
3.3 The risk allocation clause 189
4 Contractual allocation of risk breaches with Chapter 7 192
CHAPTER 10 APPLYING KNOCK-FOR-KNOCK IN GERMANY 193

1 Introduction 193
2 Use of knock-for-knock clauses in the German market 193
3 Validity issues under German law 194
3.1 Court review of general terms and conditions 194
3.1.1 General principle 194
3.1.1.1 Defnition of general terms and conditions 195
3.1.1.2 Content test 195
3.1.1.3 What is “individually negotiated”? 196
Negotiation of terms demands more than debating 196
Additional security by a “caveat” clause? 196
3.2 Are knock-for-knock clauses general terms and conditions? 197
xv
detailed contents
3.3 Applying the reasonableness test 197
3.3.1 Commercial contracts versus consumer contracts 198
3.3.2 Fairness test – what is market practice? 198
3.3.3 The “risky” terms 198
3.4 Invalid terms under German civil law – no exclusion of
liability for intent 199
3.4.1 General principle of liability under sec. 276 BGB 200
3.4.2 No agreements to the detriment of third parties 201
3.4.3 Consequences for knock-for-knock clauses under German law 201
4 Available case law 202
5 Practical consequences and advice for use of knock-for-knock
clauses under German law 202
CHAPTER 11 KNOCK FOR KNOCK UNDER BRAZILIAN LAW 204

1 Introduction 204
2 Overview of civil liability regime in Brazil 205
2.1 Contractual liability and tort liability 205
2.2 Fault-based liability and strict liability 206
2.3 Indemnifable losses 207
3 Limitation of liability clause 209
4 Knock-for-knock clauses in Brazil 211
5 Conclusion 212
CHAPTER 12 THE IMPLICATIONS OF LITIGATION POST

DEEPWATER HORIZON ON KNOCK-FOR-KNOCK
CLAUSES IN US LAW 214
Cindy Matherne Muller, Jeanne Amy and Jennifer David
1 Introduction 214
2 Applicable law 214
2.1 Federal court jurisdiction 215
2.1.1 Federal question jurisdiction 215
2.1.2 Diversity jurisdiction 215
2.1.3 Admiralty jurisdiction 216
2.2 Applicable law 217
2.2.1 Outer Continental Shelf Lands Act 217
2.2.2 State waters 219
2.2.3 Maritime law 219
2.2.3.1 Maritime contracts 219
2.2.3.2 What is a vessel? 221
2.2.4 Maritime law incomplete, supplemented with state law 223
2.2.5 Tension between OCSLA and maritime law 223
3 Indemnity in maritime 224
3.1 Enforceability of indemnity provisions in a maritime contract 224
3.2 Indemnity under the Longshore and Harbor Workers’
Compensation Act 226
xvi
detailed contents
3.3 Insurance obligations and additional assured status 227
4 Application of state law 229
4.1 The Louisiana Oilfeld Indemnity Act 229
4.2 Texas Oilfeld Anti-Indemnity Act 233
5 Conclusion 234
Appendix A 235
CHAPTER 13 MUTUAL HOLD HARMLESS CLAUSES IN

FRANCE AND FRANCOPHONE CIVIL LAW
SYSTEMS 238
1 Introduction 238
1.1 Francophone civil law systems 238
1.1.1 The French Civil Code of 1804 239
1.1.1.1 French law of obligations: contract law and
civil liability 239
1.1.1.2 Francophone African legal systems 239
1.1.2 Recent developments 240
1.1.2.1 French law 240
1.1.2.2 Francophone African legal systems 240
1.2 Mutual hold harmless clauses: a contractual re-allocation of liability 241
1.2.1 Origins and growing prevalence in French law 241
1.2.2 A note on English terminology 241
2 The legal nature of mutual hold harmless clauses under French law 242
2.1 The legal framework for mutual hold harmless clauses 242
2.1.1 Principles of the law on civil liability under French law 242
2.1.2 Contractual mechanisms for allocating risk between
the parties 243
2.1.3 Contractual mechanisms for allocating third-party liability 243
2.2 Legal characterisation of mutual hold harmless clauses
under French law 244
2.2.1 Allocation of risk between the parties 244
2.2.2 Allocation of third-party liability 244
2.2.2.1 Related third parties 244
2.2.2.2 Unrelated third parties 245
3 The legal regime governing mutual hold harmless clauses under French law 246
3.1 Exclusions of liability and waivers of recourse 246
3.1.1 Validity and exceptions 246
3.1.1.1 Exceptions linked to the law of obligations 246
3.1.1.1.1 Limiting non-contractual liability 246
3.1.1.1.2 Essential obligations 246
3.1.1.2 Specifc legislative exceptions 247
3.1.2 Efects of waiver of recourse and exclusions of liability
clauses under French law 247
3.1.2.1 Strict interpretation 247
3.1.2.2 Exceptions linked to the nature of the fault 248
xvii
detailed contents
3.1.2.3 Exceptions linked to the nature of the loss 248
3.2 Indemnifcation of third-party losses 249
3.2.1 Validity of clauses indemnifying third-party losses 249
3.2.2 Limitations 249
CHAPTER 14 LESSONS FROM THE APPLICATION OF

KNOCK-FOR-KNOCK CLAUSES UNDER
MALAYSIAN LAW 251
1 Introduction 251
1.1 General overview of the Malaysian legal system 251
1.2 The law of contract in Malaysia 252
1.2.1 Statutory and judicial approach on indemnity
clauses in Malaysia 255
1.3 Empirical study of oilfeld service contracts in Malaysia 261
1.4 Research design 262
1.4.1 Research process 262
2 Description of the case studies 263
2.1 Findings 265
2.2 Perception of the contractual formation process 265
2.3 Perception of risk allocation and indemnity clauses 268
3 Analysis of the indemnity and hold harmless clauses of
Operators A, B and C 274
3.1 Liability regarding personal injury or death of
employees and loss of property of the parties 276
3.1.1 Under LOGIC 276
3.1.2 Under FIDIC 277
3.1.3 Under Operator A’s contract 278
3.1.4 Under Operator B’s contract 280
3.1.5 Under Operator C’s contract 281
3.2 Claims by a third party 281
3.2.1 Third-party claims under LOGIC 282
3.2.2 Third-party claims under FIDIC 283
3.2.3 Third-party claims under Operator A’s contract 284
3.2.4 Third-party claims under Operator B’s contract 284
3.2.5 Third-party claims under Operator C’s contract 285
3.3 Liability regarding pollution 285
3.3.1 Pollution liability under LOGIC 285
3.3.2 Pollution liability under FIDIC 286
3.3.3 Pollution liability under Operator A’s contract 287
3.3.4 Pollution liability under Operator B’s contract 287
3.3.5 Pollution liability under Operator C’s contract 288
3.4 Insurance coverage 288
4 Conclusion 289
Index 291
xviii
C ON T RI BUTO R S
Jeanne Amy, Trial Attorney in Aviation, Space and Admiralty Litigation, Torts
Branch, Civil Division, US Department of Justice, Jeanne.L.Amy@usdoj.gov.
Felipe T. Boechem, Partner in the Energy and Infrastructure Department and Head
of the Oil and Gas practice at Lefosse, felipe.boechem@lefosse.com.
Hanieh Bolourian, Legal Research Intern at the University of Exeter, haniehbln@
gmail.com.
Erik Brannsten, Attorney at law at Bane NOR (state-owned company responsible for
the Norwegian national railway infrastructure), erik.brannsten@banenor.no.
Lauren E. Burk, Of Counsel: Baker, Donelson, Bearman, Caldwell & Berkowitz, PC.
Jennifer David, Partner in the New Orleans ofce of Broussard + Williamson, jen@
bwtriallawyers.com.
Greg Gordon, Head and Professor of Law at Aberdeen University School of Law,
Fellow at the Higher Education Academy, g.w.gordon@abdn.ac.uk.
Knut Kaasen, Professor of Law at the Scandinavian Institute of Maritime Law, Uni-
versity of Oslo; Moderator in negotiations between construction industry and oil
companies/owners to establish agreed standard contracts for ofshore and onshore
development projects; and Arbitrator and Mediator in commercial disputes, knut.
kaasen@jus.uio.no.
Henrik Lando, Professor of Law and Economics at the Copenhagen Business School,
hl.law@cbs.dk.
Bertrand Montembault, Partner in the energy, resources and infrastructure team in
Herbert Smith Freehills’ Paris Ofce, Bertrand.Montembault@hsf.com.
Paul Morton, Of Counsel in the energy, resources and infrastructure team in Herbert
Smith Freehills’ Paris Ofce, Paul.Morton@hsf.com.
Cindy Matherne Muller, Partner in the construction, energy, environmental and natu-
ral resources, and maritime industry teams at Jones Walker, cmuller@joneswalker.
com.
Kyriaki Noussia, Associate Professor of Law at the University of Reading, and
Arbitrator, k.p.noussia@reading.ac.uk.
xix
contributors
Márcio Opromolla, Partner in the dispute resolution practice at Lefosse, marcio.
opromolla@lefosse.com.
Gideon Parchomovsky, Professor of Law at Hebrew University of Jerusalem Faculty
of Law and University of Pennsylvania Carey Law School, gparchom@mail.huji.
ac.il.
Jay R. Sever, Partner, Regional Practice Coordinator, and Head of the insurance
and reinsurance practice group in the New Orleans Ofce of Phelps Dunbar, jay.
sever@phelps.com.
Endre Stavang, Professor of Law at the University of Oslo, Department of Private
Law, endre.stavang@jus.uio.no.
Kristofer Svendsen, Associate Professor of Law at the University of Stavanger and
Kristiania University College; Senior Research Fellow at Tulane Center for Energy
Law, Tulane Law School; and Associate Member at Aberdeen University Centre
for Energy Law, kristofer.svendsen@uis.no.
Uisdean Vass, Managing Director at VassPetro Limited, uisdean@vasspetro.com.
Eckehard Volz, Partner and Head of the Hamburg admiralty and energy and ofshore
team at Clyde & Co, eckehard.volz@clydeco.com.
Anna-Sophie Waldmann, Senior Associate at Clyde & Co, anna-sophie.waldmann@
clydeco.com.
Trine-Lise Wilhelmsen, Professor of Law at the Scandinavian Institute of Maritime
Law, University of Oslo; Chairman of the Standing Revision Committee for the
Nordic Marine Insurance Plan; Chairman of the board of Norsk Legemiddelans-
varsforsikring (drug liability insurance); Chairman of the board of Finanskla-
genemnda, skade; Chairman of the Petroleum Normprice Board; and Member
of the board in Nordic Ofshore and Maritime Arbitration Association,
t.l.wilhelmsen@jus.uio.no.
Wan Mohd Zulhafz Wan Zahari, Deputy Dean (Academic and Internationalisation)
and Assistant Professor of Law at Ahmad Ibrahim Kulliyyah of Laws, Interna-
tional Islamic University Malaysia (IIUM), and IIUM Company Secretary, wzul-
hafz@iium.edu.my.
xx
E D I T ORS ’ P REFAC E
The knock-for-knock indemnity: concepts, efects and questions

Knock-for-knock indemnity clauses are a well-established feature of contracts in the
oil and gas service sector. In using such clauses, the parties agree that for certain
forms of potential liability – typically property damage, personal injury to employees,
and sometimes other heads of claim such as consequential loss – any loss arising
will be absorbed by the party who sufers it: “you look after your losses, I’ll look
after mine.” It is an apparently simple, pragmatic and neat solution to the question
of who bears liability: a risk allocation model so straightforward that it was described
by one experienced English judge as “crude”.1
At frst sight, one might wonder why such a simple concept might need to have a
book written about it. However, as so often in life, the knock-for-knock indemnity
clause gives rise to more – and more important – issues than might at frst appear.
The frst clue that there is more to this concept than might meet the eye lies in the
length of some of the knock-for-knock indemnities commonly used in the oil and
gas industry. For instance, the indemnity clause in LOGIC’s General Conditions of
Contract for Ofshore Decommissioning2 extends to 11 sub-clauses spanning a little
under two full pages and making use of no fewer than eight defned terms. A further
half page is given over to a separate mutual indemnity for consequential loss.3
Why does an apparently crude risk allocation require so much precision and expo-
sition? There are many reasons. Firstly, and perhaps most importantly, these clauses
are of huge practical importance. In almost any oil and gas contract, there exists a
remote but ever-present risk that an error in the undertaking of the contracted work
could result in a disaster where the potential liabilities could easily run to hundreds of
millions worth of dollars.4 If the parties are going to use knock-for-knock indemni-
ties to distribute liability, then they are going to want to be as sure as possible that
1 Smit International (Deutschland) GmbH v Josef Mobius Bau-gesellschaft GmbH [2001] CLC 1545 at
[19]–[20] per Mr Justice Morison.
2 LOGIC, General Conditions of Contract (including Guidance Notes) for Ofshore Decommissioning
(2018) available online from www.logic-oil.com/content/standard-contracts0, Clause 22, pp. 25–27.
3 Ibid., Clause 25, p. 29.
4 Consider, for example, the Piper Alpha disaster which occurred on 6 July 1988, which led to the loss
of 167 lives, the complete destruction of the Piper Alpha installation and the shutting-in of a signifcant
volume of production capacity upstream of the destroyed platform. So great was the loss of production
that the United Kingdom as a whole went from being a net oil exporter to a net oil importer for more than
a year after the accident.
xxi
editors’ preface
these clauses are going to work as intended at the critical moment when they are
required.5 This requires clarity in drafting. It also requires the parties to anticipate
any objections that the legislature or the courts may have to the use of such clauses,
and to use the contractual wording to ensure that they comply with legislative or
judicial barriers to enforcing the clause as drafted.
Why should such legislative or judicial barriers exist? Because, in addition to being
important, these clauses are controversial. The knock-for-knock indemnity is a radical
instrument. These clauses do not tinker at the periphery of liability by, e.g., imposing
a mere limitation of liability for economic damages payable under contract in the
event of defective performance of the work obligation. They involve a fundamental
re-allocation of risk, using the parties’ contract as a means of cutting across both
contractual and delictual liability, divorcing liability from fault and wedding it instead
to the parties’ conception of efciency, found in what one of the authors describes
as the identifcation principle.6 This may be efcient, but what is the cost of severing
the link with fault? Is it morally right? Is it safe?
Legislatures in some jurisdictions have taken specifc steps to prohibit or at least
control the use of knock-for-knock indemnities.7 And while other jurisdictions have
taken no specifc steps to regulate the use of these clauses, they may nevertheless
have to comply with the requirements of more general unfair contract legislation.8
And even in jurisdictions which adopt a relatively relaxed approach to legislation,
the courts may still play a signifcant role in how or if these clauses can be enforced.
Should a clause which reallocates liability in a way that has no regard to fault in
danger of being struck down as contrary to public policy? If so, is this rule triggered
when a party seeks to be excused from the consequences of its own negligence, or
only if it claims an indemnity against the consequences of its own gross negligence
or wilful misconduct? And what if the indemnifed and the indemnifer are both at
fault? Diferent jurisdictions might answer these questions in diferent ways. And
even jurisdictions in which the courts hold that there is no bright-line rule against
the enforceability of knock-for-knock indemnities may apply technical or strained
rules of interpretation to knock-for-knock indemnities, not prohibiting, but in a
sense limiting, the manner in which the clauses may be used. It is not entirely clear
if English law would at present take such an approach,9 but it plainly has in the
past, and this has left a legacy upon the way in which knock-for-knock indemnities
are drafted: the explanations, repeatedly found in knock-for-knock indemnities,10
that the clause is intended to provide an indemnity against the consequences of the
5 Or at least, the party that, on the facts of the case, has a claim under the indemnity will want to be sure
that it will work as intended. As we shall see, the party called upon to pay under the indemnity (and perhaps
more particularly, its insurers) may feel it has a strong fnancial incentive to go looking for problems with
the drafting, meaning that the indemnity cannot be used against it.
6 See Chapter 2 of this book, p. 27.
7 See, e.g., the discussion of the principal US oil and gas jurisdictions contained in Chapter 12 of this
book.
8 See, e.g., the discussion of the Unfair Contract Terms Act 1977 in Chapter 2 of this book, p. 34.
9 Arnold v Britton [2015] UKSC 36, discussed further in Chapter 2 of this book, p. 40.
10 See, e.g., LOGIC, Ofshore Decommissioning, ibid. n. 2, Cl 22.7: indemnities to apply “irrespective
of cause and notwithstanding the negligence or breach of duty (whether statutory or otherwise) of the
indemnifed party”.
xxii
editors’ preface
indemnifed party’s own negligence are not there by accident, but in response to
mishaps in previous cases.11
The editors hope that this brief introductory preface has demonstrated that the
knock-for-knock indemnity clause gives rise to a broad range of signifcant legal
questions that merit careful thought. Throughout the remainder of this work, these
questions will be addressed from a variety of perspectives. The majority of the chapters
address the questions from the perspective of a particular jurisdiction, and there is
coverage of a broad and representative sample of jurisdictions drawn from around
the world. Other chapters are concerned with public international law and legal his-
tory. However, it is perhaps fair to say that it is the law and economics perspective
adopted by Parchomovsky and Stavang that lies at the heart of this book. It was
this paper, frst discussed at a round-table event hosted by the Institute of Advanced
Legal Studies in 2016, and the discussion which followed it, which gave the editors
confdence that the humble knock-for-knock indemnity could sustain a full edited
collection. We hope the reader will agree that that confdence was not misplaced.
The editors would also like to thank our editorial assistants, Tulane Law students
Andrew J. Pomaville and Jack C. Foster.
Kristofer Svendsen, Endre Stavang and Greg Gordon
11 E. E. Caledonia v Orbit Valve [1995] 1 All ER 174.
xxiii
TAB L E OF CA SES
France
Cass. 1e civ, 2 août 1951, JCP 1951, II, 6592 ......................................................................245
Cass. 1e civ. 15 avril 1961, JurisData no. 1961–000198........................................................249
Cass. 3e civ., 2 mars 2010, no. 09–12.272.............................................................................247
Cass. 1e civ., 31 mars 1987, no. 79–17.129 ...........................................................................248
Cass. Ch. mixte, 22 avr. 2005, no. 03–14112 ........................................................................248
Cass. Civ, 3, 27 juin 2001, no. 99–21.017 99–21.284 ............................................................248
Cass. Civ. 2e, 19 November 1964 .........................................................................................242
Cass. Civ, 2, 28 février 2013, no. 12–12.813 .........................................................................248
Cass. Civ, 2, 10 novembre 2021, no. 19–12.660....................................................................248
Cass. Com, 29 juin 2010, no. 09–11841 ...............................................................................248
Cass. Com. 24 October 2018, no. 17–25.672 ........................................................................242
Cass. Req. 19 janv. 1863, DP1863, 1, p. 248 ........................................................................246
Civ. 3e, 10 January 1978, no. 76–11.111...............................................................................249
Cour de Cassation, 11 January 1989, 86–17.323 ..................................................................242
Germany
BAG, judgement dated 12. 11. 1998–8 AZR 221–97 = NJW 1999, 966 .............................201
BeckOGK/Mäsch, 1.1.2021, BGB § 328 Rn. 123 ................................................................201
Beck OK BGB/Janoschek, 58. Ed. 1.5.2021 Rn. 5, BGB § 328 Rn. 5 ................................201
Beck OK BGB/H. Schmidt, 58. Ed. 1.5.2021 Rn. 6, BGB § 307 Rn. 96 ............................196
Beck OGK/Schaub,1.3.2021,BGB§276Rn.4 .........................................................................200
BGH, judgement dated 13. 11. 1953 – I ZR 140/52 _ NJW 1954, 229 ...............................200
BGH, judgment dated 20.3.2014 – VII ZR 248/13 = NJW 2014, 1725 ..............................197
BGH, judgment dated 20.3.2014 – VII ZR 248/13 = NJW 2014, 1725 ..............................196
BGH order dated 20.11.2012 – VIII ZR 137/12, BeckRS 2013, 5597 Rn. 7 .......................196
BGH, order dated 19.3.2019 – XI ZR 9/18, NJW 2019, 2080 ............................................196
BGH, judgement dated 12.11.1980 – VIII ZR 293/79 = NJW 1981, 275 ...........................201
BGH, judgement dated 02.10.1985 – IV a ZR 18/84 = NJW 1986, 1100 ...........................201
BGH, judgement dated 09.07.1992 – VII ZR 7/92 = NJW 1992, 3158 ...............................200
BGH, judgement dated 18.12.1996 – IV ZR 321/95 = NJW 1997, 1012 ............................201
BGH, judgement dated 20. 11. 2012 – VI ZR 268/11 = NJW-RR 2013, 550 .....................200
BVerfG, judgement dated 23. 04. 1986–2 BvR 487/80 = NJW 1987, 827 ...........................201
HK-BGB/Reiner Schulze, 10. Auf. 2019, BGB § 276 Rn. 6 ...............................................200
Jauernig/Stadler, 18. Auf. 2021, BGB § 276 Rn. 1 ..............................................................200
xxv
table of cases
MüKoBGB/Basedow, 8. Auf. 2019, BGB § 305 Rn. 35, § 310 Rn. 11 ...............................196
MüKoBGB/Grundmann, 8. Auf. 2019, BGB § 276 Rn. 1–2 ..............................................200
MüKoBGB/Gottwald, 8. Auf. 2019, BGB § 328 Rn. 261 ...................................................201
MüKoBGB/Wagner, 8. Auf. 2020, BGB § 823 Rn. 47 ........................................................200
Smit International (Deutschland) GmbH v Josef Mobius
Baugesellschaft GmbH [2001] CLC 1545 ..................................................................... xxi
Wolf/Lindacher/Pfeifer/Pfeifer, 7. Auf. 2020 Rn. 36, BGB § 305 Rn. 36 ..........................196
Malaysia
American International Assurance Co Ltd v Koh Yen Bee (f) [2002] 4 MLJ 301 ......254, 290
Amman Singh v Vasudevan [1972] 1 LNS 7 ........................................................................255
Anderson v Fitzgerald (1853) 4 HLC 484; (1853) 10 ER 551 .............................................258
Ayer Hitam Tin Dredging Malaysia Bhd v YC Chin Enterprises Sdn Bhd [1994]
2 MLJ 754, 755 .............................................................................................................257
Azman bin Mahmood & Another v SJ Securities Sdn Bhd [2012] MLJU 660...................260
Chin Hooi Nan v Comprehensive Auto Restoration Service Sdn Bhd [1995]
2 MLJ 100.....................................................................................................................258
CIMB Bank Bhd v Maybank Trustees Bhd [2014] MLJU 117 ...........................................257
Fui Lian Credit & Leasing Sdn Bhd v Kim Leong Timber Sdn Bhd[1991] 1
CLJ 522.........................................................................................................................253
Hotel Anika Sdn Bhd v Majlis Daerah Kluang Utara [2007] 1 MLJ 248 ..........................258
Lee Kam Wah v Associated Asian Securities (Pte) Ltd [1991] 3 MLJ 286 .........................257
Macon Works & Trading Sdn Bhd v Phang Hon Chin & Anor [1976]
2 MLJ 177, HC.............................................................................................................255
Ooi Boon Leong v Citibank N.A. [1984] 1 MLJ 222 ..................................................252, 272
Polygram Records Sdn Bhd v The Search [1994] 3 MLJ 127 ..............................................254
Saad Marwi v Chan Hwan Hua & Anor [2001] 3 CLJ 98 ..........................................254, 290
Sabah Shell Petroleum Co Ltd & Anor v The Owners of and/or any
other Persons Interested in The Ship or Vessel (The Borcos Takdir)
[2012] 5 MLJ 515, 565 ..........................................................................................256, 257
Sekawan Guards Sdn Bhd v Thong Guan Sdn Bhd [1995] 1 MLJ 811 ...................... 259–261
Siong Electronic Industries (1981) Sdn Bhd v Sanyo Sales & Service
Sdn Bhd [1997] MLJU 162 ...........................................................................................260
South East Asia Insurance Bhd v Nasir Ibrahim [1992] 2 MLJ 355; [1993]
1 SCR 89, SC ................................................................................................................255
Syarikat Uniweld Trading v The Asia Insurance Co Ltd [1996] 2 MLJ 160 ......................258
Sze Hai Tong Bank Ltd v Rambler Cycle Co Ltd [1959] MLJ 200 ....................................259
Norway
Arbitral Award 20 August 1973 (referred by Knudtzon in Nordisk
Skibsrederforening Medlemsblad Spesialnummer A 1984, s. 66) .................................141
Arbitral Award “Mørland 7” (ND 1988 p. 263 f.) .......................................................142, 146
Case 2013–06–25. LG-2012–077280 .....................................................................................158
“Chainco” judgement (ND 1991 p. 180 ...............................................................................146
Rt-2004–1545, para. 52 (Norwegian Supreme Court)..........................................................137
Supreme Court decision in Rt. 1994 p. 626 .........................................................................158
“Trans Tind” Arbitral Award (ND 1984 p.404) ...........................................................144, 145
xxvi
table of cases
United Kingdom
A Turtle Ofshore SA v Superior Trading Inc [2008] EWHC 3034 (Admlty)
[2008] 2 CLC 953 ........................................................................................................... 22
Alderslade v Hendon Laundry [1945] KB 189 ..................................................................... 14
Addax v Acadia [2000] 1 Lloyd’s Rep 493 .............................................................................. 8
Ailsa Craig Fishing Co Ltd v Malvern Fishing Co Ltd [1983] 1 WLR 964;
[1981] UKHL 12 (1981) ............................................................................................. 9, 45
Andar Transport Pty Ltd v Brambles Ltd (2004) 217 CLR 424 .......................................... 13
Arnold v Britton [2015] UKSC 36 ...................................................xxii, 37, 38, 40, 45, 48, 60
BHP Petroleum Ltd v British Steel plc and Dalmine SpA [2000] 2
Lloyd’s Rep 277.........................................................................................................59, 61
British Sugar P.l.c. v N.E.I. Power Projects Ltd (1997) 87 BLR 42; (1998) 87
BLR 42 CA ................................................................................................................ 8, 60
Caledonia North Sea Ltd v London Bridge Engineering Ltd
(London Bridge) [2000] SLT 1123 (Ct of Sess) ............................ 6, 18, 24, 27–29, 33, 35,
38, 41, 43, 44, 46, 48, 49, 59–61
Caledonia North Sea Ltd. v British Telecommunications Plc; Same v Kelvin
International Services Ltd.; Same v London Bridge Engineering Ltd.;
Same v Norton (No. 2) Ltd (In Liquidation); Same v Pickup No. 7 Ltd.;
Same v Stena Ofshore Ltd. same v Wood Group Engineering Contractors
Ltd. –[2002] UKHL 32; [2002] 1 Lloyd’s Rep 553; [2002] SC (HL)
117; [2002] 1 All ER (Comm) 321 ......................................................... 27–30, 33, 35, 38,
43, 44, 46, 48, 49, 60, 61, 110, 270
Caledonia North Sea Ltd v London Bridge Engineering Ltd and Others
(The Piper Alpha) [2002] UKHL 4, 32; [2002] 1 Lloyd’s Rep
553, 562, 563 (HL) .......................................................................... 3, 4, 9, 10, 29, 35, 61,
172, 277
Campbell v Conoco (UK) Ltd [2003] 1 All ER (Comm) 35 .....................................29, 46, 47
Canada Steamship Lines v The King (Regem) [1952] 1 All ER 305; [1952]
AC 192 .........................................................................................................14, 22, 41– 45
Castellain v Preston (1883) 11 QBD 380 .............................................................................. 21
Chartbrook Ltd v Persimmon Homes Ltd [2009] UKHL 38; [2009] 1 AC 1101 ................. 40
Croudace v Cawoods [1978] 2 Lloyd’s Rep 55 ........................................................................ 8
Currie v Misa (1875) LR 10 Exch 153 .................................................................................255
Czamikow v Koufos [1969] 1 A.C. 350 ................................................................................... 7
Deepak Fertilisers and Petrochemicals Corporation v I.C.I. Chemicals & Polymers Ltd
(1999) 1 TCLR 200 ........................................................................................................ 60
Donoghue v Stevenson [1932] AC 562 .................................................................................. 36
E. E. Caledonia Ltd (formerly Occidental Petroleum (Caledonia) Ltd) v
Orbit Valve Co Europe Plc [1995] 1 All ER 174, afrming [1994] 1 WLR
221; [1993] 4 All ER 165; [1993] 2 Lloyd’s Rep. 418 QBD
(Commercial Court)...................................................................... xxiii, 14, 18, 41–44, 46,
47, 109
Farstad Supply A/S v Enviroco Ltd [2011] UKSC 16; (2011) 15
EdinLR; [2010] UKSC 18; 2010 SCLR 379; (2010) 14 Edin LR 102 ......... 21–23, 27, 31,
34–36, 49, 110
Firma C-Trade SA v Newcastle Protection and Indemnity Association
(The Fanti) (No 2) [1991] 2 AC 1 (HL) ......................................................................... 22
Gillespie Bros & Co Ltd v Roy Bowles Transport Ltd [1973] QB 400 ................................. 41
xxvii
table of cases
Glencore Energy U.K. Ltd v Cirrus Oil Services Ltd [2014] EWHC 87 (Comm)................ 60
Glen’s Traders v Lancashire and Yorkshire Accident Insurance Co Ltd
(1906) 8 F 915................................................................................................................ 37
Hadley v Baxendale (1854) 9 Exch 341 ...........................................................7, 8, 60, 61, 109
Hart v O’Connor [1985] A.C. 1000 ..............................................................................253, 254
HIH Casualty & General Insurance Ltd v Chase Manhattan Bank [2003]
UKHL 6 [2003] 2 Lloyd’s Rep 297 ..................................................................... 39, 44–46
Hinton v Dibbin (1842) 2 QB 253 .......................................................................................109
Hollier v Rambler Motors (AMC) Ltd [1972] 2 QB 71 ....................................................... 45
Hotel Services Ltd v Hilton International Ltd [2000] 1 All E.R. (Comm.) 750 ................... 60
ICS. See Investors Compensation Scheme Ltd. v West Bromwich Building Society—
Investors Compensation Scheme Ltd. v West Bromwich Building Society
[1997] UKHL 28; [1998] 1 WLR 896 .......................................................................37, 39
Jodrell, Re (1890) 44 Ch D 590............................................................................................. 37
Karsales (Harrow) Ltd v Wallis [1956] 2 All ER 866, 869 ..........................................257, 258
London Bridge. See Caledonia North Sea Ltd—
London Bridge & Others British Telecommunications plc. See Caledonia
North Sea Ltd. v British Telecommunications Plc—
Lloyds Bank v Bundy [1975] QB 326 ...................................................................................254
Markerstudy Insurance Co Ltd v Endsleigh Insurance Services Ltd [2010]
EWHC 281 (Comm) ...................................................................................................... 60
McCain Foods G.B. Ltd v Eco-Tec (Europe) Ltd [2011] EWHC 66 (TCC) ........................ 60
Millar’s Machinery Co v David Way & Son (1935) 40 Com. Cas. 204 ................................ 60
Motours Ltd v Eurobell (West Kent) Ltd [2003] EWHC 614 (QB) ..................................... 59
Multiservice Bookbinding Ltd v Marden [1979] Ch. 84 ......................................................253
Nelson v Atlantic Power and Gas Ltd, 1995 SLT 102 .................................................... 41–43
Pentecost and Anr. v London District Auditor & Anr [1951] 2 KB 759, 763.....................109
Photo Production Ltd v Securicor Transport Ltd [1980] AC 827 ........................................ 59
Polypearl Ltd v E. On Energy Solutions Ltd [2014] EWHC 3045 (QB) .............................. 60
Prenn v Simmonds [1971] 1 WLR 1381 ................................................................................ 39
Reardon Smith Line Ltd v Hansen-Tangen [1976] 1 WLR 989 ........................................... 39
Robb v Salamis M & I Ltd, 2007 SLT 158 ......................................................................46, 47
Saga Cruises BDF Limited & anor v Fincantieri SPA [2016] EWHC 1875 (Comm) .......... 60
Saint Line v Richardsons [1940] 2 KB 99............................................................................. 60
Scottish & Newcastle plc v G D Construction (St Albans) Ltd [2003] EWCA Civ 16 ........ 22
Scottish Power UK plc v BP Exploration and Drilling [2015] EWHC 2658 ........................ 60
Slessor v Vetco Gray, unreported, 7 July 2006, (Ct of Sess, Outer
House); [2007] SLT 400........................................................................... 25, 29, 38, 46–48
Smith v South Wales Switchgear Co Ltd [1978] 1 WLR 165 ..........................................45, 46
Smith v UMB Chrysler (Scotland) Ltd 1978 SC (HL) 1 ...................................................... 41
Starsin, The [2003] UKHL 12; [2004] 1 AC 715 ................................................................... 40
Thompson v T Lohan (Plant Hire) Ltd [1987] 1 WLR 649 ................................................. 35
Transocean Drilling UK Ltd v Providence Resources plc [2016]
EWCA Civ 372 ...................................................................................................16, 17, 60
Victoria Laundry v Newman [1949] 2 KB 528 ....................................................................... 7
Walters v Whessoe Ltd and Shell Refning Co Ltd [1968] 2 All ER (All England
Law Reports) 816 ........................................................................................................... 14
Wessanen Foods Ltd v Jofson Ltd [2006] EWHC 1325 (TCC) ............................................ 59
WesternGeco Ltd v ATP Oil and Gas (UK) Ltd [2006] EWHC 1164 (Comm) ................... 63
xxviii
table of cases
United States of America
Am. Home Assurance Co. v Chevron, Inc 400 F.3d 265, 269 (5th Cir. 2005) ....................113
A.M.C. Lifeboats Inc. v Apache Corp., 2008 WL 217177 (E.D. La. 2008) ........................105
Amoco Prod. Co. v Lexington Ins. Co., 98–1676 (La. App. 1 Cir. 9/24/99); 745
So. 2d 676, writ denied, 99–3553 (La. 2/25/00); 755 So. 2d 253 ...................................231
Baker v Director, Ofce of Workers’ Compensation Programs 834 F.3d 542
(5th Cir. 2016) .......................................................................................................222, 223
Barker v Hercules Ofshore, Inc., 713 F.3d 208, 2013 AMC 946 (5th Cir. 2013) ........222, 223
Barnett v Am. Constr. Hoist, Inc., 2011 1261 (La. App. 1 Cir. 02/10/12);
91 So. 3d 345 ..........................................................................................Appendix A, 236
Barrios v Centaur, LLC, 942 F.3d 670 (5th Cir. 2019) ................................................220, 221
BW Ofshore USA, LLC v TVT Ofshore AS, Case No. 14–1052, 2015 U.S.
Dist. LEXIS 153840 (E.D. La. Nov. 13, 2015) .....................................................222, 223
Champagne v Tetra Applied Technologies, Inc., No. 05–299, 2006 U.S. Dist.
LEXIS 37457, *13–14 (S.D. Tex. Feb. 6, 2006) ............................................................219
Chevron Oil Co. v Huson, 404 U.S. 97 (1971) .............................................................218, 219
Cimarex Energy Co. v CP Well Testing, L.L.C., 26 F.4th 683 (5th Cir. 2022) ....................104
Cleere Drilling Co. v Dominion Expl. & Prod., Inc., 351 F.3d 642, 647
(5th Cir. 2003) ...............................................................................................................233
Daughdrill v Ocean Drilling & Exploration., 665 F. Supp. 477, 481, 482 (E.D.
La. 1987) ...............................................................................................................112, 225
De Lovio v Boit, 7 Fed. Cas. 418 (C.C.D. Mass. 1815) (No. 3776) ....................................216
Deepwater Horizon, Re, 470 S.W.3d 452 (Tex. 2015), opinion after certifed
question answered, No. 12–30230, 2015 WL 13918242 (5th Cir. June 9, 2015) ...........103
Demette v Falcon Drilling Co., 280 F.3d 492, 2002 AMC 686 (5th Cir. 2002) ...........222, 223
Emery Waterhouse Co. v Lea, 467 A.2d 986 (Me. 1983) ..............................Appendix A, 236
Empire HealthChoice Assur., Inc. v McVeigh, 547 U.S. 677, 689–690 (2006) ....................215
Engel v Davenport, 271 U.S. 33, 37 (1926) ..........................................................................215
Erie R.R. Co. v Tompkins, 304 U.S. 64 (1938) ....................................................................216
Exxon Corp. v Cent. Gulf Lines, Inc., 500 U.S. 603 (1991) ................................................216
Fields v Pool Ofshore, 182 F.3d 353 (5th Cir. 1999) ...........................................................222
Fontenot v Mesa Petroleum Co, 791 F.2d 1207. (5th Cir. 1986); 493 A.2d 946 (Del
Sup Ct. 1985) ................................................................................................................. 29
Gulf Ofshore Co., Div. of Pool Co. v Mobil Oil Corp., 453 U.S. 473, 478 (1981) ............215
Gunka v Consolidated Papers, Inc., 508 N.W.2d 426 (Wis. Ct. App.
1993)........................................................................................................Appendix A, 237
Horizon, Re, 470 S.W.3d 452 (Tex. 2015).............................................................................229
Houston Exploration Co. v Halliburton Energy Services Inc., 269 F.3d 528
(5th Cir. 2001) ...............................................................................................................112
J.A.R., Inc. v M/V LADY LUCILLE, 963 F.2d 96 (5th Cir. 1992) ....................................216
Ken Petr. Corp. v Questor Drilling Corp., 24 S.W.3d 344, 346 (Tex. 2000) ........................233
Klaxon Co. v Stentor Elec. Mfg. Co., Inc., 313 U.S. 487 (1941).................................216, 219
Kokkonen v Guardian Life Ins. Co. of Am., 511 U.S. 375, 377 (1994) ..............................215
Kossick v United Fruit Co., 365 U.S. 731, 735 (1961) .......................................................216
Laredo Ofshore Constructors, Inc. v Hunt Oil Co., 754 F.2d 1223, 1230
(5th Cir. 1985) ...............................................................................................................219
Larry Doiron, Inc., Re 879 F.3d 568 (5th Cir. 2018) (en banc), cert. denied,
138 S. Ct. 2033 (2018) ................................................................................... 219–221, 223
xxix
table of cases
Lightering LLC v Teichman Group, LLC, 328 F. Supp. 3d 625, 627–629
(S.D. Tex. 2018).............................................................................................................220
Livings v Service Truck Lines of Texas, Inc., 467 So. 2d 595, 599
(La. App. 3 Cir. 1985) ...................................................................................................230
Lloyds of London v Transcontinental Gas Pipe Line Corp., 38 F.3d 193, 196 (5th
Cir. 1994) ......................................................................................................................230
Longmire v Sea Drilling Corp., 610 F.2d 1342 (5th Cir. 1980) ..........................................227
Lozman v City of Riviera Beach, 568 U.S. 115, 121 (2013) ........................................221, 222
Marcel v Placid Oil Co., 11 F.3d 563, 569–570 (5th Cir. 1994) ............. 231, Appendix A, 236
Martin v Fab-Con, Inc., 9 F. Supp. 3d 642, 649 (E.D. La. 2014) .......................................222
Matte v Zapata Ofshore Co., 784 F.2d 628 (5th Cir. 1986) ................................................218
Meloy v Conoco, Inc., 504 So. 2d 833 (La. 1987) ...............................................................230
Mendez v Anadarko Pet. Corp. 466 F. App’x 316 (5th Cir. 2012) ......................................222
Merrell Dow Pharmaceuticals, Inc. v Thompson, 478 U.S. 804, 808 (1986) .......................215
Mobil Oil Corp. v Schlumberger, 598 So. 2d 1341 (Ala. 1992) .....................Appendix A, 235
Mooney v W&T Ofshore, Inc., 2013 AMC 1480 (E.D. La. 2013) .............................222, 223
M/V Lady Lucille, 963 F.2d at 98 ........................................................................................219
Norfolk Southern Ry. Co. v James N. Kirby, Pty Ltd., 543 U.S. 14, 25 (2004)
(citing Exxon Corp. v Cent. Gulf Lines, Inc., 500 U.S. 603 (1991)) ...................216, 219,
220
Ofshore Co. v Robinson, 266 F.2d 769 (5th Cir. 1959)...............................................222, 223
Ofshore Logistics, Inc. v Tallentire, 477 U.S. 207, 222–223 (1986) .....................................215
Ogea v Lofand Bros., 622 F.2d 186 (5th Cir. 1980) ...................................................227, 228
Oil Spill by the Oil Rig “Deepwater Horizon,” Re, 841 F. Supp. 2d at 988,
1000, 1003–4, 6 (E.D. La. 2012) ................................................................... 111, 224, 225
Oil Spill by the Oil Rig “Deepwater Horizon,” Re, 745 F.3d 157 (5th Cir. 2014) .......222, 223
Owen Equipment & Erection Co. v Kroger, 437 U.S. 365, 374 (1978) ................................216
Riley v Alexander/Ryan Marine Servs. Co., 983 F.Supp.2d 884, 887 (S.D.Tex.2013) ..........221
Roberts v Energy Development Corp., 104 F.3d 782, 784 (5th Cir. 1997) ..................219, 230
Rupp v Am. Crystal Sugar Co., 465 N.W.2d 614 (N.D. 1991) ......................Appendix A, 237
Ryan Stevedoring Co. v Pan Atlantic Steamship Corp., 350 U.S. 124 (1956) .....................227
Seas Shipping Co. v Sieracki, 328 U.S. 85 (1946) ................................................................227
Stewart v Dutra Constr. Co., 543 U.S. 481, 489 (2005) (citing 1 U.S.C. § 3) ......................221
Stewart v Grand Isle Shipyard, Inc., Case No. 10–4016, 2011 U.S. Dist. Lexis
148603 (E.D. La. December 23, 2011) ..........................................................................224
Storage & Processors, Inc. v Reyes, 134 S.W.3d 190, 192 (Tex. 2004) .................................233
Tateosian v Vermont, 945 A.2d 833 (Vt. 2007)..............................................Appendix A, 237
Texas Co. v Gianfala, 222 F.2d 382 (5th Cir.), reversed on other grounds, 350
U.S. 879 (1955) ......................................................................................................222, 223
Theriot v Bay Drilling Corp., 783 F.2d 527 (5th Cir. 1986) ................................................224
Tullier v Halliburton Geophysical Services, Inc., 81 F.3d 552, 1996 AMC 1561
(5th Cir. 1996) ...............................................................................................................228
Union Pac. Resources Co. v Dolenc, 86 P.3d 1287 (Wyo. 2004) ..........................Appendix A,
237
Union Texas Petroleum Corp. v PLT Engineering, Inc., 895 F.2d 1043, 1050
(5th Cir. 1990) .......................................................................................................218, 223
Verdin v ENSCO Ofshore Co., 104 F. Supp. 2d 682, 690 (W.D. La. 2000), af’d,
255 F.3d 246 (5th Cir. 2001) .........................................................................................230
Voisin v O.D.E.C.O. Drilling Co., 744 F.2d 1174, 1175, 1176 & n. 1, 1177, 1179
(5th Cir. 1984) .......................................................................................................226, 227
xxx
table of cases
Walter Oil & Gas Corp. v NS Group, Inc., 867 F. Supp. 549, 553 (S.D. Tex. 1994) ..........218
Warrior Energy Servs. Corp. v ATP Titan, 551 F. App’x 749 (5th Cir. 2014) ............222, 223
Wilburn Boat v Fireman’s Fund Ins., 348 U.S. 310, 1955 AMC 467 (1955) .......................223
Other jurisdictions
Brazillian Special Appeal No. 1.076.465–SP (2008/0160567–4), published on 21
November 2018 (Brazil) ................................................................................................115
Darlington Futures Ltd v Delco Australia Pty Ltd (1986) 161 CLR 500 ............................ 13
Guardian Assurance Co Ltd v Condogianis (1919) 26 CLR 231 ........................................258
xxxi
TA B LE OF S TAT U T E S A N D MO D EL
F ORM C ON TR AC TS
Brazil Civil Code (Napoleonic Code)

Civil Code (BCC) ...........11, 115, 207, 210 1804 ...........................................239, 240
Arts. 186, 187 ................................ 115 Book III......................................... 240
Arts. 393, 398, 402, 493 ................. 206 Title III .......................................... 239
Art. 432 ......................................... 115 Title IV .......................................... 240
Art. 475 ......................................... 206 Title V ........................................... 239
Art. 884 ......................................... 207 Art. 16–1........................................ 248
Art. 927 ................................... 11, 207 Art. 1170.................................246, 247
Art. 931 ......................................... 207 Art. 1217 et seq ............................. 240
Code of Civil Procedure: Federal Arts. 1231–1, 1240–1244 ................ 242
Law no. 13.105 of 16 March Arts. 1382, 1383 ............................. 245
2015 .................................................. 207 Ordinance no. 2016–131 of 10 February
Art. 373, I ...................................... 207 2016, in force 1 October 2016 ......... 240,
Federal Constitution............................ 211 244, 246
Art. 5, II ........................................ 209 Code of Commerce
Federal Law no. 10.406 of 10 January Art. 133–1 ...................................... 247
2002 .................................................. 206 Law No. 2018–287 of 20 April 2018,
Hydrocarbons Law No.9.478 of 6 ratifying Ordinance no.
August1997....................................... 114 2016–131 .......................................... 240
Law No. 6,938/81 on Environmental Law no. 2019/035/AN of 4 July 2019
Matters ............................................. 207 (replacing the 1983 Civil Code) ....... 240
Law No. 8,078/1990 on Consumer Transportation Code
Rights ............................................... 207 Art. L. 5422–12, 15 ....................... 247
Law No. 13,874/2019 on the Validity Uniform Act on General Commercial
of Limitation of Liability Law .................................................. 241
provisions ......................................... 210 German Civil Code
New Civil Code (BCC) 2002 ............... 207 (BGB) ............ 12, 13, 194, 199, 200, 203
Pronouncement No. 444, V Jornada s. 276 .............................................. 200
de Direito Civil ................................ 208 s. 276(1), sent.1 .......................200, 201
s. 276(2) ......................................... 200
Denmark s. 276(3) ..................................200, 201
Danish Contract Act (Aftaleloven) s. 305 et seq ....................194, 195, 197
s. 36(1) ............................................. 15 s. 305(1) ......................................... 195
Danish Product Liability Act ................ 14 s. 305(1), sent.1 .............................. 195
s. 305(1), sent.3 .......................196, 197
France s. 307 .......................................195, 198
Amended law no. 63–62 of 10 July s. 307(1), (2) ................................... 195
1963 .................................................. 240 s. 308 ...............................195, 198, 199
xxxiii
table of statutes and model form contracts
s. 309.......................................195, 198 s. 36 (1983) ............ 119, 122, 134–130,
s. 309(7) ..................................199, 199 137–140, 143, 145, 146,
s. 310(1), sent.2 .............................. 195 148–150, 176, 178–181
s. 36, ND 1989.225 ........................ 180
Guinea s. 36, ND 1990.204 ........................ 179
Civil Code............................................ 240 s. 36, ND 1991.180 ........................ 180
s. 36, ND 2000.240 ........................ 178
Malaysia s. 36, NL 5–1–1 ............................. 176
Application of Laws Ordinance s. 36, NL 5–1–2 ...... 176–178, 180, 181
(Sabah) 1951 .................................... 252 s. 36, Rt-906.717 ............................ 176
Application of Law Ordinance s. 36, Rt-1994.626 .......................... 180
(Sarawak) 1949................................. 252 s. 36, Rt-2000.806 .......................... 179
Civil Law Act (formerly the Civil s. 36, Rt-2001.603 .......................... 179
Law Ordinance last revised 1972) s. 36, Rt-2002.1155 ........................ 179
1956.................................................. 252 s. 36, Rt-2008.969 .......................... 179
ss. 3, 5 .....................................252, 290 s. 36, Rt-2010.1345 ........................ 179
Civil Law Ordinance 1956 ................... 252 s. 36, Rt-2013.388 .......................... 179
Consumer Protection (Amendment) s. 36, Rt-2013.769 .......................... 179
Act 2010 (2013) 1 MLJ 1................. 253 s. 36, Rt-2014.351 .......................... 179
Contract Act 1950 ............................... 289 s. 36, Rt-1915.840 .......................... 176
Part III........................................... 253 s. 36, Rt-1926.712 .......................... 176
ss. 77, 78 .........................255, 260, 289 s. 36, Rt-1948.370 .......................... 176
Federal Constitution.....................251, 252 s. 36, Rt-1948–3700 ....................... 138
Arts. 4, 74, 121 .............................. 252 s. 36, Rt-1994–626,
Oilfeld Anti-Indemnity Act’ (2020)..... 256 (“quay- inspector”)..... 138, 141, 146,
Petroleum Act...................................... 274 176, 180
Straits Settlements (fnal version of s. 36, Rt-1959–849
the Civil Law Ordinance) 1878 ........ 251 (“the apprentice”) ...................... 141
s. 36, Rt-1999.922 .......................... 178
Mali s. 36, U 1993.851 ........................... 180
Malian law no. 87–31/AN-RM of 29 s. 36, U 2005.................................. 180
August 1987 establishing the General s. 36, U 2005.243, 632, 243............ 180
Regime on Obligations..................... 240 s. 36, U 2006.632 ........................... 180
Norwegian Contracts Act , 1989—
Norway s.784 ............................................... 176
Compensation for Damages Act Damages Compensation Act (Lov
(relating to compensation in om skadeserstatning 13 June 1969
certain circumstances; 13 June Nr. 26).............................................. 184
1969 No. 26) .............................165, 185 Fabrication Contract 1987
s. 2–1.......................................130, 165 (NF 87) .............................120, 128, 189
Construction All Risks Fabrication Contract 2007 (NF 07)..... 191
(CAR) ...................... 132, 133, 137, 140, Art. 1.10 ........................................ 190
149, 155,157, 159–163 Fabrication Contract 2015
s. 29(2) ........................................... 149 (NFC 15)..........................120, 153, 164,
Sch. B .....................................161, 162 188, 190
Contracts Act , 31 May 1918, Part 1—
No. 4 .........................................150, 176 Art. 1.14 ........................................ 167
s. 4–9.............................................. 150 Art. 1.16 ........................................ 168
s. 4–9(2) ......................................... 150 Art. 1.22 ........................................ 167
xxxiv
Art. 1.26 ........................................ 168 s. 10–1(1)........................................ 132
Art. 23 ........................................... 190 s. 10–8 ............................................ 169
Art. 23.5 ........................................ 191 s. 10–9............. 154, 165, 167, 170, 183
Art. 30 ............................189, 190, 192 Petroleum Regulations (FOR-1997–06–
Art. 30.1 ......................... 167, 189–191 27–653) ......................................132, 154
Art. 30.2 ................. 167, 169, 189–191 s. 72................................................ 154
Art. 30.2(2) .................................... 169 s. 73.................................132, 154, 158
Art. 30.3 .........................169, 190, 191 s. 73(2) ........................................... 132
Art. 30.3(b) .................................... 191 Pollution Control Act 13 March
Art. 31.2(a), (b) ............................. 191 1981 Nr. 06 ...............................183, 184
Hagstrøm, Ansvarsfraskrivelse (TfR) 1996— §. 6 ................................................. 186
ss. 422–518 ..................................... 134 §. 53 ............................................... 183
Insurance Contracts Act (ICA) §. 61 ............................................... 186
(No. 69/1989) s. 4–9(1) ......149, 171, 177 Sale of Goods Act 1988
s. 1–3(1) ......................................... 166 s. 13 (1) .......................................... 157
s. 1–3(2)(c), (d)............................... 166 s. 23 (1) .......................................... 135
s. 4–9(2) ......................................... 177 Subsea Contract 2005 (NSC
s. 7–1.............................................. 171 05) ............................. 120, 153, 188–190
Maritime Act ....................................... 142 Art. 1(i) .......................................... 153
Maritime Code 1994 no. 39 Art. 23 ........................................... 190
(MC) .........................................165, 183 Art. 23.5 ........................................ 191
Ch. 10 ............................................ 185 Art. 30 ............................189, 190, 192
PA §. 7–3........................................ 165 Art. 30.1 ................................. 189–191
PA §. 7–4........................................ 166 Art. 30.2 ................................. 189–191
PA §. 7–4 (a) ................................. 166 Art. 30.3 .................................190, 191
MC §. 191 ...................................... 165 Art. 30.3(b) .................................... 191
MC §. 193 ...............................166, 187 Art. 31.2(a), (b) ............................. 191
MC §. 193 (c) ............................... 166 Tort Act
§. 507 ............................................. 183 s. 4–2 .............................................. 148
Norske Lov 5–1–2 ............................... 139 s. 4–3 .............................................. 148
Onshore Construction Standard) NS 8407— Total Contract 2015 (NTK 2007) ........ 124
cll. 40.3, 42.5.................................. 137 Art. 24.2(2), (3).............................. 124
Petroleum Act (repealed) 22 March Art. 25.4(1) .................................... 124
1985 Nr. 11 ...................................... 183 Appendix B.................................... 124
Petroleum Act 1996 no. 72 5MNOK ........................................ 131
(PA) ........................... 154, 165, 183–186 Total Contract 2015 (NTK 15)........... 120,
Ch. 7 ....................... 182, 183, 185, 192 128, 131, 132, 134–136, 139,
Ch. 8 .......................................183, 185 145, 146, 151–154, 159, 188
s. 1–6a, d.................................185, 186 Arts. 1.12, 14, 15 ........................... 153
s. 5–4 .............................................. 183 Art. 1.16 ........................................ 156
s. 7–1 ...............................183, 185, 186 Art. 1.21 ........................................ 156
s. 7–2 .............................................. 185 Arts. 1.23, 1.24 .............................. 153
s. 7–3 .......................................184, 186 Art. 1.26 ........................................ 130
s. 7–4 ...............................186, 187, 191 Arts. 2, 3 ........................................ 120
s. 7–5 ....................... 187, 188, 191, 192 Art. 3.1 ...................................142, 158
s. 7–5(c) .......................................... 186 Art. 8.1 .......................................... 123
s. 7–7 .............................................. 184 Art. 8.4 .......................................... 158
s. 7–8 .............................................. 184 Art. 8.5 .......................................... 127
s. 8–1 .......................................183, 185 Art. 8.5b ........................................ 131
xxxv
Arts. 9.2, 14.1 ................................ 123 Art. 30 ............................189, 190, 192
Art. 15 ........................................... 131 Arts. 30.1, 2 ............................ 189–191
Art. 15.1 ........................................ 124 Art. 30.3 .................................190, 191
Art. 16.1(2) .................................... 123 Art. 30.3(b) .................................... 191
Arts. 17.1, 3, 5 ............................... 121 Art. 31.2(a), (b) ............................. 191
Art. 19.1 ........................................ 157 Total Contract (Modifcation; 2015
Art. 22.1 .................................123, 149 (NTK (MOD) 15) ............................ 120
Art. 22.2 ........................................ 123 WELCAR 2001 ................................... 160
Art. 23 ........................................... 190 WELCAR 2012 ............................140, 155
Art. 23.5 ........................................ 191
Art. 24.2 .........................122, 124, 137 Senegal
Art. 24.3 .................................121, 124 Senegalese Code of Civil and
Art. 24.4 ........................................ 125 Commercial Obligations................... 240
Art. 25.1, 2 .................................... 137
Art. 25.3 .................................121, 122 Sweden
Art. 25.3(2) .................................... 122 Swedish Contract Act .......................... 146
Art. 25.4 ........................................ 124
Art. 25.5 .................................122, 125 United Kingdom
Art. 26.2 .................................121, 123 Consumer Rights Act 2015 ............. 23, 34
Art. 27 ........................................... 121 Contract (Third Party Rights)
Art. 29 ............................128, 129, 156 (Scotland) Act 2017 ........................... 36
Art. 29.1 .........................125, 149, 157 Contracts (Rights of Third Parties)
Art. 29.1(1) .................................... 135 Act 1999 ....................................... 36, 55
Art. 29.2 ................. 125, 135, 157, 160 Law Reform (Miscellaneous
Art. 30 ............ 128, 143, 189, 190, 192 Provisions) (Scotland) Act 1940......... 48
Art. 30.1 ........ 131, 134, 146, 156, 158, s. 3 ................................................... 48
189–191 Marine Insurance Act 1906
Art. 30.1(1) .................................... 136 s. 1 ................................................... 21
Art. 30.2 ................ 131, 134, 156, 158, Unfair Contract Terms Act
189–191 (UCTA) 1977 ...........xxii, 23, 27, 34–36,
Art. 30.2(1) .................................... 136 59, 62, 63, 110,
Art. 30.3 ................ 131, 134, 156, 159, 255, 268
190, 191 s. 2(1) ......................................... 34, 35
Art. 30.3(b) .................................... 191 s. 2(2) ....................................... 35, 110
Art. 30.3(3) .................................... 136 s. 2(3) ............................................... 35
Art. 31 ........................................... 132 s. 2(4) ......................................... 23, 34
Art. 31.1 .................................158, 159 s. 2A............................................... 110
Art. 31.1(b), (c).............................. 159 s. 4 ................................................... 34
Art. 31.1(3) .................................... 148 s. 4.1................................................. 63
Art. 31.2 .................................132, 158 ss. 12, 16(1) ...................................... 34
Art. 31.2(a), (b) ............................. 191 s. 16(1)(a) ................................... 34, 35
Art. 31.2(3) .................................... 148 s. 16(1)(b), 17 ................................... 35
Art. 31.2(4) .................................... 125 s. 18.................................................. 34
Art. 32.1 .................................121, 125 Unfair Terms in Consumer Contracts
Art. 32.2 ........................................ 124 Regulations 1994, SI 1994/ 3159 ........ 59
Art. 34.2 ........................................ 123 Unfair Terms in Consumer Contracts
Art. 24.2 ........................................ 122 Regulations 1999, SI 1999/ 2083 ........ 59
Total Contract (Modifcation; 2015
(NTC (MOD) 15) ............................ 188 United States
Art. 23 ........................................... 190 Anti-Indemnity Statutes, by
Art. 23.5 ........................................ 191 State (Mostly Appendix A)....... 235–237
xxxvi
Alaska. Stat. §. 45.45.900..................... 235 New Hampshire. Rev. Stat. § 338-A:1,
Arizona. Rev. Stat. §§. 32–1159, 338-A:2 ............................................. 236
34–226, 41–2586 ............................... 235 New Jersey. Stat. Ann. §. 2A:40A-1..... 236
Arkansas. Code §. 4–56–104, 104(b), New Mexico. N.M. Stat. Ann.
§ 22–9–214, 214(b) ........................... 235 §. 56–7–1, 56–7–1(1) ......................... 236
California. Civ. Code §. 2782............... 235 New Mexico Oilfeld Anti-Indemnity
Colorado. Rev. Stat. §§. 13–50, 50.5, Act (NMOAIA), N.M. Stat. Ann.
5–102, 102(8)(a), 13–21–111.5, §. 56–7–2 ........................................ 229
111.5(6)(b) ........................................ 235 New York. Gen. Oblig. Law
Connecticut. Gen. Stat. §. 52–572k, §. 5–322.1, 1(1)–(2) .................113, 237
572k(a) ............................................. 235 North Carolina. Gen. Stat. Ann.
Delaware. Code. Ann. 6 §. 2704, §. 22B-1 ............................................ 237
2704(a) ............................................. 235 Ohio. Rev. Code Ann. §. 2305.31 ........ 237
Florida. Stat. Ann. §. 725.06, Oklahoma. Stat. title 15 §. 221 ............ 237
725.06(1)........................................... 235 Oregon. Rev. Stat. §. 30.140,
Georgia. Code Ann. §. 13–8–2(b) ........ 235 140(1)–(2) ......................................... 237
Hawaii. Rev. Stat. §. 431:10– 222......... 235 Rhode Island. Gen. Laws §. 6–34–1 .... 237
Idaho Code Ann. §. 29–114................. 235 South Carolina. Code Ann.
Illinois. 740 ILCS §. 35/1 ..................... 235 §. 32–2–10......................................... 237
Indiana. Code Ann. §. 26–2–5 ............. 235 South Dakota. Codifed Laws
Iowa. Code Ann. §. 537A.5 ................. 236 § 56–3–18.......................................... 237
Kansas. Stat. §. 16–121, 121(b) ............ 236 Tennessee. Code Ann.
Kentucky. Rev. Stat. §. 371.180............ 236 §. 62–6–123 ....................................... 237
Louisiana Civil Code Article Texas Construction Anti-Indemnity
2004 .................................................. 113 Act (TCAIA).................................... 234
Louisiana Oilfeld Indemnity Tex. Ins. Code §. 151.102........234, 237
Act (LOIA) La. Rev. Stat. ........ 229–231 Tex. Ins. Code §. 151.103........217, 237
R. S. §. 9:2780.................113, 229, 236 Tex. Ins. Code §. 151.104............... 234
R. S. §. 9:2780.1 ......................232, 236 Tex. Ins. Code §. 151.103.............. 234
R. S. §. 9:2780.1(A)(1) ................... 231 Tex. Ins. Code §. 151.105............... 234
R. S. §. 9:2780.1(A)(2) ................... 231 Texas Oilfeld Anti-Indemnity Act
R. S. §. 9:2780.B .....................113, 230 (TOAIA) 1973– 2001 IELTR
R. S. §. 9:2780.F............................. 113 18 ............................................... 34, 233
R. S. §. 9:2780.1(I) ..................232, 236 Tex. Civ. Prac. & Rem. Code
R. S. §. 9:3306................................ 231 §. 127.003 ................................... 233
R. S. §. 38:2216(G) (Public Tex. Civ. Prac. & Rem. Code
Contracts) .................................. 236 §. 127.005 ................................... 233
Maryland. Code Ann. §. 5–401, Tex. Civ. Prac. & Rem. Code
401(a)(1)–(2) ..................................... 236 §. 127.005(b), ............................. 233
Massachusetts. Gen. Laws Ann. Tex. Civ. Prac. & Rem. Code §.
ch. 149 § 29C ................................... 236 127.005(c) ................................... 233
Michigan. M.C.L.A. § 691.991 ............ 236 Tex. Civ. Prac. & Rem. Code §.
Minnesota. Stat. §§. 337.01, 337.02 ..... 236 127.001(4)................................... 114
Mississippi. Code. Ann. § 31– 5–41 ..... 236 Tex. Civ. Prac. & Rem. Code §.
Missouri. Rev. Stat. §. 434.100 ............ 236 127.002(5)................................... 114
Montana. MCA §. 28–2–211 ............... 236 Tex. Civ. Prac. & Rem. Code §.
Nebraska. Rev. Stat. §. 25–21, 187, 127.003 ....................................... 229
187(1) ............................................... 236 Tex. Civ. Prac. & Rem. Code
Nevada. N.R.S. §. AB 125, §. 2 §. 127.003 ............................113, 114
(applies to residential construction Tex. Civ. Prac. & Rem. Code §.
contracts) ........................................ 236 127.004(1), (2), (3)...................... 114
xxxvii
Utah. Code Ann. §. 13–8–1 ................. 237 Instruments and Maritime
Virginia. Code Ann. §. 11–4.1 ............. 237 Lien Act
Washington. Rev. Code. §. 4.24.115, 30 C.F.R. § 585.100 ....................... 217
115(1)(b) ........................................... 237
West Virginia. Code Ann. International Statutes and Model
§. 55–8–14......................................... 237 Form Contracts
Wyoming Oilfeld Anti-Indemnity BARGEHIRE .......................................... 4
Act (WOAIA), Wyo. BIMCO ................................ 2, 4, 5, 11, 12
Stat. §. 30–1–131 .....................229, 237 BIMCO WINDTIME ............12, 195, 197
Carriage of Goods by Sea Act, 46 U.S.C.S. BIMCO LOGIC .................................... 12
§. 30701 .......................................... 216 BIMCO Newbuildcon.......................... 120
Art. III........................................... 215 Conditions of Contract as a Model
Art. III, §. 2 ................................... 215 for an International Construction
Art. III, §. 2, cl. 1 .......................... 215 Contract’ (2011) 1 BMR 32 ............. 271
Clean Water Act (CWA) 33 U.S.C. ...... 94, cl. 4.18 ........................................... 286
224, 225 Convention on Civil Liability for Oil
§§. 1251 et seq. ............................... 224 Pollution Damage Resulting from
§. 1321(b)(3), (4), (7)...................... 225 Exploration for and Exploitation
Commercial Instruments and Maritime of Seabed Mineral Resources,
Lien Act (formerly Ship Mortgage London (1 May 1977) ...................... 182
Act 1920), 46 U.S.C ......................... 216 FIDIC Conditions of Contract for
§ 31301 et seq. ............................... 216 EPC Turnkey Projects
Constitution ......................................... 215 (Silver Book), 1999 .......................... 275
1 U.S.C. §. 3................................... 221 FIDIC Federation International
28 U.S.C. §. 1331 ........................... 215 Des Ingenieurs-Conseils. ..........195, 275,
28 U.S.C. §. 1332(a) ....................... 215 277, 279,
28 U.S.C. §. 1333 ........................... 215 286, 289
Energy Policy Act of 2005................... 217 cl. 17 .............................................. 288
Harter Act 1893, 46 U.S.C. ................. 216 cl. 17.1 ....................................277, 278
§§. 30104, 30105 ............................. 215 cl. 18.1 ........................................... 289
§§. 30701–30707 ............................. 216 cl. 18.3 ........................................... 278
Lands Act cl. 18.3(d)(i)–(iii) ............................ 278
§. 1333 (a)(2) .................................. 218 General Contracting Principles (FAIR
Longshore and Harbor Workers’ Principles)–International Marine
Compensation Act (LHWCA) Contractors Association
33 U.S.C ....................................226, 227 (IMCA) 2011............................275, 276,
§. 901 et seq. .................................. 226 281, 285
§. 905(b) ..................................226, 227 HEAVYCON ....................................... 4, 6
§. 905(c) ......................................... 227 2007 Revision ....................................... 7, 8
National Defense Authorization Act IADC ..................................................... 19
for 2021 ............................................ 217 IMHH Deed. See Mutual Indemnity and
Outer Continental Shelf Lands Hold Harmless Deed
Act (OCLSA), 43 U.S.C...........111, 214, IMCA General Contracting Principles
217–219, 223 (Rev. 1)
§. 4 ................................................. 227 (n199) ............................................. 282
§. 1331 et seq ................................. 214 (n628) ............................................. 285
§. 1333 (a)(1) ...........................111, 217 LOGIC............... 15, 19, 23, 24, 26–28, 33,
§. 1333(a)(2)(A) .............................. 217 45, 47, 49, 52–54, 57–59, 63,
Ship Mortgage Act of 1920. See 108, 109, 130, 195, 197, 275,
United States Commercial 276, 279, 282, 283, 285, 289
xxxviii
LOGIC Construction Contract, 2nd cl. 11 .......................................... 54, 55
Edition, 2003.................................... 275 cl. 23 ................................................ 59
cl. 1.2 ............................................. 283 Sch. 1 ............................................... 52
cl. 1.4 ............................................. 130 Sch. 2 ............................................... 54
cl. 1.9 ......................................130, 283 New York Anti-Indemnity Law........... 113
cl. 17(1)(a)...................................... 284 Nordic Countries Contract Act............. 15
cl. 18.3 ........................................... 283 s. 36 (General Clause)...................... 15
cl. 18.3(d)(i)–(iii) ............................ 283 Nordic Freight Forwarder
cl. 19 .......................................108, 109 Agreement (NSAB).......................... 180
cls. 19.1, 2 ...................................... 108 NSAB 1975, s. 25 .......................... 180
cl. 19.2c.......................................... 130 NSAB 2000, s. 22(5) ...................... 180
cls. 19.3, 4, 5, 6, 9, 10, 11.............. 108 NSAB 2015, s. 23 .......................... 180
cls. 20, 21.1 .................................... 109 Nordic Marine Insurance Plan
cl. 21.2 ........................................... 108 2013 Version 2019 (NP) ............166, 171
cl. 22 .............................................. 277 cl. 3–33 .......................................... 171
cl. 22.1 ........................................... 286 cl. 5–14 .......................................... 171
cl. 22.1(a), (b) ................................ 286 cls. 7–1, 8–1 ................................... 171
cl. 22.1(c) ........................277, 282, 283 cl. 18–1 (i)(1), (2) ........................... 171
cl. 22.2 ........................................... 286 Oil & Gas UK Joint Operating
cl. 22.2(a), (b) ................................ 286 Agreement (JOA)
cl. 22.2(c) ................................277, 283 [standard form] ................................ 110
cl. 22.2(d)....................................... 283 cl. 6.2.4(b)...................................... 110
cls. 22.3, 4 ...................................... 286 OFL Proposal ............................... 167–169
cl. 22.5(b) ................................277, 286 1.1 .................................................. 167
cl. 22.6 ........................................... 277 1.1(3), (10) ..................................... 168
cl. 25 .............................................. 277 8.1 .................................................. 167
cls. 35, 36 ....................................... 137 8.1(a), (b) ....................................... 166
Mutual Indemnity and Hold 8.2 .................................................. 167
Harmless Deed (IMHH Deed) 8.2(a), (b) ....................................... 166
2022.........................................52–59, 62 8.3 .................................................. 169
cl. 1 .................................................. 62 8.4 ...........................................167, 169
cl. 1.1 ................................... 52, 55, 57 8.5 .................................................. 169
cl. 2 ...................................... 53, 55, 56 8.6–8.9 ........................................... 167
cl. 2.1 ......................................... 53, 55 8.12(1–4) ........................................ 170
cl. 2.2 ............................................... 53 8.12(7)............................................ 171
cl. 2.3 ............................................... 54 PROJECTCON........................................ 4
cl. 3.2 ............................................... 56 SUPPLYTIME (BIMCO) 2005 ......2, 5, 6,
cl. 3.2(ii), (iii) ................................... 56 195, 197
cl. 3.4 ............................................... 56 Supplytime Part II Defnitions ...... 168
cl. 4.1 ................................... 53, 54, 57 Boxes 2, 3 ...................................... 167
cl. 4.2 ......................................... 53, 54 1989 Revision................................. 7, 8
cl. 4.3 ............................................... 53 2005 Revision................... 7, 8, 12, 164
cl. 4.4 ......................................... 52, 53 2017 Revision....................... 6, 12, 164
cl. 5 .................................................. 54 Subclauses 9(c), (e) ............................ 5
cls. 5.1, 5.2 ....................................... 55 Clause 14 ............................... 3, 4, 199
cl. 5.3 ............................................... 56 Subclause 14(a).................................. 5
cl. 6.1(i)............................................ 55 Subclauses 14(a)(i), (ii) ...... 5, 166, 167
cl. 6.1(ii)........................................... 53 Subclause 14(c) .................................. 5
cl. 7.1 ......................................... 52, 54 Subclause 14(d)...................... 5, 8, 167
cls. 7.2, 7.3 ....................................... 56 Subclause 15 .................................. 167
xxxix
Subclause 16 ...................................... 5 Treaty with Great Britain
Subclause 17(a) .............................. 170 on Marine Transportation
Subclause 17(a)(ii) ......................... 171 and Litigation, U.S.-Gr. Brit.,
Subclause 18(c) .................................. 5 Dec. 4, 1942, 56 Stat. 1780 (1942) ..... 68
SUPPLYTIME 2005 Charter Party TUGHIRE............................................... 8
TOWCON ....................................2, 5, 6, 8 UKOOA, IMCA, OCA and WSCA...... 58
cl. 18 ................................................ 22 Uniform Intermodal Interchange and
TOWHIRE ......................................2, 5, 6 Facilities Access Agreement ............. 231
xl
PART I
CHAPTER 1
The development of knock-for-knock

clauses in the last 15 years
1 Introduction
1.1 Defnition
“Knock-for-knock” (hereinafter K4K) clauses operate in various forms. In its most
basic form, the clause provides that both parties will cover their own losses, i.e.,
the “knocks”, that are sufered by their respective property or personnel. As such,
neither party can sue the other for the loss sufered. This arrangement minimises
the costs that such “knocks” would otherwise incur, by saving the parties the
valuable time and money that would typically be spent litigating fault-contingent
liability.
1.2 Essential meaning of a K4K clause

Under a K4K scheme, the loss lies where it falls, irrespective of fault and without
recourse to other parties.1 It is often accompanied by a series of mutual indemnity
clauses, wherein parties agree to indemnify the other contracting parties against
injury to, or death of, their own personnel, loss or damage to their property, and
any other specifed losses. The primary advantage of K4K clauses is that liability is
predetermined, thus saving insurance costs and time in attributing fault, from both
a factual and legal perspective. This leads to faster compensation to injured parties,
and thus makes K4K clauses very attractive to industry, especially where multiple
parties are involved.
Essentially, K4K clauses underpin a simple, consensual scheme of mutual risk
allocation. Moreover, they function as contractual exclusion clauses, with each
party seeking (1) to exclude its own liability, even if caused by its own fault; and
(2) to obtain an indemnity from other parties for any liability to which it may be
exposed.2 Hence the need for their terms to be reciprocal, and to cover similar
liabilities, as many jurisdictions prohibit one-way liability exclusions, most notably
US maritime law.2
1 A. McCooke, Shipowners: The False Economy of Amending Knock for Knock Clauses, www.
shipownersclub.com/the-false-economy-of-amending-knock-for-knock-clauses/, accessed 10 August 2021.
2 Ibid.
DOI: 10.4324/9781003206798-1 1
kyriaki noussia and hanieh bolourian
1.3 Historical genesis of the K4K clause
The phrase and general meaning of “knock for knock” originated in the US car
insurance industry during the early 1900s.3 Early automobile insurers commonly
found themselves trapped in the nonideal position of having to pay for their custom-
ers’ car repairs frst, and then having to engage in costly, drawn-out litigation, some-
times spending years arguing as to which party was at fault. To avoid losing the time
and money required to litigate, they began to enter agreements with one another
that, after a collision, each insurer would simply pay their policyholder’s losses and
seize the continuation of the claim. K4K clauses were efectively devised and designed
as a contractual tool to limit the uncertainty of liability exposure stemming from
inherently hazardous activities.4
Additionally, K4K was brought into the maritime insurance context in the United
Kingdom during World War II as a mechanism for reducing litigation costs arising
from frequent naval accidents.5 Responding to the threat of German submarines,
English ships sailed in the dark, with all lights switched of, in a very tight cluster.
This tactic reduced their exposure to German submarines but increased the overall
rate of collisions, hence the parties decided to subject themselves to the K4K prin-
ciple, essentially agreeing that each party will bear its own costs.6
K4K clauses entered the energy sectors in the 1960s with the commencement of
oil and gas exploration in the North Sea.7 This method of risk allocation was also
common in towage agreements in the United States, contractually restricting the
allocation of liability between the two parties.8 Each party, i.e., the tug and the tow,
would bear their own risk arising from the towing operation. Additionally, such a
contract formulation was viewed as more likely to be able to be relied upon in court.9
By the 1970s, K4K clauses were being used in some ocean or international towage
agreements. In 1985, BIMCO, one of the largest international shipping organisations,
and an authority on industry best-practices, released two standard form towage agree-
ments, TOWCON and TOWHIRE, with K4K at the heart of their liability regime.
Moreover, SUPPLYTIME, a BIMCO-drafted time-charterparty contract designed for
supply vessel and ofshore work, had a K4K clause added as standard in its 1989
revision. Other standard form contracts used in the ofshore industry, and many
charterers’ in-house charterparties, also apply a K4K allocation of liabilities. This
industry-wide use, along with widespread judicial acknowledgement of its enforce-
ability, has efectuated the K4K clause’s predominance as a mechanism by which
coventurers can streamline the allocation of risk and liability in ofshore operations.10
3 P. Saraceni, N. Summers, “Reviewing Knock for Knock Indemnities: Risk Allocation in Maritime and
Ofshore Oil and Gas Contracts” (2016) 30 ANZ Mar LJ, 28–43, 28.
4 Ibid.
5 See Parchomovsky and Stavang, Chapter 3 of this book.
6 Ibid.
7 P. Saraceni, N. Summers, “Reviewing Knock for Knock Indemnities: Risk Allocation in Maritime and
Ofshore Oil and Gas Contracts” (2016) 30 ANZ Mar LJ, 28–43, 28.
8 A. McCooke, Shipowners: The False Economy of Amending Knock for Knock Clauses, www.ship-
ownersclub.com/the-false-economy-of-amending-knock-for-knock-clauses/, accessed 10 August 2021.
9 Ibid.
10 Ibid.
2
development of knock-for-knock clauses in the last 15 years
2 Issues in using, implementing and enforcing K4K clauses
In accordance with general professional liability standards, under both contract
and tort, negligent performance of a professional service typically leads to liability
for damages on behalf of the negligent party. The main issue in connection with
the K4K regime is whether the professional liability for negligently conducted or
omitted provisions of service – this being the fundamental obligation of a contract
for services – can be limited, or exempted, by an indemnity clause and, if so, to
what extent.
The efect of a contractual K4K scheme is to reverse the commonly accepted prin-
ciple whereby the at-fault party is held liable for the ensuing damage. As the K4K
provision creates a novel legal relationship, departing from centuries of default tort
law, the contract absolving parties of this liability must do so properly, or it may not
be readily enforceable. As a general matter, courts do not typically favour standard
form contracts, or worse, contracts of adhesion, and will interpret such provisions
against the drafting party, under contra proferentem. Nevertheless, K4K clauses are
regularly included in standard form contracts in the maritime law domain, in vari-
ous types of charterparties and ofshore contracts.11 Such ubiquitous use of K4K
clauses in the wider maritime sector makes logical and economic sense to frms, as
they limit the parties’ exposure to unpredictable liability.12 Furthermore, participants
in the oil and gas industry, specifcally, regularly choose to contract out of standard
tort liability, suspending the rules of negligence and strict liability in their dealings
with one another, and embrace the K4K rule instead. Conclusively, the industry’s
adoption of K4K eliminates liability for harms incurred through the course of their
inter se relationships. Hence parties internalise the cost of the harms they may sufer
at the hands of other contractual parties. As for insurance, K4K operates in terms
of frst-party insurance, an intriguing feature that runs contrary to contemporary
theory and practice.13
3 Various uses
3.1 SUPPLYTIME 2005

K4K clauses express the parties’ contractual language. Clause 14 of the SUPPLY-
TIME 2005 charter party excludes from the K4K indemnity any undeclared dangerous
cargo or hazardous or noxious substances shipped by the charterers on board the
vessel, any pollution claims and any general average. Such claims must be resolved
under traditional fault-based liability regimes. As such clauses are exclusion clauses,
there must be unambiguous wording to exclude one’s liability. The scope of K4K
clauses extends to protect the indemnifed party from gross negligence, wilful
11 See, e.g., Caledonia North Sea Ltd v London Bridge Engineering Ltd and Others (The Piper Alpha)
[2002] UKHL 4; [2002] 1 Lloyd’s Rep 553 (HL).
12 M. Mudrić (2015). “The Guardcon Contract, Knock-for-Knock Clauses, DCFR and Unfair Terms
(Part I)”, JIML 21 (2015), 51–62, 55.
3
misconduct, material breach of contract, statutory or strict liability, consequential
loss and proportionate and concurrent liability.14
3.2 Insurance and P&I clubs

In the insurance sector, P&I Clubs will often review and, if appropriate, approve
K4K clauses if they are balanced and mutual and so long as the P&I Club member
has not waived any right to limit liability under any applicable law. Unbalanced K4K
provisions are not “poolable” in respect to any liabilities of which the member would
not have been exposed in the absence of the contract. Also, for the K4K liabilities
to be “poolable”, they must incorporate indemnities, protecting members if they are
sued by a third party who is not bound by the contract, and members must contract
on terms that do not expose them to disproportionate liabilities.15
K4K clauses are a fundamental part of the maritime and ofshore oil and gas
sectors. Recognised in English and Australian jurisprudence, as well as in other
legal systems, and in various bodies of case law, such hold harmless and indemnity
clauses play a signifcant role in commercial operations within those sectors. Also,
courts carefully examine the wording of K4K clauses in an efort to determine their
meaning, scope, operation and efciency; hence clarity in drafting such clauses is
absolutely necessary to make sure that they serve the scope for which they have
been drafted.16
3.3 Use in ofshore contracts

The arrangement provided by K4K clauses has been used extensively in upstream
oil exploration activities, as such operations typically involve many contractors and
subcontractors. Furthermore, in the case of a potential loss – e.g., the Deepwater
Horizon incident in the Gulf of Mexico in 2010 – multiple potential defendants
would be involved. To avoid such complexities, the parties agree to accept responsi-
bility for loss or damage to their own property and for injury or death to their own
employees, irrespective of blame. Such arrangements were described by Lord Bingham
in the Piper Alpha17 case as a market practice that has developed in order to accom-
modate the peculiar features of ofshore operations. Additionally, K4K clauses are
often used in downstream oil and gas contracts, as the parties to such contracts are
also usually parties to the contracts upstream or are subcontractors who serve in the
same role that they serve in the upstream contracts. BIMCO has produced a series
of specialised charter forms that are designed to cater for the specialised requirements
of the ofshore industry, such as HEAVYCON, BARGEHIRE, PROJECTCON,
14 P. Saraceni, N. Summers, “Reviewing Knock for Knock Indemnities: Risk Allocation in Maritime
and Ofshore Oil and Gas Contracts” (2016) 30 ANZ Mar LJ, 28–43, 32–33.
15 Ibid., 41.
16 Ibid., 43.
17 Caledonia North Sea Ltd v. London Bridge Engineering Ltd and Others (The Piper Alpha) [2002] 1
Lloyd’s Rep 553 (HL).
4
TOWCON, TOWHIRE, and SUPPLYTIME, all of which contain K4K clauses, the
commercial purpose of which is to make it clear who is to bear the risk.18
4 Modern use
Today, the K4K principle is acknowledged as being at the very core of the SUPPLY-
TIME form-contract, and K4K clauses in SUPPLYTIME mean that each party should
bear responsibility for any damage or loss to its own property, or injury to its own
personnel, even if the damage, loss or personal injury is caused by the act, neglect or
breach of duty (whether statutory or otherwise) of the other party. When the other
party compensates the claimant, even if the loss of compensation fled by the claimant
is caused by the other party’s act, neglect or breach of duty – be it statutory or oth-
erwise – the party shall still indemnify the other party for the compensation submitted
by the claimant, and the compensation made by the party against the other party
shall be called K4K indemnity. Again, the essence of K4K clauses is that the parties,
based on simple apportionment of risks and liabilities, replace the original fault liability
by mutual agreement for several exempted and non-exempted items.19
4.1 The BIMCO SUPPLYTIME 2017

In 2017, the Documentary Committee of BIMCO adopted the revised SUPPLYTIME
2017, bringing it more in line with current practices in the ofshore sector. The objec-
tive of BIMCO’s 2017 revision of SUPPLYTIME was to better balance the interests
of the owners, and the interests of the charterers, in relation to the K4K liability
regime. Apart from that, there were also several smaller changes from the previous
iteration of the contract. In relation to the K4K clause (Clause 14) and the liability
regime, this has been strengthened in the 2017 edition, because the defnitions of the
“Charterers’ Group” and of the “Owners’ Group”, which afect the operation of the
K4K liability regime, have been improved and moved to the “Defnitions” section,
so that the usage is consistent throughout the charterparty. Subclause 14(a)(i) describes
the losses that fall on the owner, as any loss or damage to property belonging to the
Owners’ Group, or personal injury and death of anyone in the Owners’ Group, are
to be borne by the owners, even if they are caused by the negligence or default of
the Charterers’ Group, subject to three exceptions stated in subclauses 9(e), 14(c) and
18(c). Subclause 14(a)(ii) describes the losses that fall on the charterers. It is not
identical to the subclause governing the owners, and this is due to the diferences in
their respective obligations. Nevertheless, there is an equivalent strengthening of the
K4K regime for charterers, subject to two exceptions in subclauses 9(c) and 16.
18 R. Williams, “Knock for Knock Clauses in Ofshore Contracts: The Fundamental Principles”, 53–67,
in: B. Soyer, A. Tettenborn, (Eds) (2014) Ofshore Contracts and Liabilities, Informa Law.
19 Y. Han, “On Knock-for-Knock Principle: Analysis of SUPPLYTIME 2017 Clause 14(a)”, 2019
China Oceans L. Rev. (2019) 4, 128–149, 128; S. Rainey, “The Construction of Mutual Indemnities and
Knock-for-Knock Clauses” 68–107, 70–71, in: B. Soyer, A. Tettenborn, (Eds) (2014) Ofshore Contracts and
Liabilities, Informa Law.
5
Notably, the description of the included causes of loss has been expanded to include
any loss
arising out of, or in any way connected with, the performance or non-performance of
this Charter Party whatsoever, and in any circumstances, even if such loss, damage, or
personal injury or death is caused wholly or partially by the act, neglect, breach of duty,
be it statutory or otherwise.
Hence SUPPLYTIME 2017 has retained the key features that made this particular
standard form one of the most popular in the ofshore sector, while making several
changes to modernise it, and to ensure that it keeps up with the changing market
practice; strengthening the K4K liability regime by balancing liabilities and indemni-
ties; and by reducing of the number of exceptions to the regime.20
5 A historical encounter of the K4K clauses in the shipping

and in the ofshore oil and gas sectors
Early on, K4K fourished in contracts for the shipping industry, and the ofshore oil
and gas industry, in exploring the North Sea. With time, certain risks were excluded
from the K4K regime, as evidenced by the evolution of the BIMCO SUPPLYTIME
form. When frst published in 1975, the form represented a contract where all intricacies
of the sector were therein depicted, and as a result, many aspects of the regime favoured
owners, especially in regard to a charterer’s liability for loss or damage negligently caused
to the vessel or to her owners. Such provisions were only looked upon and amended in
1989, when the SUPPLYTIME form was redrafted to encompass fairer provisions, and
to achieve balance between the interests involved. The K4K regime provisions for mutual
risk allocation were popular in the ofshore industry forms of contract such as HEAVY-
CON, TOWCON and TOWHIRE, drafted by BIMCO; the LOGIC Construction
Contract; and the Norwegian Subsea Contract 05 (known as NSC 05). Under those
contracts, K4K operates fairly, as liability and indemnity provisions are balanced.
As oil-and-gas activities involve risky undertakings at virtually all stages, the use
of the clause has become popular for a number of reasons. Fault, though traceable
through the traditional allocation of liability under the common law, becomes a
matter of contractual stipulation for who bears the burden under K4K. Allocation
of liability therefore becomes simpler.
The commercial or economic imperative for mutual indemnifcation was indeed
recognised in the case of Caledonia North Sea Ltd v London Bridge Engineering Ltd
(London Bridge).21 There, Lord President Rodger noted as relevant evidence from
one of the witnesses, that the practice is fundamental economics of the business, and
thus, to the North Sea operation. The use of the clause, however, is not a foolproofy
guard against litigation for the purpose of tracing which party is liable in the event of
damage or loss, as experience indicates. A possible corollary is that parties could assess
risk and accept liabilities more easily. Indemnity clauses also reduce the scope and
20 Kennedys Law, “Notes from the Bar: BIMCO SUPPLYTIME 2017, 18 October 2017, https://ken-
nedyslaw.com/thought-leadership/article/notes-from-the-bar-bimco-supplytime-2017/.
21 Caledonia North Sea Ltd v London Bridge Engineering Ltd [2000] S.L.T. 1123 at 1150 (London Bridge).
6
cost of insurance available for parties. The extent of possible risk instances that may
occur, e.g., in an oil platform due to the act or omission of a contractor, is infnite.22
6 The evolution of K4K clauses in standard contracts and in English case law
In cases where contracts contain a K4K regime which either adopts standard industry
wording or forms part of a standard form contract, courts have shown a willingness
to give efect to the object and purpose of K4K clauses. By conforming to a K4K
clause in ofshore contracts, in the event of loss and damage to respective parties’
property, or the subsequent death of employees, companies uphold personal liabili-
ties.23 Under the BIMCO SUPPLYTIME revisions of 1989 and 2005 formats, and
the revised HEAVYCON 2007 contract, in the event of “consequential loss” the
exclusion clause becomes a provision of the K4K clause. The term “consequential
loss” has been defned in a contract setting, whereby the normal loss can be stated
as the market value of the property, as the money or services that the claimant should
have received under the contract, less either the market value of what he does receive
or the market value of what he would have transferred, but for the breach having
occurred. Indeed, consequential losses have been perceived to encompass anything
beyond such a normal measure, e.g., profts lost, or expenses incurred, through the
breach, and they are recoverable if not too remote. Because misassumption in the
classifcation of losses has instigated contractual difculties in energy construction
and supply ofshore contracts, a solution was sought. Recoverable losses in contracts
were started from the “limb one” and “limb two” categories of losses established by
the rules adopted in the seminal case of Hadley v Baxendale.24 The English court
rejected the mill’s loss of proft claim for late delivery of a mill part, because of the
lack of clarity in the mill’s subsequent loss of proft in the event of a late restart in
production. Following this case, the ruling was re-stated in Victoria Laundry v New-
man25 and in Czamikow v Koufos26 for which the Victoria Laundry ruling27 explained
the type of knowledge that parties must possess in order to recover losses. It has
been stated that knowledge “possessed” is of two kinds, one imputed, and the other
actual, and that the reasonable person is taken to know the ordinary course of things
and, consequently, what loss is likely to result from a breach of contract in that
ordinary course. Indeed, this was taken to be the subject matter of the “frst rule”
in Hadley v Baxendale28 or any other case containing knowledge of special circum-
stances outside the ordinary course of things, such that a breach given those special
circumstances would be likely to cause more loss. Finally, it was remarked that such
a case instigated a “second rule” such that additional losses are recoverable.
22 C. Ugwuanyi, “Examining the Exclusionary Nature of Oil and Gas Contract Mutual Indemnity Hold
Harmless Clauses”, I.E.L.R. 2012, 4, 136–146.
23 A. Iyer, “Excluding ‘Consequential Loss’ in Ofshore Contracts”, 23 October 2020, www.iylegal.com/
excluding-lsquo-consequential-loss-rsquo-in-ofshore-contracts.
24 Hadley v Baxendale (1854) 9 Exch 341.
25 Victoria Laundry v Newman [1949] 2 K.B. 528.
26 Czamikow v Koufos [1969] 1 A.C. 350.
27 Victoria Laundry v Newman [1949] 2 K.B. 528.
7
Recoverable indirect losses, or “limb two” losses, under English law must stem from
the breaching party’s actual knowledge of the loss, independent of the scale, which
the party was aware of at the time of contract that such a loss would result from a
breach. The difculty is that this distinction between consequential losses and all
other losses is not the same as the distinction between the “frst” and “second” “limbs”
in the Hadley v Baxendale29 rule, as consequential loss may well fall within the “frst
limb” as a direct loss which was a natural consequence of the breach. Hence it may
not be entirely accurate to say that the Hadley v Baxendale30 “second limb” covers
only, or every, consequential loss.
Under English law, the scope of certain K4K clauses can turn upon the interpreta-
tion of the term “consequential loss” provided in exclusion clauses. In ofshore con-
tracts, wherein a loss is deemed to be incurred as a “direct loss” following a breach of
contract, the parties can still be found liable to meet any number of “consequential
losses”. An example of such could be a loss of proft. English courts will construe
exclusion and limitation clauses strictly against the party seeking to rely upon them,
pursuant to the contra proferentem rule, which may – in the context of consequential
loss clauses in ofshore contracts – bring other consequences.31
The widely cited case Croudace v Cawoods32 determined that the use of the word
“consequential” did not cover any loss which directly and naturally resulted in the ordi-
nary course of events from late delivery; and in Addax v Acadia,33 non-consequential
losses associated with incurred “hedging costs” were found recoverable as “direct”
losses. In British Sugar Plc v NEI Power Projects,34 regarding increased production
costs and loss of profts for the claimant, it was held that since the faulty power
station equipment supplied by the defendant had directly led to such losses, those
losses would not be encompassed within an exclusion of liability for “consequential”
losses and could thus be recovered as direct losses.
By itself, inserting the term “consequential losses” into the exclusion clause of an
ofshore contract may be insufcient to protect the owner from costly claims for loss
of proft, production or business interruption. Nevertheless, the main hiring contracts
(TUGHIRE, TOWCON, SUPPLYTIME 89 [and revised 05] and HEAVYCON 07)
each have consequential loss provisions which purport to protect the parties from
claims for “consequential losses” specifcally.
Clause 23 of the 2007 revision of HEAVYCON (HEAVYCON 07) marked the frst
introduction of a consequential damages clause into HEAVYCON, identical to clause
14(c) of SUPPLYTIME 05. In interpreting the language of such a clause, courts may
likely rule that all loss of proft, whether foreseeable or not, must necessarily include
losses falling within the frst limb of Hadley v Baxendale35 as well as those falling
within the “second limb”, and that the lack of any reference to any other direct losses
29 Ibid.
30 Ibid.
32 Croudace v Cawoods [1978] 2 Lloyd’s Rep 55.
33 Addax v Acadia [2000] 1 Lloyd’s Rep 493.
34 British Sugar Plc v NEI Power Projects (1998) 87 BLR 42 CA.
8
would tend towards such an interpretation. However, there is still a risk that, if the
specifc term used in those clauses is “consequential loss”, courts may interpret the
clause as excluding only any “second limb” losses and not any losses falling under
“limb one”, assuming they are “direct” or “natural losses”.36
A distinction also appears to be drawn between the construction of exclusion
clauses on the one hand and limitation clauses on the other. In Ailsa Craig Fishing
Co Ltd v Malvern Fishing Co Ltd,37 the House of Lords held that a limitation clause
need not be construed in accordance with the same principles that apply to exclusion
or indemnity clauses, and that the actual absence of a reference to negligence in the
limitation clause did not prevent it from protecting against the contractor’s liability
for negligence. However, in E. E. Caledonia Ltd v Orbit Valve Co Europe,38 wherein
the indemnity clause was expressed generally and without reference to negligence,
the court found that the parties’ right to sue each other for negligence had been
preserved. For K4K to operate, an indemnity clause should expressly and clearly
refer to negligence; as this was not the case in Orbit Valve, the K4K regime did not
extend to liability caused by a party’s own negligence.39
6.1 The signifcance of the Piper Alpha litigation

The basis of “mutual hold harmless” or K4K clauses can be seen in the House of
Lords’ landmark decision in the case of the Piper Alpha litigation.40 The Piper Alpha
disaster of 6 July 1988 resulted in 167 dead, 62 injured, and total destruction of the
Piper Alpha oil platform of the coast of Scotland. The operator of the platform,
Caledonia North Sea Ltd (formerly Occidental Petroleum (Caledonia) Ltd), along
with its co-venturers, and its insurers, settled the fatal accident and personal injury
claims using a formula agreed upon within months of the disaster. In the House of
Lords’ decision, Lord Mackay reinforced the view in relation to primary liability on
the contract in this case by stating that this was not entirely typical, as the relevant
clause required the contractor to indemnify the operator in respect of injury to or
death of persons employed by the contractor, irrespective of any contributory neg-
ligence, whether active or passive, of the party to be indemnifed, unless such injury,
death, damage, loss or destruction was caused by the sole negligence or wilful mis-
conduct of the party which would otherwise be indemnifed. This case demonstrates
that an operator may voluntarily choose to insure against claims for which it has a
contractual right of indemnity, and that choice will not operate to the advantage of
the party contractually obliged to pay the indemnity.41 Accordingly, this case is notable

37 Ailsa Craig Fishing Co Ltd v Malvern Fishing Co Ltd [1983] 1 WLR 964.
38 E. E. Caledonia Ltd v Orbit Valve Co Europe [1994] 1 WLR 221.
40 Caledonia North Sea Ltd v. London Bridge Engineering Ltd and Others (The Piper Alpha) [2002] 1
Lloyd’s Rep 553 (HL).
41 T. Hewitt, “Who Is to Blame? Allocating Liability in Upstream Project Contracts”, Journal of Energy
& Natural Resources Law, 26:2, 177–206.
9
for its contribution to the emergence of the mutual hold harmless indemnity regime.
The fact that the operators had indemnity did not stop the court from holding the
contractors responsible for indemnifcation of their staf and crew members. Hence
coming to this point in law, where there is contributory negligence – versus sole
negligence on the part of the operator – the contractor needs to apply its own share
of indemnity, regardless of the operator’s contributory negligence.42 The insurers,
therefore, would have been subrogated to the claims of the operator against the
contractors, as the House of Lords upheld in the previous judgment. The indemnity
was to have applied in favour of the operator, regardless of any negligence in the
part of the operator; and death or injury to the contractor’s employees was to be
covered by the mutual hold harmless regime rather than the standard fault-based
regime.
7 The contribution of other jurisprudence to the historical

evolution of K4K clauses
7.1 US jurisprudence
K4K clauses, when present in an oilfeld agreement, must be balanced and compre-
hensive for several reasons, as, but for their existence, it would be extremely difcult
to determine who holds fault in drilling site accidents.43 Some US states (e.g., Louisiana,
Texas, New Mexico and Wyoming) have Limitation Acts prohibiting certain indemnity
agreements. This means that any contractual clause that liberates the contractor or
the indemnifed party from taking responsibility for loss or damages caused by its
own negligence or fault, such as K4K, is to be considered void. Some American courts
have also issued rulings rejecting contractual exculpation from responsibility for loss
or damage caused by gross negligence or wilful misconduct.44 However, it remains a
question as to how exactly the K4K regime works out in jurisdictions that are weary
of broad liability limitation, as it may not be a practical insurance tool in onshore
operations, where there is more clarity in terms of potential risks as compared to
ofshore sites.45 Pursuant to such terms, K4K clauses need to specify when laying out
health and safety guidelines, and each party’s responsibilities and liabilities, in light
of those guidelines. There also should be no carve-out for “gross negligence” or “wil-
ful misconduct” as they can lead to less certainty in allocation of liabilities unless
such clause is thoroughly defned. The K4K regime must also apply to all the parties
involved, such as other contractors and subcontractors, to promote fairness and equal-
ity in the legal treatment provided.46 In some American states (e.g., Texas and Loui-
siana), the anti-indemnity legislation is widely seen as a reaction to oil companies’
attempts to force local service and product providers to accept all liabilities, even if
42 Caledonia North Sea Ltd v London Bridge Engineering Ltd [2002] UKHL 32; [2002] 1 Lloyd’s Rep
562, 563, HL.
43 T. Middis, “Knock for Knock Indemnities – Are They Appropriate for On-Shore Infrastructure
Projects?” 7 May 2015, Addisons, 1–3.
44 Ibid.
45 Ibid.
46 Ibid.
10
the loss stems from the oil company’s fault. However, an enforceable K4K clause in
an oil or drilling contract can be achieved, despite such statutory limitations, given
that the language of the contract is well constructed and clear. Summarily, contractual
agreements including K4K clauses are generally enforceable in the United States,
absent any statutory or judicial precedent to the contrary.47
7.2 Deepwater Horizon

On 20 April 2010, the Macondo exploratory well was being drilled ofshore by the
Deepwater Horizon platform when it sufered a blowout, causing a fre and resulting
in 5 million barrels of oil spilled into the Gulf of Mexico. Thousands of legal suits
were fled against BP, and on 20 May 2015, BP settled the multibillion-dollar lawsuits
with its co-venturers Transocean, Halliburton and Cameron. Importantly, Transocean
and its contractors relied on K4K clauses in the respective contracts to exclude their
liability for loss and damage other than to their own employees and property.48
7.3 Brazil
In Brazil, contractual terms in the ofshore sector would fall under two main contrac-
tual regimes: (1) international contract forms, such as BIMCO standard charter
contracts, incorporating the K4K clause principle; or (2) Petróleo Brasileiro SA
(Petrobras). The latter, a Brazilian state oil contract representing 95% of Brazilian
ofshore contracts, instigates its own contractual terms under the Brazilian Civil Code.
Pursuant to Brazilian civil law, Petrobras contracts avoid K4K indemnity through
provisions limiting indemnifcation amounts and excluding indirect damages or loss
of earnings. A basic principle of Brazilian civil law is that any person who causes
damage, whether through employees or subcontractors, to another must indemnify
the aggrieved party in a form proportional to the damage sufered. Under Section
927 of the Brazilian Civil Code, this obligation arises regardless of fault, in the cir-
cumstances specifed by the Code; or when the activity that caused the damage included
a risk to the environment, or to third parties. Nevertheless, under Brazilian law, the
addition of a K4K clause is possible since terms and conditions of contracts and
clauses are negotiable between parties so long as they do not contravene public order
and third-party interests. For example, a limitation of liability clause under a contract
of carriage is considered by the Brazilian courts as contrary to Brazilian law and
therefore null and void. It must be noted that in the context of carriage of goods
contracts, for instance in relation to standard bills of lading, the Brazilian courts’
interpretation, and position on the limitation of liability clause included in the bill of
lading, is considered as onerous to the receiver and thus not valid. Also, considering
the nature of ofshore contracts in the oil sector – the equal bargaining strength,
and fexibility for more open and reciprocal negotiation of contractual terms –
47 L. Lambert, “Knock-for-Knock Contracts Are Enforceable in the US”, Standard Bulletin: Ofshore
Special Edition, November 2011.
11
dissipates the Brazilian civil courts’ position regarding the potential for unfair ofshore
contracts, and thus could potentially yield to the inclusion of a K4K clause. Such a
clause has yet to be challenged in a Brazilian court, which considers the clauses an
adaptation of foreign law and jurisdiction regulations.49
7.4 Germany
The growth in recent years of ofshore wind farm projects in the Exclusive Economic
Zone (EEZ) of Germany has given rise to the possibility of potential disputes related
to their underlying contracts in German courts and arbitral tribunals. German courts
generally lack practical experience in interpreting contractual forms such as the BIMCO
SUPPLYTIME 2005 or 2017, the BIMCO WINDTIME or LOGIC contracts, all of
which have been developed against the background of English law. The validity of
K4K clauses under German law would depend on the court’s evaluation of whether
(1) the clause is part of general terms and conditions; (2) it can be considered reason-
able; and (3) irrespective of general terms and conditions, whether it contains specifc
liability exclusions disallowed under German Law. Under German law, K4K clauses
in standard form industry contracts, such as BIMCO SUPPLYTIME, will be consid-
ered as general terms and conditions unless the particular clause has been individually
negotiated between parties. The term “negotiating” as per the German Civil Code
requires the user of general terms and conditions to allow the contractual partner to
actively protect his own interests and demonstrate his willingness to negotiate terms
in principle. Previously, “rider clauses”, pursuant to a contractual agreement, including
a reasonable and fair allocation of rights and duties, and highlighting the known
nature of every clause as part of the individual negotiations, were accepted in German
jurisprudence. Later case law shows an understanding, in the German federal courts,
that the inclusion of a rider clause does not automatically prevent a contract from
still being regarded as general terms and conditions, because whether a contract is
agreed on the basis of general terms and conditions or individually is a question of
fact, which cannot be contractually regulated by the parties; hence such a rider may
still be considered to be part of the general terms and conditions.50
The liability regime of a K4K clause is contrary to any German civil liability regime
and thus considered foreign by many courts. Such contracts containing a K4K clause
must ensure fairness between the parties to keep these particular clauses from being
invalidated by German courts. The main deterring factor in a court’s decision would
be the consideration of the practices and customs that apply in general German
business dealings, for which the ofshore wind industry’s rather young “status quo”
would likely not be a determinative factor. Hence it is likely that a German court or
arbitral tribunal will render a K4K clause invalid if it attempts to exclude liability for
damages arising from personal injury to life and body or health, wilful or intentional
49 G. Mendes Vianna, J. Furtado Senna, “Knock for Knock Clauses under Brazilian Law” (ofshore
charter contracts), Standard Bulletin: Ofshore Special Edition, November 2011.
50 E. Volz, A. Waldmann, “Knock-for-Knock Liability – Does It Work in the German Market?”,
3 September 2019, https://www.mondaq.com/uk/contracts-and-commercial-law/842380/knock-for-knock-
liability--does-it-work-in-the-german-market, accessed 18 October 2022.
12
misconduct and/or gross negligence. Consequentially, in such circumstances, a court
or arbitral tribunal may not only consider the K4K clause but rather the entire clause
to be invalid, even if only a part of such clause is afected. In efect, in case of any
damage or loss, as per the German statutory liability regime of the German Civil
Code (which adopts a fault-based liability regime), the party claiming damages would
be entitled to damages without any limitation, provided the other party was at fault.
Under German law, certain restrictions apply to all contract clauses, providing for the
provision on civil liabilities. If the exclusion of liability is for damages resulting from
wilful or intentional misconduct, even if the clause has been individually negotiated,
the clause will be invalidated, unless a carve-out regarding exclusions and limitations
of liability is in place to be used for this particular type of damages.51
7.5 Australia
In Australia, in commercial contracts, K4K clauses are construed in accordance with
ordinary rules of contractual construction. In Darlington Futures Ltd v Delco Australia
Pty Ltd,52 the High Court of Australia held that an exclusion clause is to be construed
according to its “natural and ordinary” meaning; and in Andar Transport Pty Ltd v
Brambles Ltd,53 the High Court applied the traditional rule of construction: that
indemnities should be construed strictly against the guarantor.
7.6 Nordic countries – the history and validity of K4K clauses from a
comparative perspective
In Nordic countries, the validity of K4K clauses, and the liability exclusions that they
contain, is often analysed comparatively, between the law of their tradition of origin,
i.e., common law, especially English law; and Nordic civil law, especially Norwegian
and Danish law, where such agreements are also frequently used, namely in the context
of oil extraction activities in the North Sea.54 The subjective characteristics of any
K4K arrangement depends on each jurisdiction’s respective stance, hence this becomes
one of the most important issues for the contracting parties during their contractual
negotiations, as the establishment of which national law applies will also determine
the applicability parameters of the contracts, and the extent and treatment of liability.
Standard K4K clauses often establish that the K4K liability allocation applies unless
the party causing the loss has acted with gross negligence or wilful misconduct. Fur-
thermore, the insurer might be able to exclude such qualifed forms of faulty behaviour
from the ofered coverage, so that the party causing the damage is contractually
exempted from liability but cannot recover its loss from the insurer or require the
claiming party to frst pay a deductible.55 Also, certain types of damages, such as
51 Ibid.
52 Darlington Futures Ltd v Delco Australia Pty Ltd (1986) 161 CLR 500.
53 Andar Transport Pty Ltd v Brambles Ltd (2004) 217 CLR 424.
54 S. Cavaleri, “The Validity of Knock-for-Knock Clauses in Comparative Perspective” (3 October
2017), European Review of Private Law, Vol. 25, 2017, University of Copenhagen Faculty of Law Research
Paper No. 2017–50, https://ssrn.com/abstract=3063280, accessed 27 May 2021.
13
pollution damages, may warrant exclusion from the K4K regime, and may simply be
governed by the usual principles of tort liability, since allocation of liability for those
certain types of damages is imposed by mandatory law.56
Contract validity is dependent on whether the parties had intended that a certain
behaviour of the responsible party, or a certain type of loss, is to be covered by the
exclusion. Furthermore, the liability exclusion or limitation must be compatible with
statute and public policy. English courts have traditionally assumed that, unless the
contract contains clear wording to the contrary, an exclusion or limitation of liability
does not cover negligence, and have ruled that it is “established that indemnity will not
lie in respect of loss due to a person’s own negligence or that of his servants unless
adequate and clear words are used or unless the indemnity could have no reasonable
meaning or application unless so applied”.57 In Canada Steamship Lines v Regem,58
Lord Morton developed a three-step contract interpretation test in determining the
criteria for parties to be relieved of their negligence or the negligence of their “ser-
vants”. The applicability of this three-step test was later confrmed by the Queen’s
Bench Division of the High Court of England and Wales in their interpretation of
the ofshore indemnity clauses in E. E. Caledonia Ltd v Orbit Valve Co Europe.59 This
case concerned the question of whether a contract containing a mutual indemnity
clause entitled the operators of an oil platform to an indemnity for the compensation
paid to the descendants of an engineer who had died because of a fre on the plat-
form caused, at least in part, by the negligence of the platform operators. The court
held that the operators were not entitled to indemnifcation because the indemnity
clause did not include a reference to negligence or any similar word, and that the
mere reference in the disputed clause to “any” claim or liability was not sufcient.
In Danish law, with regards to liability exclusions and limitations, the validity of a
contract depends on a court’s interpretation of four contractual factors (1) whether
a clause has not been validly adopted by the parties;60 (2) whether a clause is restric-
tively interpreted, sometimes to the efect that the clause does not apply at all (e.g.,
contra proferentem, wherein the party who drafted the clause must sufer the conse-
quence of its lack of clarity);61 (3) whether the theory dictating the consequence of
fundamental changes of circumstances (forudsætningslære) applies; and (4) whether
circumstances justifying the application of the clause are no longer present, or that
it would be contrary to the parties’ expectations to apply the clause in accordance
with its wording.62 The Danish Product Liability Act contains a prohibition against
clauses purporting to exclude or reduce liability for death or personal injuries falling
56 T-L. Wilhelmsen, “Liability and Insurance Clauses in Contracts for Vessel Services in the Norwegian
Ofshore Sector – the Knock for Knock Principle”, SIMPLY 2012, 88.
57 Alderslade v Hendon Laundry [1945] KB 189 at 192; see also Walters v Whessoe Ltd and Shell Refning
Co Ltd [1968] 2 All ER (All England Law Reports) 816 per Sellers SJ.
58 Canada Steamship Lines v Regem [1952] 1 All ER 305, § 16 to 19.
59 E. E. Caledonia Ltd v Orbit Valve Co. Europe [1994] 1 WLR (Weekly Law Reports) 221.
60 B. Gomard, “Obligationsret” 2. del, 4th ed., 2011, 311–312; MB. Andersen, J. Lookofsky, “Lærebog
i obligationsret I”, 4th ed, 2015, 411.
61 B. Gomard, “Obligationsret” 2. del, 4th ed., 2011, 313–314; MB. Andersen, J. Lookofsky, “Lærebog
i obligationsret I”, 4th ed, 2015, 411.
62 MB. Andersen, J. Lookofsky, “Lærebog i obligationsret I”, 4th ed, 2015, 413–415; V. Hagstrøm,
“Obligasjonsrett”, Universitetsforlag, 2nd ed. 2011, 658.
14
within the scope of application of this Act, i.e., caused by products within the
meaning of the Act. Furthermore, common to all Nordic countries, a clause can
be invalidated on the ground of unreasonableness pursuant to § 36 (the “general
clause”) of the Contract Act (CA) which states that a contract can be modifed or
set aside, in whole or in part, if it would be unreasonable or incompatible with the
principle of good faith to enforce it.63 The validity of a clause, and the gravity of the
fault leading to damage (gross negligence or intentional breach), are the main focus
of relevant Danish court decisions; whereas Norwegian court decisions also place
an emphasis on the hierarchical position of the person who caused the damage, as
gross negligence by a member of management can be treated diferently than similar
behaviour by a subordinate employee.64
7.7 Later judicial trends

Especially from the early 2010s, oilfeld operators have shown a real tendency to
chisel away at the K4K indemnity model, or abandon it altogether, to allocate more
responsibility to the involved service companies for all sorts of well damages, e.g.,
pollution, blowouts/wild well events, loss of oil/gas, to the extent that such damages
are caused by the negligence of the service company. Arguably, one way to do this,
i.e., to expand the range of service companies’ liabilities, is to embed a “gross neg-
ligence” exception provision in mutual hold harmless indemnity agreements.65 In
terms of contracts, many industry players use model form contracts produced by
not-for-proft industry associations such as the LOGIC standard form model contract.
LOGIC publishes a set of “standard” contracts, and in most of these contract models,
the “mutual hold harmless” clause is incorporated.66 Often, the “mutual hold harm-
less” clause and the allocation of liability vary between diferent types of contracts
and may serve one party more than the other, depending on the circumstances under
which the contract terms were negotiated. International and state oil companies usu-
ally have certain global policies for insurance in place, which exceed the minimum
requirements set by applicable jurisdictions. However, the beneft of such “mutual
hold harmless” clauses is – given that there is a clearer and simpler allocation of
liability in K4K – that liability becomes a matter of contract. Moreover, there is no
need for an investigation or assignment of fault, allowing parties to avoid lengthy
litigation. In the absence of such a regime, each contractor will need to insure against
all risks, e.g., destroying the whole facility or injuring each one of the personnel.
Often, one aspect to consider while drafting the “mutual hold harmless” indemnity
clause is to defne whose property is intended to be covered, i.e., whether it is property
63 Section 36, section 1 of the Danish Contract Act (Aftaleloven), free translation. “36.-(1) A contract
may be modifed or set aside, in whole or in part, if it would be unreasonable or at variance with the principles
of good faith to enforce it. The same applies to other juristic acts.” https://pure.au.dk/ws/fles/32350272/
Danish_20Contracts_20Act.pdf, accessed 18 October 2022.
64 V. Hagstrøm, “Obligasjonsrett”, Universitetsforlag, 2nd ed. 2011, 658.
65 S. Johanson, “Is It Really Knock-for-Knock?” 21 June 2019. Boyar Miller, www.boyarmiller.com/
is-it-really-knock-for-knock/, accessed 27 May 2021.
66 Ibid.
15
or equipment rented to the operator by the contractor, or if it is under the contrac-
tor’s or supervisor’s control but it is operated by the employees of the contractor.67
As of late, the “mutual hold harmless” regime is less commonly used due to changes
in the infrastructure of oil and gas operations, and developments in insurance attempting
to adapt to the complexity of oilfeld operations and their inherent risks. Historically, a
key factor in the emergence of the mutual indemnity regime has been access to insur-
ance. Previously, insurance policies were not well suited for the complexity and risks
associated with oil and gas exploitation. However, as the insurance market developed
a deeper understanding of the oil and gas industry, insurance policies became more
comprehensive, and as a result, afected the industry’s preferred allocation of liability.
Moreover, the recent shift in collaboration between operators and contractors has led
to the necessity for a diferent approach to risk and liability allocations. With collabora-
tive projects, tasks are tackled through bringing in specifc skilled contractors, and if
using a diferent reward system, rather than basic reimbursement, allocation of liability
would be diferent. Cases such as Transocean Drilling UK Ltd v Providence Resources
PLC68 highlight the importance of tailoring the indemnity regime to a specifc project,
rather than using a standardised version of the liability system in use.69
7.8 The way forward

The 2016 Court of Appeal (EWCA) decision in Transocean Drilling UK Ltd v Provi-
dence Resources PLC70 brought about signifcant discussion on K4K provisions in
oil and gas contracts. The case turned on the interpretation of the mutual indemnity
provisions, covering third-party losses, and consequential loss carve-outs within the
contract. The provisions in dispute were relatively typical in such contracts and have
been used in the ofshore industry for well over 25 years. Originally, these contracts
were used in recognition of the fact that costs from certain loss events, such as the
destruction of an entire oil platform, could not be borne by a single contractor. In
addition, the manner in which ofshore work has been carried out, historically, has
meant a signifcant overlapping of roles and responsibilities among groups of con-
tractors. Furthermore, a key justifcation in introducing mutual indemnity provisions
was the availability of insurance, and where such insurable risks lay. In April 2011,
Providence had contracted with rig owner Transocean for the hire of a semi-sub-
mersible drilling rig to be deployed in the Barryroe Field, Ireland, to drill an appraisal
well. The High Court held that the rig had not been delivered in good working
condition, and that Transocean was in breach of contract for the resulting delays.
Transocean’s appeal, however, focused solely on the High Court’s decision to allow
Providence to recover the “spread costs” incurred as a result of the delay, which were
the usual wasted costs of third-party personnel, equipment and services. Transocean
contended that their liability for such costs was excluded under the contract’s standard
67 Ibid.
68 Transocean Drilling UK Ltd v Providence Resources PLC [2016] EWCA Civ 372.
69 P. Murray, “Is It Time for ‘Knock-for-Knock’ to be Knocked Out?” 12 May 2018, News and Views,
Ledingham Chalmers Solicitors.
70 Transocean Drilling UK Ltd v Providence Resources PLC [2016] EWCA Civ 372.
16
consequential loss clause. The Court of Appeal reviewed the risk management and
liability provisions in the contract, focusing on the complexity of the series of indem-
nities, noting that such provisions were clearly designed to complement each other
and were efectively a detailed and sophisticated scheme for apportioning responsibility
for loss and damage of all kinds, backed by insurance. The Court also closely exam-
ined the wording of the disputed language, including critical words such as “loss of
use” (including, without limitation, loss of use or the cost of use of property, equip-
ment, materials and services including without limitation, those provided by contrac-
tors or subcontracts of every tier or by third parties). The Court also found that the
use of the phrase “without limitation” twice within the clause clearly indicated the
parties’ intention to emphasise the width of the limitation, and that this wording was
“plainly apt” to cover the “spread costs” claimed by Providence. The Court of Appeal
disagreed with the High Court’s reasoning in reaching its decision and went on to
note that a court can reinterpret the contract, so long as the parties’ intent is pre-
served, giving the words that they have used their ordinary and natural meaning. In
so ruling, the court repeatedly emphasised the choice of the parties to accept respon-
sibility for losses that might have otherwise been recoverable as damages for breach
of contract. In viewing the contract as a whole, the court stated that, in their view,
there is no reason to not give efect to the agreement. The Court of Appeal reafrmed
that parties can be rightfully bound by properly executed K4K agreements. As a
consequence, parties must be prepared to accept that, by the operation of these
contracts, claims which may have otherwise been recoverable at law may well be
excluded, as was the matter in this case.71 It is also important to note that over the
last few years, there have been signifcant changes in the ownership of key pipelines
and terminals moving to independent ownership. Consequentially, there has been a
shift in operators’ willingness to undertake certain risks with some commentators
predicting the end of the K4K regime. However, this demise will likely not happen
anytime soon, and K4K clauses remain viable, so long as proper drafting and insur-
ance arrangements can be made between parties, negotiating in good faith.72
8 Conclusive critique
It is an established rule that clauses which tend to exclude liability must be clear and
precise; courts can also hold that such clauses, where they are exclusionary, must
employ specifc and express terms. Limitation clauses are distinguishable from exclu-
sion clauses, and so should indemnity clauses be too, even if backed by emphatic
words, but especially where there is established usage/practice. The word “indemnify”
logically suggests initial liability, usually to a third party. “Hold harmless” goes fur-
ther; an explicit statement of what the party obliged to indemnify the party at loss
is to do (hold them harmless). This chapter is meant to highlight the concept of
mutual indemnifcation in UK ofshore oil industry contracts. The use of indemnity
71 C. Kumarasinghe, “Case Update: Transocean v Providence”, May 2016, HFW, www.hfw.com/Trans-

ocean-Drilling-UK-Ltd-v-Providence-Resources-PLC-May-2016, accessed 27 May 2021.
72 P. Murray, “Is It Time for ‘Knock-for-Knock’ to Be Knocked Out?” 12 May 2018, News and Views,
Ledingham Chalmers Solicitors.
17
clauses has become underscored with words such as “save”, “defend”, “hold harm-
less” and so forth. Nevertheless, it is the position argued here that the meaning of
the clause among industry parties remains: the reciprocal indemnifcation of liability
in accordance with contract terms. This becomes more important when the interests
of a third party to the contract are involved. Where parties intend to exclude liability,
then none exists ab initio; fault will lie where it falls, and there will be no need to
employ the word “indemnify”.73
The contractual practice of knock-for-knock indemnities has been reviewed by
the English courts. In a number of cases, they have examined the meaning of key
concepts such as “negligence”, “gross negligence” and “wilful misconduct” in the
context of the enforcement of liabilities. Their analysis of such contracts begins with
the presumption that a party to a contract is unlikely to intend to absolve another
party entirely from the consequences of the latter party’s negligence. It is therefore
important to provide an indemnity that expressly covers liability in negligence. In
E. E. Caledonia Ltd (formerly Occidental Petroleum (Caledonia) Ltd) v Orbit Valve
Co Europe74 and in Caledonia North Sea Ltd v London Bridge Engineering Ltd,75 the
contractor’s liability arising from negligence had to be clearly provided for in the
indemnity clause or it would not be so construed.
Although the K4K indemnity structure is produced by contract parties, it has been
noticed by the courts within the context of ofshore drilling contracts. In Caledonia
North Sea Ltd v London Bridge Engineering Ltd,76 the court accepted this as an industry
practice which was known to, and accepted by, the courts. In essence, the House of
Lords noted that the indemnity said nothing about requiring the contractors to be
liable to their employees in order for the indemnity to operate; and that it imposed a
general liability to indemnify, with an exception only in cases where the accident was
attributable to the sole negligence or wilful misconduct of the operator. Otherwise,
courts have had little difculty in upholding indemnity provisions which were clearly
worded, and in which the parties had made express provisions for the allocation
of liability. In justifying this position, the court approached the mutual indemnity
structure, not just as a risk allocation mechanism, but also as a tool through which
the real intentions of the contract parties could be discerned, and upheld, especially
when the wording of the indemnity provisions is clear and unambiguous.77
It is also noted that the wording of indemnity clauses is notoriously uncertain with
respect to their enforceability. Insurance can also be inadequate or even non-existent.
The result is that substantial residual liability can rest with the contractor. The insur-
ance market post–Deepwater Horizon appears to be nervous about perceived contrac-
tor liability risk, especially for gross negligence, and capacity is less than $1 billion
and increasingly expensive. Insurance also carries general exclusions for catastrophic
73 C. Ugwuanyi, “Examining the Exclusionary Nature of Oil and Gas Contract Mutual Indemnity Hold
Harmless Clauses”, I.E.L.R. 2012, 4, 136–146.
74 E. E. Caledonia Ltd (formerly Occidental Petroleum (Caledonia) Ltd) v Orbit Valve Co Europe Plc
[1994] 1 W.L.R. 221; [1993] 2 Lloyd’s Rep. 418 QBD (Commercial Court).
75 Caledonia North Sea Ltd v London Bridge Engineering Ltd [2000] S.L.T. 1123.
76 Ibid.
77 P. Cameron, “Liability for Catastrophic Risk in the Oil and Gas Industry”, I.E.L.R. 2012, 6, 207–219.
18
risks from the well, such as blowout and pollution, which are traditionally viewed as
operator risks, indemnifed to the contractor.
With regards to the oil and gas industry, in practice, the latter has what is essen-
tially a global template for the allocation of liability, evident in much of the stan-
dard documentation in the industry, e.g., LOGIC and IADC standard contracts.
Many internationally operating oil companies have master services agreements with
contractors, which can often be global in character. The principal exception to this
is the practices of national oil companies. Such model contract templates provide a
balance between risk and reward and “command and control” realities. However, it
is of note that recent regulatory actions against contractors are throwing this bal-
ance substantially of.78
78 Ibid.
19
CHAPTER 2
An introduction to risk allocation in oil and

gas contracts from an English law perspective1
Greg Gordon
1 Introduction
The oil and gas industry is a hazardous one. The activities involved in exploring
for, producing, transporting and processing volatile hydrocarbons are attended by
a whole host of risks: to people, property, the environment and the valuable com-
modity itself. These activities also frequently require co-operative working by large
numbers of contractors whose interests may be adversely afected in the event of
an accident causing damage to property or personnel, or economic loss, and the
commercial framework of contracts governing these operations is complex. Not
all of these contractors will be in a direct contractual relationship with each other,
and not all of them will necessarily be in a direct contractual relationship with
the operator. The oil and gas industry has developed a number of contracting
practices to allow it to manage these physical2 risks. Generally speaking, up- and
midstream oil and gas contracts seek to depart quite radically from the common
law’s presumptions about how such risk should be allocated.3 As Parchomovsky
and Stavang note in Chapter 3 of this work, “the gas and oil industry has largely
opted out from standard tort liability”. Three vehicles are commonly used to
achieve this re-allocation of risk: (1) indemnity and hold harmless clauses; (2)
clauses which exclude or limit liability for what are commonly, if rather loosely,
described as “consequential losses”; and (3) overall limitations on liability. Each
will be discussed in turn.
1 This chapter is based upon G Gordon, “Risk Allocation in Oilfeld Service Contracts”, in G Gordon, J
Paterson and E Usenmez, UK Oil and Gas Law: Current Practice and Emerging Trends Volume II: Commer-
cial and Contractual Aspects, 3rd Edition, Edinburgh University Press, 2018, pp. 175–234, and is reproduced
with kind permission of Edinburgh University Press.
2 This chapter is concerned with risk of accident and is not concerned with other risk factors such as
political or geological risk.
3 The terminology used to describe these provisions varies. Sometimes they are called “risk allocation”
provisions and sometimes “liability allocation” provisions. Little turns on this, but throughout this chapter,
the expression “risk allocation” shall be preferred simply because at the point when the contracts are drafted
and entered into, the parties are looking at the matter prospectively and are concerned with the potential
consequences of events which may or may not come to pass. While the risk of future liability is inevitable,
actual existence of liability is not; this only crystallizes after a loss-causing event has occurred some time
after the contract entered into force.
20 DOI: 10.4324/9781003206798-2
risk allocation in oil and gas contracts in english law
2 Indemnifcation and related concepts
2.1 Introduction to the concept of indemnifcation

Courtney defnes a contractual indemnity as “a promise to protect another against loss
from an event or events, or set of circumstances”.4 An alternative formulation previously
ofered by the present author is that a contractual indemnity5 is “a contractual provision
whereby the indemnifer agrees to make a payment to the indemnifed party in the event
that the indemnifed party sufers loss as a result of the occurrence of a specifed event”.6
Although there is considerable commonality between the two defnitions, they difer in
at least two respects. Courtney’s defnition relies upon promissory theory, whereas the
alternative does not;7 and Courtney suggests that the obligation is to “protect” against
the fnancial losses associated with an event. This may involve the payment of a sum
or money by the indemnifer to the indemnifed, but may also embrace some other form
of prevention: for instance, the waiver of a claim that would otherwise have existed.
The present author’s formulation, by contrast, is focused upon payment, meaning that
if some form of protection beyond payment is required, this must be obtained by a
device other than indemnity. It is submitted that although Courtney’s defnition is con-
sistent with the way that indemnity has tended to be expressed, the present author’s
formulation is more consistent with the approach adopted in the leading UK Supreme
Court case on the specifc risk allocation provisions used in the oil and gas industry.8
If this is correct – and it must be acknowledged that the conclusion builds out from
only one case, albeit a Supreme Court one – then it means that in order for a risk
allocation clause to operate so as to deal with both waiver and payment, the obligation
needs to be expanded out from a mere “indemnity” by using words such as “indemnify
and protect” or (as is more common in the oil and gas industry) “indemnify and hold
harmless”. The signifcance of this will be discussed further below.
As we shall see below, indemnifcation is a contractual device which the courts have
at times treated with considerable suspicion. The fact that the device subverts – in
fact, frequently inverts – the common law’s default approach to liability has caused
concern.9 It is nevertheless a commonly encountered legal concept, by no means par-
ticular to oil and gas contracts. Indemnifcation lies at the heart of the law of marine10
and fre11 insurance. It also features in construction contracts, where contractors will
4 W Courtney, Contractual Indemnities, Hart, 2014, para 1–2.

5 Not all indemnities are contractual: some arise as a result of operation of law. See Courtney, above n
4, paras 1.5–1.8.
6 G Gordon, “Risk Allocation in Oil and Gas Contracts”, in G Gordon, J Paterson and E Usenmez, Oil
and Gas Law: Current Practice and Emerging Trends, 2nd Edition, Dundee University Press, 2011, para 14.3.
7 For a theoretical account of contract which recasts traditional promissory theory in terms of consent,
see R. Barnett, “Contract Is Not Promise: Contract is Consent”, in G Klass, G Letsas and P Saprai (eds),
Philosophical Foundations of Contract Law, Oxford, 2014, 42–57.
8 Farstad Supply AS v Enviroco Ltd [2010] UKSC 18.
9 I refer here to the common law simply because English law is the focus of this chapter. It is not intended
to imply that civilian jurisdictions are less likely to exercise scrutiny of a knock-for-knock indemnity. See,
e.g., the discussion of German law in Chapter 10 of this book.
10 Marine Insurance Act 1906, s 1: “A contract of marine insurance is a contract whereby the insurer
undertakes to indemnify the assured, in manner and to the extent thereby agreed, against marine losses.”
11 Castellain v Preston (1883) 11 QBD 380, per Brett LJ at 386.
21
greg gordon
commonly provide their employer with indemnities against personal injury or death,
or damage to property, in any way associated with the work contracted for,12 as well
as in the aviation and shipping industries.13 The concept will also be familiar to the
company lawyer: in a corporate acquisition and disposal, the seller will frequently
be asked to provide an indemnity in respect of liabilities incurred by the company in
the period between the deal’s conclusion and its completion.14 Many other examples
could be given.15 For the remainder of this chapter, the one-sided16 indemnity clauses
just described will be referred to as simple indemnity clauses in order to diferentiate
them from mutual indemnity clauses, which will be discussed in greater detail below.
2.2 Indemnity and hold harmless clauses

As we shall see in greater detail below, the oil and gas industry makes extensive use of
a particular form of wording in its risk allocation clauses. A party will very often ofer
not merely to indemnify but to “indemnify and hold harmless” the other party.17 Before
the UK Supreme Court decision in Farstad Supply AS v Enviroco Ltd,18 commentators
on the industry’s risk allocation practice considered that the words “and hold harmless”
added little to the content of such clauses, which were often described simply as “indem-
nity clauses”.19 It is submitted that this approach was both rooted in the long-standing
legal usage of the terms “indemnity” and “hold harmless”20 and accurately mirrored that
12 See the Joint Contracts Tribunal Standard Form of Contract With Quantities available for download
from www.jctltd.co.uk/product/standard-building-contract-with-quantities, (accessed 15 December 2022)
cll 6.1 and 6.2 respectively. For an illustration of cl 6.1.2 in operation, see Scottish & Newcastle plc v G D
Construction (St Albans) Ltd [2003] EWCA Civ 16.
13 In this connection, see, e.g., S Rainey, The Law of Tug and Tow and Ofshore Contracts, 3rd Ed,
Informa, 2011, pp. 155–196.
14 See, e.g., J Young and J Kitching, “Buying and Selling a Business: Warranties and Indemnities” 1995
6(10) ICCLR 336.
15 For instance, one of the leading cases on the construction of indemnities is concerned with a lease:
Canada Steamship Lines v The King [1952] AC 192.
16 The term “one-sided” is here used not to suggest that simple indemnity clauses are necessarily unfair
or biased, merely to denote the fact that in a simple indemnity clause, the indemnity travels in one direction
only – from the indemnifer to the indemnifed party. No reciprocal or mutual indemnity is provided. The
structure is simple in that (as distinct from a mutual indemnity clause) one can tell in advance who will be the
indemnifed and who the indemnifer, in the event that the clause is triggered. Such certainty is not possible
in the case of a mutual indemnity, where the roles of the respective parties will not be known until after an
accident has happened and where the losses lie has become evident.
17 That said, clauses which use the word “indemnify” are not unknown in an oil and gas context, par-
ticularly in contracts relating to maritime matters such as towage: see, e.g., TOWCON cl 18 referred to in A
Turtle Ofshore SA v Superior Trading Inc [2008] EWHC 3034 (Admlty) [2008] 2 CLC 953, a case concerning
the towage of an ill-fated drilling rig.
18 [2010] UKSC 18, 2010 SCLR 379 (hereinafter Farstad v Enviroco).
19 See, e.g., G Gordon, UK Oil and Gas Law (Dundee, 2007), chapter 13; G Gordon, “Indemnifcation
and Contribution: Farstad Supply AS v Enviroco Ltd” (2010) 14 Edin LR 102; T Hewitt, “Who Is to Blame?
Allocating Liability in Upstream Project Contract” (2008) 26 JENRL 177 at 182; T Daintith, G Willoughby
and A Hill, United Kingdom Oil and Gas Law (3rd edn, looseleaf, 2000–date), para 1–845; D Sharp, Ofshore
Oil and Gas Insurance (1994) at p. 108.
20 See, for instance, Firma C-Trade SA v Newcastle Protection and Indemnity Association (The Fanti)
(No 2) [1991] 2 AC 1 (HL) per Lord Gof at 35: “a promise of indemnity is simply a promise to hold the
indemnifed person harmless against a specifed loss or expense.” Courtney, for example, considers the
expressions “indemnity”, “save harmless” and “keep harmless” to all fulfl the same function within the
indemnity clause. See further Courtney, para 8–11.
22
of the oil and gas industry. The oil and gas industry routinely used the terms “indemnify”,
“hold harmless” and “indemnify and hold harmless” interchangeably in its risk allocation
arrangements. Thus, in the LOGIC Standard Contracts for the Oil and Gas Industry,
clauses by which a party indemnifes and holds harmless the other are described simply
as “Indemnities”.21 Similarly, while the industry’s attempt to put in place a contractual
risk allocation regime between ofshore contractors, who would not otherwise have a
contractual relationship, is known throughout the industry as the Industry Mutual Hold
Harmless Deed, the Deed is formally titled the “Mutual Indemnity and Hold Harmless
Deed”.22 Although the Deed’s central risk allocation clause uses the wording “indemnify
and hold harmless”, the clause is titled “Indemnities by the Signatories”23 and the Deed
is itself referred to in the Deed of Adherence (by which parties other than the original
signatories can enter the scheme) as the “Indemnity Deed”. However, in Farstad v Envi-
roco, the Supreme Court held that a clause whereby the owner of a vessel under charter
agreed to “indemnify and hold harmless” the charterer against all liability resulting from
loss of or damage to the vessel was not a mere indemnity clause but a mixed risk alloca-
tion provision containing elements of (on the one hand) indemnity and (on the other)
an exclusion or waiver of liability.24 Whether it operated as an indemnity or an exclusion
would depend upon whether the clause sought to determine who was to bear responsibil-
ity for “third-party exposure” (in which case the clause would be an indemnity) or whether
it resolved “direct exposure to the other contracting party” (in which case the “hold
harmless” dimension would be activated, and it would be an exclusion or waiver of
liability).25 On the facts of the case in question, the owner had sufered damage to his
own property. The case was therefore seen by the Supreme Court as one of “direct
exposure”; hence the clause was, on this occasion, to be seen as an exclusion of liability
clause.26 This would suggest that there will be occasions when such clauses will have to
comply with the provisions of the Unfair Contract Terms Act 1977.27
21 See, e.g., LOGIC, Supply of Major Items of Plant and Equipment (3rd edn), available online at www.
logic-oil.com/content/standard-contracts-0 (accessed 27 April 2017), cl 21.
22 LOGIC, Mutual Indemnity and Hold Harmless Deed, available for download from www.logic-oil.
com/imhh/documents (accessed 2 April 2022).
23 Ibid., cl 2.
24 In so holding, their Lordships laid considerable emphasis upon the fact that the parties to the
contract had titled their clause “Exceptions/Indemnities”. This, thought Lord Clarke (delivering a speech
concurred in by Lord Phillips), was a feature of “particular importance”, and strong evidence of the par-
ties’ intentions (Farstad v Enviroco per Lord Clarke at para 22; see also Lord Mance at para 56). However,
their Lordships seem to have failed to notice that there were a number of indications of a contrary inten-
tion within the clause, including the obligation to exchange “mutual hold harmless indemnities” with other
parties in certain circumstances. This wording might tend to suggest that the words “hold harmless” were
intended only to describe a particular type of indemnity clause. Speculation is to an extent idle, but given
the importance apparently attached to the title of the clause, one cannot help but wonder how their Lord-
ships’ decision would have difered had the parties followed the form of the LOGIC contracts and titled
their clause “Indemnities”.
25 Farstad v Enviroco per Lord Mance, para 59.
26 Ibid. Lord Clarke at para 29 and per Lord Mance at para 59.
27 Although the Unfair Contract Terms Act 1977 has been superseded by the Consumer Rights Act 2015
in the context of consumer contracts, UCTA continues to govern the regulation of exemption clauses in the
business to business context. See, e.g., UCTA s.2(4).
23
greg gordon
These clauses will frequently be supplemented by a provision stating that the indemni-
fer28 will not just indemnify and hold harmless the indemnifed party, but also defend
claims taken against the indemnifed party.29 This has the efect of imposing upon the
indemnifer the burden of conducting the defence of any litigation that may arise, but
also of conferring upon the indemnifer the right to control the manner in which the
defence is conducted. Many indemnifers consider that the beneft of the right to control
the conduct of the defence outweighs the burden of conducting it. Best practice is now
thought to be not to rely solely upon the word “defend” but to include a conduct of
claims clause expressly stipulating the way in which claims are to be handled. Such
a clause is absent in the present draft of the LOGIC standard form contracts, but is
commonly revised into contracts based upon the LOGIC standard forms.
2.3 Mutual indemnity and mutual indemnity and hold harmless clauses
A mutual indemnity – sometimes also called a “reciprocal indemnity”, a “cross-
indemnity” or a “knock-for-knock” indemnity – is a contractual device where the
parties with the one hand give and with the other hand take an indemnity in respect
A accepts liability for its (A’s)

people & property and grants B an
indemnity & hold harmless
provision in respect of any losses B
may suffer in respect of those people
and property, howsoever caused
A B
B accepts liability for its (B’s)

people & property and grants A an
indemnity & hold harmless provision
in respect of any losses A may suffer
in respect of those people and
property, howsoever caused
Figure 2.1 Diagrammatic representation of a mutual indemnity and hold harmless provision
relative to personal injury.
28 Although, as we have seen, the Supreme Court held in Farstad v Enviroco that the words “hold harm-
less” add an additional element to an indemnity clause, meaning that indemnity and hold harmless clauses
will in certain circumstances operate not merely as indemnities but also as exclusion clauses, for the sake of
brevity the parties giving and receiving these clauses will be described as the “indemnifer” and the “indemni-
fed” throughout this chapter.
29 See the observations by Lord President Rodger in the Inner House phase of Caledonia North Sea Ltd v
London Bridge Engineering Ltd 2000 SLT 1123 at 1155. The standard contracts for the oil industry developed
as part of the CRINE initiative and now maintained by LOGIC contain such a provision: see, e.g., LOGIC,
General Conditions of Contract for Services (On- and Of-Shore) (3rd edn, 2014), available for download from
www.logic-oil.com/content/standard-contracts-0 (accessed 27 April 2017) (hereinafter LOGIC, Services) at
cl 19.1: “The contractor shall . . . Save, indemnify, defend and hold harmless”.
24
of related species of loss. A simple example relative to liability for loss arising from
personal injury to employees is given in Figure 2.1.
A mutual indemnity therefore difers from a simple indemnity, where one party
consistently has the burden of giving the indemnity (i.e., is the indemnifer) and the
other party consistently has the beneft of being indemnifed. In a mutual indemnity,
each party is simultaneously both an indemnifer (in relation to one species of loss) and
the indemnifed (in relation to a diferent, but related, species of loss). In the example
given above, both parties are potentially the benefciaries and the recipients of an
indemnity relating to loss arising from personal injury to employees. Which (if any) of
the parties eventually bears liability as the indemnifer or which claims as indemnifer
will depend upon (1) whether an accident occurs and (2) whose employee is hurt.30
It is important to appreciate that, to be efective, a mutual indemnity or mutual
indemnity and hold harmless clause must not be drawn so as to provide that each
party indemnifes (or indemnifes and holds harmless) the other against the occur-
rence of exactly the same loss. To illustrate the point by way of example, let us
imagine that A and B enter into an arrangement where A grants B an indemnity
against B’s house burning down, and B also grants A an indemnity against B’s
house burning down. In such a situation, all the clause succeeds in achieving is
a position where the losses arising if the house burns down is passed from one
party to another ad infnitum. Such a clause (sometimes described as a “circular
indemnity”31) is inefectual32 and leaves the risk it purports to allocate to be borne
by the parties in the way provided for by the law at large. If, however, the example
is altered so that A indemnifes B in respect of the losses incurred by B if B’s
house burns down, and B in turn indemnifes A against the losses that A sufers if
A’s house burns down, circularity is avoided. This is because the species of loss in
respect of which the indemnity is given, although conceptually related (both pertain
to the losses sufered when houses burn down), are not exactly the same: A agrees to
accept the losses to B’s property, and B agrees to accept the losses to A’s property.
3 Indemnity and hold harmless provisions in the oil and gas context
3.1 Introduction to simple indemnity and hold harmless clauses in oil and
gas contracts
Simple indemnity clauses are used in oil and gas contracts in at least two ways.
First, the petroleum industry sometimes just provides the commercial context for
30 In the oil and gas context, it is usual for the parties to enter into not just mutual indemnity provisions
but into mutual indemnity and hold harmless clauses, thus bringing into play the further conceptual issues
described at section 2.3, above.
31 Slessor v Vetco Gray, unreported, 7 July 2006, Court of Session, Outer House, available for down-
load from www.scotcourts.gov.uk/search-judgments/judgment?id=2fbb86a6-8980-69d2-b500-f0000d74aa7
(accessed 15 December 2022). See the submissions of counsel summarised by Lord Glennie at para 6. On
the facts, the court rejected the argument that the indemnity was circular.
32 It is also commercially unrealistic: why on earth would B indemnify A against the loss of B’s own house?
However, the example is given because in practice one does, from time to time, encounter circular indemnities –
almost invariably they arise by accident, when something has gone wrong in the drafting of the indemnity clause.
25
greg gordon
the kind of situation where an indemnity would be commonplace.33 Oil and gas
contracts therefore commonly contain a number of simple indemnity clauses of a
type no diferent to those routinely found in commercial agreements. In addition,
some, but by no means all, up- or midstream oil and gas contracts will also contain
one or more simple indemnity and hold harmless clauses designed to allocate between
the parties some of the risk factors specifc to the petroleum industry. In drilling34
or well services contracts,35 the indemnity and hold harmless provisions are often
more complex than those associated with other operator-to-contractor contracts.
The operator will frequently provide a simple indemnify and hold harmless provi-
sion in favour of the contractor, protecting against risks such as loss of or damage
to the hole,36 blowout, fre, the well becoming uncontrollable, or damage to the
reservoir, geological formation or underground strata, howsoever caused.
The courts have sometimes viewed indemnity clauses with suspicion, on the basis
that when they are found in a contract, this is because a dominant party has imposed
them upon a weaker one.37 However, in UKCS operator-to-contractor agreements,
one-sided indemnity and hold harmless clauses are most commonly granted by the
operator to the contractor.38 Such indemnities are not given because the operator
is weak, but because it is strong. The losses that could accrue in the event that the
well is lost or damaged are potentially very substantial – so large that it might not
be economic, or perhaps even possible, for contractors to obtain insurance against
these contingencies. However, the operator requires the well to be drilled if he is to
produce from the discovery, and is – or traditionally has been; given the changing face
of the industry, this may no longer be universally true – of a sufcient size to absorb
the losses if they come to pass.39 It is (or has been) therefore willing to accept them.
3.2 Introduction to mutual indemnity and hold harmless

clauses in oil and gas contracts
As has already been noted, the oil and gas industry is by no means alone in making
use of indemnity clauses. However, the oil and gas industry utilises indemnity and
hold harmless clauses in a more thoroughgoing way than most other industries. This
33 See the discussion at section 2.1, above.

34 See, e.g., LOGIC, General Conditions of Contract for Mobile Drilling Rigs (2002), available for
download from www.logic-oil.com/content/standard-contracts-0 (accessed 15 December 2022) (hereinafter
LOGIC, Mobile Drilling Rigs) at cl 18.
35 Ibid. at cl 19.
36 However, these indemnities may difer somewhat in their precise content. Note, for instance, that in
LOGIC, Well Services, cl 19.9(a), the operator ofers a full indemnity in respect of loss of or damage to
the hole, while in LOGIC, Mobile Drilling Rigs, cl 18.6(a), an indemnity is given subject to a (quite tightly
confned) carve-out provision in respect of damage to hole caused by contractor’s negligence.
37 See, e.g., the discussion of the Orbit Valve case at section 6.1, below.
38 Perhaps the major exception to this is in the case of ofshore construction works, where the contractor
will commonly be asked to provide a one-sided indemnity to the operator in respect of the recovery, removal
or marking of any wreck or debris associated with the work under the contract.
39 As the UKCS matures and a more diverse set of companies become operators, there will be a greater
need for operators to carry insurance against such risks.
26
is borne out by, for example, the amount of time invested by the court in Caledonia
North Sea Ltd v London Bridge Engineering Ltd (London Bridge)40 in examining the
particular features of the oil industry which give rise to what is still viewed as an
unusual and rather counter-intuitive practice.41 At bottom, the commercial purpose
of an indemnity and hold harmless clause is quite simple. The parties are allocating
(more properly, re-allocating42) between themselves the risk of loss associated with the
occurrence of a particular event. As we have already seen, in the context of a simple
indemnity and hold harmless clause, one party is agreeing that (within the parameters
of this particular contract) it is better placed than the other to bear the risk of a
particular type of loss. By contrast, in the case of a mutual indemnity and hold harm-
less clause, the parties are generally saying that neither of them should have sole
responsibility for a particular species of risk – for instance, the risk that people engaged
on the contract may be injured or killed – but that it is appropriate to divide between
themselves the responsibility for that type of risk. The division does not follow fault-
based principles, but instead follows what might call the identifcation principle, with
the loss being allocated to the party most closely identifed with it. The clause will
therefore commence with the party identifying the aspect(s) of a given type of loss
for which it is willing to take responsibility, and those in respect of which it is not.
Each party then agrees to indemnify and hold harmless the other in respect of the
element of the potential loss that it has accepted, and in return receives the beneft
of an indemnity and hold harmless provision relative to the aspect of the potential
loss accepted by the other party. So if A and B are respectively an oil company and
a contractor who have entered into a contract, in a typical mutual hold harmless
indemnity provision pertaining to the risk that personal injury or death will befall one
or another of the parties’ personnel while engaged on the contract,43 A will confrm
that it accepts responsibility for any injuries or fatalities sufered by A’s own personnel,
however caused, and that it will indemnify and hold harmless B in respect of that
category of loss. B agrees the converse: that it will accept responsibility for any injuries
or fatalities sufered by B’s own personnel, irrespective of how these were caused, and
that it will indemnify and hold harmless A in respect of such loss.44
40 Rather confusingly, as a number of defenders settled the claims against them and dropped out of the
case in the period between the Inner House appeal and the case’s hearing in the House of Lords, the House
of Lords phase of the case is reported as Caledonia North Sea Ltd v British Telecommunications plc. In the
interests of consistency the case will be referred to as “London Bridge” throughout.
41 In the Inner House of the Court of Session (London Bridge, 2000 SLT 1123) see, e.g., section 2.5 of
Lord Rodger’s speech, from 1150, Lord Sutherland at 1174E–F and L, and Lord Gill at 1213F–H. In the
House of Lords (London Bridge, orse British Telecommunications plc 2002 SC (HL) 117 [2002] 1 All ER
(Comm) 321), see Lord Bingham at paras 7–9 and Lord Hofmann at paras 81–82.
42 This exercise does not take place in a vacuum; the law has a pre-existing view on how, in the absence
of agreement, such risks should be borne.
43 Such as may be found throughout the suite of LOGIC Standard Conditions: see, e.g., LOGIC, General
Term and Conditions of Contract for Supply of Major Items of Plant and Equipment (3rd edn, 2015), available
for download from www.logic-oil.com/content/standard-contracts-0 (accessed 15 December 2022) (herein-
after LOGIC, Supply of Major Items), cll 21.1(b) and 21.2(b). For a discussion on the potential impact, fol-
lowing Farstad v Enviroco, of the Unfair Contract Terms Act 1977 upon such clauses, see section. 4.1, below.
44 For a discussion of drafting issues relative to “personnel” and cognate phrases, see the discussion at
section 6.5, below.
27
greg gordon
Mutual indemnity and hold harmless provisions are not used only for cases of
personal injury but will also commonly be agreed in relation to other categories
of loss. An operator-to-contractor agreement will typically also contain such a
clause in respect of loss of or damage to property, where A confrms that it accepts
responsibility for loss of or damage to A’s property, however caused, and that it will
indemnify and hold harmless B in respect of that category of loss; and B agrees
to accept responsibility for loss of or damage to B’s property, howsoever caused,
and indemnifes and holds harmless A relative to such losses.45 Pollution risk will
also sometimes be divided up along similar lines, with the contractor accepting
certain kinds of pollution (commonly, that emanating from its own equipment)
and the operator accepting other kinds (typically, all other instances);46 however,
it is important to note that indemnities for pollution risk are more likely than
those already discussed to be cut into by a qualifcation. Consequential losses
are sometimes dealt with as by way of exclusion or limitation of liability clauses,
but can alternatively be the subject of indemnity and hold harmless provisions.47
3.3 The rationale for a mutual indemnity and hold harmless regime in the oil
and gas context
A number of factors have been advanced as the reason for the oil and gas industry’s
use of mutual indemnity and hold harmless clauses. In the leading work on UK
ofshore oil and gas insurance, the rationale for is presented thus:
If an individual is injured he will expect to have a right to sue any party who may have
been guilty of negligence leading to the circumstances which caused the injury. This party
may be another contractor, the Principal or his employer, or any combination of all three.
The issue can become complicated by reason of contributory negligence. Determining
liability and awarding costs can be a lengthy process in these circumstances, and this can
only add to the anguish of the injured party, or the dependents of the deceased who may
have been the sole breadwinner. The employer therefore accepts a responsibility to provide
for his employees and will generally give the party with whom he is contracting a full
indemnity in respect of any suit or action brought against that other party.48
Sharp’s justifcation was accepted by several of the judges in London Bridge,49 the
main piece of litigation to arise out of the Piper Alpha disaster.50 However, it is only
45 For an example of such sub-clause, see LOGIC, Supply of Major Items, cll 21.1(a) and 21.2(2). For a
discussion of drafting issues concerning the defnition of “property”, see section 6.5, below.
46 See LOGIC, Mobile Drilling Rigs, cll 18.3 and 18.4.
47 See, for instance, LOGIC, General Conditions of Contract (including Guidance Notes) for Ofshore
Decommissioning (2018) available online from www.logic-oil.com/content/standard-contracts0, Clause 25,
p. 29.
48 D Sharp, Ofshore Oil and Gas Insurance (1994) at p. 108. London, Witherby.
49 In the House of Lords phase of the case, reported at 2002 SC (HL) 117 [2002] 1 All ER (Comm) 321, see
the dictum of Lord Bingham at para 7 and that of Lord Hofmann at para 82. In the earlier Inner House phase,
reported at 2000 SLT 1123, see LP Rodger (who is rather more agnostic about Sharp’s justifcation than his
colleagues) at 1150L–1151B, Lord Sutherland at 1174F, Lord Coulsfeld at 1202K and Lord Gill at 1213J–K.
50 The Piper Alpha installation was destroyed in a series of fres and explosions on 6 July 1988 with
the loss of 167 lives. For a detailed account of the cause of the disaster, and its impact upon regulatory law
in the United Kingdom, see J. Paterson, “Health and Safety Law Ofshore”, in G Gordon, J Paterson and
28
partially convincing. By defnition, it can only serve to explain why the industry
adopts such an approach in relation to personal injury; it cannot explain why the
industry takes a virtually identical approach in relation to damage to property,51 or
a broadly similar approach in relation to other matters such as pollution costs and
consequential loss. It is certainly true that by and large the industry prefers swift and
certain resolution to its disputes, and that it does not generally favour time-consuming
and costly litigation.52 But in so far as Sharp suggests that the primary reason for the
existence of the mutual indemnity regime is the industry’s desire to give an efectual
remedy to, or diminish the anguish of, the injured party or his dependants, he would
seem to overstate his position. If that truly were the industry’s intention, it has not
implemented it very efectively. The mutual indemnity and hold harmless regime is
not a system of strict liability so far as the injured worker is concerned; while it
determines who will ultimately pick up the bill for a personal injury claim, for there
to be a personal injury claim in the frst place, the injured worker must establish that
someone is legally liable for the injury. Liability must be established before it is ported
about to its fnal resting place. It is, moreover, unrealistic to suggest that such altruistic
concerns lie at the very heart of the industry’s approach to risk allocation. Diminu-
tion of levels of anguish is more likely to be a fortunate side efect of the practice
than its raison d’être. The principal reasons for the mutual indemnifcation regime
are far more likely to be business ones. This was recognised by Lord President Rodger
in the Court of Session phase of London Bridge when he noted that the practice of
indemnifcation was “fundamental to the economics of the North Sea operation”.53
Insurance (and the broader but related concept of risk management) is the economic
driver that makes this so.54 It may at frst sight be surprising that something which
E Usenmez, UK Oil and Gas Law: Current Practice and Emerging Trends Volume I: Resource Management
and Regulatory Law, 3rd Ed, Edinburgh University Press, 2018, pp. 187–230.
51 While, tragically, it makes perfect sense to talk of the anguish of the family in the context of a fatal
injury, it makes no sense at all to describe property losses in these terms. When an oil tool is lost over the
side of a vessel, it does not leave a grieving spouse and family behind.
52 See, e.g., D Peng, “Mutual Indemnities in North Sea Contracts – Liability and Insurance Clauses” in D
Peng, Insurance and Legal Issues in the Oil Industry, 1993, p. 157, University of Dundee, Centre for Petroleum and
Mineral Law and Policy. Among the main reasons given for the practice are “that it permits the parties to assess and
accept the risks more easily” and that “it avoids delays in claim settlement and it reduces the fghting of lawsuits”.
However, it is not immediately apparent to the present author that indemnifcation reduces disputes. Clauses are
scrutinised carefully before claims are accepted and, if there is disagreement between the parties (or more particu-
larly, between the parties’ insurers) about the proper construction of the clause, litigation will follow which may
prove to be time-consuming and costly: see, e.g., London Bridge. The Piper Alpha disaster occurred on 6 July 1988.
The proof began on 3 March 1993; in all, 391 days of evidence were heard. The case was not fnally concluded until
judgment was handed down in the House of Lords on 7 February 2002. See also the comments of Circuit Judge
Brown in Fontenot v Mesa Petroleum Co, quoted by Lord President Rodger in London Bridge, 2000 SLT 1123 at
1151C–F. It may be that in some cases the fact of indemnifcation brings a quicker resolution to the claim of the
injured party: this seems to have occurred in both London Bridge and Campbell v Conoco (UK) Ltd [2003] 1 All ER
(Comm) 35 at para 6. Even this, however, does not seem to be a universal truth: see the experience of the pursuer
in Slessor v Vetco Gray. The pursuer sufered severe injuries in an accident in May 2003. Liability was in principle
established on 23 March 2007 (see 2007 SLT 400) but even then a number of issues remained outstanding, among
them the construction of the contractual indemnity clause, discussed further at section 6.3, below.
53 London Bridge 2000 SLT 1123 at 1150I.
54 See T Daintith, G Willoughby and A Hill, United Kingdom Oil and Gas Law (3rd edn, looseleaf, 2000–
date), para 1–845: “The client will, any event, normally carry insurance cover for his own employees and
his own property and the cost of this insurance would not be reduced if the particular contractor was also
required to be insured against the same risks. It is thus normal for the client and the contractor to assume
29
greg gordon
seems to be an ancillary matter should be so fundamental. However, in a high-risk
endeavour such as the ofshore oil and gas industry, insurance premia are not mar-
ginal costs but major expenses, made all the more so by the involvement of many
contractors and subcontractors.55 Oil platforms are not stafed wholly, or even mainly,
by the operator’s own personnel. At any given moment in time, one can reasonably
expect there to be representatives from upwards of 20 other companies on board.56
If a large proportion of these companies were required to carry insurance against
the fairly remote, but potentially catastrophic, risk that they might cause or contribute
towards the destruction of the platform and/or widespread injury or loss of life
among those on board,57 then, always assuming that such insurance cover could be
obtained, the cumulative cost of doing so would be very considerable. In addition,
parties’ separate policies would simply run in parallel, and in the event of a cata-
strophic event (assuming that the cause of the calamity could be identifed, and was
attributable to one contractor) only one policy would be claimed upon, and the
remaining contractors’ policies, and that of the operator, would prove to have been
surplus to requirements. This would add considerably to the cost of operations
without adding any value to them. Counter-intuitive as it may at frst appear, on
analysis it can be seen that there are sound economic reasons for the practice of
mutual indemnifcation.
3.4 Back-to-back indemnity and hold harmless provisions

in oil and gas contracts
Although at any given time there may be upwards of 20 to 50 contractors on a
producing platform, only a handful of these parties will be in a direct contractual
relationship with the operator. So for instance, in the production phase of a platform’s
life, the operator will ordinarily enter into a handful of lead contracts through which
it will entrust important parts of the platform’s functions to three or four contractors –
typically a rig services manager, a well services supervisor, and a drilling company.
Most of the other “contractors” represented on the platform will be in a direct con-
tractual relationship not with the operator, but with either the layer of contractors
just described, or with their subcontractors. Viewed from the operator’s perspective,
these parties will be subcontractors, sub-subcontractors, and so on. Thus, at any
given time,58 there will be a number of chains of contractual relationships in place.
full liability, and give each other mutual indemnities, for claims arising out of death of or injury to their own
employees and for loss or damage to their own property . . ., regardless of any negligence or default on the
part of the other party or its employees, agents or sub-contractors.”
55 2002 SC (HL) 117 [2002] 1 All ER (Comm) 321 per Lord Bingham of Cornhill at para 2.
56 For instance, of the 165 people on the platform who lost their lives in the Piper Alpha disaster (167
people died in all; 2 were crew of the fast rescue ship Westhaven), 31 were employed by the operator. The
remaining 134 were employed by 24 diferent contractors: see London Bridge 2002 SC (HL) 117 [2002] 1 All
ER (Comm) 321 per Lord Bingham of Cornhill at para 2.
57 Not all contractors would necessarily be in a position to cause catastrophic loss. It is hard to imagine,
for example, that the catering contractor could cause the total loss of a platform.
58 Although the example focuses on the case of a producing platform, the position is similar in other
phases of the platform’s life. When it is being constructed, overhauled or decommissioned, the usual position
30
The signifcance of the above is that every link in the chain is a contract in which
the parties have to agree how to allocate risk as between themselves. The full eco-
nomic benefts of instituting an indemnity and hold harmless regime do not accrue
if only some of the parties are included in it; moreover, if some parties are part of
the regime and others are not, there is a serious risk of misunderstandings as to
who bears which risk, and of accompanying litigation, gaps in insurance cover and
uninsured losses.59 The general practice is therefore for each of the contracts within
the chain to contain so-called back-to-back provisions. The overall intent of such a
set of provisions is generally that, when all the clauses are read together across the
set of contracts, they should have the efect that, in respect of the risk element(s)
with which the indemnity clauses deal, each party in the chain bears the loss or
damage directly identifed with it, and such loss only. However, achieving this result
requires something of a leap of faith. To make the losses migrate to the appropri-
ate point in the chain, a contracting party has to, in the anterior contract, assume
responsibility not just for the losses identifed with itself, but also for the losses of
the parties below it on the chain. However, in the posterior contract, it will require
to be indemnifed and held harmless by its subcontractor in respect of all losses
identifed with the subcontractor, and any subcontractors lying down the chain of
whatever level. Equally, in the posterior contract, the subcontractor will demand an
indemnity and hold harmless clause from the contractor in respect not just of the
contractor’s losses, but also those of the parties lying above him in the chain. The
contractor will give that in the knowledge that in the anterior contract, he should
already have obtained an indemnity in respect of the losses identifed with the par-
ties above him in the chain. Thus if a set of back-to-back mutual hold harmless
indemnities pertaining to personnel and property operate as the parties intended,
the operator will ultimately carry the risk of injury to or death of its own personnel
and loss of or damage to its own property, but not any like losses sufered by the
lead contractor or its subcontractor; the lead contractor will bear the risk of injury
or death in respect of its own personnel and loss of or damage to its own property,
but not any like losses sufered by the operator or the subcontractor; and likewise
the subcontractor will accept risk in relation to its own property and personnel only.
This is shown diagrammatically in Figure 2.2.
3.5 Qualifed indemnity and hold harmless provisions

As has already been noted, when parties agree to a comprehensive mutual indemnity
and hold harmless regime they agree to bear the risk of loss not on the basis of who
was at fault, but on the basis of who is best placed to insure against the loss or
otherwise absorb it. Sometimes, however, the parties will deviate from this paradigm
and one or more of their provisions will be made subject to qualifcations or so-called
is for the operator to contract with a limited number of lead contractors and for them to let out parcels of
work to appropriate subcontractors.
59 Farstad v Enviroco demonstrates the dangers which can be posed by the interaction between the con-
tractual risk allocation provisions and the statutory law of contribution: see G Gordon, “Indemnifcation,
Exclusion and Contribution: Farstad in the Supreme Court” (2011) 15 EdinLR, pp. 259–265.
31
greg gordon
Contractual provision Intended Effect

Operator
O On the facts given (see Note, below) O carries risk in

relation only to its own personnel and is at the head of
the contractual chain.
O offers C a hold harmless indemnity
in respect of personal injury or death C Under this contract, C carries risk in relation to its
of O’s personnel own personnel and also that of all subcontractors. Thus
in any dispute between O and C, O will expect C to bear
any losses pertaining to those personnel. C, however,
C offers O a hold harmless indemnity has the opportunity, in its contract with SC, to take an
in respect of personal injury or death indemnity back in respect of the sub-contractors, thus
of C’s personnel and that of C’s Contractor attempting to move ultimate liability down the
subcontractors of whichever level contractual chain.
C: Under this contract, C takes the hold harmless

indemnity from SC referred to above; thus when the two
contracts are read together he should be carrying
liability only in respect his own personnel, not the
subcontractors’. However, as regards any dispute
C offers SC a hold harmless between himself and SC, he has had to assume risk in
indemnity in respect of personal relation to his own personnel and those of O. However
injury or death of C’s personnel and the risk to him of so doing is reduced by the terms of the
that of O hold harmless indemnity he has entered into with O in
the contract above.
SC offers C a hold harmless Sub-
SC: The fact that SC obtains an indemnity from C in
indemnity in respect of personal contractor relation to injury to both C and O’s personnel means he
injury or death of SC’s personnel and is able to accept these losses in the contract below as, if
that of SC’s subcontractors of SSC claims, he may pass these losses up the chain to C.
whichever level However, under this contract, in any matter between SC
and C, C is made to carry risk in relation to his own
personnel and those of all SSCs. He therefore needs to
ensure that his agreement with SSC contains a valid hold
harmless indemnity from SSC in relation to this risk.
SC: In any matter between SC and SSC, SC bears the

risk in relation to his own personnel; also that of O
SC offers SSC a hold harmless and C. To avoid ultimate responsibility for these
indemnity in respect of personal Sub-sub- losses SC must ensure that it has the benefit of an
effectual hold harmless indemnity from C in relation
injury or death of SC’s personnel and
that of O and C
contractor to C and O’s potential losses in its contract with C,
above.
SSC offers SC a hold harmless SSC carries risk in relation to his own personnel only
indemnity in respect of personal as, on the facts, there are no sub-contractors beneath
injury or death of SSC’s personnel him in the chain. If there were, SSC would need to
and that of SSC’s subcontractors enter into a further back to back hold harmless
(of which, on the facts given, there indemnity with them.
are none).
Figure 2.2 Simplifed example of a set of back-to-back mutual hold harmless indemnity provi-
sions in respect of personal injury.
Note: This is a simplifed representation of one quite short contractual chain. With regards to any given
producing platform, there will be more than one such chain: the operator is likely to have entered into
direct contracts with at least a handful of parties. The relationship between the parties within this chain
and those in other chains gives rise to further complications discussed at section 7, below.
32
carve-outs. Qualifcations are commonly encountered in the provisions pertaining to
responsibility for injury to or death of third-party personnel, or damage to third-
party property. Here, the parties often state that the indemnity and hold harmless
provision will be ofered only to the extent that the injury, death or damage was
caused by the negligence or breach of duty of the indemnifying party.60 At frst sight
this is a major deviation from the standard indemnity and hold harmless regime. It
is, however, justifable in the case of third-party liability, as unlike the situation where
one takes responsibility for one’s own people and property come what may, neither
party has a close association with a true third party such as would justify a deviation
from the law’s default position on how risk should be allocated.
At the level of contracts between oil companies, it is common for such mutual
indemnity and hold harmless provisions as are granted to be qualifed by the state-
ment that they will not apply in the case of wilful misconduct or gross negligence. The
objective of such a clause is to protect the company from acts of deliberate sabotage
or conduct which falls well below the standard of care which would ordinarily be
expected in such operations. Neither “wilful misconduct” nor “gross negligence” has
a wholly settled meaning in English or Scots law; as a result it is prudent, when the
terms are used, to defne them in the agreement.61 Moreover, some companies (particu-
larly some in the United States) have a corporate policy of not accepting indemnity
and hold harmless provisions which operate in such a way as to permit a contractor
to escape from the consequences of its gross negligence or wilful misconduct. His-
torically, some companies sought to exclude from the ambit of the indemnity and
hold harmless provisions losses attributable to the “sole negligence”62 of the other
party or parties, but this particular carve-out appears to be less common in current
practice. Given the complexity of oil and gas operations and the interdependent way
in which the various parties work, it is, in practice, quite rare for one party’s actions
to be the sole cause of an accident; and even when it is, the time, efort and money
which may have to be expended in order to establish that fact may be very consider-
able. Moreover, prior to the occurrence of the incident, at the point when parties
are mapping their potential liabilities and purchasing the necessary insurance cover,
it is impossible to know if it will be caused by the sole negligence of one or another
of the parties. Thus sole negligence clauses would seem to increase the prospect of
uninsured losses.
Qualifying indemnity provisions is not without its benefts. One can readily under-
stand why an operator would wish an obligation to re-drill to be “carved out” of
the general indemnity and hold harmless provision which a drilling contractor will
usually enjoy relative to loss of hole. However, the practice also has drawbacks. It
adds considerably to the complexity of what are already rather awkward clauses. It
also undercuts the economic benefts provided by the clause: the greater the number
of exceptions carved out of the indemnity and hold harmless regime, the greater the
60 See, e.g., LOGIC, Supply of Major Items, cll 21.1(c) and 21.2(c).
61 Among the matters to be dealt with in the defnition is, e.g., the issue of whose gross negligence or
wilful misconduct is relevant to the clause: for instance, all personnel or senior management only?
62 The indemnity clauses litigated in the London Bridge case were in such terms: see the extracts from the
relevant contracts reproduced at London Bridge, 2000 SLT 1123 at 1126–1129.
33
greg gordon
risk the contractor is exposed to, and the more insurance cover it must purchase.
Deviating from standard practice can also lead parties into error concerning the
precise forms of insurance cover which they require on this particular project. But
ultimately, the purpose of the risk allocation regime in any contract is to express the
parties’ intent, and the relative importance of these various factors, are matters for
the parties to themselves.
4 Selected further issues in indemnifcation law and practice in the UKCS
4.1 Statutory control of indemnity and hold harmless clauses

Unlike some other jurisdictions – for instance, several of the petroleum-producing
states of the United States63 – the United Kingdom has not imposed any specifc
statutory controls on the use of indemnity and hold harmless clauses in the oil and
gas industry. So far as the general body of commercial law statutes is concerned, the
Unfair Contract Terms Act 1977 as amended (hereinafter UCTA) imposes some
relevant restrictions.64 The restrictions imposed by UCTA on the use of indemnity
clauses apply only when the indemnifying party deals as a consumer.65 This, together
with the belief that indemnity and hold harmless clauses were mere indemnities, led
the present author to state in the frst edition of this work that “there are therefore
no statutory controls in force in the United Kingdom which impact upon the parties’
ability to make use of indemnity clauses in commercial oil and gas contracts”. Fol-
lowing Farstad v Enviroco, however, this claim can no longer be made. As indemnity
and hold harmless clauses have now been held to operate as exclusions when they
operate in the context of “direct exposure to the other contracting party”66 (as opposed
to third-party losses), then the restrictions imposed by UCTA relative to exclusion
clauses also need to be considered. Of greatest concern to the oil and gas industry
would be the rule contained in ss 2(1) and 16(1)(a) UCTA,67 that any attempt by a
party to restrict its liability for death or personal injury resulting from negligence
will be inefectual. At frst sight, this provision would seem to be triggered by an
indemnity and hold harmless clause which pertains to losses associated with personal
injury or death and which applies irrespective of negligence; and as Hewitt notes,
“[t]he application of the section appears on its face to be strict and it does not appear
63 So-called anti-indemnity statutes have been enacted in Texas, Louisiana, New Mexico, Wyoming and
Oregon: see P Gerald and H Williams, “Injuries to Third Parties Arising From Oil and Gas Operations:
An Analytical Framework for Examining Indemnity and Additional Insured Issues”, 15 J Nat Resources &
Envtl L 21 at 28–31. For a detailed discussion of the position in Texas, see T Fox, “Return to Certainty in
Risk Assessment, Management and Transfer: The Journey of the Texas Oilfeld Anti-Indemnity Act”, 2001
IELTR 18. See also Chapter 12 of this book.
64 The more extensive provisions of the Consumer Rights Act 2015 can be discounted as they apply only
to consumer contracts. UCTA continues to govern exemption clauses in business-to-business contracts: see,
e.g., UCTA s.2(4).
65 1977 Act, ss 4 (for English law and that of Wales and Northern Ireland) and 18 (Scots law). A party
deals as a consumer when he does not contract in the course of a business but the other party does: see s 12.
66 Farstad v Enviroco per Lord Mance at para 59.
67 Section 2(1) applies in English, Welsh and Northern Irish law and s 16(1)(a) in Scots law.
34
to be possible for the parties to contract out of the Act”.68 However, Hewitt goes on
to note:
indemnities concerning death and personal injury in the context of the oil and gas industry
(even where they are caused by negligence) have been upheld, notably in London Bridge.
This may be because the clauses concerned in London Bridge were not construed as
exclusions of liability for death or injury but rather exclusions of liability for claims by
third parties in respect of the same and were therefore not covered by the strict prohibi-
tion in section 2(1).
Hewitt is, it is submitted, wrong to characterise the mutual hold harmless and indem-
nity clauses in London Bridge as clauses which involved an exclusion of liability. It is
submitted that the clauses were, on their true construction, concerned not with extin-
guishing liability, but with its re-allocation.69 He is, however, surely correct in concep-
tualising the London Bridge clauses as being essentially concerned with third-party
claims. This would seem to mean that they do not seem to operate so as to, using
Lord Mance’s Farstad v Enviroco formulation, “direct exposure to the other contract-
ing party”. Indeed, it would seem to be possible to go further. It would not seem that
any indemnity and hold harmless clause appertaining to liability for personal injury
or death could ever be classifed as “direct exposure”. A company, being an incorporeal
corporation, can never itself sufer personal injury or death. Thus it would appear
that, by defnition, all deaths and personal injuries sufered as a result of the negligence
of a company must be third-party losses for the purposes of Lord Mance’s Farstad
v Enviroco formulation. Thus, in this context of personal injury, indemnity and hold
harmless clauses would seem to inevitably operate as indemnities, not exclusions.
Where, however, the case where the clause does seek to regulate direct exposure to the
other contracting party – as in Farstad itself, where the claim pertained to property
belonging to one of the parties – the clause would seem to have exclusionary efect,
and UCTA (although not argued in Farstad v Enviroco) would appear to be relevant.
However, the provisions engaged would not be the bright-line prohibition upon exclu-
sion contained in s 2(1), but the (less strict) reasonableness requirements contained
elsewhere in the Act.70 The party having the beneft of the indemnity and hold harm-
less clause has less to fear from these provisions than from ss 2(1) and 16(1)(a). Given
the widespread use of indemnity and hold harmless clauses within the industry and
the economic benefts of the practice, one would not expect such a clause to be struck
down by the court other than in very unusual circumstances.
4.2 The law’s default settings: the normal presumptions

about the distribution of risk
The risk distribution exercise that the parties carry out by entering into an indemnity
and hold harmless regime does not take place in isolation. It occurs against the back-
ground of the assumptions which, in the absence of specifc agreement, the law of
68 T Hewitt, (n. 19, above), at p. 205.

69 Thompson v T Lohan (Plant Hire) Ltd [1987] 1 WLR 649. See also Lord Mance’s “direct exposure”
formulation in Farstad v Enviroco, discussed later in this section.
70 I.e. ss 2(2) and 3 (in English, Welsh and Northern Irish law) and ss 16(1)(b) and 17 (in Scots law).
35
greg gordon
contract and tort/delict71 makes about how certain risks are to be allocated. In contract
law, the broad expectation, under both Scots law and the law of England and Wales, is
that, subject to considerations such as remoteness of damage and the need for the non-
breaching party to take reasonable steps to mitigate its loss, the party in breach is obliged
to make good the losses sufered by the non-breaching party.72 This may have the efect
of liability greatly exceeding the commercial value of the contract. If a loss occurs
without either party breaching the contract, then, viewed from the perspective of contract
law, the loss will lie where it falls. In tort/delict, the breach of a statutory duty that
causes a party loss may in some circumstances found an action in reparation. And the
founding principle of the tort/delict of negligence is that where a person owes another
a duty of care and breaches that duty, causing a loss, he is under an obligation to make
a payment of compensatory damages to make good that loss.73 But if there is no breach
of statutory or common law duty, neither party can sue and the loss lies where it falls.
It can therefore be seen that the general law’s default position is that liability follows
breach of contract or breach of duty. Under this model, liability is wedded to fault. By
contrast, as we have seen, carve-outs apart, the indemnity and hold harmless regime is
predicated on the basis that liability should not fow along these lines, but that it should
be accepted by the party best placed to insure against or otherwise absorb that particular
type of loss. Thus the common law and the contractual regime that the industry creates
to govern such matters are not closely aligned. The signifcance of this is that when the
contractual regime fails for some reason – typically, because a clause is not sufciently
clearly drafted, but perhaps, post-Farstad, because it is deemed to be an exclusion clause
and fails one of the tests set out in UCTA – and risk fails to be allocated under the
general law, it will likely be distributed in a way which is very diferent from that which
the parties intended. This may very well mean that one or both of the parties will fnd
themselves facing losses against which they are not insured. The stakes are therefore
high when one is drafting an indemnity and hold harmless clause.
4.3 The position of third parties

While the law of both England and Scotland now permits the beneft of a contract to be
extended to a third party,74 neither jurisdiction permits a contract to impose obligations
upon someone who is not party to it. It is therefore not strictly correct to say that a well-
drafted indemnity and hold harmless clause modifes or overrides the law’s conventional
approach to risk allocation. It does so only so far as the parties to the contract are concerned.
A stranger to the contract – usually described as a third party – who is injured or otherwise
71 The Anglo–American expression “tort” is not used in Scots law, which prefers “delict”. However, at
least in the feld of the tort/delict of negligence, there is little to distinguish the laws of England and Scotland.
72 For the position in England and Wales, see, e.g., G Treitel, The Law of Contract (14th edn, 2015) (here-
inafter Treitel, Contract) at 926 f. For the Scots position, see W McBryde, The Law of Contract in Scotland
(3rd edn, 2007), Scottish Universities Law Institute / W. Green and Son, chapter 22.
73 See Donoghue v Stevenson [1932] AC 562.
74 In England, where the doctrine of privity of contract was traditionally very strong, this is as a result
of the Contracts (Rights of Third Parties) Act 1999. In Scotland the long-standing common-law doctrine of
jus quaesitum tertio permitted third-party rights, albeit subject to some rather onerous qualifying conditions.
Following a review by the Scottish Law Commission, the Scots law of third-party contractual rights was
placed on a statutory footing by the Contract (Third Party Rights) (Scotland) Act 2017.
36
sufers a loss as a result in the course of the execution of the contract works will seek
compensation through the time-honoured route of suing the person or persons whose
negligence and/or breach of statutory duty caused the loss. Sometimes the person against
whom the claim is directed will, by pure coincidence, happen to be the indemnifying party.
Where, however, the third party sues not the indemnifer, but the indemnifed party, a
claim will be made by the indemnifed party under the indemnity and hold harmless
clause. So, if the standard mutual indemnity and hold harmless provision is in place
between companies A and B and C, one of company A’s employees is injured by the
negligence of one of contractor B’s employees, he can be expected to sue B for negligence
and/or breach of statutory duty.75 B cannot defend C’s claim on the basis that it has an
indemnity, as that is irrelevant so far as C, a third party, is concerned. However, B can
make a claim against A as a result of their contractual arrangements. The fact that the
third party is disinterested in the indemnity and hold harmless clause has the potential to
have serious implications for the parties, particularly in the event of insolvency. Unless it
is fortifed either by meaningful guarantees or by adequate levels of insurance, a right to
be indemnifed is only as strong as the fnancial covenant of the company providing it.76
5 Interpreting indemnity and hold harmless clauses
5.1 The traditional approach to interpretation
The general rules

Although the precise formulation of the tests used in the jurisdictions traditionally
difered, it is possible to say that both English and Scots law have traditionally taken
an objective approach to the interpretation of contracts.77 Although objective, neither
jurisdiction has been wholly literal in approach:78 both jurisdictions require the clause
or phrase under discussion to be interpreted not in isolation, but in the light of the writ-
ten document as a whole.79 Both have permitted evidence as to the objective factual
background to be adduced with a view to allowing the court to determine the aim
or thrust of the agreement,80 albeit there are diferences between the two systems
surrounding precisely when this is to be permitted.81
75 Depending on the facts, the injured employee may in addition be able to raise a case directly against
A in respect of A’s breach of its non-delegable “duty to take reasonable care to see that its employees are
safe”. See J Steele, “Employers Liability” in M Jones et al, Clerk and Lindsell on Torts, Sweet and Maxwell,
23rd Ed, 2020, para 12.14.
76 See, e.g., T Taylor, Knock for Knock Revisited, Blog-Post, Clyde and Co website, 20 February 2013,
available at www.clydeco.com/blog/energy/article/knock-for-knock-revisited; P. Roberts, Petroleum Con-
tracts: English Law and Practice (2nd Ed) Oxford, 2016 para 13.11.
77 For the traditional position in England, see K Lewison, The Interpretation of Contracts (2nd edn, 1997)
(hereinafter Lewison, Interpretation) at para 1.05. For the position in Scotland, see Scottish Law Commission, Scot
Law Com No 160, Report on Interpretation in Private Law, available for download from: www.scotlawcom.gov.uk/
fles/1512/7989/6878/rep160.pdf (accessed 27 April 2017) (hereinafter SLC, Report on Interpretation) at para 2.3.
78 SLC, Report on Interpretation, para 2.1.
79 See, e.g. (in the law of England and Wales), Re Jodrell (1890) 44 Ch D 590; for a Scots example, see
Glen’s Trs v Lancashire and Yorkshire Accident Insurance Co Ltd (1906) 8 F 915.
80 See Lewison, Interpretation at para 2.10.
81 In England, even before Lord Hofmann’s restatement in Investors Compensation Scheme Ltd. v
West Bromwich Building Society [1997] UKHL 28 (discussed at p. 39 below), and particularly thereafter, a
37
greg gordon
Interpretation contra proferentem
In addition to the generalities expressed above, both English and Scots law have
traditionally deployed the contra proferentem rule to interpret exemption, limitation
and indemnity clauses.82 An ongoing problem in the law of contractual interpretation
is the extent to which individual canons of construction, such as contra proferentem,
have continued to survive the upheavals in the general law of contractual interpreta-
tion. For present purposes, a precautionary approach will be adopted which assumes
that such rules continue to be in efect and that contractual draftsmen should continue
to draft as if such rules remain in efect. However, this is intended as an indication
of prudent drafting practice, rather than an indication that it is the optimal way of
interpreting indemnity and hold harmless clauses. The current author would argue
for a more naturalistic approach which recognises the normalcy of indemnity and
hold harmless clauses in the oil and gas context and does not call for artifcial or
strained interpretations of such clauses, at least in that particular context.
Contra proferentem means that “[a]mbiguous words in exemption clauses [will be]
construed in the way least favourable to the party relying on them”.83 In the context of
a simple indemnity, the party relying on the clause will be the one who stands to receive
the beneft of it – i.e., the indemnifed party. But who is the proferens when the clause
is mutual? The prevailing view appears to be that whichever party has the misfortune
to have a claim directed towards it (and therefore requests indemnifcation under the
contractual risk allocation provisions) is to be treated as the proferens.84 This, however,
reduces the question to a matter of happenstance. The arrangement is mutual – indem-
nity and hold harmless provisions travel in either direction – and had the facts been
diferent the indemnity might have run in the opposite direction. It seems to be artifcial,
in such a situation, to contend that there is a proferens at all. This fact was recognised
in the Scottish Outer House case of Slessor v Vetco Gray, where Lord Glennie stated:
I accept Mr Armstrong’s submission that the contra proferentem approach, which in any
event only applies in a case of ambiguity, has much less impact where the exemptions
and indemnities are mutual or reciprocal. Both parties are, in a sense, the proferens; and
it makes little sense to construe the clause against each one of them leaving the possibility
of a hole in the middle.85
willingness to admit “matrix of fact” evidence can be seen: see Lewison, Interpretation (n. 77, above), para
2.10. In Scotland, evidence as to surrounding circumstances is admissible only in the case of ambiguity, or
if the contract was unintelligible without it: SLC, Report on Interpretation, para 2.3; see in particular the
materials cited in n 7 therein. In Arnold v Britton [2015] UKSC 36, which has supplanted ICS as the leading
English case in this area, the Supreme Court clearly wishes the courts to focus primarily upon the natural
meaning of the words used by the parties and to have recourse to external factors such as commercial com-
mon sense only in “unusual” cases where the wording is not otherwise intelligible: Arnold v Britton, per Lord
Neuberger at para 17. Thus it may be that the efect of Arnold v Britton will be to close the gap between the
Scots and English approach.
82 For an excellent account of this area, see E Peel, “Whither Contra Proferentem?”, in A Burrows and
E Peel, Contract Terms (2007), pp. 61–75, Oxford University Press.
83 Treitel, Contract (n. 72, above), p. 221.
84 Orbit Valve per Steyn LJ at 182g–h; London Bridge, (Inner House) 2000 SLT 1123 per LP Rodger at
1148C–L and per Lord Sutherland at 1174L. The House of Lords passed no concluded view upon the mat-
ter as their Lordships took the view that the clauses were clear and unambiguous: London Bridge (House of
Lords) 2002 SC (HL) 117 [2002] 1 All ER (Comm) 321) per Lord Mackay of Clashfern at para 43.
85 Slessor v Vetco Gray at para 12.
38
It is respectfully suggested that this approach is logical and takes cognisance of the
realities of the parties’ contractual arrangements. It therefore has much to commend
it. However, on the balance of current authority, it cannot be said to be the estab-
lished view.
Towards contextualism (and back again)

The traditional approach referred to above requires now, at least in England, and possibly
also in Scotland,86 to be considered in the light of subsequent developments. In Investors
Compensation Scheme Ltd v West Bromwich Building Society,87 Lord Hofmann ofered
a dictum which has come to be known as Lord Hofmann’s restatement of the English
law of contractual interpretation. A comprehensive examination of the restatement is
outside the scope of this chapter. However, putting the matter shortly, Lord Hofmann
suggested that two cases decided by the House of Lords in the 1970s88 had efected a
quiet revolution: a shift away from formalism towards contextual interpretation. He
considered the efect of these cases had been to discard “[a]lmost all of the old intellectual
baggage of ‘legal’ interpretation”.89 In particular, Lord Hofmann confrmed that con-
tractual interpretation was still in a sense an objective exercise, in that it continued to be
a matter of ascertaining what a reasonable and objective bystander would make of the
parties’ communings. However, he added that such a bystander should be presumed to
be aware of at least most of90 the background knowledge and circumstances (otherwise
the “matrix of fact”) which comprised the setting in which those communings took place.
Moreover, as “the meaning of the document is what the parties using those words against
the relevant background would reasonably have been understood to mean”, the bystander’s
knowledge of the background circumstances might from time to time prompt him to
conclude that the parties must have used the wrong words. Such a conclusion would be
reached rarely, as the courts would not readily accept that the parties had made linguistic
mistakes; however, when an appropriate case did arise, the reasonable bystander should
not attribute to the parties an intention which they plainly could not have had.91 This
test, as expressed in ICS, involved a greater emphasis upon matrix of fact evidence than
had previously been seen. In ICS itself, the approach led the court to conclude that
something had gone wrong with the drafting of the clause in question, as a result of
which the court was willing to re-order the wording of the clause in order to give it
greater commercial sense. Other striking examples of what might be described as remedial
86 The extent to which the restatement has been received into Scots law is presently unclear: see D
Cabrelli, Commercial Agreements in Scotland: Law and Practice (2006) at paras 2.03–2.24 and McBryde,
Contract (n. 72, above) (3rd edn), paras 8.25–8.27.
87 Investors Compensation Scheme Ltd v West Bromwich Building Society (hereinafter ICS) [1998] 1
WLR 896.
88 I.e., Prenn v Simmonds [1971] 1 WLR 1381 and Reardon Smith Line Ltd v Hansen-Tangen [1976] 1
WLR 989.
89 At frst sight, this might appear to be a reference to formal legal rules such as contra proferentem: see
E McKendrick, Contract Law: Cases, Text and Materials (7th edn, 2016) at 381; Peel, “Whither Contra Pro-
ferentem?”, p. 61, (n. 82, above). However, as McKendrick notes, appearances can be deceptive: see further
the discussion of HIH Casualty & General Insurance at section 6.1, below.
90 Previous negotiations and declarations of the parties’ subjective intent continue to be excluded from
consideration.
91 Investors Compensation Scheme [1998] 1 WLR 896 per Lord Hofmann at 912–913.
39
greg gordon
re-drafting were seen in The Starsin92 and in Chartbrook Ltd v Persimmon Homes Ltd.93
In the now-leading case of Arnold v Britton,94 the Supreme Court took the opportunity
to restate the restatement. While Arnold v Britton does not render matrix of fact evidence
irrelevant in all circumstances, the court strongly emphasised the centrality to the inter-
pretative exercise of the words used by the parties themselves, and relegated the role of
external factors such as commercial context to supplementary criteria to which regard
could only be had in “very unusual” circumstances.95 In spite of Lord Hofmann’s appar-
ent belief that his dictum was not a radical departure, but merely a convenient synthesis
of what the law already said, the restatement proved to be controversial, and there can
be no doubt that Arnold v Britton represents a conscious reining-in of the contextual
approach. It should be noted, however, that what Arnold v Britton promotes instead of
contextualism is literalism – the fnding of meaning primarily from the words used by
the parties themselves – rather than canonical interpretation. Thus, while following Arnold
v Britton, a court may be slower than previously to have regard to factors such as com-
mercial common sense when interpreting an indemnity clause, Arnold v Britton provides
no obvious support for the practice of adopting strained or artifcial constructions of
clauses in order to comply with the old canons of construction. Yet neither are these old
rules expressly swept away. In these circumstances, a penumbra of doubt remains, and
the only prudent approach is for contractual draftsmen charged with the task of drafting
indemnity and hold harmless clauses to attempt to cross the highest barrier to imple-
mentation that they face, and to proceed on the basis that indemnity and hold harmless
clauses will continue to be construed contra proferentem. Drafting in this manner should
minimise the prospect of a dispute arising as to the true meaning of the clause. However,
in the event that an indemnity clause is challenged on the ground that it does not satisfy
one of the tests established by the traditional canons of interpretation, very careful
consideration should be given to advancing the argument that all that is needed for the
clause to operate is for this to be the ordinary consequence of the wording used by the
parties.96 In addition, if the contract in question does not accurately express the parties’
intentions, advice should be taken on the possibility of rectifying the deed.97
6 Some known problems of drafting and interpretation
6.1 Contra proferentem and the problem of negligence and

breach of statutory duty
As has already been noted, when the parties to an oil and gas contract opt to include
an indemnity and hold harmless regime in their contract, they are choosing to
92 [2003] UKHL 12; [2004] 1 AC 715.

93 [2009] UKHL 38; [2009] 1 AC 1101.
94 [2015] UKSC 36.
95 Per Lord Neuberger at para 17.
96 The recent evolution of the rule of contra proferentem is further considered immediately below.
97 For a general discussion on rectifcation of deeds, see (from the standpoint of English law) Lord
Mackay of Clashfern (General Editor), Halsbury’s Laws of England (4th edn, 2003 Reissue) (hereinafter
Halsbury), vol 32, paras 52–69. For the position under Scots law, see J Murray and S Wolfe, “Judicial and
Other Remedies”, The Laws of Scotland: Stair Memorial Encyclopaedia, vol 13 (1992), paras 69–70.
40
superimpose their own views on risk allocation upon the one provided by the com-
mon law. As we have seen, the common law generally expects liability to follow breach
of contract or breach of duty. The courts have therefore considered it to be “a
fundamental consideration in the construction of contracts of this kind that it is
inherently improbable that one party to a contract should intend to absolve the other
from the consequences of his own negligence”,98 and, have considered it to be, if
anything, even less likely that one party would go so far as to accept liability for
another party’s negligence.99 While the courts stop short of saying that the parties
cannot use the terms of their contracts to re-allocate risk in the way in which they
see ft, they have traditionally invoked the contra proferentem rule to say that the
parties must use clear language when they do.100 Canada Steamship Lines Ltd v The
King101 is authority for the proposition that efect will be given to a term which
expressly states that the indemnifed party is to be relieved from the consequences
of its own negligence. This could be described as the primary route to enforceability
under Canada Steamship – express provision.102 However, the same case also provides
that, if the clause does not expressly indemnify in respect of negligence or some
synonym for it, it will operate to relieve the indemnifed party from the consequences
of its own negligence only if (1) the words used in the clause are “wide enough in
their ordinary meaning to cover negligence” on the part of the indemnifed party,
and (2) no alternative ground of liability which is neither too remote or too fanciful
to be in the parties’ contemplation might exist. A party able to satisfy both elements
of this test could be described as having identifed a secondary route to enforceability
under Canada Steamship. The question of whether or not an indemnity provision
will protect the indemnifed against the consequences of its own negligence is of
considerable practical importance. Unless there have been clear instructions that the
indemnifed party’s own negligence is to be excluded from the ambit of the indemnity
and hold harmless clause, a provision which is inefectual against the indemnifed
party’s own negligence is unlikely to be what a party to an oil and gas industry
contract wanted. Worse, it is unlikely to match with the risks that the party has
insured against. The Orbit Valve case demonstrates the risks involved, while the Nelson
case illustrates an alternative approach to this problem. In Orbit Valve, a case deter-
mined under English law, the employers of a man killed in the Piper Alpha disaster
were sued by the operator, who sought reimbursement of the compensation which
the operator had paid to settle a claim by the deceased’s family. The operator, who
accepted that the negligent actions of one of its employees had at least contributed
towards the disaster, claimed that it was entitled to recover under the contractual
98 Gillespie Bros & Co Ltd v Roy Bowles Transport Ltd [1973] QB 400 per Buckley LJ at 419.
99 Smith v UMB Chrysler (Scotland) Ltd 1978 SC (HL) 1 per Viscount Dilhorne at 7.
100 See, e.g., London Bridge (Inner House) 2000 SLT 1123 per LP Rodger at 1148K–L. See also Orbit
Valve [1993] 4 All ER 165 (afrmed by the Court of Appeal: see [1995] 1 All ER 174) per Hobhouse J at 173f:
“The parties are always able, by the choice of appropriate language, to draft their contract so as to produce
a diferent legal efect. The choice is theirs. In the present case, there would have been no problem in drafting
the contract so as to produce the result for which the plaintifs have contended; however, the contract was
not so drafted and contains only generally working and is seriously lacking in clarity.”
101 Canada Steamship Lines Ltd v The King [1952] AC 192 (hereinafter Canada Steamship).
102 Ibid. per Lord Morton of Shuna at 208.
41
greg gordon
indemnity and hold harmless clause. However, the contractor contended that the
clause was inefectual: the operator was seeking to be reimbursed for a loss incurred
at least in part as a result of its own negligence; however, the clause did not expressly
state that it would operate in these circumstances. Applying Canada Steamship, the
Court of Appeal held that the clause contained words which were wide enough to
have the efect of conferring an indemnity upon the indemnifed party relative to the
consequences of its own negligence. However, it refused to give efect to the clause
on the basis that the second limb of the Canada Steamship test was not satisfed.
The words were not necessarily directed towards the indemnifed party’s negligence,
but could just as easily have been intended to exclude another head of claim, namely
breach of statutory duty.103 This head of claim was not either too remote or too
fanciful to be in the parties’ contemplation. The claim therefore failed.
The result in Orbit Valve has received some support. Rainey, a leading commentator
and practitioner in the feld, recently endorsed the question posed by Lord Steyn in
Orbit Valve: “Why do draftsmen not take note of the impact of a clear and consis-
tent line of judicial decisions?”104 That is a fair question. It would certainly make life
easier if draftsmen were aware of these rules and followed them. However, it could be
argued that the more fundamental questions to arise from Orbit Valve pertain to the
reasoning of the court, rather than the drafting of the parties. Is it really appropriate
for the court to take a clause that is, in its own terms, perfectly clear, deem it ambigu-
ous because it does not expressly deal with a matter likely to be of no signifcance to
the parties in question, and thereby render the clause unusable? How are the interests
of justice served when an artifcial construction is applied to perfectly clear wording,
leading to an unexpected result and uninsured losses? Does the occasional and highly
specifc use of canonical construction really make the law clearer or fairer when the
law generally favours an approach based upon textual analysis?
In the Scots case of Nelson,105 the court was faced with a factual situation and con-
tractual clause that was very similar to that in Orbit Valve. A worker employed by a
contractor was injured on an ofshore installation. The operator settled the employee’s
personal injury claim and intimated a claim under the indemnity against the employer
of the injured worker. The indemnity and hold harmless provision envisaged that the
employer would take responsibility for “any and all losses, claims, suits, demands . . .
and causes of action in respect of death or of injury to [the employer’s] personnel . . .
howsoever caused”. The employer argued that the indemnity did not apply in cir-
cumstances where the indemnifed party was seeking to be protected against its own
negligence. However, the court rejected this argument. At frst instance, the judge
attached great weight to the fact that the integrity of the reciprocal arrangement
of risk allocation would be seriously undermined if the employer’s argument were
successful; there was no evidence that the parties considered the issue of negligence
to be especially important in their system of risk allocation, and to give primacy to
this issue would disrupt the parties’ own carefully considered scheme. On appeal, the
103 Orbit Valve [1995]1 All ER 174 per Steyn LJ at 181h–j.

104 S Rainey, “Construction of Mutual Indemnity and Knock-for-Knock Clauses”, in B Soyer and A
Tettenborn (eds), Ofshore Contracts and Liabilities (2015), 67–107 at 107, Informa.
105 Nelson v Atlantic Power and Gas Ltd, 1995 SLT 102.
42
Inner House of the Court of Session upheld his judgment. The Inner House endorsed
the reasoning of the Lord Ordinary, and considered that it fortifed its view that the
clause should be given efect to.106 However – while there is an intriguing hint that
the court would have been willing to distinguish Canada Steamship and decide on
the basis of the Lord Ordinary’s analysis had this been necessary107 – the primary
basis for the Inner House’s decision was its interpretation of the requirements of
the Canada Steamship test.108 The Inner House considered that where (as here) the
indemnifed party’s actings causing the loss could be characterised both as negligent
and a breach of statutory duty, then while Canada Steamship might prevent the
operation of the indemnity in relation to negligence, it did not preclude the indemnity
from taking efect in relation to the breach of statutory duty.109 This is a markedly
diferent approach to Canada Steamship than that taken in Orbit Valve.
The House of Lords had the opportunity to consider the construction of oil and
gas risk allocation clauses in London Bridge, a Scots case which (like Orbit Valve)
arose out of the Piper Alpha disaster. Unfortunately, however, the court did not have
the opportunity to settle which of the two competing approaches was to be preferred,
as the wording of the contractual risk allocation clauses was materially diferent to
those discussed in Nelson and Orbit Valve.
In London Bridge, the operator sought to use indemnity and hold harmless provi-
sions in order to recover from the various employers of personnel injured or killed
in the disaster sums in respect of damages paid in order to settle the personal injury
claims of the injured and the families of the deceased. After proof, it was held that
the accident had occurred as a result of both the negligence of the operator and
that of an employee of a specialist valve contractor.110 The operator was again met
with the argument that the clause did not provide it with a remedy as it required
to be interpreted contra proferentem,111 and did not deal sufciently clearly with the
issue of negligence to satisfy the Canada Steamship requirements. As noted above,
this time, the indemnity clauses112 were drafted diferently to the clauses in Orbit
Valve and Nelson. The clauses were rather inelegantly expressed and did not address
the question of negligence as expressly as one might have desired. However, they
did make express provision for how issues of contributory negligence were to be
resolved.113 This fact was sufcient to permit the court to conclude that the parties
had intended that the mutual indemnity and hold harmless provision would take
efect even in circumstances where one of the parties was seeking protection from
106 Ibid. per the Opinion of the Court at 104.

107 Ibid.
108 Ibid., pp. 102–103.
109 Ibid., 103.
110 This fnding was signifcant as the relevant clause contained a “sole-negligence” carve-out, and had
liability not been at least partially the fault of another party, the pursuer’s claim would have failed: London
Bridge (Inner House) 2000 SLT 1123 per LP Rodger at 1131.
111 See also the discussion at section 6.1, above.
112 The case involved a multiplicity of parties; seven separate sets of indemnity provisions were under
consideration.
113 See the extracts from the relevant contracts reproduced at London Bridge 2000 SLT 1123 at
1126–1129.
43
greg gordon
consequences of its own negligence.114 The claim for indemnity therefore succeeded.
Although London Bridge was concerned with a clause that was markedly diferent
from that in Orbit Valve, it can still be noted that the overall tenor of London Bridge
difered from that in Orbit Valve, with the court in London Bridge being led more by
the wording of the clause itself and showing markedly less concern that the clause
was a trap for the unwary.
Away from the oil and gas context, the House of Lords has had an opportunity
to consider the continued role for the rules in Canada Steamship following Lord
Hofmann’s restatement. HIH Casualty & General Insurance Ltd v Chase Manhattan
Bank115 (HIH Casualty & General Insurance) was concerned with the interpretation of
an insurance policy designed to pay out to the insured (who were investors in flms)
in the event that the flms in which they had invested did not make enough money
to permit the repayment of the investor’s loans. This rather specialised “high risk,
high-premium”116 insurance product was designed and marketed by a frm of brokers
who also acted as the insured’s agent when presenting the proposal to the insurers.
The policy contained a “truth of statement” clause which, put shortly, provided that
the insured (i.e., the investors) “would have no liability of any nature to the insur-
ers for any information provided by any other parties”, and in particular that any
information either provided or not disclosed by the brokers would not be a ground
by which the insurers could avoid or escape their liability to make payment under
the policy. This clause was included in the policy in order to protect the insured
against the risk of unwittingly becoming responsible for anything said or known by
the many other players involved in the procurement of fnance for any of the flms in
question.117 In the event, there were substantial shortfalls in the flms’ revenue and the
investors claimed under the policies. The insurers repudiated liability on the grounds
of misrepresentation and non-disclosure, either fraudulent or negligent, on the part
of the broker. The insurers contended that the wording of the “truth of statement”
clause was not sufciently specifc to meet the terms of the Canada Steamship rules
in that although the wording was sufciently wide to extend to negligence or fraud,
it could just as readily be intended to exclude other causes of action which it was
not fanciful to imagine the parties to have had in contemplation, such as innocent
misrepresentation or non-disclosure. The House of Lords agreed relative to fraud,
holding, by majority, it to be “a thing apart”,118 a species of liability which “must be
114 See London Bridge (House of Lords) 2002 SC (HL) 117 [2002] 1 All ER (Comm) 321 per Lord
Mackay of Clashfern at paras 40–43.
115 [2003] UKHL 6 [2003] 2 Lloyd’s Rep 297. For an excellent summary, see E Peel, “Whither Contra
Proferentem?” (n. 82, above), at pp. 61–64.
116 Lord Hofmann at para 25. This paragraph contains an evocative description of the flm investment
industry which the author commends to the reader.
117 A multiplicity of parties were involved, including the broker: see HIH Casualty & General Insurance
per Lord Hofmann at paras 26–33. The insured needed to be protected against the risk (not a fanciful one,
given the manifold duties generally imposed by the law upon an insured, the contract being one of utmost
good faith) of being deemed to have constructive knowledge of things known by persons deemed to be the
insured’s agent.
118 HIH Casualty & General Insurance per Lord Bingham at para 15. His Lordship continued: “Parties
entering into a commercial contract will no doubt recognise and accept the risk of errors and omissions in
the preceding negotiations, even negligent errors and omissions. But each party will assume the honesty and
good faith of the other; absent such an assumption they would not deal.”
44
excluded in clear and unmistakeable terms on the face of the contract”.119 Importantly,
however, the House of Lords unanimously rejected the insurers’ argument relative to
negligence. Although, as Peel notes, “[t]here was no suggestion from their Lordships
that [the possibility that the parties might have had in contemplation innocent mis-
representation or non-disclosure] was regarded as fanciful”,120 the so-called Canada
Steamship rules were relegated to the status of helpful but non-determinative guid-
ance121 as opposed to a code. As Lord Bingham put it:
The passage does not provide a litmus test which, applied to the terms of the contract,
yields a certain and predictable result. The courts’ task of ascertaining what the particular
parties intended, in their particular commercial context, remains.122
In turning to that task, the House of Lords had no hesitation in holding that the
parties had used “comprehensive language, clearly chosen to give [the investor] an
extended immunity”.123 It found “nothing commercially surprising in this
interpretation”,124 as in a transaction of this kind “the possibility that [the broker]
might make and fail to correct a representation which was later held to be both
untrue and negligent would be very real”.125 Thus, while it would appear that contra
proferentem has survived the purge of “[a]lmost all of the old intellectual baggage
of ‘legal’ interpretation”, it has not emerged unchanged from its brush with Lord
Hofmann’s restatement. Fraud continues to be “a thing apart”; negligence, it would
seem, is not. As HIH was decided at the height of the contextual era, its authority
might be thought to be questionable following Arnold v Britton. However, while the
case does, at points, advert to considerations of commercial common sense, the judg-
ment is fundamentally a close textual analysis of the wording of the insurance policy
in question. As such, it is submitted that it is compatible with Arnold v Britton.
The indemnity and hold harmless clauses presently in wide use within the oil and
gas industry make express reference to the parties’ intention that the clauses will be
efectual irrespective of negligence or breach of duty, whether statutory or otherwise.126
It would seem to be foolhardy for contractual draftsmen deliberately to deviate from
that practice. However, in the event that a clause is, upon examination following inti-
mation of a claim, found to have been drafted in a way which is less explicit on this
119 Ibid. per Lord Bingham at para 16.

120 E Peel, “Whither Contra Proferentem?” (n. 82, above), at 62.
121 As Lord Hofmann notes at paras 61–63, they had also been characterised as mere guidelines not to
be mechanistically followed in Smith v South Wales Switchgear Co Ltd [1978] 1 WLR 165; Hollier v Ram-
bler Motors (AMC) Ltd [1972] 2 QB 71; and Ailsa Craig Fishing Co Ltd v Malvern Fishing Co Ltd [1983]
1 WLR 964.
122 HIH Casualty & General Insurance per Lord Bingham at 11.
123 Ibid. per Lord Bingham at 12.
124 Ibid. See also Lord Hofmann at 67: “There is no inherent improbability in such an intention. As
Rix LJ said, in a case like this the question of negligence can never be all that far from the contemplation
of the parties.”
125 HIH Casualty & General Insurance per Lord Bingham at 13. See also Lord Hofmann at 67: “And it
seems to me that the commercial objective of the Truth of Statement clause would be substantially under-
mined if Chase’s right to the policy monies depended upon an inquiry into whether Heaths had or had not
taken reasonable care in checking the truth of representations or deciding which facts should be disclosed.”
126 See, e.g., LOGIC, Mobile Drilling Rigs at cl 18.8; see also the discussion the IMHH Deed at section
7.2, below.
45
greg gordon
point than one might wish, London Bridge and HIH Casualty & General Insurance
suggest that the courts may be willing to take a somewhat more forgiving approach
than that seen in Orbit Valve. More radically – and more speculatively – if Lord Glen-
nie’s dictum in Slessor127 fnds favour, then the contra proferentem rule will be found
to be wholly irrelevant in the case of mutual indemnity and hold harmless clauses.128
6.2 Words delimiting the circumstances in which the indemnity and hold
harmless provision will take efect
Another important issue which arises in the drafting of an indemnity and hold harm-
less clause is the need to delimit the set of circumstances in which the indemnity is
to take efect. The indemnity and hold harmless provision will not be being given
generally, for all times and for all circumstances. Instead, it is given because it has
been rendered necessary by the fact that one party intends to carry out work under
a particular contract. Words need to be included in the clause which make this clear
and state the circumstances in which the provision is to take efect. If this is not
done, the provision is in danger of being either struck down as a result of its inde-
terminate scope, or of applying in circumstances not intended by the parties. However,
and again, exercising prudence as a result of the possible continued application of
the contra proferentem rule, draftsmen must give careful thought to the words they
use. It is dangerous to draw too narrow a connection between the operation of the
indemnity and the scope of work under the contract. This is so because there are
many circumstances in which property may be lost or personnel injured or killed
while they are on an operator’s platform. Taking the example of an injury on an
ofshore platform, such an event may befall a worker as a result of an accident directly
connected to the scope of work to be done under the contract. However, it may also
occur as a result of an accident occurring while he was carrying out work, but for
reasons wholly unconnected to it, or while he is on the platform but not engaged in
this work.129
Smith v South Wales Switchgear is authority for the proposition that if an indem-
nity is drawn so that it takes efect only if injuries are sufered during the “execution
of this Order”, it will capture only injuries occurring as a result of the doing of the
contractual work.130 The dangers that this judgment poses to the party claiming the
indemnity are clear. In Campbell v Conoco (UK) Ltd,131 it was argued that a clause
which provided that the subcontractor would indemnify and hold harmless the lead
contractor132 against all injuries sufered by the subcontractor’s employees “as a result
127 Discussed at section 5.1, above.

128 Lord Glennie’s dictum would not prevent the rule from continuing in operation in the case of simple
indemnities.
129 A worker may, for example, be at rest in the accommodation module when a fre breaks out, as hap-
pened to many of those killed or injured in the Piper Alpha disaster. Or the injury or illness may be the result
of a more mundane accident. For instance, the employee may be injured falling from his bunk (as occurred
in Robb v Salamis M & I Ltd, 2007 SLT 158, albeit no indemnifcation issue was litigated here).
130 Smith v South Wales Switchgear [1978] 1 WLR 165 per Lord Keith at 178.
131 Campbell v Conoco (UK) Ltd [2003] 1 All E R (Comm) 35 (hereinafter Campbell).
132 The terminology used throughout the case is rather confusing: because the case concerns a subcon-
tract which incorporated the terms of the lead contract mutatis mutandis, the contract describes the parties
46
of or arising out of or in connection with the performance or non-performance of
the contract” did not take efect in circumstances where the employee sufered injuries
after being struck in the back by a blast of compressed air discharged from equip-
ment which had nothing to do with the job he was doing, but which happened to be
positioned close to the area in which he was working. The subcontractors contended
that the indemnity would be triggered only if the injuries were directly attributable
to the performance of work under the contract.133 The Court of Appeal held that
the wording of that particular clause did not bear any such construction, and that
the phrase in question was wide enough to encompass at least a situation where the
injury occurred while the worker was engaged in his work,134 and may well have been
wide enough to encompass a situation where injury occurred while the worker was
at rest.135 However, the court also observed that “each contract depends on its own
wording and context”. Thus, if they wish the clause to be efective for the whole
period when the worker covered is on the platform, drafters of indemnity and hold
harmless clauses must be careful not to use words which tie the triggering of the
clause too closely to the performance of the contract or its works.136
6.3 Multi-party issues

Sometimes, for instance, in a consortium or alliancing agreement, there will be more
than two parties to the contract. This does nothing to alter the central concept of
the indemnity and hold harmless clause. However, the addition of further parties
does add a further layer of complexity to the drafting exercise. Slessor v Vetco Gray
illustrates the difculties that can be caused by attempting to provide for an unfamiliar
situation. This case concerned the interpretation of an indemnity clause contained
not in a standard services contract but within a multi-party consortium agreement.
The clause was in the following terms: “The Parties hereto mutually and irrevocably
undertake to release, defend and indemnify each other for damage to any property
and/or injury to/or death of the personnel of the others, arising out of or in con-
nection with the Work, howsoever caused”.137
At frst glance, this may look very like a fairly standard mutual indemnity and hold
harmless clause (albeit one in which the parties use the expression “release” rather
than “hold harmless”). However, it difers from such a clause in one crucial respect.
Ordinarily, A will indemnify B in respect of any injuries sufered by A’s personnel.
This clause was held to achieve something quite diferent: A here indemnifes B and
as operator and contractor. However, in context this is to be read respectively as contractor (Amec) and
subcontractor (Salamis): Campbell at paras 6–7.
133 Ibid. at para 10.
134 Ibid. per Rix L J at paras 18–19.
135 This point was left open as it was not necessary for the court to decide it: see Campbell per Rix LJ
at para 24. In Orbit Valve, similar wording was held not to limit the indemnity only to occasions when the
injured party was actively carrying out work under the contract but to extend also to injuries occurring while
the injured party was on the platform but at rest: Orbit Valve [1995] 1 All ER 174 at 186.
136 The modern LOGIC standard contract wording, “arising from, relating to or in connection with the
performance of or non-performance of the CONTRACT”, is a good example of a broadly drawn formula-
tion: see, e.g., LOGIC, Services, para 19.2(b).
137 Slessor v Vetco Gray (n. 31, above), at para 3.
47
greg gordon
C for any injury sufered by B and C’s personnel while B gives a like indemnity
to A and C in respect of injury or death to their personnel, and party C similarly
indemnifes A and B. The court declined to interpret the clause contra proferentem,
for the reasons which were given above, and recognised that it would be unusual for
the parties intentionally to draft a clause in these terms.138 However, the court felt
that the words used by the draftsman were clear and unambiguous, and that efect
must be given to them.139 Given that the wording achieved the diametric opposite
of one what would ordinarily expect an indemnity clause to be seeking to achieve,
one cannot help but wonder if this may have been one of the occasions referred
to by Lord Hofmann, where “the language has gone wrong”. However, this was a
Scottish case, and no attempt was made to present an argument in accordance with
Lord Hofmann’s restatement. Moreover, as we have already seen, following Arnold
v Britton,140 English law has retreated from the high water mark of Lord Hofmann’s
approach and would seem to be broadly in sympathy with the Scots approach to
interpretation.
6.4 “Full and primary”

In London Bridge, one of the arguments advanced in the attempt to defeat the opera-
tors’ claim under the indemnity was that, as the operators’ insurers had already settled
the claims of the contractor’s employees or their representatives, the operators had
sufered no loss in respect of which they required to be indemnifed. It was also
argued that the operators’ insurers could not claim under the indemnity either, on
the basis that both the insurers and the contractual indemnifers had ofered primary
indemnities, and the insurers were therefore not entitled to rights of subrogation, but
only to contribution;141 the argument ran that subrogation rights were enjoyed only
by a secondary indemnifer who has paid “out of order”, to allow it to recover from
the primary indemnifer who has a stronger obligation to settle the claim. The con-
tract was silent on this aspect of the nature of the indemnity. On the particular facts
of London Bridge, it was held that the contractual indemnity was primary, and the
insurers were secondary indemnifers entitled to their right of subrogation. Thus,
they could claim to be reimbursed in full. However, it must be emphasised that this
decision does not settle all such debates once and for all; for instance, the decision
was based at least in part upon the fact that there was no contractual obligation
138 Slessor v Vetco Gray per Lord Glennie at para 11. The court observed that the usual regime could
easily have been put in place by substituting the contentious phrase with “their own personnel”. A yet safer
way of drafting the clause might have been to individually list out each party’s responsibilities to the others.
This can be cumbersome, particularly where multiple parties are concerned, but in such matters elegance
is less important than clarity. Inverting the intentions of the parties, which seems to be what occurred in
Slessor, may seem an extreme drafting error, but there is always a risk, when trying to draft indemnities in
too concise a form, of inadvertently creating ambiguity, or of failing to consider which result would obtain
under the clause in any given factual situation.
139 Slessor v Vetco Gray at para 11.
140 [2015] UKSC 36.
141 The Scots law of contribution is concerned with the respective liabilities of joint wrongdoers and is
contained in s 3 of the Law Reform (Miscellaneous Provisions) (Scotland) Act 1940.
48
upon the operators to insure.142 How the court would have decided the matter if there
had been such an obligation is therefore an open question. Best practice (which does
not appear to have been followed in the LOGIC Standard contracts, but which is in
evidence in the Industry Mutual Hold Harmless Agreement) is therefore to expressly
state that the indemnities are “full and primary”.
6.5 Defnitional issues
Company groups
In practice, it is common for the parties to intend the efect of the indemnity clause
to extend beyond the immediate parties so as to include their respective groups. Thus,
in an operator-to-contractor contract, a contractor will commonly agree to indemnify
and hold harmless the operator’s group against loss of or damage to the contractor
group’s property, and the injury or death of the contractor group’s personnel; likewise
the operator143 will usually agree to indemnify the contractor group against loss of
or damage to the operator group’s property, and the injury or death of the operator
group’s personnel. Typically, an operator group will be defned so as to include the
operator itself, its co-ventures, its and their respective afliates, and its and their
respective directors, ofcers and employees and personnel.144 A contractor group will
generally include the contractor, its subcontractors and its subcontractors’ own sub-
contractors of whatever level, and its and their respective afliates and its and their
respective directors, ofcers and employees and personnel.145 This is done in recogni-
tion of the fact that contractors and operators will commonly use other members
of their groups to carry out the activities and operations envisaged by the contract,
and, if the benefts of the indemnity regime discussed at paras 6.16–6.18 are to accrue
fully, it is necessary for the efect146 of the indemnity provisions to extend to all parts
of the group. It will usually be the intention of the parties to establish an indemnity
regime on a grouped basis. However, parties need to be mindful in such cases not
just to scrutinise the indemnity clause but also carefully to check the defnition of
“group” or whatever cognate expression is being used in the contractual documenta-
tion, and to ensure that the implications of accepting those indemnity provisions,
read together with that defnition, are what is truly intended.147
142 London Bridge per Lord Hofmann at para 97.

143 The operator is generally referred to, in such contracts, as the “company”, and its group as the
“company group”. However, in the interests of keeping terminology consistent the term “operator” is used
throughout this chapter.
144 See, e.g., LOGIC, Services, cl 1.2.
145 See, e.g., ibid., cl 1.6.
146 The term “efect” is used because persons who are not party to a contract cannot be directly bound
by it; the grouping provisions do not directly bind the non-signatory members of the group, but achieve their
efect circuitously. If there is a contract between A and B, and a dispute arises between A and a company
(B2) which is part of B’s group, if A and B’s contract states that neither will be liable for the relevant species
of loss and provide each other with indemnities in respect of it, then, while B2 is still entitled to sue A if B2
sufers such a loss, A can claim an indemnity in respect of it from B.
147 A second Farstad v Enviroco case arising from the same incident as the frst but argued in the English
courts and pertaining to the defnition of “subsidiary” of a parent company group is Farstad Supply A/S v
Enviroco Ltd [2011] UKSC 16.
49
greg gordon
Employees and personnel
As we have seen, many indemnity clauses are concerned with dividing up responsibil-
ity for injuries, illnesses or fatalities sustained by people during the course of the
project. The expectation will ordinarily be that each party takes responsibility for its
own people. Great care needs to be taken in how this intention is expressed. In par-
ticular, it should not be expressed too specifcally or narrowly, by using a word which
has a technical legal meaning, such as “employee” or even “worker”. Many people
working on a platform will not be “employees” in the formal legal sense, but contrac-
tors. If the indemnity refers only to “employees” it is most unlikely to refect the will
of the parties. The current general practice is to frame the indemnity using a less
technical term, such as “personnel”, and then to defne that term broadly.
Property
Similarly, many indemnity clauses will be concerned with apportioning responsibility for
damage to property. Again, it is important for the parties – most particularly, contractors
and any subcontractors entrusted with expensive pieces of equipment which they do not
own – not to be lured into a false sense of security by observing that the contract contains
a clause of this nature, but to give critical thought as to its precise terms, and the terms
of any accompanying defnition. Contractors will usually want “company property” to
be given an expanded defnition so that it encompasses not just items owned by the
company but all items on the platform which the contractor might be called upon to use
and/or which might reasonably be afected by the works carried out by the contractor.
The operator will not necessarily own all the expensive or important items of property
on the platform: some may be hired, or subject to retention of title clauses, or may be
the property of other contractors, for instance the drilling contractor.
7 The problem of multiple parties
7.1 Introduction to the problem

As we have already seen, at any given time there is likely to be a multiplicity of par-
ties with personnel and/or property on a production platform. We have already
discussed the special difculties caused by ensuring that all subcontractors within a
given contractual chain have appropriate risk allocation terms in place. This, however,
is only part of the picture. Although the great majority of the undertakings with
personnel on a platform at any given time are likely to be in a contractual relation-
ship with someone, they will not all be in a contractual relationship with each other.
They will all be involved in a contractual chain; however, they will not all be involved
in the same contractual chain. This situation is demonstrated in Figure 2.3.
If, for example, subcontractor D2 negligently causes an explosion which injures
personnel from D, A, D1A and B2, then there will be a direct contractual link (and
presumably an indemnity and hold harmless provision) between D2 and D. Moreover,
it is likely that back-to-back provisions will exist which, issues such as insolvency and
drafting errors apart, should have the efect of apportioning risk throughout D2’s
contractual chain, i.e., from A, through D and D2 down to D2A. However, there is
no contractual connection, and therefore no contractual indemnity, between D2 and
50
Operator
Lead
contractor
B C D
Sub-
contractor
B1 B2 C1 C2 D1 D2
Sub-sub-
contractor
B1A B2A B2B C1A D1A D1B D1C D2A
Figure 2.3 Simplifed example of contractual relations on a production platform.
D1A, or between D2 and B2. Unless the parties are willing take the view that such
risk will be dealt with by the law at large – a decision that would be surprising, given
that it would undercut much of the economic benefts of the indemnity regime – a
means needs to be found of “flling in the blanks” in the indemnity regime. How can
this be achieved? A system whereby the parties entering into individual mutual hold
harmless agreements with all other contractors with whom they would otherwise
have no contractual link is a theoretical possibility, but it would require signifcant
administrative efort, expense and co-ordination and would have the potential to go
awry, leaving dangerous gaps in insurance cover if, for instance, a new contractor
or subcontractor not previously known to operate on the platform was awarded
a contract and information about this was not promptly disseminated among the
contracting community on that platform. This potential solution is therefore unat-
tractive in practice. Until the early 2000s, the practice in the UKCS was generally that
the operator and contractor would grant indemnities on a grouped basis, with the
operator’s group defned so as to include “other contractors”148 and the contractor’s
148 This term would be defned so as to include all contractors involved in operations other than the con-
tractor’s own subcontractors of whatever level. If one refers to Figure 2.3, if the operator (A) was contracting
51
greg gordon
group defned so as to include its afliates and all subcontractors of whatever level.
The operator provided the contractor with an indemnity in respect of all claims
relative to the people and property not just of the operator, but also of all contrac-
tors with whom the contractor did not itself have a contractual relationship, and by
the contractor providing an indemnity in respect of its group’s people and property
not just to the operator, but also to the other contractors. The risks would then be
further broken up and re-allocated along the back-to-back principles above. Thus
the problem of multiple parties was dealt with by the operator acting as a fulcrum
in the contractual arrangements. This provided a workable solution to the problem.
However, the practice was difcult and time-consuming for operators to administer.
It was also risky: if, for some reason,149 the back-to-back provisions broke down, then
the operator would commonly fnd itself responsible for a greater degree of liability
than it had anticipated or intended. For both of these reasons a more ambitious
solution, modelled upon what had become the practice of certain operators,150 was
instituted in 2002. This is discussed further below.
7.2 The Industry Mutual Hold Harmless Agreement (IMHH)

The problems referred to above, and the fact that the industry perceived the potential
for efciencies and savings,151 led to the creation by LOGIC’s Standard Contracts Com-
mittee of an Industry Mutual Hold Harmless sub-committee which, after extensive
consultation, recommended the institution of a Mutual Hold Harmless Scheme and
prepared a draft of the relevant Deed.152 This scheme was initially efected by the
Mutual Indemnity and Hold Harmless Deed (the “IMHH Deed”)153 which had been
signed by 93 parties154 when the scheme became efectual on 1 July 2002.155 A signifcant
number of other contractors have subsequently entered the scheme. The scheme is
presently administered by LOGIC156 (which is itself the benefciary of extensive simple
indemnity provisions relative to losses which might arise as a result of the manner in
with contractor D on this model, A would defne its Group so as to include all contractors of whatever level
within the chain of contractual relationships fowing from its contracts with B and C.
149 E.g., drafting error such that the frst level of indemnifcation worked but the back-to-back provision
was inefectual; or a failure to ensure that a back-to-back arrangement was entered into.
150 E.g. Shell.
151 See LOGIC, IMHH – Introduction and Background, available for download from: https://www.
logic-oil.com/imhh (accessed 20 November 2022): “The IMHH Scheme has been created to manage the risks
inherent in the industry in a much more comprehensive and efective manner. It operates on the premise that
a company is in a much better position to protect and insure their own people and equipment. In doing so,
companies can be more certain of the risks they need to insure and this, in turn, reduces multiple policies
insuring the same risk.”
152 LOGIC, About: see text under the heading, “Introduction”. Note that although the word “indem-
nity” is not used by LOGIC in its introductory discussion of the Deed, as we shall see, the core risk allocation
provision of the Deed is clearly a mutual hold harmless and indemnity clause.
153 Although the Deed bringing the scheme into existence is clearly titled the “Mutual Indemnity and
Hold Harmless Deed”, it is known throughout the industry as the “IMHH Deed”. This chapter will follow the
prevailing industry practice. The Deed is available for download from: https://www.logic-oil.com/sites/default/
fles/documents/IMHH%20DEED%202022%20FOR%20WEBSITE_1.pdf (accessed 20 November 2022).
154 Listed in Sch 1 to the IMHH Deed.
155 IMHH Deed, cl 7.1, read together with the date of the Deed.
156 IMHH Deed, cl 1.1, defnition of “Administrator”. The Administrator may be replaced when a
majority of signatories believe it has carried out its duties incompetently: see IMHH Deed, cl 4.4.
52
which it carries out its administrative function: see IMHH Deed, cll 4.2 and 4.3 read
together with cl 6.1(ii)).157 LOGIC’s primary functions are to act as the signatories’
attorney in executing Deeds of Adherence with new parties, receive and intimate to
the remaining parties any notices of withdrawal, ascertain and act upon the will of
the signatories upon the occurrence of any of the events which allow signatories to
vote on whether to terminate the scheme, and to maintain the IMHH website. The
initial IMHH Deed expired at the end of 31 December 2011. A replacement 2012 Deed
entered into force immediately upon the expiry of the 2002 Deed. It has since been
superseded by a further replacement, the 2022 Deed, which entered into force in Janu-
ary 2022. All references to the IMHH Deed are to the 2022 version.
The Deed’s core indemnity provisions

The central purpose of the Deed is to provide an efcient and practical means of
bringing into existence an efectual set158 of hold harmless indemnities between par-
ties who, save for the purposes of entering into an agreement about the allocation
of physical risks, would have no commercial imperative to contract with each other.
The IMHH Deed achieves this by providing:
the Signatories shall be solely responsible for and shall defend, indemnify and hold harm-
less the other Signatories and the other members of their respective Groups against all
Claims arising from, out of, or relating to the Services in connection with:
(i) personal injury to or sickness, disease or death of Personnel of the Indemnifying
Signatory or any other members of its Group; and
(ii) loss of, recovery of, or damage to any Property of the Indemnifying Signatory
or any other members of its Group; and
(iii) Consequential Loss sufered by the Indemnifying Signatory or any other members
of its Group.159
The “indemnities given pursuant to this Deed”160 are expressed as being “full and
primary” and to apply “irrespective of cause and notwithstanding the negligence or
breach of duty (whether statutory or otherwise) of the Indemnifed Party and shall
apply irrespective of any claim in tort, under contract or otherwise at law”.161
Taking the main indemnity and hold harmless clause phrase by phrase, in connec-
tion with all claims arising from, out of, or relating to the services that it covers,162
it provides a full and primary mutual indemnity and hold harmless regime drawn
157 LOGIC, is provided with a right to outsource the performance of its administrative duties: cl 4.4
states that the administrator shall be “entitled to discharge any of its obligations and/or duties under this
Deed by procuring that such obligations or duties are performed on its behalf by another person”, to be
known as the Service Provider. The service provider, like the administrator, may be replaced if the majority
of the parties to the Deed so wish. The service provider does not, through the operation of the IMHH Deed
itself, obtain a right to share in the simple indemnity and hold harmless provision granted to but one would
presume that any person agreeing to act in that capacity would insist upon obtaining the beneft of a back-
to-back indemnity from the administrator in any contract appointing them to the position.
158 As we shall see, the Deed does not provide for a wholly comprehensive set of indemnities: matters
such as pollution are not included.
159 IMHH Deed cl 2.1.
160 Note that this term is habile to include not just the main cl 2.1 indemnity but also others that may
exist throughout the Deed – e.g. the indemnities given to the Administrator by cl 4.2 and 4.3.
161 IMHH Deed cl 2.2.
162 Note the important exceptions discussed below.
53
greg gordon
on a grouped basis, operating irrespective of negligence or breach of duty, whether
statutory or otherwise, and covering (1) personal injury, sickness, disease or death
of an expansively drawn set of people defned as “Personnel”; and (2) loss of, recov-
ery of, or damage to “Property” and (3) a class of loss defned as “Consequential
Loss”. Thus, the IMHH Deed provides indemnities and a hold harmless provisions
in respect of many, but by no means all, of the areas where the industry commonly
uses mutual indemnity clauses.163 The indemnity clause is supported by a clause
providing for a waiver of rights of subrogation164 and is a meticulously drafted
clause which avoids the errors and ambiguities of many of the clauses which have
been litigated in the past.
Other important provisions

The IMHH contains a number of other signifcant provisions which will be discussed
briefy in turn.
Entry into force

As has already been noted, the 2022 IMHH Deed entered into force as between
its original signatories on 1 January 2022.165 The Deed becomes efective as between
those parties from that date.166 However, many parties have entered the IMHH
scheme since it was instituted. Such parties do so by entering into a Deed of
Adherence167 between themselves and all the other current members of the scheme,
who sign per the Administrator as attorney.168 Such parties are known as “New
Parties”169 and become members of the scheme only from the date on which a
Deed of Adherence signed by that Signatory has been dated by the Administrator”.170
The IMHH website displays the date when each party to the Deed became a
member of the scheme.171 The Deed expressly states that the core indemnity provi-
sions contained in cl 2 do not apply and are not enforceable “in respect of any
Claims arising out of events occurring prior to the date on which that Signatory
became a Signatory”.172
163 The Deed does not cover, for instance, pollution: this is because “[t]he prevailing consensus in the
industry was that the scheme should apply to personal injury, property damage and consequential losses
only. Other areas of risk such as pollution were considered, but were ultimately discounted. One of the main
reasons for having a mutual hold harmless arrangement in respect of a Signatory’s own property, personnel
and consequential loss is that the Signatory is best placed to assess the value at risk and, if required, make
the appropriate insurance arrangements. Pollution risks are less quantifable and hence would have created
a complication to the IMHH provisions which was not widely welcomed”. See LOGIC, IMHH General
Guidance, available for download from: www.logic-oil.com/imhh/general-guidance (accessed 27 April 2017)
(hereinafter LOGIC, General Guidance).
164 IMHH Deed, cl 5.
165 IMHH Deed, cl 7.1, read together with the date of the Deed.
166 Ibid., cl 11.
167 Ibid., cl 4.1; see also Sch 2 to the Deed for the prescribed for of the Deed of Adherence.
168 IMHH Deed, cl 4.2.
169 Ibid., cl 4.1.
170 Ibid., cl 11.
171 See www.logic-oil.com/imhh/signatories (accessed 27 April 2017). To view the dates, click on the
hyperlink embedded within the name of each of the parties.
54
Geographical extent
The Deed applies not only to services provided on or in the UK Continental Shelf
and/or between the United Kingdom low water mark and the innermost boundary
of the UK Continental Shelf, but also to those carried out within Irish territorial
waters and upon the Irish Continental Shelf.173 As further noted below, the Deed
does not apply to landward areas.
Order of precedence
The Deed expressly provides that it “shall not take precedence over, amend, modify,
or apply to the terms of any agreement between Signatories entered into prior to,
on or after this Deed becoming efective in relation to such Signatories”.174 It is worth
emphasising that the Deed refers to “any” agreement. So where a particular incident
involves a dispute between signatories to the Deed who are at the material time in a
contractual relationship germane to the provision of Services as understood by the
Deed,175 the specifc contract will take precedence over the general Deed. This is so
even where the specifc contract is silent – either by accident or design – on the ques-
tion of risk allocation. Thus, the IMHH Deed will not ride to the rescue if signatories
to it are in a direct contractual relationship and omit to include indemnity provisions,
or to include provisions which are inefectual. This is to provide signatories with the
freedom to enter into diferent liability arrangements if they so wish.
Extension of benefits to groups

In common with many oil and gas indemnity clauses in the UKCS, the indemnity
and hold harmless provisions in the IMHH Deed are drawn so as to be for the beneft
only of the Deed’s immediate signatories but also to “the other members of their
respective Groups”.176 The Deed defnes “Group” broadly so as to include inter alia
“the Signatory in question and its respective Afliates, the Personnel of all of the
foregoing and their Invitees”,177 and cl 6.1(i) utilises the provisions of the Contracts
(Rights of Third Parties) Act 1999 to extend the benefts of the core, cl 2 indemnity
provision, and also the waiver of subrogation clause, discussed immediately below,
to all members of a Signatory’s Group.
Waiver of rights of subrogation

The IMHH Deed provides that the signatories thereto shall procure that their insur-
ers shall waive their rights of subrogation or any other rights they may have to
proceed against the other signatories or members of those signatories’ groups in
relation to the matters covered by the Deed.178 Signatories are entitled to require each
other to exhibit evidence of the waiver.179 The Deed also states that, in the event that
173 IMHH Deed cl 1.1, under the defnition of “Services”.

174 Ibid., cl 11.
175 See the defnition in the IMHH Deed, cl 1.1.
177 For the full defnition, see IMHH Deed, cl 1.1, “Group” read together with the other defned terms
which themselves feature within that defnition.
179 Ibid., cl 5.2.
55
greg gordon
the parties fail to procure the relevant waiver, the failing party’s rights to enforce its
rights and benefts under the IMHH Deed shall be suspended.180
Right to defend
As has already been noted, the obligation to indemnify and hold harmless is separate
and distinct from the right and obligation to defend a claim; however, it is relatively
common within the UKCS for the indemnifer to accept (or demand) responsibility
for conducting the defence of any claim made against the party it is indemnifying.181
Under the IMHH Deed, the indemnifying signatory (i.e., the party who is, in the event
of a claim, liable under the Deed to indemnify whosoever the claim is initially directed
towards by the party sufering a loss) is entitled, but not obliged, to take over the
conduct of the defence to any such claim.182 Where this option is exercised, the indem-
nifying signatory accepts certain obligations concerning the provision of information
to,183 and consulting with,184 the indemnifed signatory, and is in turn entitled to rea-
sonable assistance from the indemnifed signatory in the conduct of the defence.
Parties to the IMHH Deed enter into a contractual relationship which radically re-
allocates the legal liabilities they carry. Where a company is a member of the IMHH
scheme, other members of the scheme are likely to feel entitled to refrain from enter-
ing into mutual hold harmless indemnity contracts, and/or to purchase less insurance
cover which they would otherwise have been likely to believe they required. It is
therefore imperative that signatories to the Deed know with certainty which parties
are members of the scheme at any given time. For this reason, the right to withdraw
from the scheme is quite heavily regulated. Upon signature, the parties are locked in
to membership of the scheme until 31 December 2031,185 subject only to potential
escape points occurring on 31 December 2024 and 2027, but only in the event that
they have provided the Administrator with 60 days’ notice of this intention.186
Exceptions
A number of exceptions are carved out of the IMHH Deed’s core indemnity and hold
harmless provisions or introduced by way of the Deed’s defned terms. Each of these
has a very signifcant practical impact upon the Deed’s scope. Each is discussed below.
Transportation by air excluded

None of the cl 2 indemnities applies or is given in relation to “any Claims arising from,
out of or relating to the transportation by air of any member of a Signatory’s Group”.
Thus, transportation by air falls wholly outside the scope of the IMHH Scheme.
180 Ibid., cl 5.3.

181 See para 14.8.
182 IMHH Deed, cl 3.2. Where it does not take up this option, the Indemnifed Party assumes an obliga-
tion not to settle the case without the consent of the Indemnifying Signatory: IMHH Deed, cl 3.4.
183 IMHH Deed, cl 3.2(ii).
184 Ibid., cl 3.2(iii).
186 IMHH Deed, cl 7.3. The Administrator is obliged by the same clause to notify all other parties in
writing.
56
Landward areas
Despite the fact that, in practice, onshore works will quite commonly be the subject
of indemnity provisions similar to those governing ofshore,187 the IMHH Deed
applies only to services and/or supplies “carried out on or in the United Kingdom
Continental Shelf and/or between the United Kingdom low water mark and the
innermost boundary of the United Kingdom Continental Shelf ”.188 Thus, services
or supplies carried out on the landward side of the UK low water mark are excluded.
This has the efect of excluding onshore services, and will also have the efect of
excluding any services or supplies carried out in the part of the territorial sea which
lies on the landward side of the low water mark.189 It was decided to so limit the
efect of the IMHH Deed because
The general industry view was that extending the IMHH to cover onshore activities as
well as ofshore activities could have inadvertently led to the application of the IMHH
in circumstances where it was not intended (e.g. motor vehicle accidents) and, as such,
in order to avoid any ambiguity it was limited to ofshore activities only.190
Operators
Although operators played a substantial part in bring it into being, the IMHH Deed
is designed for, and signed by, the contracting sector. It is not well suited to operator-
to-operator situations191 and, given that the operator sits at the top of the contractual
pyramid, operators should already be in a direct or at least indirect contractual
relationship with anyone who steps onto their platform for commercial purposes, and
therefore in a position to negotiate whatever contractual risk allocation regime the
parties to those contracts choose.192
Commentary and conclusions on the IMHH scheme

Some brief concluding words on the scheme are ofered below.
The contractual nature of the scheme

It is important to appreciate that the Deed is not a document in the nature of a piece
of legislation, imposing conferring rights and imposing obligations by virtue of the
mere fact that it exists. It is a contractual solution to a problem; it is just not the
same contractual solution that the oil and gas industry previously deployed. It is
axiomatic, therefore, that it takes efect only between those parties who have executed
it, or those who have entered into a Deed of Adherence. It needs to be emphasised
that it cannot simply be assumed that all oil and gas contractors are adherents to
the Deed. In particular, not all drilling contractors have signed the deed, although
187 LOGIC, Services is intended to be applied both on- and ofshore. Note also that the pursuer in Sles-
sor v Vetco Gray was working onshore when he was injured.
188 IMHH Deed, cl 1.1, defnition of “Services”.
189 A surprisingly large area, although not one in which large amounts of oil and gas operations have
been undertaken: see para 4.1.
190 See LOGIC, General Guidance.
191 The possibility of an operator-to-operator IMHH Deed is discussed at para 14.65.
192 See LOGIC, General Guidance, under the heading “Can Operators Sign?”
57
greg gordon
the numbers to have done so are gradually increasing. It is understood that the drill-
ers’ general reluctance to adhere to the Deed is because the potential losses which
may be sufered by drilling contractors, whose equipment is extremely valuable, and
who will commonly have relatively large numbers of personnel on a rig or platform,
is substantially larger than most contractors, and the drilling community have, by
and large, yet to be persuaded of the value of the scheme. Other contractors will
need to take appropriate steps to accommodate the risk caused by this refusal, either
by seeking an indemnity from the operator, entering into a mutual hold harmless
and indemnity arrangement directly with the driller, or simply by accepting the
exposure and insuring accordingly. In addition, and as noted above, the Deed’s order
of precedence clause means that the Deed is rendered irrelevant if there is any direct
contractual relationship bearing upon “Services” as understood by the Deed.
Potential dangers of the IMHH

The IMHH was launched with considerable fanfare and broad industry backing193
and has quickly been accepted by the majority of the contracting community in the
UKCS. It provides a solution to a problem which has caused considerable difculties
and inefciencies for some time. There is no reason to believe that in general terms
the scheme has been, or will continue to be, anything other than a success. However,
the IMHH initiative – or perhaps more properly, the initiative’s application in practice – is
not without its potential pitfalls. The Deed, while by no means long, is of necessity
relatively complex. While many of the contractors who operate within the UKCS are
large and sophisticated companies possessing the understanding and resources to
ensure that their personnel understand the full import of the scheme, this is not true
of all contractors. It is least possible that, in spite of the eforts which have been
made by LOGIC to promote an understanding of the Deed, not all signatories to
the scheme are fully conversant with all its subtleties. Although experienced oilfeld
contractors should be aware of the Deed’s provisions, new contractors, some of whom
work in the ofshore feld only semi-regularly, are signing up all the time. Moreover,
the natural turnover of staf within even the more established players means that
ongoing training and education is essential. In addition, and rather perversely, as the
Deed becomes more broadly accepted, there is the danger that familiarity may breed
contempt. Procedures may become sloppy, and parties may simply assume that
everyone working on a particular rig or platform with whom they are not in a con-
tractual relationship is a subscriber to the IMHH Deed. As we have already seen,
however, this is unlikely to be true of the driller, and may not be true of all other
contractors. If the contractor is to minimise risk it is incumbent upon it to ascertain
whose property and personnel is or will be on the platform at the material time and
to check that those on board fall within the ambit of the IMHH scheme, or to obtain
a binding contractual undertaking from the operator that – apart from true third
parties – only subscribers to the IMHH will be permitted onto the platform. None
193 It was launched under the auspices of LOGIC and with the support of UKOOA, IMCA, OCA and
WSCA.
58
of this is intended to attack the principles underlying the IMHH Deed but simply
to urge due care in the scheme’s application.
8 Liability for “consequential” loss

In addition to containing the indemnity and indemnity and hold harmless provisions
described above, a wide variety of oil and gas contracts will also include clauses
which exclude, limit194 or put in place an indemnity and hold harmless regime195 in
respect of liability for so-called consequential or indirect losses.196 Such provisions
are commonplace in operator-to-contractor contracts197 and are seen in some – but
by no means all – contracts between oil companies.198 Where such clauses are used,
their purpose is to protect parties from the full consequences of causing certain types
of loss, typically delays in or loss of production, loss of use, loss of profts, and
business interruption. The quantum of such losses is generally difcult to anticipate
in advance and may be hard to quantify even retrospectively.199 However, they tend
to accrue on a daily basis and, if a particular feld, facility or critical piece of infra-
structure is “knocked out” for a sustained period of time, these losses can rise to
extraordinary and alarming levels. As such, they are frequently either excessively
expensive or simply impossible for a contractor to insure against, and the industry
has by and large accepted that such losses should lie where they fall. Thus, we can
see that economic factors very similar to those that drive the indemnity regime
underpin the exclusion or limitation of consequential losses.
8.1 What is “consequential loss”?

There is considerable uncertainty over what, if any, fxed meaning the term “conse-
quential loss” conveys. The prevailing interpretation under English law has, for some
194 In so far as these contracts “exclude” or “limit” such loss they are subject to the provisions of the
Unfair Contract Terms Act 1977, but not the 1999 Regulations. However, the courts have repeatedly held
that, while they will always consider full circumstances of the case (see, e.g., Motours Ltd v Eurobell (West
Kent) Ltd [2003] EWHC 614 (QB)), it will be only in unusual circumstances that they will hold a business
– business exclusion of consequential loss clause unreasonable: see Wessanen Foods Ltd v Jofson Ltd [2006]
EWHC 1325 (TCC); Photo Production Ltd v Securicor Transport Ltd [1980] AC 827.
195 A carefully drafted mutual indemnity regime is thought necessary where the intention is to extend
the efect of the consequential loss provisions to groups: if there is a contract between A and B, and a dis-
pute arises between A and a company (B2) which is part of B’s group, a simple exclusion or limitation of
consequential loss will not be binding upon B2. However, if A and B state that neither will be liable for the
other group’s consequential losses and provide each other with indemnities and hold harmless provisions
in respect of such loss, then, while B2 is still entitled to sue A if B2 sufers a consequential loss, A can claim
an indemnity back from B.
196 The precise terminology varies: sometimes “consequential loss” or “indirect loss” is used on its own.
Sometimes “damages” is used instead of “loss”. See the discussion in London Bridge (Inner House) 2000
SLT 1123 per LP Rodger at 1156D–E.
197 They appear throughout the LOGIC standard contracts series. See, e.g., LOGIC, Supply of Major
Items at cl 23.
198 See A Jennings, “FPSO Agreements” and J Izod, “Satellite Tie-Back and Processing Agreements”,
both in M David (ed), Oil and Gas Infrastructure and Midstream Agreements (1999), at 209f and 170,
respectively. London, Langham Legal Press.
199 Note the difculties which occurred in this regard in BHP Petroleum Ltd v British Steel plc and
Dalmine SpA [2000] 2 Lloyd’s Rep 277 (hereinafter BHP Petroleum).
59
greg gordon
time, been that the expression is almost a term of art, which usually denotes a refer-
ence to the category of losses encompassed by the second limb of the rule in Hadley
v Baxendale.200 On this analysis, if a contract excluded liability for consequential loss,
the parties would still be able to recover the sort of damages envisaged by the frst
limb of Hadley v Baxendale, which is to say, losses “such as may be fairly and rea-
sonably be considered . . . [as] arising naturally, i.e. according to the usual course of
things, according to the breach itself ”. However, the parties would not be entitled to
recover in respect of the second limb, which is to say, for damages
such as may reasonably be supposed to be in the contemplation of the parties at the time
they made the contract as the probable result of the breach. Now if the special circumstances
under which the contract was actually made were communicated by the plaintifs to the
defendants, and thus known to both parties, the damage resulting from the breach of such
a contract, which they would reasonably contemplate, would be the amount of injury which
would ordinarily follow from a breach of contract under those special circumstances.201
If this line of reasoning was ever sound, one could question whether it is consistent
with the modern approach to contractual interpretation. Imposing a fxed meaning
to a phrase smacks of the application of the canonical approach of an earlier age,
rather than the approach currently ascendant in English law, which is fundamentally
directed towards the import of the text used within its contractual (but not necessar-
ily commercial) context.202 There are some – albeit equivocal – signs that the English
judiciary may be preparing to adopt a diferent view to this question.203 If so, it may
come to follow the approach to this question taken in Scotland, which rejected the
“consequential loss as second limb” approach in London Bridge.204 In London Bridge,
the court held that it was not appropriate to attach “a special, technical meaning” to
the phrase “indirect or consequential loss” so that the phrase would be “interpreted
as referring to losses which would be regarded as indirect or consequential under
200 (1854) 9 Ex 341. The line of reasoning sprang from Millar’s Machinery Co v David Way & Son (1935)
40 Com. Cas. 204 and Saint Line v Richardsons [1940] 2 K.B. 99 and has subsequently been consolidated
by decisions such as British Sugar P.l.c. v N.E.I. Power Projects Ltd (1997) 87 B.L.R. 42; Deepak Fertilisers
and Petrochemicals Corporation v I.C.I. Chemicals & Polymers Ltd (1999) 1 T.C.L.R. 200; Hotel Services
Ltd v Hilton International Ltd [2000] 1 All E.R. (Comm.) 750; McCain Foods G.B. Ltd v Eco-Tec (Europe)
Ltd [2011] E.W.H.C. 66 (T.C.C.); Markerstudy Insurance Co Ltd v Endsleigh Insurance Services Ltd [2010]
E.W.H.C. 281 (Comm.); Glencore Energy U.K. Ltd v Cirrus Oil Services Ltd [2014] E.W.H.C. 87 (Comm.);
Polypearl Ltd v E. On Energy Solutions Ltd [2014] E.W.H.C. 3045 (Q.B.).
201 (1854) 9 Ex 341, 355–356 per Alderson B.
202 See the discussion of Arnold v Britton at section 5.1, above.
203 In Scottish Power UK plc v BP Exploration and Drilling [2015] EWHC 2658, Mr Justice Leggatt
stated obiter (at para 180) his concurrence with a leading commentator’s critical analysis of the “consequen-
tial loss as second limb” approach. In Transocean Drilling UK Ltd v Providence Resources plc [2016] EWCA
Civ 372, Lord Justice Moore-Black, delivering a judgment concurred in by Lord Justices McFarlane and
Bridges, noted (at para 15) that “the expression ‘consequential loss’ has caused a certain amount of difculty
for English lawyers”, and thought it “questionable” whether some of the “consequential loss as second limb
cases” would be decided in the same way today. In Saga Cruises BDF Limited & anor v Fincantieri SPA
[2016] EWHC 1875 (Comm), Deputy High Court Judge Sara Cockerill QC considered (at para 198) “the
consequential loss as second limb” line of authority to be “well-established”, although she did qualify this
by adding the rider “on the current state of the law”.
204 As we shall see, the detailed reasoning is contained in the Inner House phase of the case. However,
this reasoning was expressly approved of by Lord Mackay of Clashfern in the House of Lords: 2002 SC
(HL) 117 [2002] 1 All E.R. (Comm) 321 at para 69. Lord Mackay’s judgment was in turn concurred in by
all the other judges in the House.
60
the rules in Hadley v Baxendale and subsequent cases”.205 Instead, while prior court
decisions were certainly a relevant consideration, they were not to be thought of as
determinative. The court’s primary task was to interpret the words in the context in
which they were used.206 When such an approach was taken, the Court of Session
held, and the House of Lords confrmed, that the disputed loss in London Bridge207
did not fall to be considered as “indirect or consequential” but was a direct loss.208
8.2 Commentary and conclusions on consequential loss

It is wholly understandable that the parties to oil and gas contracts should seek to
exclude, place limits upon the recovery of, or seek indemnities in respect of, certain
types of potentially exorbitant losses. It is, however, perhaps unfortunate that the
term “consequential loss” was ever adopted as the vehicle for attempting to do so.
Even when the term was thought to be a synonym for the losses envisaged by the
second leg of the Hadley v Baxendale test, it created difculties. It is not always easy
to ascertain whether a particular type of loss is going to fall within limb one of the
test or limb two. There was a clear danger that the contractual draftsman would
assume that the expression was a convenient shorthand for all of the types of poten-
tially exorbitant or hard to insure losses that he wished to exclude. However, it was
not. Many of the most onerous losses which might be sufered during oil and gas
operations do not clearly fall under the rubric of the second limb of Hadley v Bax-
endale, but may instead be captured by the frst limb of the test. This can be illustrated
by considering the factual circumstances described in BHP Petroleum. In this case,
a contract was entered into for the procurement of steel for a gas reinjection pipeline.
The purpose of the project was to improve production from a producing oilfeld.
The steel was in due course found not to conform to specifcation. It required to be
replaced and the gas reinjection programme was delayed. In such circumstances, it
is not just the cost of replacing the pipe that “[arises] naturally, i.e. according to the
usual course of things, according to the breach itself ”. It is also plainly apparent
that production from that oil feld will be delayed for a period of time while these
works are carried out. Production lost during that period of time would therefore
appear to be a direct loss, not a “consequential or indirect” one. So, a clause which –
without saying more – excludes “consequential or indirect” loss is unlikely to exclude
the loss described. And while the London Bridge approach is, it is submitted, a just
205 Lord Rodger at 1154.

206 London Bridge (Inner House) 2000 SLT 1123 per LP Rodger at 1156.
207 The circumstances were rather unusual. Because the operator had a material connection to Texas,
they feared (with considerable justifcation) that if the damages claims of the persons injured and those of
the families killed in the disaster were not settled on a generous basis, they might be sued in Texas, where
the level of awards of damages were signifcantly higher than in Scotland. The claims of the injured parties
and the families of the deceased were therefore settled on a “mid-Atlantic” basis, which is to say, at a level
somewhere between traditional Scots awards and those which could be expected in the state of Texas. The
defenders accepted that the element of the Scots element of the damages award was a direct loss, but that
the “mid-Atlantic” augmentation was an indirect or consequential loss.
208 For a fuller account see G Gordon, “The Exclusion of Consequential Loss in the Piper Alpha Case:
A Study of Lord Rodger’s Method of Textual Analysis”, in Simpson, Styles, Wilson and West, Pragmatism,
Continuity and Change: Essays in Honour of Prof Angelo Forte, 2016 Aberdeen University Press pp. 434–465.
61
greg gordon
and sensible one, it does little to make the situation more straightforward. Although,
as with the contextual approach more generally, it may very well make the outcome
of any given litigation fairer, the uncertainty that it engenders arguably increases the
chances of litigation occurring in the frst place and may very well make the scope
of enquiries of any litigation which does result more wide-ranging.
Against this rather shifting background, the safest way in which to construct a clause
excluding or limiting the type of losses which the oil industry has tended to design as
“consequential or indirect” is to make the working presumption that those words will of
themselves convey very little reliable protection, and to defne all essential content into
them. This means that the draftsman must give consideration to the types of uninsurable
or potentially exorbitant losses that might befall the project and must list them out in the
clause. This is the approach taken by most of the contractual documents in broad use
in the industry just now. There is some disagreement on whether the best practice is to
defne consequential loss by reference to a closed list, or to state that consequential loss
“includes but is not limited to” the potential heads of loss items listed. Jennings favours
the former approach on the basis of clarity,209 but the latter approach is the one which
is used more commonly throughout the industry’s standardised contractual documen-
tation. Thus, the Industry Mutual Hold Harmless Deed defnes consequential loss as:
(i) consequential loss under applicable law; and
(ii) loss and/or deferral of production, loss of product, loss of use and loss of revenue,
proft or anticipated proft (if any) whether direct or indirect to the extent that these
are not included in (i), whether or not foreseeable at the date of execution of this Deed.210
Finally on this topic, where, as will frequently be the case, the contract contains both
an indemnity regime and either an exclusion, limitation, or an indemnity in respect
of consequential loss, thought needs to be given to the interaction between these
provisions. They may very well overlap. It will generally be intended that the conse-
quential loss clause will take precedence over the indemnity provisions.211 Whatever
the parties intend, clear words should be used to express their wishes.
9 Overall limitation of liability

An overall limitation of liability212 – sometimes described as a “liability cap” – is a clause
that seeks to limit a party’s liability not by reference to particular species of loss, but
by reference to a total sum payable. Such caps have the beneft of providing at least a
level of certainty as to overall liability213 and facilitate the obtaining of insurance.214
209 A Jennings, “FPSO Agreements” in M David (ed), Oil and Gas Infrastructure and Midstream Agree-
ments (1999) at p. 210, (n. 198, above).
210 See the IMHH Deed, cl 1.
211 See, e.g., LOGIC, Supply of Major Items, at cl 23.
212 A straightforward liability cap is of course a limitation clause and is therefore subject to a number
of the controls imposed in UCTA, discussed at paras 14.26 and 14.66. For the reasons given at para 14.66,
it is thought unlikely that the courts would decline to enforce a cap on liability in all but the most unusual
circumstances. However, it is also possible to draft these clauses as indemnities: for reasons similar to those
given in para 14.63, this is necessary if the intention is to extend the efect of the cap to groups.
213 See the discussion below relative to third-party issues.
214 P Roberts, (n. 76, above), para 13.61.
62
In any given clause, the limit or cap may be expressed in one of a number of
diferent ways: either as a fxed sum of money (often the maximum amount of
insurance available at an economically viable rate) or as a proportion (or multiple)
of the sum payable for the job. However it is expressed, the level at which the cap is
to be set is essentially a commercial matter to be agreed between the parties, albeit
if the cap is set too low, there is a risk that it could be deemed unreasonable for the
purposes of the Unfair Contract Terms Act 1977.215 Liability caps are commonplace
in oil and gas contracting practice.216 Such clauses are less frequently written about,
simpler to understand and somewhat less technically demanding to draft than either
indemnity or consequential loss provisions.217 However, they are not wholly without
legal difculty. As we have already seen, an inappropriate level of liability could
lead to difculties with enforcement under UCTA 1977; and it would be prudent for
the clause to clarify if it is intended to take efect even against the proferens’ own
negligence. Moreover, consideration needs to be given to the interaction between
the liability cap and other risk allocation provisions, and in particular, third-party
claims. In WesternGeco Ltd v ATP Oil and Gas (UK) Ltd,218 the liability cap was
held to apply only internally to the parties and not to have the efect of limiting
the amount payable as a result of an indemnity granted against third-party claims.
Contractors therefore need to understand that if the liability cap is intended to
take efect against an indemnity against third-party claims, very clear and specifc
wording will have to be used.
Liability caps are particularly important in circumstances where the parties are not
prepared to agree to a mutual hold harmless indemnity and/or consequential loss
regime.219 In these cases, the liability cap is the only clause which stands between a party
and the prospect of unlimited – and potentially exorbitant – liability. The importance
of the clause is perhaps most clearly demonstrated in the context of applications for
access to infrastructure, where it is the expectation of the industry’s regulator that
the infrastructure owner will be prepared to set a reasonable liability cap.220
A liability cap is arguably less important in cases where the parties have agreed
upon an indemnity regime and/or exclusion of liability for consequential loss. The
combined efect of these clauses is, or should be, to shut out a large number of
the heads of claim which may result in losses disproportionate to the gains that a
party had in contemplation when commencing on the project. However, even here a
contractor should seek to obtain a liability cap for two reasons. First, some oil and
gas projects are so valuable that even the potential losses which are unequivocally
direct may be more than a contractor can comfortably bear. And second, even if
215. See the discussion of UCTA at section 4.1, above. See also P Roberts, (n. 76, above), para 13–81.
216. See, e.g., LOGIC, Supply of Major Items, cl 35.
217. This is not to say that such clauses create no difculties. One sometimes encounters liability caps
which are subject to exclusions – where this is so, proper risk management requires that the efect of the
exclusion must be clearly understood before the clause is agreed to. In addition, the interrelationship between
the liability cap and the warranties provisions in the contract needs to be closely considered.
218. [2006] EWHC 1164 (Comm).
219. Indemnities are not uniformly accepted in all oil and gas contracts: see R Palmer, “Tie-In Agree-
ments”, in M David, Oil and Gas Infrastructure Agreements, (n. 198, above), p. 231 at 239.
220. See “North Sea Transition Authority, Guidance on Disputes over Third Party Access to Upstream
Oil and Gas Infrastructure”, NSTA 2022, para 79.
63
greg gordon
the losses which the contractor fears should, strictly speaking, be shut out by the
indemnity and/or consequential loss clause, there is always a risk that those clauses
might for some reason be susceptible to challenge. A liability cap therefore performs
the valuable function of providing a second line of defence against indeterminate or
disproportionate liability. Any embarrassment felt by asking for a second line of pro-
tection should be swiftly overcome by considering just how dreadful the consequences
are likely to be if something has gone wrong with the drafting of the indemnity or
consequential loss clause.
64
CHAPTER 3
Contracting around tort defaults

The knock-for-knock principle and accident costs
1 Introduction
Unknown to most, the gas and oil industry has largely opted out from standard tort
liability. For several decades now, members of the gas and oil industry have contrac-
tually suspended the rules of negligence and strict liability in their dealings with one
another. The industry has endorsed the knock-for-knock rule under which losses lie
where they fall and govern their internal interactions. As a result, gas and oil com-
panies are subject to a dual liability regime: vis-à-vis third parties, industry participants
are subject to standard tort liability. Vis-à-vis each other, however, they operate under
the knock-for-knock principle.
The knock-for-knock principle eliminates the option of bringing private tort suits
for harms incurred by gas and oil companies in their inter se relationships. Instead,
it forces parties to incur the cost of the harms they sufered at the hands of other
contractual parties. From an insurance perspective, the knock-for-knock principle
relies exclusively on frst-party insurance, an intriguing feature that runs contrary to
contemporary theorizing and practice.1 Given the moral hazard problem generated
by the knock-for-knock regime, one would not expect it to see it in the real world.
Yet, it exists.
At frst glance, the gas and oil industry’s adherence to the knock-for-knock system
is peculiar. Economically-minded tort theorists would expect the knock-for-knock
system not to come into existence or unravel the moment a major accident occurs.
By its very design, the knock-for-knock rule gives rise to a serious moral hazard
problem. Left to their own devices, proft-maximizing actors who operate under
this system are expected to underinvest in precautions and fall short of the optimal
standard of care. Foreseeing this problem, rational actors would be expected to shun
the knock-for-knock principle and subject themselves to the formal tort system while
buying third-party insurance. Yet, insiders in the gas and oil industry maintain that
1 See Jefrey O’Connell & John Linehan, Neo No-Fault Early Ofers: A Workable Compromise Between
First and Third-Party Insurance, 41 Gonz. L. Rev. 103, 107–108 (tracing the rise in the focus on third-
party insurance in our tort framework). See also Nora Freeman Engstrom, An Alternative Explanation for
No-Fault’s “Demise”, 61 DePaul L. Rev. 303 (2012) on the history of the American no-fault experiment
explaining why automobile no-fault faltered as an efective legislative movement shedding light also on tort’s
durability.
DOI: 10.4324/9781003206798-3 65
gideon parchomovsky and endre stavang
the knock-for-knock system is efcient and vital to the successful operation of the
gas and oil sector.2
In this chapter, we explore the content and structure of the knock-for-knock regime.
Furthermore, we specify the conditions under which it may be preferable to standard
tort liability coupled with third-party insurance. Our ultimate goal, however, is to
conduct a comprehensive examination of the welfare implications of the knock-for-
knock principle. We conclude that despite the gas and oil industry endorsement of the
knock-for-knock principle, from a societal standpoint the case is at best questionable.
The chapter commences by tracing the historical origins of the knock-for-knock
principle and explaining its efects. The discussion then analyzes the knock-for-knock
principle from a social perspective, focusing on the potential for negative spillovers.
Although gas and oil insiders consider the knock-for-knock system efcient because
it economizes litigation costs, it is far from clear that it is also efcient from a social
perspective. From a social vantage point, the system is desirable if and only if the
private cost savings it afects are not outweighed by the moral hazard problem the
system creates.3
Hence we review the various legal and non-legal mechanisms capable of ame-
liorating the moral hazard problem arising from the knock-for-knock regime. The
four principal mechanisms are (1) contractual provisions that exclude intentional
harm and harm resulting from ultra-hazardous activities from the knock-for-knock
system; (2) deductibles and mandatory precautions requirements in insurance poli-
cies; (3) state and federal regulation; and (4) continuous interactions among repeat
players that lead to cooperative behavior. We conclude that the knock-for-knock
regime would be socially efcient if the activities of the contracting parties that may
result in harm could be kept totally distinct from activities that may result in harm
to third parties (we coin this condition “the separating equilibrium hypothesis”), or
alternatively, if the aforementioned mechanisms were powerful enough to force the
contracting parties to adopt the socially desirable level of precautions in all of their
activities, both inter se and vis-à-vis third parties (we term this option “the optimal
pooling equilibrium hypothesis”).
The chapter seeks to contribute to two distinct literatures. First, we hope to enrich
tort theory by studying an unusual liability regime that has evaded the searching eyes
of tort theorists. It is important to explain in this context that the knock-for-knock
2 See, e.g., Cary A Moomjian, Contractual Insurance and Risk Allocation in the Ofshore Drilling
Industry, Drilling Contractor, Jan./Feb. 1999, at 19–21; Trine-Lise Wilhelmsen, Liability and Insurance
Clauses in Contracts for Vessel Services in the Norwegian Ofshore Sector – The Knock-for-Knock Principle,
Scandinavian Yearbook of Maritime Law (SIMPLY) Conference paper for the Oslo/Southampton/Tulane
Conference, 2 October 2012, at 101. Wilhelmsen provides, on the basis of several decades of Maritime Law
scholarship in Norway on legal relationships involving knock-for-knock clauses, a very good account of the
system and its rationale. Id. at 87–94, 95–101. The leading treatment however is (albeit in Norwegian) Hans
Jacob Bull, Tredjemannsdekninger i forsikringsforhold (Oslo 1988).
3 Guido Calabresi famously termed this component “tertiary cost”. Guido Calabresi, The Cost of
Accidents (1970); see also infra note 45. On the relevance of various types of administrative costs for legal
rule design in this area, see also Ronald H. Coase, The Firm, the Market, and the Law 114–119 (1988);
Steven Shavell, Economic Analysis of Accident law, 262–291 (1987); Michael Faure & Göran Skogh,
The Economic Analysis of Environmental Policy and Law, 141–239 (2003); William D. Landes &
Richard A. Posner, The Economic Structure of Tort Law (1987); and Richard A. Posner, Catastro-
phe – Risks and Response (2005).
66
contracting around tort defaults
system is not a pure no-liability regime. Nor is it a no-liability regime with frst-
party insurance. Rather, it is a hybrid, or a dual-liability system under which parties
are subject to a no-liability regime and negligence or strict liability regime. Thus,
the knock-for-knock system creates unique incentives for the parties that cannot be
found in standard liability regimes. It also allows one to test the standard economic
models of tort liability in a complex real-world setting.
The second line of scholarly inquiry into which this chapter fts is the literature
on private ordering or private production of law. Building on Robert Ellickson’s
pathbreaking study of the extra-legal norms among neighbors in Shasta County con-
cerning trespass by cattle,4 scholars have set out to study communities that opted out
of the legal system and replaced it with a private set of norms. At the risk of mild
overgeneralization, it can be said that as a rule those studies endorse private order-
ing as the superior form of rulemaking. Our case study is unique in several critical
respects. To begin with, our case study involves only a partial opt-out from the legal
system. As importantly, accidents in the gas and oil industry often afect multiple
third parties, unlike previous studies which focused on largely self-contained commu-
nities or interactions, such as disputes among neighbors in an isolated community,5
or among merchants in the diamond industry,6 or fshermen in Massachusetts.7 One
need only recall the British Petroleum (BP) oil spill in the Gulf of Mexico to recall
this point. Relatedly, the harm resulting from an accident in the oil industry may be
enormous and irreparable. Finally, the oil industry constantly expands its operations
to new places. At the time of this writing, new sites are being developed of the coasts
of Mexico, Brazil and north of the border between Russia and Scandinavia. Every
such expansion extends the knock-for-knock regime to new territories, efectively
transplanting it into new legal systems.
At the end of the day, our analysis sounds a cautionary note as to the social desir-
ability of the knock-for-knock principle. And while we stop short of calling for a
legislative or regulatory ban, we believe that there is a strong prima facie case against
extending the knock-for-knock principle to other industrial settings.
Structurally, the chapter unfolds in four main sections. In section 1, we introduce
the knock-for-knock principle and explore its historical origins. In section 2, we posi-
tion the knock-for-knock principle within the theoretical literature on private order-
ing and social norms. In section 3, we specify the conditions under which a liability
regime, such as knock-for-knock, that requires parties to bear their own loss may
be socially desirable. In section 4, we illuminate the hidden costs of the knock-for-
knock rule and explain why it is unadvisable to extend it beyond its present scope.
A short conclusion ensues.
4 Robert C. Ellickson, Of Coase and Cattle: Dispute Resolution among Neighbors in Shasta County, 38
Stan. L. Rev. 623 (1986).
5 See, e.g., Glen O. Robinson, Communities, 83 Va. L. Rev. 269, 306–309 (1997).
6 Lisa Bernstein, Opting Out of the Legal System: Extralegal Contractual Relations in the Diamond
Industry, 21 J. Legal Stud. 115 (1992).
7 André Verani, Community-Based Management of Atlantic Cod by the Georges Bank Cod Hook Sector:
Is It a Model Fishery?, 20 Tul. Envtl. L.J. 359, 369–370 (2007).
67
2 The knock-for-knock principle
George Fletcher famously theorized that tort liability is predicated on the principle
of average reciprocity of risk.8 On this principle, tort liability should not attach to
actors who accidentally expose each other to risks roughly similar in their gravity,
while making reasonable eforts to avoid risky behavior. According to Fletcher,
such risks are reciprocal in nature and accordingly their materialization should not
trigger tort liability.9 In contrast, if an actor exposes another to a risk to which
she is not herself exposed and harm follows, tort liability should be imposed because
the risk was not reciprocal.10 Of course, whether a risk is reciprocal depends on
the parties relevant positions towards one another and the nature of their activities.
For example, an accident’s risk is reciprocal between two drivers, but non-reciprocal
between a driver and a pedestrian.11 Hence if an accident occurred in the former
case there would be no liability, whereas in the latter case liability would attach to
the injurious driver.12
As it was frst developed, the knock-for-knock system seemed to perfectly embody
Fletcher’s “reciprocity” principle. It was developed in London during World War II
to reduce litigation costs from frequent naval accidents. Responding to the threat of
German submarines, English ships sailed in the dark with their lights switched of
in tight clusters.13 This reduced ship exposure to German submarines but increased
the rate of collisions. Instead of engaging in expensive and prolonged litigation
over those harms, the parties decided to subject themselves to the knock-for-knock
principle, where each party agreed to bear its own costs.14 In this original setting, the
risk each ship exposed to the other was reciprocal in nature, albeit not identical.15
The knock-for-knock principle was extended to other settings. Until the 1960s, the
knock-for-knock principle governed automobile accidents in the United Kingdom,16
forcing drivers to buy frst-party insurance against accidents or bear their own losses.
This system was subsequently abolished.17 The knock-for-knock principle was also
extended to gas and oil companies and to diferent segments of the general ship-
ping industry. In both industries, the knock-for-knock principle continues to persist.
However, only in the gas and oil industry it became ubiquitous.18
8 George Fletcher, Fairness and Utility in Tort Theory, 86 Harv. L. Rev. 537 (1972).
9 Id. at 542.
10 Id. at 548.
11 See Gideon Parchomovsky, Fair Use, Efciency and Corrective Justice, 3 Legal Theory 347 (1997)
(using this example as an illustration of a reciprocal risk).
12 Id.
13 See generally Frank H. Shaw, The Convoy Goes Through (1942), available at www.convoyweb.org.
uk/extras/index.html (recounting frsthand experiences on World War II naval convoys).
14 See, e.g., Treaty with Great Britain on Marine Transportation and Litigation, U.S.-Gr. Brit., Dec. 4,
1942, 56 Stat. 1780 (1942). This is perhaps the best example of Fletcher’s “average reciprocity of harm.” See
Fletcher, supra note 8, at 542.
15 The precise risk to which each ship was exposed depended on its position in the feet. Ships that sailed
at the center of the feet were exposed to a greater risk.
16 Richard Lewis, Insurer’s Agreements Not to Enforce Strict Legal Rights: Bargaining with Government
and in the Shadow of the Law, 48 Mod. L. Rev. 275, 285–286 (1985).
17 Insurance companies, however, still use knock-for-knock arrangements in their inter se relationships
to economize on administrative costs. Id.
18 Wilhelmsen, supra note 2.
68
2.1 The industrial setting
Mature petroleum provinces, such as the North Sea and the Gulf of Mexico, contain
various patterns or mixes of gas and oil deposits and intricate infrastructure for
extracting and transporting the resource. The rig is the system’s heart and is used
for drilling. Surrounding the rig is a series of complex pipelines that transports the
extracted gas and oil. Ships are often used as an alternative means of resource trans-
portation, as well as a means for transporting workers and equipment. This intercon-
nected infrastructure is connected to the outside world through ports and onshore
pipeline endpoints, typically with some large industrial facility for handling, treatment
and further transportation on land.
The gas and oil industry has various positive and negative economic efects on its
surroundings. On the positive side, gas and oil platforms create jobs and generate
revenues for local businesses and communities. Furthermore, the gas and oil industry
often directly invests in the local communities. On the negative side, however, the
operation of gas and oil sites may cause adverse externalities. For example, it can
harm the environment through degradation of marine resources, discourage tourism
and possibly dislocate local populations.
The ownership interests of pipeline infrastructure are managed by joint venture
agreements between two or more companies.19 Typically, there are separate joint
venture agreements for the feld and for each pipeline, though in Norwegian waters
most of the pipeline infrastructure is managed through a single comprehensive joint
venture agreement among all the oil companies with an ownership stake in the
pipelines. The knock-for-knock system requires joint venture–related agreements to
contain knock-for-knock clauses that bind all the parties involved in the venture.
2.2 Illustrating the operation of knock for knock

As stated, the general thrust of knock-for-knock is to expose the contracting parties
to as little tort liability as legally possible and replace it with frst-party insurance.20
To illustrate the efect of the knock-for-knock rule, let’s assume that a drilling com-
pany contracts to perform drilling services for an oil company, as was the case in the
Deepwater Horizon blowout that occurred in 2010. Both the drilling company and
the oil company had insurance against harm. Each company used subcontractors,
who employed other subcontractors. At any given time, there was a mix of the drill-
ing company’s and oil company’s employees, as well as employees of the various
subcontractors on the drilling platform. Thus, an accident will sometimes initially
19 The joint ventures are typically not deemed to be separate legal entities.
20 Specifcally, a typical agreement would employ the following language: “Contractor shall indemnify
Company Group from and against any claim concerning: a) personal injury to or loss of life of any person-
nel of Contractor Group, b) loss of or damage to any property of Contractor Group”, etc. Sometimes these
wordings would exclude liability and order indemnifcation altogether without any exceptions. At other
times, the clauses would include exceptions for intentional harms and gross negligence by the employees with
whom the company in question could be identifed with, i.e., top management people. See sources indicat-
ing such capping and restriction of the scope for knock-for-knock agreements. For a current exposition of
knock-for-knock clauses in contracts for vessel services in the Norwegian ofshore sector, see Wilhelmsen,
supra note 2, at 87–94.
69
impose costs on the oil company or its subcontractors, and at other times the cost
will most immediately fall on the drilling company or its subcontractors.
In this setting, the knock-for-knock systems accomplishes two things. First, if
the oil company’s employees or property are harmed, the knock-for-knock clauses
exclude tort liability, preventing the oil company from suing the drilling company in
torts. Since the oil company’s insurance provider (under a frst-party contract) can-
not obtain a subrogated claim against the drilling company, no tort claim arises.21
Losses are fnanced through the oil company’s insurance contract. The same holds
true if the drilling company’s employees or property are harmed. Second, in the
case of harm to the oil company’s employees or subcontractors, the oil company
undertakes to indemnify the drilling company in the event of a lawsuit against it by
a “member” of the oil company family.22 Thus, the drilling company does not have
to call on its liability insurance. In this case, too, no lawsuit would be brought, either
by the drilling company or by its insurance company (based on subrogation). The
same holds true if a member of the drilling company’s family sues the oil company
or one of its members. Therefore, knock-for-knock clauses not only bar direct tort
suits by the contracting parties, but prevent the enforcement of subrogated claims
that would produce much of the same fnancial efects as tort claims.23
The function of the knock-for-knock principle is to allocate harm to the “victim”
ex ante while spreading losses through frst-party insurance.24 By efecting cost savings
through the elimination of litigation and cheaper insurance,25 the knock-for-knock
system increases gas and oil industry’s profts, thus it has become the darling of
industry.26 But the picture painted by industry insiders is too rosy. Knock-for-knock
arrangements give rise to a moral hazard problem.27 By exempting industry partici-
pants from liability for harm, they take away their incentive to act responsibly. Hence
from a social perspective, a knock-for-knock arrangement is welfare-enhancing if it
does not raise the number or severity of accidents or if the deterrence defcit it efects
is smaller than the cost-savings it generates.28
The knock-for-knock principle only applies to the internal relationship among
the contracting parties. It does not extend to harm caused to outsiders.29 In their
interactions with outsiders, the contracting parties are subject to the standard rules
of the tort system.30 Members of the gas and oil industry cannot unilaterally negate
or modify their liability through their contractual arrangements toward third parties,
21 Id. at 93–94.
22 Id. at 91–93.
23 Id. at 85–86.
24 Id. at 83.
25 Id. at 96.
26 In Wilhelmsen’s opinion, such clauses should be fully upheld by courts during any contract enforce-
ment in the wake of any accident. Id. at 102–111.
27 Id. at 98–99.
28 Id. at 101.
29 Id. at 90–91. Even though each party may still be liable to damages negligently inficted upon third
parties, employees of the contracting parties are not considered third parties and still fall within the scope
of the agreement: “[T]he Owner/Company or Charterer/Contractor not only agrees to be responsible for
any damage that befalls the property of the company or the property or the persons of the employees, but
also assumes responsibility for such damages throughout the group.” Id. at 89.
30 Id. at 83.
70
such as fshermen, local governments, or communities. Accordingly, the activities of
industry participants that afect third parties are governed by either strict liability
or negligence. In efect, therefore, members of the gas and oil industry voluntarily
subject themselves to a dual liability regime. Vis-à-vis each other, the knock-for-
knock principle provides that each contractual party must bear the harm inficted
upon it by other contractual parties even when they behaved negligently. Vis-à-vis
third parties, industry members must abide by the rules of the tort system or else
face the consequences.31
3 Knock for knock, social norms and private ordering

At frst glance, it could be concluded that the knock-for-knock system is another
example of private ordering superiority over public ordering. This is because it con-
stitutes another example where a private system contractually produced a better liabil-
ity regime than provided by tort law. Yet, we are not at all convinced this claim is
demonstrative.
The “order without law” literature pioneered by Robert Ellickson32 might suggest
that the knock-for-knock regime is just another example of efcient private order-
ing. In his famous study of Shasta County, Ellickson studied how farmers resolved
trespass disputes. He reported that his study’s participants bypassed the legal system
and followed a “live and let live” norm.33
In another important study of the diamond industry, Lisa Bernstein found that
diamond merchants opted out of the legal system and developed their own set of
courts and substantive rules.34 Bernstein argued that the private rules are clearly
superior to the defaults provided by the legal system. Similarly, Eric Feldman, who
studied the norms of dispute resolution among tuna fshermen in Japan, discovered
that members of the fshing industry adopted a unique system of norms and institu-
tions that allows for an expedient resolution of disputes.35
Yet there is a major diference between our study and those of Ellickson, Bernstein
and Feldman. While the other studies pointed to the superiority of private ordering
over state production of legal rules, which celebrated the advantage of private enter-
prise and the prevalence of social norms over formal legal rules, our study presents
a much more challenging case. In the case of the knock-for-knock principle, it is
impossible to unequivocally conclude that it is welfare-enhancing. This is because the
operations of the gas and oil industry give rise to myriad efects that are borne by
other parties. The scale of the operations of the gas and oil industry is enormous,
and it is not nearly as self-contained as the diamond industry, the ranchers in Shasta
County or the tuna fshermen in Japan. Moreover, accidents in the gas and oil indus-
try have far-reaching consequences that go beyond the interests of other industry
31 Id.
32 Ellickson, supra note 4.
33 Id. at 15 et seq.
34 Bernstein, supra note 6.
35 Eric Feldman, The Tuna Court: Law and Norms in the World’s Premier Fish Market, 94 Cal. L. Rev.
313 (2006).
71
participants. An oil spill, unlike a wandering cow or a dispute about a diamond or
a tuna fsh, can devastate entire populations of people. Hence the knock-for-knock
principle tests the outer limits of private ordering and must therefore be subject to
careful scrutiny.
Furthermore, it is clear that Lisa Bernstein’s explanation why the diamond industry
opted out of the legal system does not apply in our case. Bernstein suggested that the
diamond industry’s reluctance to subject itself to judicial review was largely motivated
by concealment considerations.36 Industry participants wanted to keep the industry
and its practices away from the public and therefore shy away from court proceed-
ings that would lead to disclosure. Concealment is not a strategy that society should
condone, let alone discourage, in oil industry. Accidents in the gas and oil industry
often give rise to casualties and substantial environmental harms. Consequently, the
operations of gas and oil companies ought to receive – and, in fact, do receive –
close scrutiny from environmental and other regulatory agencies both ex ante and
ex post. Furthermore, when an accident occurs it will be thoroughly investigated by
the police and state or federal prosecutors.
We, therefore, proceed on the assumption that the knock-for-knock principle
maximizes gas and oil industry profts. The accepted rationale for the knock-for-
knock regime is that the savings generated in insurance and litigation costs exceed
the possible increase in the cost of accidents because state regulation of health,
safety and the environment provides an adequate deterrent against risky behavior.37
By adopting the principle, the contracting parties can then share a surplus, making
knock-for-knock mutually advantageous.38
This does not mean, of course, that the preference of the gas and oil industry is
optimal from a social perspective. The savings in tertiary costs39 may be overstated,
the costs of harm may be underestimated, and the regulatory burden resulting from
this system may be too heavy. A theoretically inclined reader may fnd an afnity
between the knock-for-knock rule in the context of accidents and the “live-and-let-
live” rule in nuisance.40 But just like the belief that nuisance law is growing organically,
harmoniously and along a socially optimal path all by itself is not self-evident, the
knock-for-knock’s contractual underpinnings have to be carefully evaluated from a
36 Bernstein, supra note 4 at 148 (noting that diamond dealers prefer arbitration to adjudication because
it is, inter alia, more secret).
37 Bull, supra note 2, 333–357; Wilhelmsen, supra note 2, 95–101.
38 From the perspective of economic theory both strict liability as well as the negligence rule may be the
basis for a preferable liability regime. But obviously a no liability regime is preferable if liability results in
relative small deterrence efect compared to costs of operating the regime. See, e.g., R.H. Coase, The Problem
of Social Cost, 3 J. L. & Econ. 1, 44 (1960); Louis Kaplow & Steven Shavell, Fairness Versus Welfare, 114
Harv. L. Rev. 961, 1011–1017 (2001).
39 Guido Calabresi defnes “tertiary costs” as those arising from administering an accident liability
regime. These include the costs of reducing “primary costs” of accidents, which are the direct harm to vic-
tims and costs of accident avoidance, and “secondary costs”, the costs of spreading risks in an economically
desirable way, such as through insurance. Calabresi, supra note 3, at 24 et seq.
40 See, e.g., Richard A. Epstein, Nuisance Law: Corrective Justice and its Utilitarian Constraints, 8 J.
Legal Stud. 49, 82–87 (1979); Robert C. Ellickson, Order Without Law: How Neighbors Settle
Disputes (1991).
72
social welfare point of view.41 It is also not the case that the social norm literature
has successfully demonstrated how private law can be replaced partly or in whole
by private ordering through informal norm formation.42 In sum, the stakes in oil
provinces like the Gulf of Mexico and the North Sea are so high that the extant
academic attention given to this issue seems unreasonably low.43
4 The efect of knock for knock on social welfare

Economic models of tort liability teach that a no-liability regime places the burden
of investing in precautions solely on the potential victim.44 As a result, the injurer
has no incentive to expend resources on accident avoidance or on mitigation of harm
because they do not bear the cost of harms to third parties. The injurer will invest
in precautions only to prevent self-harm and only if the expected harm is greater
than the cost of the precautions.45 Any expenditure that exceeds the magnitude of
the expected self-harm is a pure waste.
41 For recent nuisance scholarship taking a social welfare perspective, see, for example, Keith N. Hylton,
The Economic Theory of Nuisance Law and Implications for Environmental Regulations, 58 Case W. Res. L.
Rev. 673 (2008), Henry E. Smith, Exclusion and Property Rules in the Law of Nuisance, 90 Va. L. Rev. 965
(2004).
42 Robert Cooter, Structural Adjudication and the New Law Merchant: A Model of Decentralized Law,
14 Int’l Rev. L. Econ. 215, 226 (1994); Eric Posner, Law and Social Norms, 176, 179 (2000); and Robert
C. Ellickson, The Market for Social Norms, 3 Am. L. Econ. Rev. 1, 30–36 (2001).
43 Most articles directly addressing the benefts and drawbacks of the knock-for-knock rule are still
relegated to less-prominent journals and draw little attention as of yet. See, e.g., Nick Kangles et al., Risk
Allocation Provisions in Energy Industry Agreements: Are We Getting It Right?, 49 Alta. L. Rev. 339 (2011);
Christopher L. Evans & F. Lee Butler, Reciprocal Indemnifcation Agreements in the Oil Industry: The Good,
the Bad, and the Ugly, 77 Def. Counsel J. 226 (2010). Even three years after the Deepwater Horizon explo-
sion, only one article analyzing the knock-for-knock rule in light of the disaster has been written. Chidi
Egbchue, Reviewing Knock-for-Knock Indemnities Following the Macondo Well Blowout, 7 Const. L. Int’l
7 (2013).
44 For the canonical expositions, see Shavell, supra note 3, and John Prather Brown, Toward an Eco-
nomic Theory of Liability, 2 J. Legal. Stud. 323 (1973). Calabresi, supra note 3, is the main prior contribu-
tion. However, more than a century ago, the Austrian lawyer-economist Victor Mataja ofered the following
observations on tort liability for, e.g., environmental harms. First, he pointed out that, a negligence rule is
sometimes insufcient and strict liability is desirable on the basis of cost-internalization considerations.
Second, he noted that a system of strict liability must rely on a broad conception of compensable harm is
required, one that goes beyond pecuniary harm. Only in this case will it be safe to assume that cost inter-
nalization leads to a powerful enough incentive to take precautions. See Izhak Englard, Victor Mataja’s
Liability for Damages from an Economic Viewpoint: A Centennial to an Ignored Economic Analysis of Tort,
10 Int’l R. L. Econ. 173 (1990). The interdependence of the issues of strict liability and compensable harm
illustrated by Mataja with the help of the welfare economics of his day is confrmed by the now standard
economic model of precaution, harm, and liability. See, e.g., Endre Stavang, Explaining Welfare-Based Torts,
4 Global Jurist Topics 1 (2004). Finally, “the economic model of precaution and liability” may be seen as
a corollary of Aristotle’s famous quip about how un-owned resources are least well managed. In a similar
vein, Dari-Matiacci and Parisi remark: “[A]lthough the cost-beneft calculus is a tool of modern economists,
lawmakers have long understood the need to balance the costs and benefts produced by desirable activities.
The intellectual roots of tort law can be traced back as far as the Roman principle that a party reaping the
benefts (commoda) of an activity should also bear its costs (incommode). And in a broad sense, the eco-
nomic analysis of tort law can be seen as a mere formalization of that pre-theoretical intuition.” Amsterdam
Law School Legal Studies Research Paper No. 2013–09/Amsterdam Center for Law & Economics Working
Paper No. 2013–01, 2. In sum, the model crystallizes sustainable (contingent) normative insights.
45 Robert Cooter & Ariel Porat, Does Risk to Oneself Increase the Care Owed to Others – Law and Eco-
nomics in Confict, 29 J. Legal Stud. 19, 21–22 (2000).
73
This means accident rates and their order of magnitude should be greater under
the knock-for-knock rule than either a negligence or a strict liability rule – the two
standard default settings for tort systems. Hence, the knock-for-knock rule would be
welfare-enhancing only if one of the following conditions obtains: (1) “the perfect
separating equilibrium”, where the inter se activities of the contracting parties are
completely separate from the activities that afect the rest of society, such that the
harms resulting from these activities will not afect third parties; (2) the expected pri-
vate harm to the contracting parties from accidents exceeds the expected social harm
and the parties contractually adopted optimal measures to address the private harm
(we term this option the “perfect alignment hypothesis”); (3) the knock-for-knock rule
creates negative externalities for third parties, but those are outweighed by the gains
to the contracting parties. Otherwise, the knock-for-knock system is inefcient. We
therefore proceed to discuss the likelihood that one of the previous conditions exists.
4.1 The perfect separating equilibrium

In an ideal setting, the knock-for-knock rule could be efcient when there is perfect
separation between the activities undertaken by the contracting parties toward each
other and their activities that afect third parties. The perfect separation condition
implies the existence of two distinct sets of relationships: one between the parties to
the joint venture contract and another between those parties and the rest of the
world. It also requires that there are no spillovers between the two relationships.
Under these conditions, the erosion of the parties’ motivation to optimally invest
in precautions would not adversely afect third parties. And if one assumes – as gas
and oil insiders do – that the knock-for-knock system is internally efcient, then it is
also socially acceptable because the external risks created by the contracting parties
are governed under the tort system.
Although the perfect separating equilibrium is a theoretical possibility, it is hard
to see how it can exist. Gas and oil operations are complex, and it is unrealistic for
contracting parties to separate activities that come under the knock-for-knock system
from those that may afect third parties. In fact, the opposite is true. Experience teaches
us that any accident potentially afects insiders and outsiders alike. This is clearly true
for large-scale accidents, such as the BP leak, but it is also true for accidents whose
efects are more limited. This does not mean, of course, that all accidents in the gas
and oil industry generate third-party efects. However, this is not the relevant inquiry.
All it takes to refute the perfect separating equilibrium hypothesis is to show that
some accidents give rise to adverse third-party efects and that these accidents occur
as result of the lower level of care from the knock-for-knock system.
4.2 The perfect pooling equilibrium

The knock-for-knock system may also be socially desirable if it does not lower the
contracting parties’ investment in precautions. Their precaution regarding investments
may not change because various internal and external mechanisms – some of them
private and some of them public – that apply to them. There are four possible
mechanisms that may induce contracting parties to adopt the socially optimal level
74
of care, notwithstanding the knock-for-knock system: (1) self-regulation; (2) repeat
interactions; (3) health and safety regulation; and (4) mandatory provisions in insur-
ance policies. A rich literature suggests that each of these mechanisms may ameliorate
the moral hazard problem to which the knock-for-knock system gives rise. We examine
each mechanism in turn.
4.2.1 Self-regulation
Left unchecked, the knock-for-knock principle creates severe moral hazard problems.
This is because each party bears none of the cost it imposes on other contracting
parties but obtains the full cost of precautions. Thus, parties operating under the
knock-for-knock principle have an inherent incentive to underinvest in precautions
and take on an excessive level of risk. To address this problem, the contracting par-
ties and insurance companies introduced contractual measures to ensure parties do
not abuse the knock-for-knock system. In the main, those measures fall into two
categories: exclusions and deductibles.46
As their name implies, exclusion clauses deny coverage to parties in cases of harm
resulting from intentional acts and gross negligence. Exclusions provide an efective
mechanism for controlling the insured’s behavior when the insurer can set the proper
standard of behavior and can monitor the activities of the insured at a reasonable
cost.47 Although the degree of an actor’s negligence cannot always be readily estab-
lished, insiders in the gas and oil industry have developed shared conceptions of
behaviors that fall outside ordinary negligence.
The second standard mechanism employed to counter opportunistic behavior is
deductibles.48 Deductibles are a standard feature in all insurance contracts.49 The pur-
pose of deductibles is to prod the insured to invest in precautions up to the deductible
amount. Deductibles should ideally be set to equal the cost of optimal precautions.50
In the real world, however, insurers often do not have the information necessary to
achieve such accuracy.51 Yet, even if a deductible is set too high, it would induce the
insured to take precautions.52
Although exclusions and deductibles operate within the knock-for-knock system’s
framework, they also reduce the risk of harm to third parties. Accidents resulting in
harm to third parties may arise from negligent acts, gross negligence, or intentional
acts. It is reasonable to hypothesize that, on average, harm is greater in cases of
gross negligence or intentional behavior. If this is true, then contractual exclusions
dramatically lower the risk of harm to third parties.
Furthermore, because it is difcult to distinguish between “merely” negligent acts
and acts that constitute gross negligence, it is possible that the contracting parties
46 See, e.g., Jeffrey W. Stempel, Stempel on Insurance Contracts, 2–95 et seq. (2007).
47 Id. at 2–96.
48 Eugene R. Anderson et al., Insurance Coverage Litigation, 1–59 (2004).
49 Id.
50 B. Peter Pashigian et al., The Selection of an Optimal Deductible for a Given Insurance Policy 39 J.
Bus. 35 (1966).
51 Id. at 42.; Ralph A. Winter, The Liability Insurance Market, 5 J. Econ. Persp. 115, 122–124 (1991).
52 Given that the underlying legal system constitutes the right incentives as in Shavell, supra note 3,
at 5–47.
75
would prefer cautious conduct and refrain from questionable behavior in their internal
interactions.53 But irrespective of the veracity of these conjectures, because exclusions
and deductibles lower the probability of inter se accidents, the likelihood of harm
to the environment is reduced. Moreover, said contractual mechanisms also obviate
the need for regulatory oversight to some extent, and thereby save societal resources.
4.2.2 Repeat interactions

The moral hazard problem is further mitigated because members of the gas and oil
industry continuously interact with each other. As Robert Axelrod54 and others55
established, parties engaged in repeated interactions tend to follow the “tit-for-tat”
strategy. Specifcally, Axelrod demonstrated that in repeat interactions the dominant
strategy is to cooperate as long as the other party cooperates and to punish defection
by defection.56 An important implication is that long-term relationships between the
same actors tend to foster cooperative behavior.57
Historically, the gas and oil industry constituted a near-perfect setting for a tit-
for-tat strategy. The industry comprises a relatively small group of members that
interact with each other continuously. The companies in the gas and oil industry
are also familiar with one another, and the reputational stakes they face are high.
Consequently, members have an incentive to maintain a good relationship with their
peers and not “defect” and risk retaliation. At least historically, the industrial setting
in the gas and oil sector has been similar to the setting Ellickson found in Shasta
County. And as the setting in Shasta County led to the adoption of a cooperative
norm of “live and let live”, the conditions that prevail in the gas and oil industry are
likely to induce a similar cooperative approach among industry members.
Specifcally, we expect repeat interactions among members to exert a disciplin-
ary efect on their members’ behavior through inter se relationships. Members are
likely to pass small or short-term gains and avoid precautionary investments to reap
long-term gains resulting from cooperation. If, indeed, the contracting parties are
guided by long-term cooperation, their investment in precautions, even in their inter
se relationships, may not fall short of the socially desirable standard. Consequently,
the fear that the knock-for-knock regime gives rise to negative externalities could be
much smaller than one would think. That said, one cannot rule out the possibility
that some industry participants may not be adequately informed about the risks
involved in gas and oil operations, which may lead to distortions in the pricing of
53 See, e.g., Robert D. Cooter & Thomas S. Ulen, Law and Economics (4th ed.) (2003); John E. Calfee
& Richard Craswell, Some Efects of Uncertainty on Compliance with Legal Standards, 70 Va. L. Rev. 965
(1984).
54 Robert Axelrod, The Evolution of Cooperation: Revised Edition (2009).
55 See, e.g., Paul G. Mahoney & Chris William Sanchirico, Norms, Repeated Games, and the Role of
Law, 91 Cal. L. Rev. 1281, 1289 (2003); George J. Mailath & Larry Samuelson, Repeated Games and
Reputations: Long-Run Relationships 3–4 (2006); Robert E. Scott, Confict and Cooperation in Long-Term
Contracts, 75 Cal. L. Rev. 2005, 2026 (1987).
56 Axelrod, supra note 54.
57 See, e.g., Mailath & Samuelson, supra note 55; Jane Sell & Rick K. Wilson, Maintenance of Coopera-
tion: Expectations of Future Interaction and the Trigger of Group Punishment, 77 Soc. F. 1551, 1552–1554
(1999); Rachel E. Kranton, The Formation of Cooperative Relationships, 12 J. L. Econ. & Org. 214, 217–219
(1996).
76
risks.58 Furthermore, the desire of regulators to make the industry more competitive
by increasing the number of active frms in an oil producing locale may exacerbate
knock-for-knock–induced moral hazards without ever reaching the elusive goal of
perfect competition.
4.2.3 Health and safety regulation

The presence of health and safety regulation is another factor that ameliorates the
knock-for-knock regime’s potential moral hazard problem. Gas and oil companies
operate within a regulatory framework that afects every facet of their activities. As
a rule, regulatory standards are immutable and are not afected by private ordering.
The presence of health and safety regulations forces gas and oil companies to comply
with their content and ensure that they do not fall below the regulatory standard.
Hence if the relevant regulatory standards are set appropriately, they should guarantee
that the operations of oil and gas companies do not pose a risk to third parties.
Critically, the need to comply with regulatory standard should also afect the level
of care taken by gas and oil companies in their internal relations. Assuming, as we
do in this section, that it impossible to separate activities that may cause harm to
other contracting parties from those that jeopardize outsiders, the measures taken to
comply with regulatory standards should also yield benefts to contracting parties. In
other words, the reduction in the external risk brought about by regulation should
also diminish the internal risk faced by the contracting parties.
Consider, for example, safety regulations for the construction and operation of
drilling rigs. The regulations are designed to protect the public at large. But frst
and foremost, they protect the employees of the companies that operate the rig and
perform the drilling. Of course, accidents may, and unfortunately do, happen, but
clearly there are positive spillovers from the regulation for the contracting parties.
Clearly, the more comprehensive the regulation, the lesser the scope of the moral
hazard problem.59 In the extreme, if regulations covered every aspect of the contract-
ing parties’ operations and were perfectly enforced, there would be no moral hazard
problem. In this case, the regulation would force the parties to adopt the socially
optimal level of care and the knock-for-knock regime’s sole efect would be to econo-
mize litigation costs. Yet, in the real world, not all aspects of the operations of gas
and oil companies are subject to regulation. It must also be recognized that not all
regulatory standards are set properly and that industry participants do not always
comply with them. The administrative and information costs that are endemic to the
regulatory process often create regulatory lacunas or gaps.60 This, in turn, makes it
rational for efciency-minded regulators to rely on courts to fll those gaps.61
58 Wilhelmsen, supra note 2, at 101.

59 See, generally, on the comparative analysis of approaches to the control of risk, Shavell, supra note
3, at 277–291.
60 Shavell, supra note 3, at 282.
61 Susan Rose-Ackerman, Public Law versus Private Law in Environmental Regulation: European Union
Proposals in the Light of United States and German Experiences, in Law and Economics of the Environ-
ment 13, 19–22 (Erling Eide & Roger van den Bergh, eds. 1996). For economic modeling of the argument
that regulation and liability should be used jointly because neither policy used alone incentivizes frms to
choose the efcient level of care, see Edward L. Glaeser & Andrei Schleifer, The Rise of the Regulatory State,
77
Notwithstanding this observation, regulation is only as efective as its enforcement.
Even the best regulatory standards would fail to yield socially optimal results if they
were not properly enforced. The dynamic and highly complex operations of the gas
and oil industry call for continuous monitoring and enforcement. Naturally, doing
so is time-consuming and cost prohibitive, and even the best regulators may not live
up to this standard. Tellingly, in its fnal report, the National Commission on the
BP Deepwater Horizon Oil Spill concluded that the accident was avoidable and that
the accident’s responsibility should be apportioned between the private companies
that operated the site, “in the frst instance by BP, Halliburton and Transocean” and
“government ofcials who, relying too much on industry’s assertions of the safety of
their operations, failed to create and apply a program of regulatory oversight that
would have properly minimized the risk of deep water drilling.”62
In the real world, the scope and substance of the regulation imposed on the
gas and oil industry varies dramatically among countries. In Norway, for example,
regulation of maritime and petroleum safety is largely based on self-assessment and
self-regulation.63 The particular regulatory standards prescribe broad end-result while
containing no reference to specifc behavior. It is therefore up to the industry to
choose how to reach the performance requirements. If an accident occurs, sanctions
may be imposed on the company. This, in turn, may lead to under-deterrence on the
individual level. Moreover, the informational ex ante advantage enjoyed by frms in
assessing the costs and benefts of rules aimed at inducing safety can be used by them
to subvert the regulation and in some cases may even result in industry-wide collusions.
4.2.4 Insurance
The economic and legal literatures suggest that insurance contracts signifcantly afect
the behavior of the insured.64 Multiple insurance scholars developed this theme and
termed the phenomenon “insurance as governance.”65 Insurance contracts often require
policyholders to adopt various precautionary measures on the insured party as a
precondition for collecting the insurance money. Furthermore, insurance policies
41 J. Econ. Lit. 401 (2003); Charles D. Kolstad, Thomas S. Ulen, & Gary V. Johnson, Ex Post Liability for
Harm vs. Ex Ante Safety Regulation: Substitutes or Complements?, 80 Am. Econ. Rev. 888 (1990); Patrick
W. Schmitz, On the Joint Use of Liability and Safety Regulation, 20 Int’l Rev. L. Econ. 371 (2000); Steven
Shavell, A Model of the Optimal Use of Liability and Safety Regulation, 15 Rand J. Econ. 271 (1984). For
a study contextualizing these and other theoretical contributions through analysis of Norwegian law, see
Ekaterina Denisova, The Role of Environmental Civil Liability in Regulation of Marine Oil Pollution in Nor-
way, 383 MarIus (Scandinavian Institute of Maritime Law) 1 (2009).
62 John M. Broder, Blunders Abounded Before Gulf Spill, Panel Says N.Y. Times, 6 January 2011 at
A14, available at www.nytimes.com/2011/01/06/science/earth/06spill.html?_r=3&ref=gulfofmexico2010&.
For broader discussions of the regulatory aspects of Macondo, see, e.g., Rebecca M. Bratspies, A Regula-
tory Wakeup Call: Lessons from BP’s Deepwater Horizon Disaster 5 Golden Gate U. Envtl. L.J. 7 (2011);
Mark A. Cohen et. al. Deepwater Drilling: Law, Policy, and Economics of Firm Organization and Safety,
Resources for the Future Discussion Paper 10–65 (2011); Sandra Zellmer, Joel A. Mintz, & Robert Glicksman,
Throwing Precaution to the Wind: NEPA and the Deepwater Horizon Blowout 2 J. Energy Env. L. 62 (2011).
63 Anita Ronne, Alternative Approaches to Regulatory Agency Structures and Powers: Eastern and West-
ern Europe, 15 J. Energy & Nat. Resources L. 41, 44 (1997).
64 Steven Shavell, On Liability and Insurance, 13 Bell J. Econ. 120 (1982).
65 See, e.g., Kenneth Abraham, Four Conceptions of Insurance, 161 U. Pa. L. Rev. 653, 683–697 (2013);
Richard V. Ericson & Aaron Doyle, Uncertain Business: Risk, Insurance and the Limits of Knowl-
edge (2004); Richard v. Ericson et al., Insurance as Governance (2003).
78
employ exclusions and deductibles to ameliorate adverse selection and moral hazard
problems that plague insurance markets. Consequently, insurance markets can supple-
ment or even supplant private litigation, regulation, and social norms as a means for
guiding behavior.66
We do not dispute the general observation that insurance companies can curb risky
behavior. That said, the insurance market plays a very limited role in regulating the
operations of large gas and oil companies around the world. To begin with, frst-party
insurance contracts only target the behavior of accident victims. Obviously, frst-party
insurance contracts cannot signifcantly infuence other companies’ level of care in joint
ventures. One could nonetheless suggest that if the same insurance company were to
insure all industry members, it could afect the behavior of all parties involved and
induce them to adopt the appropriate level of care. This is not the case in the real
world, however. Oil companies and many of their contractors sometimes self-insure
and more typically own or control small insurance companies that provide them with
frst-party insurance. The insured risks then get traded on reinsurance markets. Our
conjecture is that this approach treats risk more as a basis for bets and fnancial cal-
culations in the “Protection and Indemnities (P&I)” market67 and does not necessarily
lead to a governance scheme, which is more likely to arise under a tort liability regime.
5 The hidden perils of knock for knock

Thus far, our analysis has established that the parties’ contractual measures coupled
with the regulations and norms born from repeat interactions do not ensure optimal
investment in precautions; rather, the combined efect of these measures is to amelio-
rate the moral hazard problem to some extent. In light of this conclusion, it is impos-
sible to say, in the abstract, that the private gains to the contracting parties from the
knock-for-knock rule may be greater than the marginal reduction in deterrence afected
by the rule. Hence, based on our analysis, we cannot assert with any degree of certainty
that – as industry insiders insist – the knock-for-knock rule is efcient.
In this section, we highlight four additional factors that cast further doubt as to
the desirability of knock-for-knock arrangement in the gas and oil industry. The frst
is the assumption introduced in section 4.2, namely risk interdependencies. We show
that because the gas and oil industry has a high level of risk interdependencies, rela-
tively minor oversight can rapidly evolve into major accidents. This means that it is
critical to prevent even the smallest risks from arising. The second factor is what we
call “litigation externalities.” As we will explain, because knock-for-knock clauses bar
suits by other contractual parties, they efectively eliminate the most natural plaintifs,
who are best positioned to sue for deviations from socially acceptable practices. The
third factor we highlight in this part is harm to environmental interests. As we will
66 Faure & Skogh, supra note 3, 263, 264.

67 Remarks, although not as skeptical as here, on P&I clubs, are made in Jan C. Bongaerts & Aline F.M.
De Bièvre, Insurance for Civil Liability for Marine Oil Pollution Damages, 12 Geneva Papers on Risk and
Insurance 145 (1987); Faure & Skogh, supra note 3, at 273–274; Gary D. Libecap & Steven N. Wiggins,
Contractual Responses to the Common Pool: Prorationing of Crude Oil Production, 74 Am. Econ. Rev. 87
(1984).
79
demonstrate that the knock-for-knock rule exacerbates the risk of environmental harm
relative to standard liability regime. Fourth, we will explain why ex post compensa-
tion does not provide an adequate solution for the erosion in deterrence resulting
from knock-for-knock clauses. The harms resulting from accidents in the gas and
oil industry are so grave that compensation after the fact would not be an efective
remedy. In the case of large-scale accidents, the damage amounts are so substantial
that the responsible party would prefer to fle for bankruptcy protection or pledge for
government intervention to cap its liability to victims. Ironically, perhaps, the gravity
of the harm prevents society from relying on ex post compensation, requiring instead
that society puts the premium on ex ante deterrence.
5.1 Risk interdependencies

As we explained in section 4.1, if the contracting parties agree to the knock-for-knock
arrangement and their interactions can be kept distinct from actions that afect third par-
ties, then society has no reason to worry about the efect of the regime on third parties.
By contrast, if the contracting parties’ actions, especially their precautionary investment,
adversely afect the public at large, then knock-for-knock clauses present a serious cause
for concern. In the latter case, the contracting parties’ decision to forgo certain precautions
on account of the knock-for-knock regime increases the risk to third parties.
In the gas and oil industry, the degree of interdependence among risks is high.
Relatively minor errors in the operation of oil rigs and drilling equipment may have
disastrous consequences for the public at large. The Gulf of Mexico oil spill, also
known as the “Macondo blowout”, provides a gruesome illustration. A White House
commission reported and placed the accident’s blame on the cost-cutting policies
adopted by British Petroleum and its partners. Specifcally, it stated that “many of the
decisions that British Petroleum, Halliburton, and Transocean made that increased
the risk of the Macondo blowout clearly saved those companies signifcant time (and
money).” Furthermore, the report cautioned that “absent signifcant reform in both
industry practices and government policies, [a blowout] might well recur.”
While these fndings do not directly link the knock-for-knock regime to the oil spill,
it does suggest that cutbacks on investment in precautions can have serious adverse
efects on third parties. And although we cannot prove knock-for-knock clauses caused
the insufcient precautions adopted by British Petroleum and its partners, our earlier
analysis demonstrates a characteristic of knock-for-knock clause logic is that they
are liable to efect a reduction in the level of care relative to standard tort liability.
5.2 Litigation externalities

The deterrent efect of legal rules critically depends on enforcement. In the present
context, liability is largely governed by private enforcement. Both standard tort liability
and contract liability require the fling of private enforcement suits by victims.68 If
private lawsuits are not brought, under-deterrence would result. Once litigation is
68 Shavell, supra note 3, at 262–277.
80
incorporated into our analysis, another peril of knock-for-knock clauses comes to
the fore. Knock-for-knock clauses “eliminate” the plaintifs who are most likely to
sue – namely, the other participants in the joint venture. The other members of the
joint venture are not only the most likely victims but also the parties who can most
cost-efectively sue. The private parties involved in the venture are best situated to
prove deviations from the socially desirable standard of behavior. Their daily presence
on gas and oil sites coupled with their familiarity with safety standards puts them
in a unique position to monitor the actions of their partners and collect evidence of
sub-standard behavior. Hence they are the most natural private enforcers.
The elimination of suits for negligence by other contracting parties leaves the lion’s
share of the enforcement efort to third parties, who are not bound by knock-for-
knock clauses. However, in many cases the harm to third parties is too low to justify
legal action. This is especially true in jurisdictions that do not recognize class actions
and whose laws do not incorporate other procedural mechanisms for aggregating
small claims. In this context, it behooves us to remind our readers that class actions
are largely unrecognized in Europe,69 as well as in most other countries around the
world.70 Furthermore, even in those cases in which the harm to third parties may be
substantial, it may be too difcult or costly for them to prove negligence in court. Third
parties wishing to sue for harms they sufered at the hands of gas and oil companies
would be forced to expend signifcant resources to substantiate their claims. The high
cost of litigation means that only parties who incurred a severe harm – that is, a
harm greater than the expected cost of fling suit – should be expected to sue. The
upshot is that many small and even medium-sized harms may be left unredressed.
It is likewise critical to understand that litigation costs are not symmetrical.71 Fur-
thermore, there is a substantial heterogeneity among defendants as far as litigation
costs are concerned. The cost of litigation is higher for individuals who are not repeat
players.72 Such parties must build their cases from scratch and incur expenditures
that repeat players avoid on account of economies of scope and scale.73 To give a
concrete example, compare the cost of defense for a gas and oil company that gets
sued for negligence to the cost of a frst-time plaintif. The company typically has
a legal department that is well versed in such claims after handling multiple similar
suits. The plaintif, by contrast, would have to search for an attorney, who would
have to educate herself about the facts of the case and the applicable law. This means
that defendants enjoy an inherent cost advantage over most individual defendants.
69 See generally, Deborah R. Hensler, Goldilocks and Class Actions (response to Margaret H. Lemos,
Aggregate Litigation Goes Public: Representative Suits by State Attorneys General, 126 Harv. L. Rev. 486
(2012)), 126 Harv. L. Rev. F. 56, 56–7 (2012) (noting that this issue is the subject of vigorous debate in the
European Union, where advocates for private enforcement of antitrust and consumer protection law have
struggled against those who champion traditional European reliance on public enforcement and deride
proposals for “American-style class actions”.).
70 Id. at 57 (“[T]wenty-odd countries outside the United States that have adopted class actions, most
have limited standing to represent a class to public ofcials or nonproft organizations vetted or approved
by the government.”).
71 See generally Gideon Parchomovsky & Alex Stein, The Relational Contingency of Rights, 98 Va. L.
Rev. 1313 (2012) (noting the phenomenon and analyzing its implications).
72 Id. at 1344 (explaining the cost advantage of repeat litigators).
73 Id. at 1318 (discussing the efect of economies of scale on scope on litigants).
81
The corporations that participate in the joint venture could level the litigation play-
ing feld to some extent on account of their being repeat players who are familiar
with the legal machinery of litigation. But for the knock-for-knock clauses, those
companies could bestow a positive externality on third parties and remediate to some
degree the inherent under-deterrence problem. However, the knock-for-knock regime
prevents this from happening.
5.3 Harm to environmental interests

The environmental harm caused by various gas and oil companies are well-docu-
mented, and one may surmise that there are other long-term harms that have yet to
materialize. Let us be clear: we do not dispute the evidence.74 But this is not the point.
The critical question is whether the knock-for-knock principle contributes to these
harms. Or, to state the question somewhat diferently: would those harms not occur
under a negligence or strict liability regime? Although standard liability regimes are
highly imperfect in protecting the environment, our analysis suggests that the knock-
for-knock system exacerbates the risk of environmental harm.
Environmental interests may be protected under a strict liability or a negligence
rule. At least in theory, both negligence and strict liability can lead to desirable
results, depending upon how the liability rules are operationalized by the relevant
decision-makers.75 It is important that a negligence rule set the right legal standard:
adopting either an overly lax or stringent standard would distort actors’ investment
decisions, leading to too few or too many precautions. Under a strict liability regime,
it is crucial to carefully calibrate the sanction to the harm caused by the activity.76
If the assessment of harm is too lax or too stringent, actors would be incentivized
to underinvest or overinvest in precautions.77
Legislators around the world, often under the rhetoric of the so-called “polluter
pays” principle,78 have increasingly gravitated toward strict liability as the main tort
regime in this context.79 A major rationale for adopting strict liability concerns infor-
mation asymmetries. For a negligence rule to function properly, the court weighs the
marginal social costs and benefts of various courses of action and then adopts the
optimal behavior standard. To perform this task, courts must consider (and monitor)
all the actions tortfeasors can take. If, however, certain aspects of the tortfeasor’s
74 See, e.g., James Hansen & Makiko Sato, Greenhouse Gas Growth Rates, 101 Proc. Nat. Acad. Sci.
U.S.A. 16109, 16109 (2004); Andrew K. Jorgenson, Global Warming and the Neglected Greenhouse Gas: A
Cross-National Study of the Social Causes of Methane Emissions Intensity, 1995, 84 Social Forces 1779,
1981–1982; Max Blumer & Jeremy Sass, Oil Pollution: Persistence and Degradation of Spilled Fuel Oil, 176
Sci. 1121, 1122 (1972).
75 Shavell, supra note 3 at 121. Faure & Skogh, supra note 3, at 247–248.
76 Shavell, supra note 3, at 9, 128.
77 See supra Part I.C.
78 Hans Chr Bugge, The Principles of “Polluter-Pays” in Economics and Law, in Law and Economics
of the Environment 53 (Erling Eide & Roger van den Bergh eds. 1996); Hans Chr Bugge, The Polluter Pays
Principle: Dilemmas of Justice in National and International Contexts, in Environmental Law and Justice
in Context 411 (Jonas Ebbesson & Phoebe Okowa eds. 2008); Faure & Skogh, supra note 3, at 26–30.
79 See, e.g., Ursula Kettlewell, The Answer to Global Pollution – A Critical Examination of the Problems
and Potential of the Polluter-Pays Principle,3 Colo. J. Int’l Envtl. L. & Pol’y 429, 438–441 (1992).
82
behavior cannot be monitored under a negligence rule, strict liability will tend to
be superior. The reason is that strict liability dictates the behavior, which allows the
actor to adjust any aspect of their controlled behavior, or about which they may
have superior information.80 Another problem with negligence as the sole basis for
imposing liability (as opposed to a hybrid system of negligence with partial strict
liability regime, as is in the case of liability for ultra-hazardous activities) is that neg-
ligence determinations may in practice be reduced to enforcement of settled industry
practices81 or regulatory standards.82 These sources may not necessarily represent the
social ideal but be adopted by courts nonetheless, owing to institutional constraints
and incomplete information.
Unfortunately, although many national legal systems are improving their strict
liability metrics concerning environmental harm, they are still far of the theoreti-
cal mark. This is because the legal system fails in defning compensable harm, both
conceptually and institutionally.83 Moreover, insolvency84 and damage caps85 further
complicate the task of “pricing” environmental risk through the tort system. More
generally, there are so many loopholes and imperfections in strict liability’s current
design that there is little reason to believe the tort system could fully internalize
environmental risks.86 In practice, therefore, neither negligence nor strict liability
provides optimal protection to environmental interests.
80 To elaborate, one can distinguish between two forms of precautions: level of care and level of activity.
An example, in the context of noise from trafc, is the distinction between protective measures such as thicker
walls and double paned windows (care) and the lowering of the amount of trafc (activity). Another example
in the context of wind energy is the distinction between the improvement of windmill technology (care) and
the number of hours of operation (activity). Finally, in the context of toxic releases from ships in fjords,
bays and harbors, the distinction lies between the choice of chemicals and ship paint (care) and the number
of trips (activity). In these examples, inclusion of the activity level in the determination of negligence may
be hampered for two reasons. First, such inclusion requires precise information about the use that the actor
obtains from carrying out the activity. Yet, the court may lack such information. Secondly, such negligence-
balancing requires the fnding of a certain level of activity as a matter of fact, and that factual determination
may be hard to make in practice. For these reasons, strict liability enjoys an advantage over negligence in
certain pollution contexts. For more refned analysis adding more, see esp. Steven Shavell, Strict Liability
Versus Negligence, 9 J. Legal Stud. 1 (1980).
81 Gideon Parchomovsky & Alex Stein, Distortionary Efect of Evidence on Primary Behavior, 124 Harv.
L. Rev. 518, 545–546 (2010).
82 Endre Stavang, The European Court of Justice and the Environmental Liability Directive, 5 Environ-
mental Liability 198 (2010).
83 Donald N. Dewees, The Role of Tort Law in Controlling Environmental Pollution, 28 Can. Pub. Pol’y
425, 430–431 (1992); Claus Ott & Hans-Bernd Schäfer, Widening the Scope of Environmental Liability, in
Law and Economics of the Environment 91, 99–100 (Erling Eide & Roger van den Bergh eds. 1996); Faure
& Skogh, supra note 3, at 284. On the new and so-called Ecosystem Services Approach and the worry of its
lacking legal implementation in the Macondo context, see Carrie Presnall, Laura López-Hofman & Marc
L. Miller, Can the Deepwater Horizon Trust Take Account of Ecosystem Services and Fund Restoration of the
Gulf? 40 ELR 11129 (2010).
84 Faure & Skogh, supra note 3, at 267–270, 276–277.
85 Michael G. Faure & Hui Wang,Financial Caps for Oil Pollution Damage: A Historical Mistake?32
Marine Pol’y 592 (2008).
86 See, e.g., Debra L. Baker, Bankruptcy – The Last Environmental Loophole, 34 S. Tex. L. Rev. 379, 387,
397 (1993); Catherine Tinker, Strict Liability of States for Environmental Harm: An Emerging Principle of
International Law, 3 Touro J. Transnat’l L. 155, 160 (1992); Palma J. Strand, The Inapplicability of Tra-
ditional Tort Analysis to Environmental Risks: The Example of Toxic Waste Pollution Victim Compensation,
35 Stan. L. Rev. 575, 590 (1983).
83
Enter knock-for-knock. The knock-for-knock system does not take full account of
environmental interests. Hence to some degree it widens the gap between the optimal
level of protection and that taken in practice. This implies that knock-for-knock’s
introduction marginally worsens the fate of environmental interests. However, con-
sidering the risk interdependencies introduced in section 3.2 and discussed in section
4.1, and the fact that they may create massive environmental harms that are not
always measurable and may not even be readily observed, there is reason to believe
that the knock-for-knock system presents a real threat to environmental interests.
At this point, one may invoke the precautionary principle to call for a ban on
knock-for-knock clauses. The precautionary principle maintains that if an activity
threatens environmental harm, measures should be adopted to control it even if the
precise causal relationship has yet been established.87 To rephrase, the precautionary
principle embodies the idea that regulators should err on the side of safety. Although
we are sympathetic to the precautionary principle, we do not think that its applica-
tion in this case necessitates an outright ban on knock-for-knock clauses. Properly
understood, the precautionary principle requires the imposition of regulation on
the operations of the gas and oil industry. Of course, such regulation already exists.
The precautionary principle does not prescribe the precise nature, or even the scope,
of the regulatory measures that need to be adopted. Given the highly incomplete
data on the relationship between knock-for-knock clauses and environmental harm,
we feel that currently there is not enough evidence to legally ban knock-for-knock
arrangements. But given the genuine uncertainty surrounding the social desirability
of knock-for-knock clauses, it is not surprising that in some jurisdictions standard
rules of contract enforcement are not fully applicable to such clauses.
5.4 Deterrence versus compensation

One might argue that the erosion in deterrence that results from knock-for-knock
does not pose a serious concern as long as injured parties are compensated after the
fact. Extant theorizing suggests that compensation serves a dual purpose: it restores
victims to their pre-accident state, and it deters tortfeasors from deviating from socially
accepted standards of behavior.88 Corrective justice scholars emphasize the former
purpose,89 while law and economics scholars put the premium on the latter.90
However, ex post compensation cannot always be relied on to afect the socially
desirable level of deterrence. Paradoxically, after-the-fact compensation is especially
inefective with large-scale catastrophes. In such cases, the aggregate harm typically
87 Stephanie Joan Mead, The Precautionary Principle: A Discussion of the Principle’s Meaning and Status
in an Attempt to Further Defne and Understand the Principle, 8 N.Z. J. Envtl. L. 137, 138 (2004).
88 See Michael G. Faure, Calabresi and Behavioural Tort Law and Economics, 1 Erasmus L. Rev. 75, 94
(2008) (“When [as a result of liability] an enterprise is held to compensate the costs its activity generates,
dangerous activities will become more expensive and the enterprise will, as a result of market forces, have
an incentive to increase safety.”).
89 See Mark R. Reiff, Punishment, Compensation, and Law: A Theory of Enforceability 170 (2005)
(“Most corrective justice theorists . . . contend that a person who wrongfully injures another has a moral
obligation to compensate the injured party for his loss.”).
90 See Richard Posner, Frontiers of Legal Theory 266 (2001) (arguing that deterrence is more
important than compensation).
84
exceeds the fnancial resources of the responsible parties. And although criminal law
is supposed to address the deterrence defcit, it often falls short of achieving this goal
when corporate entities are involved.91 Obviously, criminal prosecution, even when
successful, ofers little solace to tort victims.
The key point for the purpose of this discussion is that after-the-fact compensa-
tion cannot be counted on to prompt members of the gas and oil industry to adopt
the socially optimal level of precautions. In this context, the emphasis should be
put on preventing accidents ex ante. Ex ante prevention is typically the domain of
regulation.92 But given the inherent difculty in enforcing regulatory standards and
monitoring compliance and because relatively minor failures may evolve into large-
scale accident, we believe that it would be a mistake from a societal perspective to
allow the members of the gas and oil industry to contract into a liability regime
whose efect is to cause a drop in the level of care.
6 Conclusion
Can private ordering successfully replace standard tort liability in an industrial setting?
In this chapter, we looked at this question by discussing the knock-for-knock regime
that formally only governs the inter se relationships of industry participants in the
gas and oil sector. Although industry insiders are convinced that the knock-for-knock
rule is optimal for the gas and oil industry and that there is no real social downside
to its adoption,93 our analysis provides a more cautionary note. The knock-for-knock
regime lowers operation costs for gas and oil companies, but only at the cost of creat-
ing a serious moral hazard problem. From a societal perspective, the knock-for-knock
rule is efcient only if the private cost savings it generates are greater than the increase
in the rate and severity of accidents that result from its adoption. Our analysis sug-
gests that the moral hazard problem to which the knock-for-knock rule gives rise can
be efectively addressed only by far-reaching and wide-ranging regulation. In this
regard, we think it is prudent to consider that both the informational asymmetries
and the tendency to exploit it may vary greatly depending on the ownership structure
of the operating companies. For example, in Norway, 83% of the largest oil company
(Statoil) is owned by the Kingdom of Norway. This should be expected to infuence
safety when compared to felds like Macondo in the Gulf of Mexico or felds in the
developing world, where there is no similar government involvement.
More generally, although the operations of gas and oil companies are regulated
everywhere in the world, the scope and nature of the regulation varies dramatically
from one country to another. The extent of government ownership of gas and oil
companies can also afect the risk to which third parties are exposed. Where the
government is the controlling shareholder of gas and oil companies, one can fnd,
91 Assaf Hamdani & Alon Klement, Corporate Crime and Deterrence, 61 Stan. L. Rev. 271 (2008)
(challenging the conventional view concerning the deterrence value of corporate criminal liability and “dem-
onstrating that harsh entity-level penalties might discourage monitoring for misconduct and undermine
compliance incentives within professional frms.”).
92 See Yael Aridor Bar-Ilan, Justice: When Do We Decide, 39 Conn. L. Rev. 923, 932–938 (2007) (sum-
marizing the role of ex ante regulation in the law).
93 Bull, supra note 2.
85
on average, stricter regulatory standards and a higher degree of industry compliance.
It should be noted, though, that it is difcult to fnd gas and oil sites where the pre-
vious conditions fully obtain. On the other hand, it is easy to identify gas and oil
provinces in diferent parts of the globe where regulatory oversight is lacking, and
government ownership of the venture is low to non-existent. At the end of the day,
then, one can see the glass as half full or half empty.
One fnal note is in order. Given the great variation in institutional quality and
regulatory cultures around the world, a uniform approach to knock-for-knock clauses
may be misguided. We conjecture that knock-for-knock clauses produce the greatest
savings for gas and oil companies in developed countries, such as the United States,
Canada and Europe, where litigation costs are very high. These countries have
frst-rate regulatory institutions and adequate regulatory capabilities. Consequently,
the risk posed by knock-for-knock clauses is relatively moderate. By contrast, the
regulatory infrastructure in developing countries is often inadequate, and the use of
knock-for-knock clauses poses a much greater risk of harm to the environment and
third parties. This risk is compounded by the fact that the legal systems of many
developing countries do not recognize class actions or a comparable mechanism for
aggregating small claims. Correlatively, the cost savings enjoyed by gas and oil com-
panies as a result from knock-for-knock clauses are more modest. If we are right, a
tentative conclusion emerges: knock-for-knock clauses can be tolerated in developed
countries where they produce the highest private gains while arguably posing a man-
ageable public risk, and they should be banned in developing countries where, from
our analysis, they yield modest private gains while posing a signifcant risk of harm.
Acknowledgements
We are grateful to Ian Ayres, Tom Baker, Shyam Balganesh, Derek and Jane Bam-
bauer, Abraham Bell, Miriam Bitton, Ellen Bublick, Hans Christian Bugge, Fabrizio
Cafaggi, Guido Calabresi, Alon Cohen, Simon Deakin, Erling Eide, John Goldberg,
Sharon Hannes, Olav Hasaas, Keith Hylton, Jacob Nussim, Ariel Porat, Christopher
Robertson, Carol Rose, Erik Røsaeg, Peter Siegelman, Henry E. Smith, Stephen
Smith, Roy Spece, Alex Stein, and participants at the law and economics workshops
at Bar Ilan and Tel Aviv faculties of law and the frst annual meeting of the Private
Law Consortium for invaluable comments and contributions. For outstanding research
assistance, we would like to thank Levi Morris and Ananth Padmanabhan.
86
CHAPTER 4
On knock-for-knock clauses and

their optimal regulation
Henrik Lando
1 Introduction
The knock-for-knock system is a contractual system in which the parties to the system
agree to cover all their losses through frst-party insurance and to dispense with tort
liability. The system ensures that each party’s insurance contracts cover losses to the
party’s own assets and own employees, and that the frst-party insurers cannot seek
redress under tort law against the other parties. Moreover, while the system cannot
dispense with tort claims raised by third parties, the system can allocate such risks
to the better insured party (e.g., the operator of an oil rig rather than the contrac-
tors) through indemnity clauses.
Liability exclusions are not at all uncommon in the business world but the so-
called unqualifed knock-for-knock contracts exclude liability and indemnifes also
for grossly negligent acts or even sometimes for losses caused by willful misconduct.1
Such exclusions are thought by some academics and judges to be unjust and/or to
undermine incentives for precaution. This concern has been heightened after the
tragic Macondo and Alpha Piper accidents. Whether for fairness reasons or out of
a concern for precaution, courts have sometimes struck down unqualifed knock-for-
knock clauses. For instance, in a case where a vessel through gross negligence collided
with an oil-platform, a Norwegian court set aside the exclusion of liability for gross
negligence, arguing that “the lack of precaution was blameworthy and grave . . . and
the lack of precaution could have had far worse consequences”.
The same concern for proper precaution led to Parchomovsky and Stavang’s study
in Chapter 3. Their study adopted a skeptical attitude towards unqualifed knock-
for-knock contracts, especially in countries where it cannot be assumed that public
regulation substitutes for tort law in securing a proper incentive for precaution.
This chapter addresses two topics in relation to knock-for-knock contracts. The
frst topic is the nature and purpose of knock-for-knock clauses, and why we observe
them in some sectors of the economy but not in others. Why, for instance, do we
observe knock-for-knock contracts mainly in the ofshore rather than in the onshore
oil and gas sector? I will argue that the answers in the literature are not satisfactory,
although they contain elements for an explanation. Instead, I suggest a simple model
can be used to think about when liability exclusions are optimal.
1 Gross negligence is more often applied than willful misconduct and we shall focus on gross negligence.
DOI: 10.4324/9781003206798-4 87
henrik lando
The other topic was raised by Parchomovsky and Stavang: when should courts
uphold unqualifed knock-for-knock contracts? I broaden this question by inquir-
ing not only what courts should do but also what lawmakers and public regulators
should do to implement an optimal regulatory system. My point will be that civil
fnes are meant to exactly counter negative efects on third parties. Instead of strik-
ing down unqualifed knock-for knock-clauses when an injurer has acted with gross
negligence, I suggest raising fnes for such gross negligence. Fines do not jeopardize
the signifcant savings of dispute costs that are associated with certainty of enforce-
ment of knock-for-knock clauses.
First, however, it is worth positioning the discussion in the context of the literature
on exclusions of liability, part of which addresses knock-for-knock contracts.
2 On the literature
Knock-for-knock contracts are essentially mutual liability exclusions, and hence
related to to the literatures on tort law and on liability exemptions and exclusions.
Several articles address whether tort law’s deterrent efect warrants its administra-
tive costs. Part of that literature is empirical and part is theoretical. An example
of the empirical literature is from Sugarman. As for theoretical literature, one con-
tribution is that of Polinsky and Shavell who analyze when the deterrence benefts
of product liability claims outweigh their dispute costs. They fnd that for certain
categories of goods that are produced for mass markets and that are monitored for
their quality by consumer organizations and others, the case for liability is at best
tenuous. Dispute costs are high, and the efect of liability on deterrence appears
small in a market where quality is constantly monitored and where reputation works
as a strong incentive.
When liability exclusions are efcient is analyzed, e.g., in Polinsky’s textbook. A
major point is that when the prospect of being found liable diminishes the risk of an
accident, dispute costs are also hereby diminished, as the dispute costs are incurred
only when there is an accident.
However, the law and economics piece most related to this chapter is from Parcho-
movsky and Stavang in Chapter 3 of this book. They maintain the knock-for-knock
system is likely to diminish precautions, and while this may be an acceptable conse-
quence to the parties involved in the knock-for-knock system, since it is outweighed
by their administrative savings, such may not be the case for third parties who may
not be fully compensated in case of an accident. While third parties (that are not
parties to the knock-for-knock contracts) can claim compensation for economic loss
and for physical injury, a higher likelihood of accidents will harm them to the extent
that some of their losses are not covered (or not documentable) under tort law. On
this basis, Parchomovsky and Stavang advocate a skeptical attitude towards knock-
for-knock clauses, especially in countries where public regulation cannot be trusted
to maintain deterrence.
In the legal literature, Bull and also Wilhelmsen explain and discuss the advantages
and disadvantages of the knock-for-knock system. Bull is the standard legal refer-
ence, while Wilhelmsen includes a broader economic perspective by taking account
of the role of deterrence.
88
on knock-for-knock clauses and their optimal regulation
Lando and Mortensen discuss when knock-for-knock clauses will be enforced by
Danish courts, and when they should be from an efciency perspective. The present
chapter goes further by providing a simple model for when knock-for-knock clauses
are used and by considering knock-for-knock clauses in a broader perspective of
optimal regulation. Finally, Mielcarek discusses the advantages and disadvantages
of gross negligence and willful misconduct liability exclusions and stresses the role
of dispute and insurance costs. Mielcarek also makes the claim that liability exclu-
sions, by excluding disputes over claims, are likely to enhance a spirit of cooperation
among the parties.
We now turn to our frst issue, which considers the basic question that when par-
ties agree to knock-for-knock contracts, when are they useful?
3 When are knock-for-knock clauses useful?

Although exclusions of liability occur in contracts of diverse kinds and in many
diferent parts of the economy, we observe the knock-for-knock contracts pre-
dominantly but not exclusively2 in oil and gas extraction or in the construction
of wind turbines. Moreover, as a stylized fact, we observe knock-for-knock
arrangements occur more often when these activities are carried out ofshore than
onshore.
To explain these patterns, the following factors are often evoked:
• The liability system involves double insurance, both frst-party insurance and
third-party (liability) insurance. The knock-for-knock system dispenses with
double coverage.3
• Liability involves administrative cost; conficts are expensive in terms of time
and money. They may also sour relationships.
• Public regulation is intense in the oil and gas industry and can ensure that
the parties exercise efort to lower risks of accidents.
• Liability insurance contracts are costly to draft, especially for small
contractors.
• The parties’ concern for reputation maintains their incentive for precaution.
However, these explanations apply to many sectors of the economy. The liability
system usually involves the risk of double insurance and always includes administra-
tive costs. Public regulation is widely present in sectors that do not rely on knock-
for-knock clauses; liability insurance is often costly to draw up, and incentives created
through fear of loss of reputation are ubiquitous in the economy. Also, the explana-
tions seem applicable both onshore and ofshore. We need a model of the use of
liability exclusions that can account for when exactly they are advantageous.
2 The exact extent to which mutual liability exclusions for gross negligence are employed in the economy
I do not know, but it should be noted that similar arrangements can arise as when contractors and owners
together draw up an insurance that covers all damage and losses that may occur at a worksite, as in a Con-
tractors’ All Risks (CAR) Insurance.
3 Note that this argument can be criticized: due to competition between the insurers, one would expect
insurance premia in the long run to refect the actual payouts made by the insurers to the parties.
89
henrik lando
3.1 A model of the use of knock-for-knock contracts
In economic terms, liability rules fnd their rationale in their deterrent efect. If the
parties can create frst-party insurance, this is a better way for the parties of allocat-
ing risk when there is no deterrent efect, as frst-party insurance dispenses with the
costs that arise when a frst-party insurer raises claims against the injurer’s liability
insurer. Hence if rational contracting parties agree on a liability rule it must be
because they believe in its deterrent efect. Alternatively, the parties could be con-
cerned with fairness; however, if losses are placed primarily on insurance companies,
then the role of fairness is reduced. On this basis, one might conclude that the
deterrent efect is small in sectors that employ knock-for-knock contracts; however,
this is not a satisfactory explanation because it fails to explain why the deterrent
efect would be smaller in some sectors than in others, or why it would be smaller
ofshore than onshore. Both ofshore and onshore activities are subject to public
safety regulation.
Based on the above logic, we can conclude that since liability is applied in many
sectors of the economy, it must be the case that liability (often) does increase incentives
for precaution. Most likely, this is due to the activities of insurers who may require
certain precautions, graduate premia over time, or set deductibles that encourage the
injurer to be cautious.
Let us therefore assume, in a simplifed model of knock-for-knock clauses where
only one party acts, that the probability of an accident caused by that party under
a negligence rule, p, is greater than the probability when the party cannot be
held liable for any kind of negligence (as under an unqualifed knock-for-knock
system), r. Let us also assume there are costs of taking greater precautions, such
as when employees must undergo training or procedures for precaution must be
put in place, and so forth. Denote this cost by c, and let L be the monetary cost
of an accident.
Furthermore, we assume that if an accident occurs, the probability of a con-
flict is q and the average cost of resolving a potential conflict over who was
at fault is K. Often, part of this cost is borne by the insurance companies on
either side, but eventually this cost will be carried by the parties through the
insurance premium.
Based on this notation, we can derive when liability is better than exclusion from
liability if we assume that the parties wish to minimize the sum of their expected
costs. Note that rational parties will wish to do so, since they can share the saved
costs between themselves when they agree on the price of the contract.
If there is a rule of liability that leads the injurer to take precautions, total expected
costs for the two parties combined amount to:
c + p(L + qK )
If there is a knock-for-knock exclusion from liability, the expected total costs for the
parties’ amount to rL. Hence, exclusion of liability is preferable when:
c + p(L + qK ) > rL (1)
90
This equation brings out the essential factors: how efective are the extra precautions
taken under liability (p − r)L compared with their costs c, and compared with the
expected confict costs incurred under a liability rule: pqK?
This simple calculation casts light on why the knock-for-knock system is more
often employed ofshore than onshore. An essential diference between ofshore and
onshore activities is their inherent dangerousness. Let us catch this aspect by letting
the probabilities be systematically smaller onshore, that is p and r are both equally
reduced by some factor ∂. Here ∂ expresses the general operational dangerousness
of the operations, and so subtracting ∂ represents onshore activity. If the condition
(1) above is less likely to be fulflled when probabilities are reduced by ∂, we have a
simple explanation of why ofshore activities more often than onshore activities use
exclusions of liability. When probabilities are so reduced, we obtain the condition:
c + ( p − ∂)(L + qK) > (r − ∂)L
This can be rewritten as:
c + p(L + qK) − ∂qK > rL (2)
This equation is less often fulflled than (1), since ∂qK must be subtracted from the
left-hand side. Hence, exclusion is more rarely optimal when the activity is less inher-
ently dangerous, as it is onshore.
The logic of this result can be explained as follows. If the activity is dangerous in
the sense that accidents occur with a signifcant probability even when the parties take
care, and if there is a signifcant risk that there will be litigation even though care was
actually taken (i.e., qK is non-trivial), there will be signifcant costs of confict when
liability is not excluded. The high costs of confict may render it optimal to dispense
with liability. Hence, the inherent dangerousness of the activity, as measured by the
probability of accidents when the parties take proper care, may help to explain the
pattern of use of knock-for-knock contracts.
The model can be interpreted in two ways. We can either interpret the occurrence of
accidents despite due care as involving exogenous events, and the dispute between the
parties can then be about whether these events or some alleged negligence on the part of
the injurer caused the accident. Or we can think of accidents occurring due to lapses, e.g.,
momentary inattention on the part of management or employees, or perhaps inadvertent
miscommunications in large organizations. Such mistakes can occur even when parties have
shown a proper precautionary attitude, and when they do occur, they can cause accidents
and disputes about whether the lapse occurred and whether it amounted to negligence.
Another factor that, beside the inherent dangerousness of the activity, can give rise
to the knock-for-knock system is uncertainty about liability. This is the role of qK; the
higher this number, the more attractive the knock-for-knock system. One fact that springs
to mind here is that when many workers coordinate their acts on a site, doubt can easily
arise about who of many potentially culpable actors was (most) to blame for the accident.
This can also help explain the greater use of knock-for-knock clauses ofshore as work-
ers are forced into smaller spaces. Both Bull and Wilhelmsen stress this latter factor, the
difculty of ascertaining fault, as important in explaining the knock-for-knock system.
91
henrik lando
Moreover, the inherent dangerousness and the difculty of apportioning liability can
interact to create a rationale for knock-for-knock clauses: When qK is high, the inher-
ent dangerousness of the activity becomes more important, as is clear from equation
(2). Then, even if difculties of ascertaining who was negligent are the same ofshore,
they can still account for the knock-for-knock system, since the difculties arise more
often ofshore due to the inherent dangerousness of ofshore activities.
A caveat to the explanation ofered in terms of inherent dangerousness is worth
noting. It seems plausible that care may be more important when an activity is inher-
ently dangerous. If so, one might not expect the knock-for-knock system to be used
more often ofshore. The explanation based on inherent dangerousness rests on the
incentive efect being limited and not much greater when the activity is inherently
dangerous. The theory must therefore be reformulated: the knock-for-knock system is
likely to arise under inherently dangerous circumstances when attribution of fault is
difcult, and when incentives for care are established to a signifcant extent through
other instruments than tort liability.
These explanations of the knock-for-knock system fnd support in the description of
ofshore projects that employ knock-for-knock contracts. For instance, Fjaervoll writes:
At an ofshore project, the parties involved in the operations work in an environment
where there is high risk of causing damages, because high property values are situated
in a limited area where many contractors and sub-contractors work at the same time.
Considering the nature and size of ofshore projects, the threat that damages will occur
is present at all times and the parties involved are therefore exposed to substantial risks
for damage, not only caused due to their own actions but also as a result of actions by
others operating at the same project.
It is also consistent with this explanation that the knock-for-knock became known
during the World War II when the allied countries, to defend against submarine
attacks, moved their ships in convoys in which the ships could not avoid colliding.
The countries agreed to each pay for their own damage, regardless of fault. In these
convoys there would seem to be a high risk of collision even when all actors were
careful, and it seems likely it was difcult to determine whether a ship’s collision was
due to a negligent act rather than the infuence of the sea and weather but also whose
negligence was involved when both parties acted to prevent a collision.
We now address the question of whether courts should respect the exclusion of
liability for gross negligence.
4 Should courts impose carve-outs for gross negligence?

The following discussion will be based on a set of premises that are worth establish-
ing at the outset.
First, we will assume a Scandinavian context and not, for instance, the context of
developing countries where public regulation and tort law may be much less developed.
Second, we shall take for granted that if there are no externalities involved such that
all parties external to the contract are fully compensated for the losses arising as a result
of accidents that occur within the knock-for-knock contracts, there is no reason for
courts to set aside the parties’ agreements. This premise is based on the idea that if the
parties to the knock-for-knock system have agreed to a knock-for-knock contract and no
92
third party is afected, their choice of contract is Pareto-optimal and all parties then
stand to beneft in an ex ante sense from court enforcement.
This reasoning takes for granted that the parties do not wish courts to interfere in
what may be an incomplete contract. If, for instance, the behavior of a party leading
to an accident is reckless, the parties would perhaps have carved out such behavior
in their contract if they had had the resources to anticipate it and to make reference
to it in the contract. Just like courts generally interpret contracts to exclude meanings
that the parties cannot have intended, certain types of gross negligence may perhaps
justifably be struck down. However, we shall not consider this aspect but assume that
the behavior for which the parties would want courts to apply liability is so extreme
and occurs so rarely that we can omit it from analysis. Based on this assumption
we can maintain that courts should, if at all, only strike down carefully drafted and
mutually agreed knock for knock contracts when third parties are adversely afected.
More precisely, we shall assume that courts can be justifed in striking down knock-
for-knock clauses only if the clauses mute incentives for precaution and if third parties
are afected by the lack of precaution because they are insufciently compensated in
the event of an accident. There may be extreme instances of misconduct for which
the court should not uphold the liability exclusion, but such instances will not be
specifcally addressed here.
Third, we shall make the point that while employees are generally compensated
under worker’s compensation schemes (which are themselves in some countries essen-
tially knock-for-knock systems in that they do not allow insurers to seek redress),
such schemes undercompensate physical injury or death by not taking into account
the full loss for the injured person or for his or her relatives and friends.4 Therefore,
employees (and their families and friends) hold an interest in frms being incentiv-
ized to take greater care.
Fourth, as far as personal injury to employees is concerned, the knock-for-knock
system limits the incentive for care, although to a limited extent. Most economic
consequences of such harm are covered by a worker’s compensation system. Thus,
whether parties enter a knock-for-knock contract, most of the economic losses to
employees will be covered by frst-party insurance. However, tort law supplements
worker’s compensation, since not all categories of losses are covered by the worker’s
compensation system, and this supplementary tort law will be afected by the knock-
for-knock system and by whether it is qualifed or not. Therefore, a court may strike
down the indemnifcation of tort claims relating to physical harm to employees and
thereby afect incentives for precaution.
Fifth, accidents cause harm to third parties who are far from always fully com-
pensated through tort law. For instance, in large oil-spills fsheries, tourist industries
and others may see their livelihood afected. In general, they will not all be eligible
for compensation under tort law. Losses may, for instance, be spread thin and each
afected victim may not have the incentive to sue for compensation or be able to
4 Non-economic harm is part of the worker’s compensation system, but the compensations are low
compared with the utility consequences sufered when there is severe physical harm.
93
henrik lando
organize a class action suit. Or the doctrine of pure economic loss can bar claims
from people who have sufered neither personal injury nor loss of physical assets.
Sixth, we imagine a setting where O examines C’s procedures before allowing C to
take acts that may jeopardize O’s valuable assets X. Practitioners often mention that
such vetting can be extensive, and that there can therefore be an inducement for C
to gain a reputation for maintaining safe procedures. It is worth keeping this real-
ity in mind, as already alluded to above, as one might otherwise get an exaggerated
impression of the role of marginal incentives created by liability.
Under these premises we now consider a framework for understanding the extent
to which carve-outs of knock-for-knock contracts afect incentives and whether this
efect warrants court intervention.
4.1 A framework
Consider two parties, O, the operator, and C the contractor. In oil extraction, O is the
oil company (owning the concession), and C is one of often many contractors hired
to dig oil from a platform owned by O. Both parties put assets and employees at risk.
The assets of O are denoted X, whereas the assets of C are denoted Y. Typically, X is
of much greater value than Y. Third parties and employees can sufer losses that in
economic terms sum to E. Part of these, γE, are compensated under tort law (such as
the harm to privately owned land) while (1 − γE) is not compensated (0 < γ < 1).
We imagine that both O and C make economic decisions that afect the safety of
the project and that both parties may lapse. Lapses are considered exogenous mis-
takes, as when employees act without paying attention or thinking rationally about
consequences. In mathematical terms, one can think in terms of actually exercised
care z being afected by the degree of precaution and a stochastic lapse σ : z = x − σ,
where σ is a stochastic variable. This means that gross negligence may occur when a
party did not intend for it to occur, but perhaps did not do enough to encapsulate
the efects of lapses. This way of considering accidents has implications for how one
analyzes incentive efects, as will become clear.
The two parties will be insured by IC and IO, respectively. These insurance companies
provide frst-party insurance and, if relevant, liability insurance to their respective
party. We assume that the insurer, whether as frst-party or as liability insurer, can
afect a party’s negligence through requiring certain precautions or through deductibles
or adjustments of premia over time in response to accident history. However, these
instruments will be assumed to create less than frst-best incentives for precaution.
Furthermore, in part due to the incompleteness of tort compensation, we assume
that a public regulator sets fnes for unlawful behavior. Often, grossly negligent acts
can be fned under rules of public regulation; the size of such fnes is a policy instru-
ment under the discretion of Parliament and to some extent also under the discretion
of courts. Notably, in Danish law,5 neither civil nor criminal fnes can be indemnifed
under knock-for-knock contracts.
5 In the United States, civil fnes (e.g., for breach of the Clean Water Act) are multiplied by some factor
when the act has been grossly negligent.
94
One can analyze the regulation of knock-for-knock clauses either from the perspec-
tive of a court, which must take other instruments (e.g., public fnes) as given, or from
the perspective of the overall optimal regime of regulation. I shall adopt the latter
perspective and argue that the rationale for imposing a carve-out is then quite weak.
Let us frst consider the arguments for imposing a carve-out, i.e., for not allowing
the liability exclusion, and then the arguments against.
The main argument in favor of a carve-out springs from the fact that the employees
of O and C and third parties sufer losses that are not compensated under tort law.
As only γE of their losses is compensated, they are interested in a higher level of care
from both O and C. This is an externality that parties to the knock-for-knock clauses
do not take into account when choosing between the qualifed and unqualifed knock-
for-knock system, and they are therefore likely to choose a system that induces too little
precaution. The reason the level of precaution is smaller under the unqualifed knock-
for-knock system is that when there is liability for gross negligence, it is expected that
the parties will take out liability insurance (in order to lower the risk of bankruptcy) and
the liability insurers will take steps to ensure a greater level of care. Thus, three factors
will increase the level of care by the parties when they are liable for gross negligence:
1 The insurers will require deductibles (or increase the premium) in the event
of gross negligence. The frst-party insurers already set deductibles under the
unqualifed knock-for-knock system, but there will be an additional efect
of the deductibles on the liability insurance, especially for C.
2 The liability insurer will require specifc safety measures as a condition for
coverage; these requirements may be more expansive than those required by
the frst-party insurers under the qualifed knock-for-knock system.
3 The parties will anticipate added dispute costs when there will be disagreement
after an accident about whether the injurer’s level of care constitutes gross
negligence. These costs can be rendered less likely by taking greater care.
The argument for carving out gross negligence is that these three factors will beneft
third parties and employees to the extent that they are undercompensated under tort
law and the worker’s compensation system.
Note that these incentive efects will be more important the lower the level of
care taken under the qualifed knock-for-knock system. If incentives fall short by
a wide margin, a marginal increase will be likely to be more important (according
to the principle of decreasing marginal return). The extent to which incentives for
precaution fall short under the qualifed system is, however, not clear. On the one
hand, there are three reasons for expecting incentives to be insufcient as compared
with the frst best under the qualifed system. First, the compensated losses under
tort law are not the full losses sufered. Only part of E is compensated. Second,
simple negligence does not imply liability. As simple negligence is far more common
than gross negligence, this is a factor that can seriously lower the incentive for IC
to induce precaution through the liability insurance contract. Third, O and C are
only incentivized by the deductible in their liability insurances, not by the full loss.
On the other hand, public regulation, including the imposition of fnes for breach
of safety regulation, and C’s and O’s concerns for their reputation for safety, are
95
henrik lando
likely to increase their levels of care. It is therefore difcult to say whether the
marginal increase in precaution due to the three factors mentioned above is of
real importance.
Against these arguments for carving out gross negligence, two main arguments
can be put forward.
The frst concerns the added dispute cost. As mentioned in the frst part of this
chapter, in sectors where the risk of accidents is large, and where lapses are likely
to lead to accidents, there will be a considerable number of instances in which it
can be claimed that there was negligence, and in some of these cases, especially
when a signifcant accident has occurred, there may be a considerable temptation
for the frst-party insurer of the victim to claim that the injurer acted with gross
negligence. Hence, dispute costs may well be much larger in case of a carve-out,
especially because of hindsight bias: when the accident is severe, simple negligence
may appear to be gross.
The second concerns the need for C to take out a large liability insurance. The
drafting of a large liability insurance is said by practitioners to involve signifcant
costs, especially when there is considerable asymmetric information about C’s type.
Based on these arguments, it can be difcult to tell whether the benefts of carving
out gross negligence outweigh the costs. However, the following argument indicates,
I believe, that a carve-out is not part of an optimal regulatory system.
Most of the benefts and the costs of a carve-out are taken into consideration by
the parties themselves when they decide whether to carve out or to exclude liability
for gross negligence. The parties themselves must believe that the incentive efects of
liability for gross negligence are smaller than the costs of a carve-out. As mentioned,
the reason for imposing a carve-out on the parties is that third parties and employees
hold an interest in a higher level of care than that which is optimal from the point of
view of the parties, since the third parties and the employees are undercompensated
in case of an accident involving personal injury or environmental harm. However,
this interest in greater incentives for care can be accomplished by setting fnes for
breach of regulations; such breach is typically involved when the parties act with gross
negligence. Fines increase deterrence without jeopardizing the savings in dispute costs
and drafting costs that unqualifed knock-for-knock clauses accomplish. The point is
that fnes serve the role of making up for the fact that not all losses are compensated
through tort law. They serve to increase the parties’ incentives where third parties (or
employees) are less than fully compensated under tort law or workers’ compensation
system. The administrative costs of higher civil fnes are unlikely to be of the same
order of magnitude as the extra dispute costs that may arise when parties hope for
courts to strike down the knock-for-knock clause.
One more argument in favor of respecting the unqualifed knock-for-knock sys-
tem is that it should not be taken for granted that the legal system is an efcient
mechanism for resolving complex disputes, such as those arising in ofshore oil
extraction. Courts may not be able to accurately discern a cost-efective manner
what happened in an accident involving several actors, and who were ultimately
to blame. Instead, the parties may rationally decide to rely on reputational mecha-
nisms in which privately held information that is difcult to convey to a court
can play a part.
96
5 Conclusion
I have argued that knock-for-knock clauses emerge mainly in sectors where the risk
for the parties of negligently causing accidents through lapses is high even when the
parties take adequate precautions, and where it can be hard to tell who, if any, has
acted negligently. Under such circumstances, dispute costs can be much reduced by
dispensing with tort law. Such dispensing involves an externality, as employees and
third parties are generally not fully compensated for their losses. This externality may
call for courts to invalidate the exclusion of liability for gross negligence to make
room for liability insurers to increase the incentive for care. However, such court
policy will force even small contractors to draw up extensive liability insurance, which
is typically costly, and is also likely to induce disputes even when the act of the injurer
only amounts to simple negligence. The temptation on the part of the victim’s frst-
party insurer to argue that the act was grossly negligent can be great. The incentive
efect can therefore be created at lower cost by increasing the fnes levied against an
injurer who has violated public safety regulation in a grossly negligent manner. In
an incentive perspective, such fnes serve the role of exactly increasing care when
third parties are not fully covered under tort law or the workers’ compensation
system.
It may be remarked that the system which appears optimal is hence similar to that
which emerged after the Macondo accident. The Louisiana court set very high fnes
that were augmented by the fnding of gross negligence, and the court emphasized
that fnes could not be indemnifed as they served a deterrent role.6
6 See, e.g., Saraceni’s and Summer’s description of the Macondo accident and its legal repercussions here:
https://www.studocu.com/en-gb/document/university-of-london/contract-law/knock-for-knock/2603964
97
CHAPTER 5
Knock-for-knock indemnity
provisions and liability insurance
Potentially strange but always complicated bedfellows
1 Introduction
Although knock-for-knock (K4K) indemnity provisions have become common (if
not necessary or legally required) in many energy and marine insurance contracts,
in reality their actual application in insurance claims can be complex, as K4K
provisions do not always harmonize with basic general liability policy provisions.
As discussed in more detail in Chapter 2, K4K indemnity provisions involve recip-
rocal responsibilities, wherein each party agrees to be responsible, or to indemnify,
the other for damage or injury to its own property or employees, regardless of fault
for loss. The rationale for K4K provisions starts with the need to avoid delay and
disruption. In theory, these provisions will prevent the need for detailed investiga-
tion and reduce the scope (and costs) of any dispute or litigation. They can also
beneft the business relationships of the contracting parties by reducing the pos-
sibility of scorched-earth contract and tort disputes and, theoretically, encouraging
the parties to provide a safer workplace. The concept of K4K indemnity is so
fundamental to marine energy contracts that such provisions are required by the
P&I insurance clubs for certain type of insured entities to be considered in their
insurance “pooling” arrangements. In this context, any deviation from the K4K
regime by the entities will necessitate their purchase of additional extensions to
their coverage that will typically have lower limits and be more expensive. However,
outside of the context of P&I insurance clubs, other insurers may be less experi-
enced with understanding the interplay between K4K provisions and their policy
language.
Indeed, such provisions are somewhat contrary to basic underwriting concepts con-
tained in most liability insurance. These provisions are intended to make indemnifca-
tion obligations reciprocal for certain losses between two contracting parties. When
these indemnity obligations are insured, however, they may require an insurer to
indemnify a party who is not an insured and whose liabilities and operations may
be wholly foreign to the insurer. While it has become common for insurers to make
non-contracting parties “additional insureds” (AI), the indemnity scenario presented
under K4K provisions can be very diferent than a traditional AI scenario – not least
because the non-insured reciprocal indemnitee may never have direct rights under the
insurance policy. Especially when combined with “other insurance” provisions, these
K4K provisions can present hopelessly complicated scenarios for the contracting parties
98 DOI: 10.4324/9781003206798-5
knock-for-knock indemnity provisions and liability insurance
and their insurers. This short chapter will highlight and address some of those actual
and potential complications.1
2 K4K provisions and general liability insurance

K4K liability provisions are most often used in contracting environments that involve
high risk and potentially large, if not catastrophic, potential liabilities, such as oil
feld and ofshore construction and operations. Although the parties to these indem-
nity contracts are sometimes large, well-leveraged, multinational corporations, they
are just as often smaller operations that rely heavily upon liability insurance to protect
their corporate assets. Indeed, insurance must play a critical role in the success of
these indemnity schemes. Ideally, each party In a K4K indemnity contract should
be in a position to warrant that it has obtained applicable insurance to cover its
reciprocal indemnifcation obligation. Such insurers presumably would be able to
respond to any claim made by its named insured for its specifc reciprocal obligations
under the indemnity contract. However, the insurers are neither signatories to the
indemnity contracts nor, usually, part of the negotiations that result in the K4K
obligations, so there can be no guarantees that the insurance liability policies will
conform or even harmonize with the indemnity agreements. In fact, insurance policies
can potentially complicate the contractual agreements and lead to liability disputes
and coverage litigation that K4K provisions were intended to avoid.
Indeed, there is an inherent tension between the fundamental purpose of general
liability insurance and the enterprise being undertaken in K4K indemnity schemes.
General liability insurance is intended to respond to the unforeseen fortuitous respon-
sibilities of the insureds to other third parties to whom the insureds are legally liable.
Reciprocal indemnity agreements upend that arrangement inasmuch as they may
require an insured to pay for an event or costs for which the insured is not responsible.
The insured may have no involvement at all in the event for which it is responsible
under the K4K indemnity arrangement. K4K provisions undermine the basic tenet
of liability coverage – that there must be a claim by a third-party presenting potential
judgment liability to trigger coverage, because with reciprocal indemnity schemes the
claim is from the named insured to cover liability it previously agreed to undertake.
Many of the standard general liability form conditions and exclusions confict with the
intended goals of K4K clauses. For example, one standard condition is that no insured
“will, except at that insured’s own cost, voluntarily . . . assume any obligation . . . without
[the carrier’s] consent.”2 While this condition is generally only applied to assumptions
of obligations after a loss, it is a textual example of the complications of trying to force
independently negotiated and separate contracts to work together. An insured’s agree-
ing to be responsible for a risk is contrary to a standard form general liability policy’s
Conditions to coverage. Similarly, the conditions call for the insured to assist the insurer
1 This chapter does not attempt to address every relevant aspect of this subject. For example, it does not
analyze anti-indemnity statutes and their impacts on K4K provisions and liability insurance. Rather, the
intent of this chapter is to focus generally on the impacts that K4K contracts have when they are applied to
standard liability insurance provisions.
2 ISO Form CG 00 01 04 13, Section IV – Commercial General Liability Conditions (2)(d).
99
jay r. sever and lauren e. burk
in enforcing the insurer’s rights against any person or organization “which may be liable
to the insured”,3 but K4K provisions necessarily require that the insured not enforce
rights against the responsible party. And as a consequence, to be marketable, insurance
policies often must include waiver of subrogation endorsements whereby the carrier limits
is recovery rights to supports its insured’s agreement not to try to seek reimbursement
for claims paid by other insurance companies.4 Additionally, standard form general
liability exclusions are intended to avoid many of the risks that insureds undertake in
K4K provisions. As a key example, the maritime and energy industries frequently involve
the use of watercraft, but general liability policies exclude coverage for bodily injury and
property damage resulting from the “ownership, maintenance, use or entrustment to oth-
ers of any . . . watercraft owned or operated by or rented or loaned to any insured.”5
The insured’s indemnity obligations of K4K provisions may prevent application of this
standard exclusion, resulting in a carrier providing unintended coverage.
It is axiomatic that general liability policies are not meant to respond to injuries to the
insured’s employees. K4K provisions also usually include an insured’s accepting liability
for injury to its employees, which is a clearly excluded risk on a general liability policy.
General liability policies exclude such risks to avoid functioning as and overlapping
with workers’ compensation coverage.6 Again, by contracting to take responsibility for
injuries to its employees and agreeing to indemnify a potential responsible third party
for injuries to the insured’s employees, the policy may ultimately provide coverage for
the excluded risk of injury to the insured’s employees.
Notwithstanding all of these incongruities, insurers clearly have adapted to K4K
schemes. They continue to write policies to insure liabilities assumed in reciprocal
indemnity agreements. But they do so, to some degree, in an environment that continues
to present challenges and potential disputes arising out of the disconnect related to (1)
the fact that the insurance underwriters are not involved in the drafting and negotiat-
ing of the underlying indemnity contracts that attempt to govern the parties’ liabilities
as well as their insurance obligations, and (2) their policies contain language that is
often at odds with the undertakings set forth in the underlying indemnity contracts.
3 Impact of blanket AI endorsements

Inasmuch as coverage for K4K indemnity obligations is not fully based upon an
evaluation of an insured’s specifc business risks, there are some similarities to blanket
AI endorsements. Blanket AI endorsements provide automatic AI status to entities
that a named insured is required by contract to name as an AI in the named insured’s
policy.7 However, blanket AI provisions are meant only to add a contracted AI to
3 Id. at c.(4).
4 Another way insurers undertake their insureds’ K4K provisions are endorsements stating that the cov-
erage aforded is non-contributory and will not seek contribution from other insurance carriers.
5 ISO Form CG 00 01 04 13, Section I – Coverages, 2. Exclusions g.
6 Akin to note 4, supra, insurers now accept their insureds’ K4K provisions by including an exception
to the Employer’s Liability Exclusion for “liability assumed by the insured under an ‘insured contract’.”
7 ISO endorsement CG 20 33 is commonly used form, and identifes the additional insureds as follows:
any person or organization for whom you are performing operations when you and such person or organi-
zation have agreed in writing in a contract or agreement that such person or organization be added as an
100
the liability coverage, and they do not provide the kind of frst-party coverage implied
in the K4K system.
As discussed in Chapter 2, because contracting parties agree to be responsible for
losses to their employees and property regardless of liability for the injuries, the K4K
system may allow an insurer8 to review and potentially approve its named insured’s
known risks. Blanket AI endorsements have the contrary efect. Insurers have limited,
or zero, opportunity to fully evaluate the potential risk associated with AIs which
stand to gain beneft from blanket AI endorsements, because there is no limitation
on the named insured’s ability to enter into contracts and no limitation on the types
of businesses that might obtain AI status. The named insured is free to bind the
insurer to any entity with which the named insured agrees to name as an AI with no
notice to the carrier and without consideration to how such agreement might expand
the risk under the policy. Thus, although there is some symmetry between blanket
AI endorsements and K4K provisions, that does not mean that AI issues are any
less problematic under K4K scenarios. K4K provisions theoretically foster amicable
working relationships between the contracting parties by allocating responsibility for
losses before they happen. However, when an insurer and its policy terms are added
to the mix, and it challenges the agreement between the parties, the benefts of K4K
may be lessened or destroyed. Identifying the correct AI form is critical because, as is
discussed below, many AI forms preclude coverage for the specifc risks that insureds
generally agree to undertake in a K4K clause.
Blanket AI endorsements frequently require that liability be tied to the named
insured’s acts or omissions, which conficts with the K4K provision’s predetermined
allocation. For example, the commonly used AI Form CG 20 10 04 13 only applies to
injuries “caused, in whole or in part, by . . . [the named insured’s] acts or omissions
. . . in the performance of [the named insured’s] ongoing operations for the additional
insured.” Similarly, another commonly used form, CG 20 37 04 13, states that it
only applies for bodily injury and property damage “caused, in whole or in part, by
‘[the named insured’s] work’ at the location designated . . . performed for that addi-
tional insured.” Such requirements are completely contrary to K4K clauses, which
are written such that the insured is responsible for whatever losses it has agreed to
accept liability for in the K4K provision (often its employees and equipment) regard-
less of the insured’s involvement in the loss.
Another aspect of K4K provisions that can be complicated by the specifc word-
ing of the AI form at issue arises when the K4K provision’s AI clause encompasses
a contracting party’s group, or afliates, requiring AI status for all of those entities.
The afliate entities may be unknown to the insured at the time of contracting but
additional insured on your policy. . . . A person’s or organization’s status as an additional insured under this
endorsement ends when your operations for that additional insured are completed.
IRMI. “Additional Insured Status: Blanket Additional Insured Endorsements.” Available at www.irmi.
com/online/crt/ch011/1l11e000/al11e030.aspx (accessed 23 January 2022).
8 See Christopher L. Evans & F. Lee Butler, “Reciprocal Indemnifcation Agreements in the Oil Indus-
try: The Good, the Bad and the Ugly, 77 Def. Couns. J. 226, 228 (2010) (“each party [to a knock-for-knock
contractual provision] can better determine the amount and cost of the insurance they will need for the job
by simply knowing how many people will be at the work site that fall under the contracting party’s scope
of liability.”)
101
can be wide-reaching, and the insurance policy’s AI endorsement may not satisfy
that requirement, potentially resulting in the insured’s breach of the underlying
contract. For example, ISO Form L805 (05/09), an Additional Insured – Primary
and Noncontributory – Automatic Status When Required in Contract or Agreement
Endorsement (AI Endorsement) provides, in part, that the “Who Is an Insured”
provision is amended “to include as an AI any person or organization when you and
such person or organization have agreed in writing in a contract or agreement that
such person or organization be added as an AI on your policy” (emphasis added).
When an insured contracts with Company X and agrees to name X’s afliates as AIs,
such AI endorsements may run afoul of the insured’s agreement because the insured
has not contracted with any “afliates” of X and arguably no coverage would be
aforded to X’s afliates under these terms. Similarly, the AI Endorsement states that
“[s]uch person or organization is an additional insured only with respect to liability
for ‘bodily injury’ . . . caused, in whole or in part, by your acts or omissions, or the
acts or omissions of those acting on your behalf . . . [i]n the performance of your
ongoing operations for the additional insured.” If the insured is not performing
ongoing operations for X’s afliates, no coverage would be aforded to the entity,
and the insured would, again, be in breach of the contract.
Another potential confict between the underlying contract and the AI endorsement
occurs where gross negligence and willful misconduct are excluded by K4K provi-
sions. Putting aside the question of whether gross negligence or willful misconduct is
adequately defned in the K4K agreement, the question may arise whether AI status
aforded by the contract is nevertheless provided (even if reciprocal indemnity is not).
As suggested, the language of the AI provision may control this issue; e.g., if the
AI provision contains limitations as to the status of the AI or follows the scope of
the indemnity contract, it may not provide AI coverage where the purported AI is
allegedly grossly negligent or accused of willful misconduct. There is, nevertheless,
the possibility that an entity could be provided AI status even if there is no recipro-
cal indemnity due to the alleged gross negligence or willful misconduct of either of
the contracting parties. In this situation, the AI provision would ultimately provide
more coverage than was accounted for by the K4K provision.
4 “Other insurance” clauses

One of the most basic questions that arises in K4K scenarios is the following: which
insurance is meant to cover a loss where there are reciprocal indemnity obligations
and multiple potentially applicable policies? Other insurance clauses attempt to pri-
oritize and “stack” coverages when multiple policies are implicated. In general, such
clauses either attempt to make the policy in question either primary to or excess over
any other policy(ies) that provide coverage on the same basis. Whether, and how,
such clauses work to prioritize and allocate coverages is highly dependent upon the
wording and on the laws applicable to such provisions.
Unless the order of coverage is specifcally spelled out in the K4K provision, it is
possible, if not likely, that more than one liability insurance policy will potentially
respond to the claim. This will require some determination as to whether either policy
is primary, or whether both policies apply on some pro rata basis. Moreover, even if
102
the K4K agreement specifcally attempted to prioritize the insurance obligations, there
is a valid question as to whether such contractual obligation would be binding upon
the insurer, particularly if the policies’ “other insurance” provisions do not refect the
commercial agreement. Thus, unless the polices are written to harmonize with the
reciprocal indemnity scheme, “other insurance” provisions are often likely to confict
with K4K provisions, particularly when there is the possibility of gross negligence
or intentional conduct or the parties are AIs in each other’s policies. Because there
is no bright-line legal rule as to whether “other insurance”’ clauses trump the insur-
ance priority provisions in K4K provisions, there will often be confusion as to the
priority of coverage, unless there has been some coordination in the underwriting
of the potentially applicable policies.
5 Incorporation of indemnity limitations in insurance policies

The Deepwater Horizon group of decisions illustrates the potential interconnection
between the specifc reciprocal K4K indemnity and AI requirements often contained
in K4K indemnity contracts and the application of the actual language in the poli-
cies themselves. BP’s position was that it was owed both AI status and coverage under
Transocean’s $750 million of primary and excess insurance policies.9 Transocean’s
policies contained blanket AI endorsements, granting AI status to “[a]ny person or
entity for whom the ‘Insured’ is obligated by oral or written ‘Insured Contract’ to
provide insurance such as aforded by [the] Policy.”10 There was no dispute that the
drilling contract was an “Insured Contract.”
Transocean’s insurers argued that the scope of coverage was governed by the
terms of the agreement between Transocean and BP, specifcally limiting coverage
to above-surface pollution. BP’s primary argument was that a court could look only
to the four corners of the policies themselves to decide whether they cover a par-
ticular loss, and that the policies themselves did not limit Transocean’s coverage to
pollution coverage caused either above or below the surface.11 However, the Texas
Supreme Court eventually concluded that BP was not entitled to unrestrained cover-
age because the language contained in the BP-Transocean drilling contract actually
limited Transocean’s liabilities for sub-service pollution liabilities. In fact, the drill-
ing contract did not require Transocean to procure insurance to cover BP’s liability
below the surface. The Texas Supreme Court essentially incorporated the terms of
the drilling contract into the insurance policies themselves, holding that Transocean’s
policies’ reference to Insured Contract required looking at the drilling contract itself
to determine Transocean’s specifc obligations in order to resolve BP’s coverage status.
The Deepwater Horizon decision makes clear that if the wording of an insurance
policy allows for the incorporation of another contract to determine AI status and
coverage, a court must incorporate the other contract to the extent required by the
language of both contracts.
9 In re Deepwater Horizon, 470 S.W.3d 452 (Tex. 2015), opinion after certifed question answered, No.
12–30230, 2015 WL 13918242 (5th Cir. June 9, 2015).
10 Id.
11 Id. at 456.
103
Another complicated issue where the underlying contract infuences the insur-
ance contract and could result in controversy (if not litigation) involves the setting
of “minimum” insurance limits in a K4K agreement. As is the case with indemnity
contracts generally, negotiating parties sometimes will choose to require a minimum
limit of insurance to be provided by one or more of contracting party’s indemnity
contracts (e.g., Master Service Agreements [MSA]) which contain K4K provisions. In
a case involving the Texas Oilfeld Indemnity Act (TOIA), the United States Court of
Appeals for the Fifth Circuit disagreed that the word “minimum” limits the amount
of insurance actually available but ultimately found that the MSA’s insurance require-
ments capped coverage under the TOIA.12 The MSA in Cimarex Energy provided a
K4K indemnity requiring Cimarex Energy (Cimarex) and CP Well Testing (CP) to
indemnify each other against their “group.” The MSA further required Cimarex to
secure a minimum of $1 million in primary general liability coverage and $25 million
in excess, while CP was required to obtain a minimum of $1 million in primary general
liability coverage and $2 million in excess. CP actually carried a total of $11 million
in primary and excess coverage, however. After an employee connected to CP was
injured, Cimarex settled the claim for $4.5 million and looked to CP for reimbursement.
CP refused to indemnify Cimarex for more than the “minimum” $3 million required
in the contract, causing Cimarex’s insurer to cover the additional $1.5 million. That
Cimarex insurer thereafter sued CP to recover the diference, arguing that the use
of the word “minimum” does not limit the amount of insurance that a contracting
party may obtain, and that CP had ample limits to cover its reciprocal insurance
obligation. Both the district court and the Fifth Circuit initially held that, in using
the word “minimum”, the MSA agreement did not set a ceiling for the amount of
insurance available for CP’s reciprocal obligations but instead only established a foor.13
However, the Fifth Circuit agreed with the district court that while $3 million was not
necessarily the ceiling, CP’s additional $8 million in coverage was not obtained “for
the beneft of [Cimarex] as indemnity and therefore not available to Cimarex.” The
Fifth Circuit specifcally endorsed the district court’s reference to extrinsic evidence,
including language in CP’s excess policies, which limited those policies to paying the
minimum limits CP agreed to procure in an indemnity agreement. Thus, neither CP
nor its excess insurers above $3 million were required to reimburse Cimarex’s insurers.
If nothing else, the Cimarex Energy decision, and the dispute underlying that decision,
make clear that the parties’ intent when they use words like “minimum” in reciprocal
indemnity contracts may not harmonize with either the expectations of the insurers or
the language of their respective policies. It is likely that the Cimarex Energy decision
will result in a reassessment of such contract language in future K4K agreements.
6 Waiver of subrogation
Many K4K provisions require that the insurers of the reciprocal indemnity obliga-
tions waive their rights of subrogation. Generally, outside of the context of K4K
provisions, an insurers’ right of subrogation is sacrosanct, and any attempt by an
12 Cimarex Energy Co. v CP Well Testing, L.L.C., 26 F.4th 683 (5th Cir. 2022).
13 Id. at *3.
104
insured to waive such rights would be viewed as unenforceable or null and void. This
is yet another curiosity of the K4K contracting world.
The impact of subrogation waivers can be profound. In A.M.C. Lifeboats Inc. v
Apache Corp., the court was asked to harmonize two K4K indemnity contracts with
the insurance procurement provisions contained therein.14 Importantly, one of the
applicable insurance policies contained a full waiver of subrogation. The court found
that when insurers include this provision, they and their insureds “cannot recoup
from the additional insured any amount they have paid to settle a risk covered by the
policy, even on the theory that the recoupment is based on the additional insured’s
risks not covered by the policy.”15 The court further found that, read together, the
two operative contracts mandated that A.M.C. Lifeboats’ insurance policy was spe-
cifcally made primary to the policy issued to Apache. Thus, the claims made by
A.M.C. Lifeboats against Apache were disallowed.
7 Conclusion
As demonstrated herein, K4K provisions do not easily harmonize with liability insur-
ance policies. This may well relate to the common lack of communication between
(1) the parties negotiating reciprocal indemnity agreements and (2) those underwriting
the insurance policies meant to fnancially support liability arising from the opera-
tions addressed in those agreements. Despite these challenges, many businesses clearly
wish to continue purchasing liability insurance policies to support reciprocal indemnity
schemes, and many insurers wish to continue selling and issuing them. The bigger
question may well be whether insurers and insureds alike will work together to better
address some of the uncertainties and incongruities before they result in coverage
litigation.
14 A.M.C. Lifeboats Inc. v Apache Corp., 2008 WL 217177 (E.D. La. 2008).
15 Id. at 5.
105
CHAPTER 6
The efect of choice of law on

knock-for-knock clauses
Uisdean Vass
1 Introduction
Writing in the context of the upstream oil industry, I take the broad “knock-for-
knock” concept to be composed of three elements which are, beginning with the
simplest and moving to the most expansive, (1) the basic knock-for-knock clause
itself; (2) the wider system of indemnities and exclusions which, along with the
knock-for-knock clause, form the fabric of oil service responsibility allocation; and
(3) a supportive general system of law which facilitates the application of such con-
tractual devices. In order to look at the efect on knock-for-knock clauses when choice
of law changes from one law to another, one must have a default system to compare
other laws against. For reasons stated below, I propose to “adopt” English law as
being the default or “mother” system for knock-for-knock clauses.1
After having analysed knock-for-knock clauses and relevant supporting English
law immediately below, I will go on to analyse what happens, or may arguably hap-
pen, when one attempts to use the knock-for-knock modality under other systems of
law, specifcally including US maritime law, Texas law, Louisiana law, and Brazilian
law. English law allows sophisticated contracting parties very broad latitude to agree
through contract their own rules for the allocation of contractual and tortious liabili-
ties. In particular, knock for knock under English law is very common in international
oil and gas and international maritime contracts. English law knock for knock has
attained this widespread acceptance in certain high-risk industries because of the
respect it attaches to the wishes of sophisticated contracting parties expressed in
writing, to the certainty it brings to contractual construction, and to the wealth of
its body of relevant precedents and judicial commercial expertise.2
1 As a Scottish solicitor, it is worth noting that while Scotland is a separate British legal jurisdiction,
along with England and Wales, and Northern Ireland, the diferences between English contract law and
Scottish contract law are regarded as being small in substantive terms. Despite the fact that the majority
of UKCS oilfelds lie in what might be loosely described as “ofshore Scotland” English Contract Law is
the default choice-of-law for contracts involving the ofshore oil and gas industry in the United Kingdom.
2 As the authors Nadorf and Gomes state: “By far the most popular governing law for Wellsite Contracts
between IOCs and Contractors in respect of operations to be conducted in a civil law jurisdiction is that of
England and Wales. English law is favoured because it is: (i) generally regarded as user-friendly, (ii) particularly
well suited to interpreting the nuances of a contract drafted in English, (iii) fexible, pragmatic and commer-
cially minded, seeking to uphold freedom of contract, and (iv) provides a healthy body of oil and gas case law.
In addition, English courts are highly respected for their independence, efciency, predictability and probity”.
Nadorf & Gomes, “Look before you leap: are your oil patch liability clauses enforceable? (An analysis under
civil law jurisdictions with emphasis on Brazil)”, Journal of World Energy Law and Business, 2021, 14, 49, at 52.
106 DOI: 10.4324/9781003206798-6

the effect of choice of law on knock-for-knock clauses
In terms of oil industry imperatives, the moral is to think very carefully about
how choice of law may afect one’s crucial knock for knock and broader indemnity
and insurance regime.
2 Knock for knock and the full English legal treatment
2.1 The knock-for-knock clause itself

In the most basic case, two contracting parties (A and B) promise one another that
if damage accrues to one or other party arising from the performance of the contract,
Party A will bear the damage to its property and people, and Party B will bear the
damage to its property and people.
For example, see the following language in a Petrobras (Petrobras Clause) standard
form drilling contract:3
The CHARTERER undertakes to indemnify PETROBRAS Group for any and all dam-
ages involving
i. The CHARTERER GROUP personnel.
ii. goods or facilities owned by the CHARTERER Group
PETROBRAS undertakes to indemnify the CHARTER Group from any and all dam-
ages involving
i. PETROBRAS Group personnel.
ii. goods or facilities owned by PETROBRAS Group.
If we take the Petrobras Clause at face value (and leave out possible controversies
about what the word “indemnify” may actually mean),4 the intent expressed seems
to be that the Charterer will bear losses to its “people and property” expressed as
belonging to Charterer Group. “Charterer Group” will mean the Charterer company
and its directors, ofcers, and employees, along with the Charterer’s subcontractors
and their directors, ofcers, and employees.
In order to “isolate” Petrobras from damages to the people and property of the
Charterer Group, the Charterer must agree to three diferent conditions with Petro-
bras. Firstly, if the employees of the Charterer Group are damaged by the acts of
Petrobras, then the legal recourse of these individuals (who are not parties to the
Charterer/Petrobras contract) must be against the party at fault, which in this case
will be Petrobras. In order to “protect” Petrobras from losses arising from such
third-party suits, the Charterer will have to agree to compensate Petrobras for any
losses recovered by the employees of the “Charterer Group”. This kind of contrac-
tual clause is properly referred to as an indemnity. Secondly, the Charterer must also
agree to indemnify Petrobras from damages payable in suits brought by members
of the Charterer Group (other than itself) for property losses caused by Petrobras
in the performance of the contract. Thirdly, and perhaps most consequentially, the
Charterer must waive its own right to sue Petrobras for its property losses caused by
3 Quoted in Nadorf & Gomes, id. at 65.

4 See Chapter 2, section 2.1 of this book.
107
uisdean vass
Petrobras in the performance of the contract. This latter type of clause is properly
called a waiver, an exclusion, or a release.
2.2 Simple knock for knock in contractual context

At least in the oil industry, simple knock-for-knock clauses rarely appear on a stand-
alone basis. These clauses will generally appear as the principled centrepiece of a
wide range of specifc indemnities, exclusions, and sometimes limited liability clauses.5
For example, in the (UK) LOGIC Well Services Edition 2, March 2001 (LOGIC),
the indemnities and exclusions are largely contained in Clause 19, and the classic
knock-for-knock clauses are Sub-Clauses 19.1 (Contractor in favour of Company)
and Sub-Clause 19.2 (Company in favour of Contractor). The key knock-for-knock
language is “save, indemnify, defend and hold harmless”, which covers both contract
and tort. Except where expressly qualifed, this language in favour of the benefciary
(whether Company or Contractor) is held to cover “negligence or breach of duty”.6
Further, LOGIC Clause 19 sub-clauses speak to such issues as pollution and con-
tamination from the reservoir;7 pollution emanating from, inter alia, the equipment
of Contractor Group while at the wellsite, and above the rotary table or the vessel
bottom, and caused by the negligence or breach of duty of Contractor Group up to
an agreed limit;8 loss of Contractor Group property, materials, and equipment
downhole;9 repair of Contractor Group’s equipment (excluding downhole equipment)
for damage beyond fair wear and tear if such damage is caused by corrosion, erosion,
or abrasion;10 loss or damage to the well, blowout, or uncontrolled well condition
and damage to reservoir;11 and losses arising from the use of radioactive tools
downhole.12 In Clause 21.2 the Company agrees to indemnify and hold harmless the
Contractor Group for Company Group’s Consequential Losses,13 and the Contractor
5 A limited liability clause simply limits the extent to which contracting party A is required to compensate
contracting party B for damages. Indemnities may be qualifed by limitations of liability. See, e.g., LOGIC
Well Services Edition 2, March 2001, Sub-Clause 19.4 (Contractor indemnifes Company up to an agreed cap
for pollution caused at the wellsite through its breach of duty or negligence, which pollution occurs above the
rotary table or above the bottom of the vessel). In the leading case of Ailsa Craig Fishing Co Ltd v Malvern
Fishing Co Ltd [1981] UKHL 12 (1981), Lord Wilberforce stated “Clauses of limitation are not regarded
by the courts with the same hostility as clauses of exclusion: this is because they must be related to other
contractual terms, in particular to the risks to which the defending party may be exposed, the remuneration
which he receives, and possibly also the opportunity of the other part to insure”. Ailsa Craig [1981] UKHL.
6 LOGIC, Sub-Clause 19.11. For reasons discussed in section 1.3, we interpret the use of the word “neg-
ligence” in Sub-Clause 19.11 to be wide enough to cover the kind of conduct usually thought of as being
“gross negligence” but not the kind of conduct usually thought of as being “wilful”.
7 Id. Sub-Clause 19.3. Company indemnifes and holds harmless unless the damage is caused through
wilful misconduct.
8 Id. Sub-Clause 19.4. Contractor pays up to agreed limit.
9 Id. Sub-Clause 19.5. Company pays except if the loss is caused by the negligence or breach of duty of
Contractor Group.
10 Id. Sub-Clause 19.6.
11 Id. Sub-Clause 19.9. Company pays except if loss is caused by Contractor wilful misconduct.
12 Id. Sub-Clauses 19.10. Contractor pays up to an agreed limit if the losses were caused by Contractor
breach of duty or negligence, but Company indemnifes and holds Contractor harmless for all such losses
whose value exceeds the mentioned cap.
13 This is a defned term which covers “indirect losses and/or loss of production, loss of product, loss of
use, and loss of revenue, proft or anticipated proft . . . and whether or not such losses were foreseeable at
108
in turn agrees to indemnify and hold harmless the Company Group for the Contrac-
tor Group’s Consequential Losses.
LOGIC provides a frst-class model of how a knock-for-knock provision lies at the
heart of a complex structure of indemnities, exclusions, and limitations of liability
clauses in standard oil and gas service contracting. While LOGIC embodies a sig-
nifcant range of mutual indemnities, covering both contract and tort liabilities, we
should refect that overall, the main benefciary of the complex indemnity structure
of Clauses 19 and 20 is clearly the Contractor. This is because the Contractor is
the likely party-in-breach as it is the main contract-performing party, and there are
certain oil industry risks that Contractors are unable or unwilling to accept.
2.3 Knock for knock in its English legal context

My thesis is that English law establishes a strongly supportive foundation for the
type of knock-for-knock/indemnity-exclusion structure described above. Let us briefy
examine certain pertinent legal principles. At frst sight, English law, through its
contra proferentem rule, tends to interpret contractual clauses against their benefcia-
ries. Therefore, if an indemnity or an exclusion does not explicitly mention that it
covers negligent acts of the indemnitee, then courts will presume that negligence is
not covered.14 However, if knock-for-knock drafters do mention “negligence” (which
they now invariably do), then they are using a word which has a legal defnition
under English law. The word “negligence” under English law (assuming no contractual
language to the contrary) is held to cover any kind of negligent conduct, as opposed
to wilful conduct.15 This would include the type of conduct often thought of as being
“gross negligence”.
“Gross negligence” and “wilful misconduct” have no defned meaning under Eng-
lish law but are frequently used in contracts, often as “carve-outs” for indemnity or
exclusion provisions. These concepts should always be specifcally defned in contract,
though if they are not so defned, English courts will attempt to construe them
according to the wording and context of the contract.16 Provided that the meaning
the time of entering into the CONTRACT”. Id. Sub-Clause 21.1. This careful wording is designed to cover
all heads of probable “economic loss” whether such loss is deemed to be “direct” or “indirect” under the
celebrated holding of the House of Lords in Hadley v Baxendale (1854) 9 Ex. 341.
14 The classic case on this topic is E. E. Caledonia Ltd v Orbit Valve [1994] 1. W.L.R. 221. The Orbit
Valve case involved a situation where an employee of an oil service contractor was killed on the Piper Alpha
platform in the 1988 disaster. The employee’s heirs successfully brought suit against the operator, oil com-
pany E. E. Caledonia Ltd (Occidental) with the operator admitting negligence which caused the death. The
oil company in turn sued Orbit Valve for indemnifcation under its oil service contract. While the contract
contained an indemnity by Orbit Valve in favour of E. E. Caledonia this did not explicitly mention the word
“negligence”. The court found that the law in that case would presume that “negligence” was not covered.
15 As long ago as 1842, Lord Denman opined in Hinton v Dibbin (1842) 2 QB 253 that “it may be doubted
whether between gross negligence and negligence merely, any intelligible distinction exists”. In the later case
of Pentecost and Anr. v London District Auditor & Anr [1951] 2 KB 759, 763, Lynskey J opined that “gross
negligence is not known to the English common law as far as civil proceedings are concerned.”
16 A good recent analysis of English and Scottish case law construing the meaning of “gross negli-
gence” and “wilful misconduct” is contained in Pickavance & Bowling, “Exclusions from Immunity: Gross
Negligence and Wilful Misconduct” (a paper presented to the Society of Construction Law at a meeting
in London on 5 September 2017). See www.eversheds-sutherland.com/documents/services/construction/
D207-pickavance-bowling.pdf.
109
uisdean vass
of gross negligence and wilful misconduct is sufciently defned in the contract, there
is little that English courts will do to restrict corporate parties’ ability to limit or
eliminate the agreed contractual allocation of a company’s tortious or contractual
responsibility for acts of gross negligence or wilful misconduct.17 This is important
because gross negligence and (perhaps less frequently) wilful misconduct are used as
both carve-outs but also “carve-ins” in commercial indemnity/exclusion structures.18
English courts interpret the words “indemnify and hold harmless” to include both
a traditional indemnity and an exclusion.19 Furthermore, English courts do not attach
any signifcant policy diference between a traditional indemnity and a traditional
exclusion at least insofar as companies are concerned. This is a most important point
because, as we shall see below, jurisdictions such as US maritime law draw a sharp
policy distinction between indemnities and exclusions. Moreover, unlike a number of
other systems, English judges will not consider the existence (or not) of insurance in
assessing the enforceability of indemnity and exclusion clauses.20
17 No party may contractually agree to be defrauded, and exclusion clauses may not serve to eliminate
or virtually eliminate a party’s entire set of contractual obligations. Any agreement by an individual to waive
his or her rights to damages for injury or death based on negligence is void. See the Unfair Contract Terms
Act (UNCTA) 1977, Section 2A. There are also UNCTA limitations on exclusion clauses for companies, but
these are much less onerous, and the provisions of UNCTA section 2(2) (subjecting other exclusion clauses
to a reasonableness test) do not pose a threat to most of the type of commercial exclusions which might be
found in commercial knock-for-knock clauses and indemnities/exclusions. Penalty clauses in contracts are
outlawed under English law, but liquidated damages provisions are acceptable, so long as they fairly and
reasonably estimate actual losses.
18 “Gross negligence” is quite frequently used as a carve-in in oil service contracts, but “wilful miscon-
duct” is rarely, if ever, used as a “carve-in” in such contracts. As a practical matter, it is virtually impossible
to insure against acts of wilful misconduct. However, in the case of so-called commercial oil and gas con-
tracts which are usually between oil companies, it is not uncommon to see “wilful misconduct” as a carve-
in. This is perhaps the ultimate display of the extent to which English law is willing to recognise and apply
the well-written indemnity/exclusion provisions of sophisticated commercial parties. To cite one example,
Clause 6.2.4(b) of the [standard form] Oil & Gas UK Joint Operating Agreement (JOA) provides that the
Participants are required to indemnify the Operator for their share of any Consequential Loss caused by
the Operator including such Consequential Loss as is caused by Wilful Misconduct (capitalised terms used
in this sentence have their meaning stated in the JOA).
19 This was the holding of the UK Supreme Court in Farstad Supply AS v Enviroco Limited and others
[2010] UKSC 18. See also Chapter 2, section 2.1 f. of this book.
20 Perhaps the leading UK authority on the interface between indemnities and insurance is the case of
Caledonia North Sea Limited v British Telecommunications & Ors [2002] UKHL 32. Caledonia concerned the
tragic Piper Alpha disaster on 6 July 1988, in which some 165 individuals were killed. The majority of the
claims by the victims or their heirs were settled and paid by the operator’s insurers at an early stage before
the cause of the disaster was ascertained. The operator carried insurance for the death or injury of contrac-
tor personnel, even though insurance for such losses was not required by the operator’s contracts with the
contractors. The contracts had “knock-for-knock” type clauses under which the contractors were obligated
to indemnify the operator for the death or injury of employees belonging to each individual contractor. In the
wake of the claim settlements, the operator’s insurers brought suit against the contractors based on the fact
that they were subrogated to the indemnity rights of the operator. By the time of the appeal, only one con-
tractor had failed to settle. The (remaining) contractor argued that since the operator’s insurer had paid the
relevant damage claims, the insurer had no standing to sue the contractor, as such would constitute a double
recovery. The House of Lords rejected this argument holding that an insurer has a right to be subrogated to
the rights of its insured, even if the right in question is an indemnity. The mere fortuitous fact that the opera-
tor had insurance for the accident claims did nothing to afect its wholly separate right to indemnify against
the contractor. In the words of Lord Bingham of Cornhill: “Thus, the existence of such [operator] insurance,
prudent though it no doubt was in business terms, is irrelevant to the mutual [indemnity] obligations of the
operator and the contractor; in technical language, it was strictly res inter alios acta”. Caledonia North Sea
Limited [2002] UKHL at para 13. In other words, indemnities are one thing and insurance is another.
110
3 Compare the United States
3.1 The jurisdictional issue

Since my primary interest is the application of knock-for-knock to the ofshore oil
and gas industry, it is worthwhile to briefy note (as is much more fully explained in
Chapter 12 by Cindy Matherne Muller et al.) that, when operating in the Outer Con-
tinental Shelf of the United States (OCS), contracting parties do not have the legal
authority to choose their own choice of law to govern oil service contracts. Instead,
the governing law is efectively established by the provisions of the Outer Continental
Shelf Lands Act (OCSLA).21 That statute provides that federal law will mandatorily
apply to “artifcial islands” and other “installations” for the purpose of “exploring
for, developing or producing resources therefrom”.22 As explained in Chapter 12, this
federal law is in the nature of a surrogate state law, meaning that for exploration of
the OCS of the coast of Louisiana, Louisiana law will be the applicable federal law
without recourse to state choice-of-law rules. The main exception to this rule is where
the mineral operations on the OCS are being conducted by a vessel, and in that case,
federal maritime law will apply. For a detailed analysis of when a “vessel” will be held
to be involved in oil and gas operations on the OCS, see Chapter 12. In practice, this
distinction is very important to our knock-for-knock analysis, as both Louisiana and
Texas have so-called anti-indemnity statutes (see discussion below) which would severely
prejudice English-style knock-for-knock (if found to be applicable). Conversely, federal
maritime law is comparatively friendlier to the English model.23
3.2 Federal maritime law – Deepwater Horizon

The case In re Oil Spill by the Oil Rig “Deepwater Horizon”24 addressed the potential
extent of the liability of driller Transocean for the damages and death occasioned
by the infamous Macondo well blowout incident of 20 April 2010, when an explosion
and fre occurred on the mobile drilling unit Deepwater Horizon. BP was the opera-
tor that had contracted with Transocean to drill the Macondo well. The incident
occurred on the OCS and not in state waters, and all parties concurred that federal
maritime law (and not surrogate state law as federal law) applied, presumably because
the maritime status of the drilling operation was incontestable.
While the decision covered a range of matters, we will focus on one main issue. The
Drilling Contract at Article 24.2 provided that BP would “protect, release, defend,
indemnify and hold harmless” Transocean from pollution or contamination arising
below the surface of the water, whether or not the pollution or contamination was
21 Littoral states in the United States have been accorded varied extents of “state waters” of the US
coasts. In the case of Louisiana this has been fxed as three miles, but in Texas state waters extend to three
marine leagues, which approximates to 12 miles. All waters beyond state waters to the edge of the legal
continental shelf area are federal in nature.
22 43 U.S.C. § 1333(a)(1).
23 It might be refected that mandatory choice-of-law rules applicable in the OCS contrast with the
broad (but not totally unfettered) freedom enjoyed by contracting parties in the United Kingdom to select
their contractual choice of law. See discussion in Chitty, ON CONTRACTS, Paragraphs 33–001 to 33–010.
24 841 F. Supp. 2d 988 (E.D. La. 2012).
111
uisdean vass
“caused in whole or in part by the negligence or the fault of ” Transocean.25 Further-
more, the Drilling Contract went on to clarify at Article 25.1 that the words “protect,
release defend, indemnify and hold harmless” meant, inter alia, that the indemnity
extended to “gross [negligence] or any other theory of legal liability”.26 BP fled for
summary judgment arguing that while it was clearly obligated to indemnify Trans-
ocean for Transocean’s acts of ordinary negligence which may have caused the fatal
pollution and contamination, it was not liable to indemnify Transocean for other
acts of fault in connection with the pollution and contamination, including gross
negligence and acts giving rise to an award of punitive damages. BP argued that under
federal maritime law an indemnity clause which covers gross negligence is invalid as
being contra bonos mores. The Court noted that Drilling Contract Articles 24.2 and
25.1 were broad enough to cover both indemnities and releases, but the issue here
was indemnifcation. Judge Barbier, who carefully analysed earlier case law, found
that there was no controlling case in the Fifth Circuit which examined the issue of
whether the inclusion of gross negligence would serve to invalidate an indemnity,
as opposed to a release. Interestingly, the Court was crystal clear that under federal
maritime law, a contractual release for acts of gross negligence would certainly be
against public policy.27 The Court noted a signifcant policy diference between the
circumstances of (true) indemnity and (true) release. In the case of an indemnity, the
principal issue is who ultimately pays for the damages incurred by a genuinely injured
(and compensated) third party. In the case of a release, a genuinely innocent party
loses all hope of compensation, and a “guilty” party walks away free from liability.
Addressing the indemnity issue as one of frst instance, the Court found that in these
circumstances of true indemnity, BP’s contractual obligation to indemnify Transocean
for its acts of gross negligence was not contra bonos mores. The Court found that it
had to weigh two competing considerations: freedom of contract and “a reluctance
to encourage grossly negligent behaviour”.28 The Court ruled that courts should
be loath to restrict the agreements of competent parties. It noted that the Drilling
Contract also allocated liabilities to Transocean (e.g., for pollution arising above the
water level). The system of reciprocal indemnities in the Drilling Contract should
arguably have served to deter Transocean from grossly negligent acts. And Transocean
and BP were highly sophisticated parties of relatively similar economic strength.29
The Court took a diferent view on whether any indemnity in favour of Transocean
for punitive damages payable to third parties would be legally valid. Citing Daugh-
drill v Ocean Drilling & Exploration,30 Judge Barbier explained that the purpose of
punitive damages is to punish and deter wrongdoers. Given that background, any
indemnifcation for punitive damages would be against public policy.31
25 Id.
26 Id.
27 The Court cited the Fifth Circuit case of Houston Exploration Co. v Halliburton Energy Services Inc.,
269 F.3d 528 (5th Cir. 2001) for this principle.
28 841 F. Supp. 2d 1000.
29 Id.
30 665 F. Supp. 477 (E.D. La. 1987).
31 841 F. Supp. 2d 1003.
112
3.3 The anti-indemnity statutes
3.3.1 General observation

As detailed in Chapter 12, there exists a societal concern in the United States, refected
in varied state legislation, that indemnity provisions in commercial contracts, whether
these be in areas such as construction or oil and gas, can be used to force contractors
to indemnify “project owners” (whether these be builders, oil companies, or others)
against the negligence of the “owner” parties. As Muller, Amy, and David note:
“Most US states have anti-indemnity statutes, so when contracting for work to be
performed in a particular state, great care should be taken to understand the param-
eters of each state’s particular provisions”.32 Since this chapter is primarily concerned
about knock-for-knock clauses in the context of the oil and gas business, we will
briefy look at the Louisiana Oilfeld Indemnity Act (LOIA)33 and the Texas Oilfeld
Anti-Indemnity Act (TOAIA).34 Such “anti-indemnity” policy considerations are
entirely absent under English law.
3.3.2 Louisiana Oilfeld Indemnity Act and other state provisions

LOIA was enacted in 1981. According to the Fifth Circuit in Am. Home Assurance
Co. v Chevron, Inc, the “purpose of the legislature, and thus the policy interest of the
state, is to protect certain contractors, namely those in oilfelds, from being forced
through indemnity provisions to bear the risk of their principal’s negligence”.35 Simply
put, LOIA provides that any contractual provision “pertaining to a well for oil, gas
or water”36 which purports to require a contracting party to indemnify a counterparty
for death or personal injury arising from that counterparty’s negligence is invalid. As
indicated in Chapter 12, there is considerable case law on what constitutes “pertaining
to a well”, and there are certain exemptions to the scope of the statute. For example,
the provision does not apply to bodily injury or death arising from radioactivity or
operations to control a wild well.37 The statute does not apply to property damage.
LOIA seems to apply as much against relevant indemnities favouring contractors
as it does indemnities favouring oil companies. Since LOGIC style knock-for-knock
clauses and related indemnities/exclusions arguably principally favour contractors, it is
difcult to see the rationale for LOIA-style indemnity invalidation, at least in the ofshore
context. The same might be said regarding the TOAIA on indemnity invalidation.38
Additionally, Louisiana Civil Code Article 2004 provides that “Any clause is null
that, in advance, excludes or limits the liability of one party for intentional or gross
fault that causes damage to the other party”. This language would appear to rule out
32 See Chapter 12 of this book.

33 La. R. S. § 9:2780.
34 Tex. Civ. Prac. & Rem. Code § 127.003.
35 400 F.3d 265, 269 (5th Cir. 2005) (cited in Donaho, “Texas Oilfeld Indemnity Handbook” (Donaho)
(Baker Hostetler) (2019).
36 La. R. S. § 9:2780.B.
37 Id. § 9:2780.F.
38 There are precedents in other states for the invalidation of indemnities which are granted by the con-
tractor but not those favouring the contractor. An example is the New York Anti-Indemnity Law which is
general, and not oil and gas specifc, in nature. See New York Gen. Oblig. Law § 5–322.1.
113
uisdean vass
any contractual release for acts of gross negligence or wilful misconduct. Donaho’s
view is that it is unclear whether an indemnity for gross negligence is valid under
Louisiana law.39
3.3.3 Texas Oilfeld Anti-Indemnity Act and other state provisions

TOAIA was enacted in 1973 and amended in 1989 and 1999. Its purposes are similar
to that of LOIA, and its wording is also similar to some extent. In the case of
TOAIA, indemnities in an “agreement pertaining to a well for oil, gas, or water or
to a mine for a mineral” covering the negligence of a contracting party are void if
this negligence causes “personal injury or death”, “property injury”, or any loss
which “arises from personal injury, death or property injury”.40 Therefore, unlike
LOIA, TOAIA extends also to indemnities covering property damage. TOAIA lists
15 distinct activities which fall under the “well” or “mine” defnition.41 TOAIA also
lists injuries and contracts to which it is not applicable, and these include losses from
radioactivity,42 pollution43 and reservoir damage,44 and under contracts such as joint
operating agreements.45 TOAIA has a legislative “Insurance Coverage Exception”,
also referred to as a “Safe Harbor Provision”. What this means is that TOAIA will
not invalidate otherwise defcient indemnity provisions if the indemnitor provides
insurance for the indemnity obligation to the indemnitee. There are diferent rules
for “mutual indemnity obligations” and “unilateral indemnity obligations”.46 Sum-
marily, TOAIA, like LOIA, seems to apply as much against indemnities favouring
contractors as it does indemnities favouring oil companies. Also of note, Donaho
opines that it is not clear whether an indemnity covering gross negligence is against
public policy in Texas or not.47
4 Compare Brazil
4.1 General observations

Brazil, a classic “full” civil law jurisdiction, takes its place among the world’s great
ofshore oil producers. Importantly, up until the enactment of the Hydrocarbons
Law in 1997, the national oil company Petrobras SA held a monopoly over the
Brazilian upstream oil and gas industry.48 Despite the creation of a new regime of
concessions, Petrobras retained title to its signifcant producing acreage and it remains,
far and away, the main player and largest oil producer in Brazil. The Petrobras con-
tracting system is of huge importance for both Petrobras and its supplier community.
39 Donaho, supra note 35, at 46.

40 Tex. Civ. Prac. & Rem. Code Ann § 127.003 (West 2016).
41 Id. § 127.001(4).
42 Id. § 127.004(1).
43 Id. § 127.004(2).
44 Id. § 127.004(3).
45 Id. § 127.002(5).
46 See discussion of the “Safe Harbor” provisions in Donaho, supra note 32, at 8–9.
47 Donaho, supra note 35, at 46.
48 Law No. 9.478 of 6 August 1997. See generally, the wider discussion of knock-for-knock clauses under
Brazilian law in Chapter 11 of this book.
114
As alluded above, Petrobras, in addition to many private operators, is now moving
to an indemnity system based on knock-for-knock clauses. Petrobras invariably
demands that its oil service and construction contracts are subject to Brazilian law
and Brazilian adjudication, though private oil companies do frequently use English
choice-of-law clauses. At the time of writing, there is considerable interest in the
Brazilian and wider oil industry,49 and among Brazilian legal practitioners, regarding
how the knock-for-knock modality functions under Brazilian law.50
4.2 Knock-for-knock and Brazilian law

There is little specifc case law on knock-for-knock clauses as such in Brazil. The
general rule of the Brazilian Civil Code is that parties must respond for the damage
which they have caused.51 However, there is also the competing precept of freedom
of contract which can, in principle, allow contracting parties to allocate between
themselves responsibility for losses in ways at variance from the provisions of the
Civil Code. There is authority to the efect that limitation of liability clauses are valid
if they are (1) reasonably related to the prospective damage and (2) they have not
resulted from the unequal negotiating power of one contracted party.52 Nadorf and
Gomes note that the frst principle (reasonable relation to prospective damage) could
sit awkwardly with knock-for-knock clauses since indemnities and exclusions under
such clauses must, by their nature, cover unforeseeable future damages. Furthermore,
most commentators believe that Brazilian courts would not be inclined to treat clas-
sical indemnities and exclusions much diferently, especially where the parties are
sophisticated commercial players. The main Portuguese verb used in oil and gas
indemnities is indenizar, which most commentators see as covering both classical
indemnity and exclusion senses. Last, most Brazilian authorities seem to take the
view that contractual parties have the right in commercial contracts to allocate con-
tractual responsibility amongst themselves providing that this allocation does not
ofend the rules of good public order.53 Most commentators are clear that indemnities
covering wilful misconduct (dolo) would be invalid as violating good order. Arlota
grounds this view in the fact that the Brazilian Civil Code at Article 432 imposes a
duty of good faith on contracting parties, and wilful misconduct would always be a
violation of good faith.54 Accordingly to Arlota, the better, or majoritarian, view is
that indemnities covering gross negligence (culpa grave) are also inadmissible. However,
49 This author had extended discussions regarding the issue with a major oil service company based in
Aberdeen in November 2021.
50 This author gave a presentation to a Brazilian legal and industry audience on “Knock-for-Knock – A
Liberal English Journey” on 1 November 2021, organised by Lefosse Advogados.
51 Brazilian Civil Code Articles 186, 187.
52 Special Appeal No. 1.076.465–SP (2008/0160567–4), published on 21 November 2018. This case is
cited and discussed by Nadorf & Gomes, supra note 2, at 63.
53 See Alexandre Sales Cabral Arlota (Arlota), “A Cláusula Knock-for-Knock” (The Knock-for-Knock
Clause), 2019 (published in Brazil solely in Portuguese), p. 116.
54 Id. at 165–170. It might be noted that under English law, there is not, as a general proposition, a duty
to act in good faith.
115
uisdean vass
Arlota accepts that a minority view exists that argues for the admissibility of indem-
nities for gross negligence.55
5 Conclusion
Knock-for-knock clauses and the kinds of quite radical responsibility shifting through
indemnities and exclusions common in oil service contracts are likely to receive their
widest, most consistent, and strongest support under English law. Providing the
activity intended to be indemnifed/excluded is express in the contract (e.g., negligent
acts), English law has not, like Louisiana and Texas, any great concern about either
indemnities or exclusions for negligent or grossly negligent conduct. Given that an
efective knock-for-knock clause between corporations requires both indemnities and
exclusions, it is important that English law, at least at the corporate level, makes no
great policy distinction between indemnities and exclusions. The opinion of Judge
Barbier in the Deepwater Horizon case discussed above highlights how diferent the
situation is in that respect under US maritime law. And as we saw, under Louisiana
law there can be no contractual exclusion for gross negligence. Under English law,
the validity or otherwise of indemnities is not impacted by the presence or otherwise
of insurance, which is not the case, say, in Texas. It is doubtful whether indemnities
for acts even of gross negligence (or far more, wilful misconduct) are valid in Brazil.
And in the other jurisdictions which we reviewed, courts may look beyond what the
parties have written and enquire into considerations of individual fairness, bargaining
strength, and so forth.
The situation is much diferent in England and Wales. Bereft of considerations of
legal complexity, doubtful clarity, and moral agonising, oil service contracting under
English law is above all simple, certain, and not very sentimental. It might be objected
that the Gladstonian market-orientation of English commercial law might serve to
encourage irresponsible behaviour by parties covered by broad and morally dubious
indemnities. But this does not seem to be the case. The safety record of the UK oil
industry, for example, is very highly rated, and no rational contracting party is likely
to be encouraged to indulge in risky behaviour by the English knock-for-knock regime
as no-one wants the stain of personal casualties, property losses, and pollution.
55 Id. at 171–176.
116
PART II
CHAPTER 7
Indemnity clauses in fabrication and

construction contracts in Norway
DOI: 10.4324/9781003206798-7 117

CHAPTER 7A
Limiting and channeling liability under ofshore

construction contracts in Norway
Knut Kaasen
1 The problem
The purpose of this chapter is to investigate means and boundaries for regulating two
types of liabilities arising from contractual relationships: liability for default and liabil-
ity for physical damage caused in connection with the work under the contract. The
basis for these analyses lies within agreed Norwegian standard conditions of contract
for major ofshore construction projects.1 Such projects involve lengthy contract periods,
huge costs and complex deliveries due to the extensive research and development often
required. Therefore, they normally carry the risk that the contractor is not able to
fulfll all of his contractual obligations, thus exposing him to broad liability for con-
tractual breach – potentially exceeding his fnancial capabilities. As such exposure is
not considered to be efcient or rational, the parties agree on some kind of limitation
to this exposure. This is illustrated by ofshore construction contracts.2
Inherent in major construction projects is also the risk that any work related to
the contract may cause damage to life, health or property of one of the contractual
parties or a third party. Damage to third parties may result in one of the contractual
parties being held liable, implying that a loss is brought into the contractual rela-
tionship in the same way as if one party causes direct physical damage to the other.
Without provisions to the contrary, any physical damage sufered by a third party
or a contractual party would incur liability on behalf of the tortfeasor, potentially
1 Norwegian Total Contract 2015 (NTK 15), Norwegian Fabrication Contract 2015 (NF 15) and more
(see note 6 below). NTK 15 (and its “siblings”) may be downloaded at no charge from www.norskindustri.no/
contentassets/69b36f82c6f341a68c1aa7691e4f5e06/versjoner-lagt-inn-des.-2017/norsk-totalkontrakt-2015.pdf
(Norwegian and English parallel texts). The sponsors of the NF and NTK standards have changed over the
years since the frst edition in 1987, due to organizational changes. The current 2015 editions are agreed between
Norsk Industri (The federation of Norwegian Industries) and Norsk Olje og Gass (The Norwegian Oil and Gas
Association), representing Norwegian contracting companies and oil and gas companies operating on the Nor-
wegian Continental Shelf respectively. The two organizations “recommend that the standard contract is applied
when contracting for the supply of large components for the production of production of petroleum reserves on
the Norwegian continental shelf if the contractor will be responsible for engineering, procurement, construction
and potentially also installation (EPC (I))” (preface to the standard, a similar recommendation is included in the
preface to the other 2015 standards). A brief general presentation of the 2015 standards is given in Roggenkamp
et al. (ed.), Energy Law in Europe, third ed. (Oxford 2016) 11.87–11.112. Even contracts that generally are not
based on these standards will normally contain clauses to the same efect as those discussed here.
2 Clauses limiting liability for breach in contract (and for damages in tort, see below) are of course to
be found in several other types of manufacturing contracts (shipbuilding, on-shore construction, data pro-
gramming, etc.) both in Norway and internationally. The reason for concentrating on ofshore construction
contracts is provided below.
118 DOI: 10.4324/9781003206798-8

limiting and channeling liability in norway
exposing him to losses exceeding his fnancial capabilities. According to ordinary tort
law criteria for liability, it would also be necessary to establish whether the damage was
actually caused by the party, whether he caused it by an act of negligence, whether
the party sufering the damage should have avoided (some of) the loss, and so forth.
As an alternative, positive contract regulation may simplify the distribution between
the contractual parties of both the loss caused by damage sufered by one of them
(internal claims) and loss arising from third-party claims being honored by a contractual
party (external claims). There is a long tradition for including regulations to the efect
that such losses are distributed exhaustively and predictably according to clear criteria.
Such arrangements, at the same time, facilitate rational insurance coverage: risks are
identifable and expressly channeled to one of the parties, allowing for rational insurance
coverage of the risk. By adding provisions to the efect that the other party shall beneft
from such insurance (status of co-insured and waiver of subrogation), the cost and
hassle caused by double insurance can also be avoided. Hence these contracts combine
the channeling of liability for damages with the establishing of an efective insurance
coverage. In combination, these provisions constitute the knock-for-knock system.
These major construction contracts therefore illustrate systems for limiting the
parties’ exposure to two types of contractual risk: the liability for breach is to a large
extent limited, and so is liability for damage caused by contract work, but based
on a diferent mechanism. While the former is a universal, well-known contractual
technique, the latter – the knock-for-knock regime – is not as widely applied or
known in its combination of liability, indemnity and insurance provisions. In either
respect, contract provisions limiting or excluding liability may be contested: should
the agreed risk and liability systems be fully afrmed, or should they be censored
based on general rules of contract law – the test of reasonableness?3 The issue of
censoring clauses on limitation of liability for breach and clauses disclaiming liability
(and indemnifying the other party) in tort are often discussed simultaneously. As we
will see, ofshore construction contracts require a more nuanced approach.
From a practical point of view, the efect of provisions limiting liability in difer-
ent types of situations are quite similar: the party at fault will not be held liable for
any loss actually arising from a situation that he would typically be accountable for
based on general rules of law. But in legal terms, liability for breach is fundamentally
diferent from liability in tort. The diferences relate to the basis and conditions for
liability, how the loss is calculated, the method of limiting the liability, and the con-
siderations potentially resulting in “the limitation being limited”. One would there-
fore expect these two areas of liability to be kept apart in most contexts. However,
despite the diferences, taking a combined view of the liability regimes for breach and
damage ofers an opportunity to discuss two general aspects of contractual regimes
limiting risk related to construction contracts. First, they illustrate various techniques
for limiting diferent kinds of risk exposure, and second, in doing so, they ofer an
opportunity to investigate the general reasoning behind setting aside clauses that
limit liability – or giving them their intended efect.
3 The test of reasonableness refers to the criteria of the Norwegian Contracts Act sect. 36. See further
section 4.1, below.
119
knut kaasen
The agreed standard ofshore construction contracts include all of these elements.
This is illustrated when the contract object sufers damage prior to delivery. Because
title to the contract object passes to the client as the work progresses, such damage is
deemed to hit property belonging to the client. At the same time, there is an inherent
risk that contract specifcations will not be fulflled at delivery (i.e., the risk of breach).
Normally, the loss caused by the damage is covered by the project insurance, which
the client is required to establish. The situation thus ofers an interesting example
of the link between the two liability situations, including the role of contractually
organized insurance coverage. In the following, the means and boundaries of limiting
risk for default and damage are considered based on the well-established Norwegian
Total Contract 2015 (NTK 15),4 which also has infuenced contract practices and
standards far beyond its intended scope of application.5
The obvious starting point is that liability for breach and liability for damages in
tort must be handled separately – they are diferent risks that are subject to diferent
types of limitation. Therefore, we start by a separate look at each of these types of
liability and corresponding limitation. Also, the discussion of whether the clauses
can be given full efect may well take diferent routes depending on which type of
liability we look at. These diferences may, on the other hand, ofer contrasts that are
helpful when discussing the ultimate general issue: to which extent may contractual
arrangements defning extent of liability for damage and breach occurring in con-
nection with fulflling the contract be set aside by general rules of contract law that
limit the parties’ autonomy to establish such arrangements? First, however, we need
to look into the contract provisions on liability for breach (section 2) and physical
damage (section 3), both based on the regime of the “NTK family of contracts”.6
2 The contractor’s liability for breach under ofshore construction contracts
2.1 The liability
2.1.1 The need for a balanced regime

The provisions on liability for breach must, on the one hand, make it desirable for
the contractor to avoid breach. Further, they should hold the company harmless – at
least to some extent – from the breach. These aspects are discussed in this section
2.1. On the other hand, the liability should be realistic, providing a workable balance
4 NTK 15 is the (so far) latest in a row of “agreed documents” originating in the Norwegian Fabrication
Contract 1987 (NF 87). Since then, there have been fve rounds of revisions, none of them implying major
material or editorial changes. The terms have been negotiated and agreed between major oil companies
operating on the Norwegian Continental Shelf on the one side and the association of Norwegian mechani-
cal industries on the other.
5 Slightly modifed versions of NTK 15 (or its predecessors dated 2000, 2005 or 2007) are used in several
major onshore construction projects, and the standard has put clear marks on agreed contracts dedicated to
that area (NS 8405, NS 8407) as well as on shipbuilding contracts (BIMCO’s Newbuildcon) and contracts
for data programming projects.
6 In addition to NTK 15 the family comprises the Norwegian Fabrication Contract 2015 (NF 15), the
Norwegian Total Contract (Modifcation; in two versions) 2015 (NTK (MOD) 15) and Norwegian Subsea
Contract 2005 (NSC 05).
120
between incentive and exposure. Often this is achieved by capping the liability for
delay (in the form of liquidated damages) at some percentage of the contract price
per day delay and a maximum of, say, 10% of the contract price. However, with very
large contract values, this could result in liability exceeding hundreds of millions of
Norwegian kroner, which to many contractors would not make any practical sense
as a limitation. Consequently, a general cap to liability for delay could be set at a
fxed amount (related to the contract price or not), curtailing the contractor’s exposure
if the percentage system yields higher limits. Liability for defects could similarly be
maximized at a percentage of the contract value in combination with a maximum
fxed amount. It can also be limited by excluding certain costs (e.g., related to remov-
ing other parts in order to rectify the defective part).7 Some techniques for limiting
liability for breach are further presented in section 2.2.
2.1.2 Types of sanctions against breach

Contractor’s default gives rise to the traditional consequences: he has to rectify defects
and pay liquidated damages upon delay.8 In addition, default has a practical implica-
tion: delay (directly or by way of defects having to rectifed) normally entails that
payment is withheld or curtailed because payment normally is based on progress of
the work. The contractor may also be held liable for the company’s losses caused by
the default, except in cases of delay.9 Ultimately, substantial breach may result in
termination of the contract (having more drastic consequences than cancellation),
but this is utterly unusual.10
The standard does not provide an explicit basis for price reduction in the case of a
contractor’s breach. However, a similar result follows from the provisions that apply when
a contractor has not fulflled his contractual obligations upon delivery. His claim for
consideration is then reduced corresponding to the part of the work which he has not
contractually completed, either based on the percentage of original work still remaining
or in the form of a “negative VO” deleting the remaining work from the contract. In both
cases the economic result is similar to the concept of price reduction because “market
7 The standard contracts of course also contain provisions on the company’s (client’s) liability for default.
Default under payment obligations entail liability for interests. Other types of default on company’s side give
the contractor the right to request a variation order (VO) on certain conditions, including, inter alia, that the
contractor has fulflled his obligation to check company provided items upon receipt and that he has pre-
sented the variation order request (VOR) within a prescribed time limit. The practical efect of a VO in this
situation is an obligation on company to pay. The company’s liability is not subject to any limitation, except
the general provision that neither party can be held liable for the other party’s “indirect losses” (NTK 15
Art. 32.1). However, the company has a discretionary right to cancel the contract at convenience (Art. 17.1),
implying that company’s exposure in principle never exceeds the cancellation fee of 4% of the contract price
at the time of cancellation (Art. 17.3) plus contractor’s right of consideration for work actually performed at
the time of cancellation (Art. 27). This right to cancel also means that it is not an option to require specifc
performance (naturaloppfyllelse) from company. As the company’s liability for default is not limited beyond
these provisions, this aspect is not discussed in the following.
8 Liability for delay is thus not set at the actual loss sufered by the company due to the delay, but see
(c), below.
9 NTK 15 Art. 24.4 and 25.3, para. 3.
10 According to NTK 15 Art. 26.2 termination entails an ex nunc settlement, see (c), below. By contrast,
termination (for company’s convenience) takes the form of an “unfnished delivery” and the company shall
pay the corresponding part of the contract price plus the lesser of 4% of the contract price and 6% of the
remaining part of the contract price as compensation for loss of proft; see Arts. 17.5 and 17.3.
121
knut kaasen
value” for this type of performance for most practical purposes is equal to the contract
price agreed – there are no other footholds. This is diferent if the remaining work at
delivery consists of rectifcation of defects only – as opposed to contract work which has
not been performed at all. In such cases, price reduction in the classical sense is not an
option. Instead the provisions on the right and obligation to rectify exclusively govern
the situation, supplemented by the provisions on liability for certain costs incurred by
company.11 If the contract object remains faulty because the contractor does not succeed
in rectifying it or is not permitted to try, his liability is positively and negatively defned
by these contract provisions,12 no matter how it difers from what would typically follow
from the general background law rules on price reduction.
2.1.3 Specifc features

Even if the types of consequences of contractor’s breach are traditional, the details
are not.
Liability for delay is standardized in the form of liquidated damages (LD), which
accrue irrespective of whether and to which extent the company has sufered a loss
due to the delay. This simplifcation makes the reactions to delay more efective, but
at the same time it disregards the tort perspective of ordinary contract law.13 The LD
amounts to a fxed amount or percentage of the contract price per day, maximized
to a specifed amount or percentage.14
The contractor has an obligation but not a right to rectify defects: the company
may leave the rectifcation to another contractor – typically an ofshore “hook up”
contractor – even if the contractor himself would be able to carry out the work within
a reasonable time. However, in such cases the contractor’s cost exposure is defned by
the rules on calculating the work as if it had been variation work under the contract,
not by reference to the company’s costs if it had let another contractor carry out
the rectifcation.15 Also, under the ofshore standard contracts, the company has the
right to take delivery on the agreed delivery date, even if the contractor does not
want to deliver because the work has not yet been completed.16 The choice between
delay and uncompleted delivery thus rests with the company.
11 NTK 15 Art. 25.5.

12 NTK 15 Art. 25.3.
13 As a general rule, these contracts do not give the company the right to claim damages for delay based
on his actual costs. However, there is one modifcation: if liquidated damages have accrued on previous
penalty milestones, but the contractor meets the fnal delivery date, the accrued liquidated damages are
annulled, except to the extent that the company can show that it has sufered losses by such earlier delays
(NTK 15 Art. 24.2).
14 Under Norwegian law there is no basis for setting the agreed LD system aside on the basis that it is
deemed to amount to a penalty rather than a standardised compensation; there is no distinction between
“penalty” and “liquidated damages” like in some other jurisdictions. However, the efect of the agreed system
in the actual situation could in principle fail the test of reasonableness under the Contracts Act § 36, see
section 4.1, below.
15 NTK 15 Art. 25.3 para. 2. Rectifcation work will often have to be performed ofshore. Restricted
access (transportation, housing and work areas) warrants that company has to remain in charge of setting
priorities in a hectic commissioning phase.
16 NTK 15 Art. 24.3. The reasoning behind this equals that of depriving the contractor from a right to
rectify defects.
122
Termination due to breach does not imply restitution of the parties’ respective
performances. Both parties keep what they have received from the other party under
the contract as per the day of termination: the company has title to the contract
object as it presently stands and the contractor keeps the consideration that he has
received for the work so far.17 The parts of the contract scope which have not been
performed, are deleted. Termination thus takes efect ex nunc, not ex tunc as normally
is the case outside of construction contracts.
2.1.4 Breach of secondary obligations

The standards do not specify remedies for breach of secondary contractual obligations,
e.g., the contractor’s duty to keep certain information confdential, ensure opportunity
for trade union activity and request company consent prior to subcontracting parts
of the work.18 Neither do the standards exclude remedies according to general prin-
ciples of contract law in such cases. This means that such general principles supple-
ment the standards, implying for example that the contractor can be held liable for
company’s losses resulting from the contract object not being free of liens and
encumbrances at delivery.19
2.1.5 Obligation to notify – and the role of the variation mechanism

Contracts governing major construction projects normally contain several clauses
obliging the contractor to notify company of certain facts during the contract period,
for example of any defects, discrepancies and inconsistencies discovered in company-
provided specifcations and items. Breach of such duties normally entail that the
contractor is held liable for all costs incurred by the company that would have been
avoided if the duty to notify had been fulflled. Such breach therefore results in a
new payment obligation on the contractor. There is no limitation to this exposure,
so the obligation to notify is of no relevance in our context.
However, one specifc type of notifcation is relevant when considering the end
efect of the contract’s general regulation of liability for breach. By means of the
“variation mechanism” the company may (by variation order, art. 14.1) unilaterally
change the contract schedule or specifcations, amending the relevant defnition of
the contractor’s obligations under the contract and thus redefning what constitutes
“breach”. More importantly in our context, the contractor may allege that company
has instructed or caused him to perform the work diferently from the contract speci-
fcations or schedule without issuing a variation order, thus at the outset depriving
him of the right to corresponding adjustments of the contract price and schedule.
If he fails to notify accordingly within a prescribed time limit,20 he loses his right to
request a variation and consequently will have to fulfll the new requirements without
any additional compensation.
17 NTK 15 Arts. 22.1 and 26.2, respectively.

18 NTK 15 Arts. 34.2, 9.2 and 8.1 respectively.
19 NTK 15 Art. 22.2.
20 Normally 21 calendar days (see NTK 15 Art. 16.1 para. 2), but specifc time limits are applied in
certain situations.
123
knut kaasen
In these cases, there is no obligation to notify. The loss of rights resulting from
a claim not being timely notifed is not a remedy for breach of contractor’s obliga-
tions. Lack of notifcation does not introduce obligations that the contractor did not
already have21 but deprives him of rights he might otherwise have. However, these
provisions on notifcation form an important part of establishing a basis for the
efective control of complex and dynamic projects. This purpose is served by putting
pressure on the contractor’s wish to contribute to the information fow by submitting
notifcations. The pressure is efectively obtained by nullifying claims which have not
been presented in due time.
2.2 The limitation of liability

If the contractor breaches his contractual obligation, he is subject to remedies as
discussed in section 2.1. However, his liability for remedies is limited in several ways,
which are briefy presented below.
Liability for delay is maximized at a certain percentage of the contract price per day
of delay, never cumulatively exceeding a set total percentage or amount (whichever
is the lower).22 This kind of arrangement is normally conceived as a way of limiting
the liability for delay. However, it can also work the other way: the liquidated dam-
ages accrue at the defned level even if the actual loss caused to the company by the
delay is lower. The liquidated damages thus constitute a system for compensating
a “normal loss”, disregarding the actual loss. The design parameter is less the wish
to limit the contractor’s liability than to establish simple rules for assessing loss in
a complex setting.
The contractor’s obligation to remedy defects is also limited. Under no circum-
stances does it include costs relating to, inter alia, “dismantling of other objects
than the Contract Object to provide access to the Contract Object” or costs relating
to certain other activities, e.g., heavy lift operations ofshore.23 Other remedial costs
are not limited, except that the contractor’s total liability for breach of contract is
limited to a set percentage of the contract price, maximized at a certain amount.24
21 This is true even in case of company’s instructions: admittedly, the contractor has to implement these
instructions, but this follows directly from the mere instruction, not from the fact that he has not presented
a timely VOR to be compensated for the extra work allegedly implied in the instruction (see Art. 15.1).
22 Until 2015 these percentages were fxed as a part of the agreed standards, at a rate of 0.15% per day
and a maximum of 10% of the contract price (NF/NTK 2007 Art. 24.2, paras. 2 and 3). In the 2015 versions,
the liquidated damages per day are defned in Appendix B, and thus not made part of the agreed standard.
The fxing of the maximum exposure is also the result of individual negotiation, as the percentage and a
maximum amount is to be flled in. However, the normal level is discreetly indicated by the fgure of 10% in
square brackets, followed by this footnote: “As an example it is inserted the rate that was included in NTK
2007. The rate is determined on a case by case basis when the Contract is applied.” While no ofcial com-
ments have been made to these amendments, it may be assumed that they have to do with the international-
ization of the market and the fact that the contractors more frequently are large companies as opposed to
relatively minor Norwegian specialized yards previously taking on the obligations.
23 NTK 15 Art. 25.4.
24 NTK 15 Art. 32.2, leaving it for the contracting parties to fll in the fgures, but indicating an appro-
priate level by putting 25% in square brackets, with a footnote to the same efect as in relation to the open
clauses on the level of liquidated damages (see note 22 above). In the predecessor NTK 07, both fgures were
a part of the agreed standard. In addition, there was a specifc limitation to the contractor’s liability for costs
incurred in remedying defects, set at 15% of the contract price (NF/NTK 07 Art. 25.4 para. 1).
124
These clauses on liability for delay and defects are explicitly the sole remedies open
to the company against contractor for these types of breach.25
Finally, overarching these traditional distinctions, the standard contract also contains
a diferent type of limitation of liability: each party shall indemnify the other party
from indirect losses, including but not limited to “loss of earnings, loss of proft, loss
due to pollution and loss of production” (NTK 15 art. 32.1). This applies both to
losses in contract and in tort, and therefore includes all indirect losses caused by breach.
There are some additional mechanisms resulting in various types of limitations
to the contractor’s liability under the contract, but they do not relate to liability for
contractual breach. One type is the knock-for-knock regulation: these clauses limit a
party’s liability for loss caused to the other party by physical damage but also imply
that he must carry the loss resulting from damages to his own property caused by
the other party (see section 3). Also, from a practical point of view, the contrac-
tor’s right to draw on the company’s insurance coverage may be said to amount to
a limitation of his liability.26 As an example, the contractor is obliged – in line with
general principles of contract law – to take necessary actions to bring the contract
object in line with contractual requirements if it is damaged prior to delivery; in the
worst case this may imply having to reconstruct it in case of total loss. However,
major parts of the resulting costs can normally be claimed under the company’s
“construction all-risk insurance”.27
3 The knock-for-knock system
3.1 The call for regulation
3.1.1 The risks of damage

Despite restrictive safety regulations and extensive safety management systems, the
work related to ofshore construction contracts obviously entails the risk of injury
or damage being caused to personnel and property of the parties, subcontractors or
third parties without any contractual link to the parties. The contract object itself
may also sufer damage during its construction.
Such damage may be caused directly by the work, e.g., scafolding falling down at
the contractor’s employee. The link may also be more indirect, e.g., the contract object
has a weakness that causes damage to employees of the company after having been in
operation for some time. Another distinction can be drawn between damages that are
linked to the fact that the parties have a contractual relationship and those that could
just as well occur without any such background (damages caused by the contractor’s
general activities versus those caused by his performance under the contract). Finally,
the damage may result in direct losses (the broken crane has to be replaced), or in
indirect losses (loss of proft, loss related to other contractual relationships, etc.).
25 NTK 15 Arts. 24.4 and 25.5.

26 The company has a similar right under the contractor’s insurances. However, this relates to contractual
claims against the contractor, ref. NTK 15 Art. 31.2 para. 4. As seen from the company’s side, this does not
amount to a limitation of liability, but rather a security for the contractor’s ability to honor his liabilities.
27 NTK 15 Arts. 29.1 and 29.2.
125
knut kaasen
Naturally, the contractor is most likely the tortfeasor, as he is carrying out most of
the physical activities that are likely to cause damage. However, several activities are
normally subcontracted, or even the company may take part in certain work – itself
or by hiring other contractors – that may cause damage. It is also conceivable that
third parties cause damage or injury to persons or property that are related to the
contractual work.
3.1.2 Consequences of no regulation

In absence of specifc regulations in the contract, the legal handling of damages
related to the contract work would have to be based on the rules of tort law. The
tortfeasor will then be held liable for the damage provided that the ordinary con-
ditions are fulflled. In principle, this implies that the party sufering losses is
indemnifed from the loss, while the potential tortfeasor has an incentive to avoid
causing damage. Handling the situation along these lines of traditional tort law
necessitates extensive and individual investigation and evaluation of a number of
facts. For example, are there circumstances which may constitute a basis for liability
(can the tortfeasor be blamed, or is the damage caused by a kind of activity which
may give rise to strict liability?); is there a sufcient causal link between the dam-
age and these circumstances, and is the damage a foreseeable consequence of them?
These relevant facts take numerous forms, and the conclusions reached by such
investigations and evaluations may be highly discretional, resulting in a lack of
predictability. This is a challenge to both parties. To a potential tortfeasor, it will
be difcult to foresee to which extent he may be held liable for damages caused
by his activities, and the potentially injured party will not know for certain whether
he will have a claim against anyone if damage occurs. Such knowledge is a pre-
requisite for establishing rational, and thereby reasonable, insurance coverage for
such risks – including the risk that a liable party will not be able to meet his
liabilities.
3.1.3 Potential regulation

An alternative to having the assignment of damages hinge on factual and legal dis-
cretionary decisions – creating unpredictable legal positions – is to regulate the chan-
neling of the risk. The simplest method is to distinguish between who sufers the loss
or damage.
The frst class of losses comprises damage sufered by one of the contractual par-
ties themselves, by one contract party causing damage to the other party’s persons
or property. Personal injury will normally involve employees. The employers are obvi-
ously not in a position to regulate all aspects of such claims, but they may regulate
how the claim – once presented by the employee – is to be distributed between the
contract parties, not necessarily dependent on who takes the position of employer.
On the other hand, property damage may be regulated without regard to such limita-
tions, subject only to possible restrictions on the enforceability of agreements limiting
liability (see section 6). Such regulation may therefore identify calculable types of
risks (based on, e.g., the type of damage to which kind of property and the resulting
amount of loss), paving the way for rational handling of potential liability in terms
126
of insurance coverage. A specifc challenge arises in case of damage to the contract
object itself – this is an interesting example of a meeting point between liability in
tort and liability in contract.28
Second, the contract could regulate risks of damage to parties who are not con-
tractual parties themselves, but who have a direct or indirect contractual relationship
to one of the contract parties. Admittedly, the contract parties do not necessarily
have a direct control over this legal relationship (although the main contract may lay
down how certain parts of the subcontracts are to be drafted).29 However, similar to
the situation where employees sufer damage, the contract parties may nevertheless
distribute between themselves any claim which one of them has been held liable for
in the frst place, resulting in their respective exposure becoming foreseeable. Such
provisions may be supplemented by provisions that would create a contractual obli-
gation for each party to implement liability clauses in their respective subcontracts
facilitating the liability regime introduced in the main contract. We will return to the
technicalities of such systems in section 3.2.
Finally, the contract may distribute the risk of liability for damage caused to third
parties not having any direct or indirect contractual relationship to the contract parties.
Such third parties may bring a loss into the contractual relationship by holding one of
the parties liable based on ordinary tort law. The resulting risk to the parties may be
made foreseeable and therefore manageable by prior arrangements in the main contract.
3.1.4 The insurance aspects

The need for insurance stems directly from risks of liability and loss. By identifying
and defning the risk in the contract, the need for insurance coverage of that risk
is also identifed. The parties get a basis for acquiring appropriate coverage for liabil-
ity, property loss and personal injury. And they can avoid insuring against the same
risk – as both a liability issue and, e.g., a property loss under their respective
insurances.
However, if the adequacy of the insurance coverage was estimated by each contract
party unilaterally, there would be a risk that the party would not have sufcient insur-
ance in covering the risk allocated to him by the contract. The contractual channeling
of risk would then be inefective – the loss might have to be borne by the other party
despite the contract’s liability system. This other party would consequently have a need
for insurance covering this risk, undermining the benefts that were to follow from
the originally clear distribution of risks. Therefore, there is a need for both obliging
the parties to acquire the necessary insurances, and for this coverage to be rational
(i.e., not overlapping). These considerations call for the insurance arrangements to be
regulated in the contract.
There is, however, an additional need: little is gained by predetermining the distri-
bution of risk related to damages if the insurer reintroduces all the problems of tort
law by holding the party who caused the damage liable under a claim for recourse.
28 See Mestad p. 267 at note 64, and section 3.2 a), below.
29 See NTK 15 Art. 8.5.
127
knut kaasen
If the agreed channeling of risk is to be efective, the resulting insurance coverage
must be the fnal word – at least between the parties and their insurers.
3.1.5 A global view of risk, liability and insurance

The observations presented in sections 3.1.3 and 3.1.4 indicate that there is a need
to evaluate the contract’s provisions on risk, liability, indemnifcation and insurance
as a whole. Each element can only be efective if designed in light of the others. The
natural starting point is to channel the risk of losses due to damage caused in rela-
tion to the contract work as clearly and defnitively as possible to one of the parties.
This provides predictability that paves the way for simplifed handling of each incident
(because the details of a typical tort action do not have to be investigated) and
therefore, for a rational insurance coverage of the risk.
This approach implies that traditional tort law concepts of reparation and preven-
tion are signifcantly toned down. The provisions on allocation of risk and liability
do not appear to refect them at all. This is the cost resulting from cutting loose
from the demand for individual – and thus hardly predictable – evaluation of other
aspects of the damage than the basic question of where it hit (plus its link to the
contract work and the resulting amount of loss).
3.2 Provisions on liability
3.2.1 Overview
The provisions on liability in NTK 15 arts. 29 and 3030 are disconnected from tradi-
tional conditions for liability under tort law. Liability for damages arising in connec-
tion with the contract work is allocated to the parties disregarding how the damage
occurred, whether it was caused by negligence in any form, whether the tortfeasor
was an employee of one of the parties or an independent contractor, and (generally)
the amount of loss caused by the damage. Rather, the allocation is based on a simply
verifable, objective criterion: who has sufered the loss frst-hand? Naturally, the
damage must belong to a category included in the contract’s liability regime: the
damages and the contract work must be linked factually and time-wise. This condi-
tion may, of course, call for difcult analysis. However, these provisions on liability
clearly allocate the risks in a less discretionary and more predictable way than would
have been the case if ordinary tort law were to govern.
These criteria for allocating such losses are more challenging if the loss in the frst
place is sufered by someone other than the contractor or the company. The proj-
ect involves a large number of actors (companies and individuals) having diferent
types of direct and indirect contractual relationships – or even no such relationship
at all. In a liability regime allocating risks based on whether the damage hits either
the contractor or company, damage that in the frst place hits none of them must
30 Hans Jacob Bull, Tredjemannsdekninger i forsikringsforhold (Oslo 1988) p. 333 f. ofers a thorough
analysis of the system of liability and insurance provisions of NF 87, the frst agreed standard contract for
construction work on the Norwegian Continental Shelf – the “grandfather” of NTK 15. As the clauses
of NTK 15 on the relevant points are similar to those of NF 87, Bull’s analysis is still highly relevant and
provides basis for several views expressed in the following.
128
in the next round be allocated to one of them in order to be handled efectively by
the liability regime of the contract. This brings in the concept of “risk zones”:31 If
a certain person, property or party in the contractual hierarchy of the project suf-
fers damage linked to the project, that damage is held to be the risk of either the
contractor or the company, thus belonging to said party’s risk zone.
This naturally leads to distinguishing between three situations: frst, where the
damage hits one of the contract parties; second, where it hits someone else, though
contractually linked to a party (including the employees of a party); and third, when
damage hits a totally external third party. Based on this distinction, the risk zone
of the company and the contractor respectively comprises damages sufered by said
party itself or someone it has a contractual relationship to – other than via the con-
tractual relationship to the other primary party. However, since both the company
and contractor belong to the contractual hierarchy of the project, there is a need to
decide which parts of the hierarchy should be allocated to which zone. We will return
to this issue of “nuclear family” or “extended family” in (b) below.
Damages caused to third parties are allocated to the two contract parties based
on two criteria: when the claim for damages is presented and the amount of loss
caused by the damage. All this allocation of risk applies regardless of whether the
damage is caused by the other party, possibly even if he has acted with gross negli-
gence. This liability system is often termed the knock-for-knock32 system. As shown,
it consist of three main components.33 First, it implies that the parties refrain from
claiming damages from each other, even if the normal tort law conditions are met.
Second, they waive their right of subrogation against each other in cases where they,
according to contract, have covered the loss sufered by a third party, even if there
is a basis for subrogation according to ordinary rules. And third, they accept to
hold the other party harmless if he has had to cover third-party claims – including
claims from his own employees – that according to the contract shall not be borne
by him, even if neither he nor the third party has any basis in general law for such
claim for being held harmless.
The provisions on damage to the contract object (NTK 15 art. 29) constitute a class
of their own. In line with general contract law, damage to the contract object prior
to delivery is explicitly at the contractor’s risk, as is his unconditional obligation to
specifc performance (thus rendering the discussion on the extent of this rule under
Norwegian law superfuous). These provisions apply even if the damage is caused by
the company under circumstances that would normally entail liability under ordinary
tort (or even contract) law. On the other hand, this obligation is closely linked to the
company’s insurances. The end efect is that the contractor’s economic exposure by
31 Bull (note 30) p. 347.

32 The historical background for the general idea of this system is said to be an agreement between the
United States and its allies during World War II related to ship convoys transporting goods across the North
Atlantic, and even the Barents Sea, in midwinter darkness and storms, without carrying lanterns. Damage
caused to one Allied ship by another should be borne by the state owning the damaged ship, no matter
the circumstances – there were more important issues at hand. This system was, however, of course not as
detailed and refned as the one to be found in modern ofshore construction contracts. See further Ørvig,
The knock-for-knock agreement, AfS (Arkiv for Sjørett) 3.448 f.
33 Bull (note 30) pp. 346–347.
129
knut kaasen
having to repair the contract object in order to make it conform to contract specif-
cations is normally very limited.
3.2.2 The family zones

Since the liability provisions entail that each party carries the risk of damages hitting
his part of the contractual hierarchy of the project, the contract must determine
which part of the hierarchy shall belong to which party. There are several alternatives
in this respect. The company’s risk zone may comprise all actors having a direct or
indirect contractual relationship to the company (its own employees; mother, daughter
and sister companies; its contractors other than the contractor in question; and all
employees of these companies); while the contractor’s risk zone may comprise similar
actors having a contractual link to the contractor. As an alternative, the company
risk zone may be limited to the company itself; its mother, daughter and sister com-
panies; and all employees of these companies, leaving out contractors and actors
contractually linked to them.34 The latter alternative implies that the “third-party
zone” ((C) in the quote below) becomes much bigger, in that it includes the whole
set of project contractors of any tier, other than the contractor in question, as well
as all their employees, in addition to “genuine third parties” having no contractual
relationship to company or contractor. Under the frst alternative, the third-party
zone is similarly very restricted. Further, no matter the width of the risk zones, the
parties are identifed with their zone members only to the extent that the members
are involved in the construction project governed by the contract. NTK 15 art. 1.26
introduces the extended company zone.35
In order for the allocation of liability to work as intended, the liability provisions
must be supplemented by indemnifcation clauses to take care of situations where
the party sufering damage has relied on tort law to hold liable another contractual
party, rather than the party who should carry the fnal loss according to the knock-
for-knock provisions.
If for instance an employee (A) of a subcontractor (B) is injured by an employee of
another main contractor (C) than the one who engaged B, the injured employee A may
often hold C (who is the employer of the tortfeasor) liable according to ordinary tort
law.36 In this way a loss which has hit the family that A is part of has been moved to the
family to which C is a member. It is shifted back to the A family by contractually oblig-
ing the head of A’s family – the contractor in our contract – to indemnify C (or whoever
is held liable) from the loss.
34 Bull (note 30) p. 347 introduces the term “extended family solution”, as opposed to “nuclear family
solution” in contracts where the company is not identifed with its other contractors and their contractors
and subcontractors and the employees of these companies.
35 In contrast, the UK standard model contract LOGIC (can be obtained against payment from https://
www.logic-oil.com/content/standard-contracts0) is based on a nuclear solution on the company side and an
extended solution on the contractor side (Clauses 1.4 and 1.9, respectively). This means that damage caused
to another main contractor’s property by the contractor is handled as a third-party damage. This implies
that the company is only liable to the extent that the damage “is caused by the negligence or breach of duty
. . . of the Company Group” (Clause 19.2 c), also necessitating investigating whether an act of negligence on
the part of the company contributed to the damage.
36 See the Norwegian Act relating to compensation in certain circumstances (13 June 1969 No. 26) sect.
2–1.
130
If a contract party is held liable based on ordinary tort law for losses that according
to the contract shall be borne by the other family, the head of the other family shall
hold him harmless, no matter the basis of liability (negligence, vicarious liability,
strict liability, etc.) or the extent of the damage. However, the distribution within the
other family is not regulated directly in the main contract. Thus, it is conceivable
that the loss stays with the head of the family, simply because he has not arranged
any legal basis for transferring it somewhere. Normally, however, the contracts between
the family members will contain back-to-back arrangements for such situations. Main
contracts based on NTK 15 require both parties to see to it that such arrangements
are implemented in their contracts related to the project.37 When this arrangement
is in place, the party who frst-hand sufered the damage undertakes to hold the
family head harmless from losses related to the damage. When this regime is imple-
mented, there is no point in the injured party holding the tortfeasor liable based on
tort law; the “carousel of loss distribution”38 is stopped at its start.
3.2.3 The third-party zone

Under NTK 15 the risks for damage inficted on (the very limited group of) third
parties are divided between the contract parties based on the magnitude of the loss
and the time at which claims are presented. The formal main rule is that the contrac-
tor carries this risk, but his liability for loss or damage arising out of each incident
is limited to a set amount to be inserted into the general conditions of contract.39
The company shall hold the contractor harmless for liability above this level, no
matter the basis for the contractor’s liability. After issue of the acceptance certifcate,
the company shall indemnify the contractor group from and against any claims of
the kind mentioned in the frst paragraph above, regardless of any form of liability
on the part of the contractor group (art. 30.3 last para).
3.3 Insurance
3.3.1 Considerations
The provisions on insurance contained in NTK 15 are based on several types of
considerations, the fundamental one being that the insurance system should refect
the liability regime (see section 3.1(d)). This entails that the insurances should enable
each party to stand the fnancial burden following from the contract’s channeling of
risk and liability; but at the same time, double insurance coverage should be avoided.
Therefore, the contract should state which party should establish which types of
insurance coverage. On the other hand, the insurances should not undermine the
37 See Arts. 30.1 last para and 30.2 last para, and – specifcally relating to the contractor’s obligation –
Art. 8.5 b).
38 The term is introduced by Bull (note 30) p. 341; see pp. 341–342 on how the concept works.
39 In versions prior to 2015, the amount was fxed at 5 MNOK. Leaving it blank in the standard must
be seen in light of the fact that the contractors in these projects frequently are major international construc-
tion companies having insurance arrangements in place that allows them to undertake larger risks related
to third-party damage.
131
knut kaasen
liability regime. To this end the contract should require the parties to include provi-
sions on co-insurance and waiver of subrogation in their insurance policies.
A further consideration is the need for rational structuring of the insurance cover-
age: the parties’ existing insurances should be used, rather than tailoring insurance
systems for the individual contract, and the party likely to obtain the most extensive
coverage at the lowest price should buy insurance coverage.
The ofshore aspect brings in a fnal consideration. Both the Petroleum Regulations40
and the Joint Operating Agreement between the joint holders of a production license
impose a duty on the licensee to obtain specifc insurances covering activities under
the production license.41 The licensee is also obliged to ensure that the contractors
and subcontractors that he engages in his activities hold insurance for their employ-
ees to the same extent as the operator insures his own employees.42 The insurance
provisions of NTK 15 are partly explained by this background.
3.3.2 Main components

The insurance provisions of NTK 15 art. 31 comprise three elements. First, certain
specifed types of insurance are to be provided. Second, the other party is to be made
co-insured under these policies. Third, the insurer should be barred from pursuing
a right of subrogation against the other party, no matter the circumstances of the
loss covered by the policy. The core requirement is each party’s obligation to provide
certain insurance coverage. Generally, the insurance should cover all liability allocated
to the party in question. However, there are modifcations to this. By way of illustra-
tion, it would not be of much help to the company that the contractor has the risk
for damage to his own equipment if he is not in a fnancial position allowing him
to repair or replace the equipment. In this case the company obviously has a self-
interest in having a satisfactory insurance for property damage in place, but the
contract does not require the contractor to provide such coverage.43 The contract
specifes type and amount of coverage and the period that the insurance should be
valid. Obviously, there are diferences between the company and the contractor in
these respects, the most important stemming from the fact that the company usually
provides Construction All Risks (CAR, ref. (c) below) insurance, which covers the
whole development project rather than each individual contract. The contract seeks
to exploit the benefts of this by also making the CAR policy the main coverage for
the contractor’s fnancial exposure in case of damage to the contract object and
40 The Petroleum Regulations (FOR-1997–06–27–653) sect. 73.

41 The licensee is required to obtain insurance coverage for property damage insurance, wreck removal
and personal injury to employees. See further Bull (note 30), at pp. 99–103 (based on the regulations then in
force, which on the relevant points are almost equal to the present regulations). He emphasizes (pp. 99–100)
that these provisions are the only example of obligatory insurance coverage for damage to property. This fact
must be seen considering insurance’s role in the knock-for-knock system. However, it also has some link to
the Norwegian petroleum safety regulations, which state that due account shall be taken “of the safety of
. . . the fnancial values which the facilities and vessels represent, including also operational availability” (the
Norwegian Petroleum Act 1996 sect. 10–1 para. 1).
42 The Regulations sect. 73 para. 2.
43 See NTK 15 Art. 31.2, which admittedly requires all risks, hull and machinery insurance for vessels
engaged in the activities, but this is due to the liability coverage under such insurance. See Falkanger, Bull
and Brautaset, Scandinavian Maritime Law (4th ed. 2017) pp. 616, 653 f.
132
materials prior to delivery. This is obtained by requiring the CAR policy to name
the contractor as co-insured. On the other hand, the contractor is required to ensure
that the company has a right of direct action against the insurer under the policies
which the contractor has to obtain. Both parties shall ensure that their insurers
renounce any right of subrogation against the other party. And fnally, the contract
establishes systems for ensuring that required insurance coverage is maintained and
for handling incidents covered by insurance.
3.3.3 Insurance arrangements

At the core of the insurance system established by the contract we fnd the company’s
CAR insurance. The conditions of the policy are normally attached to the contract
as an exhibit, albeit often in a preliminary version. The conditions are largely
standardized.
The CAR policy covers the total project, and because of this, it may appear as
overkill in relation to each individual contract. While the whole group of licensees
under the license normally provides the CAR jointly, it also happens that individual
participants make their own arrangements. Notably, the state will be self-insured for
its participating interest in the form of the State’s Direct Financial Interest (SDFI).44
Such arrangements, however, do not afect the fnal distribution of the loss, neither
between the participants in the group of licensees (i.e., “the company”) nor in rela-
tion to the contractor.
The CAR policy provides coverage for both liability and property damage. Depend-
ing on the individual conditions (which are generally standardized to a high degree),
the property damage coverage includes physical total loss of and damage to the
property insured, rescue costs, certain costs unrelated to physical damage, and certain
elements of liability. The policy normally covers the contract object itself, components
meant for incorporation into the contract object, materials necessary for perform-
ing the work without becoming a part of the contract object, and consumables. The
coverage extends over a lengthy period of time, including a “maintenance period”
following the main period. Further, the CAR insurance covers a wide range of risks;
it takes the form of an “all risks” coverage, i.e., property insurance covering loss
arising from any fortuitous cause except those that are specifcally excluded.
The contractor’s insurances are normally less standardized. They are usually bought
in the ordinary market rather than in the “oil market” and are normally general insur-
ances running unrelated to any specifc project. By way of example, the contractor
will have occupational injury insurance covering his employees without regard to
their involvement in specifc projects, and his subcontractors may have P&I insur-
ance covering the vessel chartered for the project. However, some project-specifc
modifcations usually must be made in order to fulfll obligations under the insurance
provisions of the contract, e.g., obtaining a right of “direct action” by the company
against the contractor’s insurer.
44 For a brief presentation of SDFI, see www.petoro.no/about-petoro/establishment-of-sdf-and-petoro.

The Norwegian state-owned company Petoro AS has managed the SDFI portfolio since 2001.
133
knut kaasen
4 Setting aside provisions limiting exposure resulting from default or damage?
4.1 Authority for censoring provisions limiting exposure?

All of the provisions mentioned in sections 2 and 3, limiting a party’s exposure
when liable for default or physical damage, are clearly worded and agreed. Some
of them even state explicitly that they govern “even if the loss or damage is the
result of any form of liability, whether strict or by negligence in whatever form”
by the party relying on the limitation.45 But even in the absence of such clear
wording, there is nothing in the contract indicating that the parties have envis-
aged that there should be made exemptions to the arrangements in certain
situations.
These exclusion clauses apply mutually. Hence their enforceability should be evalu-
ated in relation to both parties. However, in practice it is prevailingly the contractor
who has the beneft of the clauses limiting his liability and holding him harmless. This
is obviously the case in relation to liability for contractual default; but also, physical
damage caused in connection with the work is more likely to have been caused by
the contractor than by the company. In the following discussion, we will therefore
concentrate on the contractor’s right to limit and be held harmless from liability for
breach and for physical loss and damage.
The starting point is that the parties are at liberty to contract out of the liability
regimes contained in the background rules of law applied when a contract is silent.
This includes the right to limit the application of such rules. But any form of
disclaiming or limiting liability under contract is met by skeptical scrutiny.46 Such
arrangements may shake the basic balance of the contractual relationship. It may
be conceived as unacceptable, should the arrangement be valid and enforceable
in all liability situations, irrespective of the actions of the liable party. Therefore,
the traditional view in jurisprudence – referring to, inter alia, pacta sunt servanda –
is that agreed disclaimers and limitations may be set aside in certain situations,
despite their clear wordings. After the introduction of the Contracts Act sect. 36,
it is natural to use this as a base for censoring such provisions. The parties cannot
contract out of this provision – it is mandatory background law. The question is
therefore whether “it would be unreasonable or confict with generally accepted
business practice” to invoke the provisions of NTK 15 that result in limitation
or exclusion of the contractor’s exposure in case of loss, damage or default for
which he is accountable. In determining this, “account will be taken not only of
the contents of the agreement, the positions of the parties and the circumstances
prevailing at the time of conclusion of the agreement, but also of subsequent
events and circumstances in general.”47
45 See, e.g., frst para. of Arts. 30.1, 30.2 and 30.3.

46 An extensive discussion can be found in Viggo Hagstrøm, Om grensene for ansvarsfraskrivelser, særlig
i næringsforhold [On the limits for disclaimers and exculpatory clauses, specifcally in commercial contracts]
(TfR 1996 s. 422–518) (hereinafter Hagstrøm, Ansvarsfraskrivelse).
47 The Contracts Act, 31 May 1918, No. 4 sect. 36. Unofcial translation.
134
4.2 Diferences and similarities between the two types of liability situations
4.2.1 Liability situations

The test of reasonableness according to the Contracts Act sect. 36 may depend on
whether the issue is the limitation of contractual liability for delay and defects48 or
the channeling of liability for losses sufered by the contractual parties or third par-
ties due to damage caused in connection with the contract work. In contrast to the
former, the latter situation does not entail limitations to the contractor’s fundamental
obligation to fulfll his contractual obligations but concerns the fnal placing of a
loss emerging from physical damage caused by the contract work.
In between these situations, we have the case of damage to the contract object prior
to delivery, i.e., while it is still in the contractor’s custody. The contractor is obliged
to deliver in accordance with the contract specifcations – this is at the core of his
contractual obligations.49 Failing to do so brings into play the ordinary provisions
on defects and delay – including the limitations to the contractor’s exposure in this
respect (and hence the question of whether the contractor may rely on these limita-
tions if he willfully refrains from taking necessary actions to remedy the damage).
In this respect we are in the frst category mentioned above. However, based on art.
29.2 the contractor may usually claim compensation in addition to the contract price
to cover most of the extra costs related to curing the damage. In practice this implies
that the risk of damage to the contract object normally does not form part of the
contractor’s general performance risk, which is basically covered by the contract price.
In this sense, art. 29.2 implies a limitation to the contractor’s unlimited obligation
to fulfll his contractual obligations at his own cost.
To which extent this right of compensation will be upheld in all situations depends
on a somewhat diferent evaluation than the one applied to the question of censoring
ordinary exemption clauses related to contractual liability. Because the compensation
of costs is based on insurance, the discussion of whether the resulting limitation of
the contractor’s exposure can be maintained must be based on the conditions of the
insurance policy and the background rules of law on insurance. These questions are
discussed in sections 6.2 and 6.3.
4.2.2 Liability regulations

When evaluating the enforceability of the provisions limiting the contractor’s exposure
under the contract, some diferences between the provisions should be noted.
48 Defned by reference to certain amounts or percentages of the contract price; see section 2.2 above.
49 This general rule of contract law is expressed in NTK 15 Art. 29.1 para. 1, stating that “If loss of
or damage to the Contract Object occurs between the start of the Work until the time when the Delivery
Protocol has been concluded . . ., Contractor shall carry out necessary measures to ensure that the Work is
completed in accordance with the Contract.” There are no explicit modifcations to this strict obligation,
as opposed to the rule contained in the Norwegian Sale of Goods Act 1988 sect. 23 (1) (stipulating the
buyer’s rights following the seller’s breach of contract): “The buyer may maintain the contract and demand
performance. This rule does not apply if there is any impediment which the seller is unable to overcome, or
if the performance would entail such considerable inconvenience or cost to the seller as to be substantially
disproportionate to the buyer’s interest in the seller’s performance.” The reason that no such reservation is to
be found in NTK 15 is undoubtedly the compensation regime stipulated by sect. 29.2 based on the insurance
regime established under the contract.
135
knut kaasen
First, only the liability clauses explicitly state that they prevail “regardless of any
form of liability whether strict or by negligence, in whatever form” on the part of the
party to be indemnifed from liability for loss or damage arising in connection with
the contract work.50 There is no similar explicit statement in the provisions dealing
with liability for breach of contract. However, neither do these provisions contain any
modifcations to the clearly implied joint view of the parties: the limitations shall
have efect to the extent permitted under the background mandatory rules of law.
Second, while the limitations to liability for breach operate to the beneft of the
contractor only,51 the provisions on limits to and indemnifcation from liability for
loss and damage are in principle mutual: the parties are to indemnify each other
from loss and damage inficted by the other party or third parties. However, this
reciprocity is not quite realized in practice; there is little doubt that the contractor
(and the other parties he is identifed with – “the Contractor Group”) for purely
practical reasons is more likely to cause damage to the company’s (and the Com-
pany Group’s) personnel and property than the other way around. On the other
hand, the contract is often used on a back-to-back basis between the contractor
and his subcontractors, and in this relation, the reciprocity is real. Another difer-
ence is now, therefore, more signifcant. The reasonableness of provisions on breach
of contract may be evaluated separate from most other elements of the contract.
Conversely, in discussing the “reasonableness” of provisions limiting exposure in
case of physical loss or damage it is necessary to consider the totality of a system
comprising elements of liability, indemnifcation and insurance. The risk for such
liability is presumed to be covered – and is in fact covered – by insurances for
the beneft of all involved parties, not just the one who has bought the coverage.
As far as this right is upheld, the party sufering the loss has no need to hold the
other party under the contract liable for damage he has caused. This is not just
a matter of taking into consideration insurance coverage which happens to be in
place; the insurance arrangement is, in all detail, made a mandatory part of the
liability regime under the contract.
The combination of liability and insurance regimes may be seen as an alternative
to traditional channeling of loss based on tort law. There are several examples in
Norwegian and Scandinavian law52 that the existence or even possibility of insurance
coverage of losses may be relevant when determining the reasonableness of agreed
channeling of contract related risks.53 In all these examples, the insurance regimes
are more loosely defned and regulated than the regime that is made an integrated
part of NTK 15.
50 NTK 15 Art. 30.1 para 1, Art. 30.2 para. 1 and Art. 30.3 para. 3.
51 The company may obviously also breach its obligations, but the liability for breach is in principle not
limited; see further note 7 above.
52 The Norwegian legislature has explicitly stated a wish to obtain Scandinavian uniformity in this area,
see the preparatory works for the Contracts Act sect. 36: NOU 1979:32 p. 40 and Ot.prp. nr. 5 (1982–83)
p. 11 f.
53 See as illustrations Jan Hellner, Speciell avtalsrätt II Kontraktsrätt, 2 häftet Almänna ämnen, tredje
upplagan (Stockholm 1996) p. 224, Hagstrøm pp. 630–631 and Erik Røsæg, Lastehåndterings- og forvaring-
stjenester (MarIus nr. 271 s. 27–48) s. 41. See also Rt-1994–626 (“quay-inspector”) at p. 630.
136
5 Are the agreed limitations of liability for breach of contract valid?
5.1 Points of departure

The provisions limiting the contractor’s liability for breach of contract are clearly
drafted and agreed without any explicit reservations. They are part of an agreed
standard based on extensive and repeated rounds of negotiations between professional
and balanced parties54 disconnected from the entering into of any individual contract.55
These provisions have been included in this type of contract since the beginning of
ofshore activities on the Norwegian Continental Shelf, and are to be found also in
other standard contracts for major construction work.56 Such clauses appear to
amount to standard practice in this sector – they are “generally accepted business
practice” (sect. 36). This indicates that the provisions are commercially sensible
responses to general needs. There is potential for major losses, and the parties need
to place the risk in a predictable way. In dramatic instances, the losses caused by
breach are so high that the breaching party is not fnancially able to cover it, which
means that exposure to unlimited liability does not make practical sense and thus
would misguide the parties. Also, specifcally relating to the contractor’s obligation
to deliver according to contract specifcations, the provisions limiting his exposure
in this respect must be viewed in conjunction with his right to call on the company’s
CAR insurance (see further section 6). Against this backdrop, setting aside the agreed
limitations to the contractor’s liability for breach requires strong arguments. The
arguments must relate to the contractor being to blame for the breach;57 if not, the
limitations cannot be set aside.
Some situations may be sorted out at the start: the agreed limitations cannot be
invoked if the breach is caused by willful misconduct on the part of the contractor’s
leading personnel.58 This liability would apply inevitably if the leading personnel
acts with the knowledge that their act or omission is likely to impose loss on the
54 If the parties to an individual contract are not in balanced positions, the contractor would normally
be the weaker party, implying that censoring the liability regime would be to the disadvantage to the weaker
party – which would not make sense. However, as the construction part of the ofshore industry becomes
more global, it is conceivable that the commercial strength of the parties tips the other way.
55 The situation is therefore very diferent from the one contemplated in the preparatory works of the
Contracts Act sect. 36 (NOU 1979:32 p. 39) as being at the core of the “reasonable” test. The test according
to sect. 36 aims mainly to protect the weaker party against the other party’s abuse of freedom of contract.
The Norwegian Supreme Court concluded in Rt-2004–1545 that provisions of an insurance bought by a
professional party were not unreasonable, referring, inter alia, to the fact that the provisions were drawn up
by a committee in close cooperation with representatives of the interested parties in the relevant business
sector (para. 52 of the judgement).
56 See, e.g. (the Norwegian onshore construction standard) NS 8407 cll. 40.3 and 42.5 and (the UK
ofshore construction standard) LOGIC cl. 35 and 36.
57 Liability for breach is not contingent on the contractor being to blame in any way for the breach, see
NTK 15 art. 24.2 and 25.2 ref. 25.1.
58 See Are Brautaset, Kontraktsreguleringen ved salg av gass (I: Brautaset et al., Norsk Gassavsetning.
Rettslige hovedelementer. Oslo 1998) p. 117, observing that an important consideration when deciding
whether a limitation to liability for breach is unreasonable according to the Contracts Act sect. 36 is that
behind every contractual obligation there should be a party having a certain minimum liability. If exemption
from all liability were to be accepted even when the breach is caused willfully by leading personnel, it would
mean an exemption from liability for all types of loss covered by the limitation. This, he holds, would entail
results that can hardly be accepted and that would be regarded as unreasonable.
137
knut kaasen
company and in such a way that their action or omission transpires as a reckless
neglect of the company’ interests (see section 5.4).59 Equally clearly, the contrac-
tor may invoke the limitations if the breach is caused by his leading personnel’s
ordinary negligence – this will not be “unreasonable” according to sect. 36. It is
also clear that there is no basis for setting the agreed limitation aside if the breach
is solely caused by ordinary or gross negligence on the part of subcontractors or
the contractor’s personnel other than the leading personnel.60 Normally the agreed
limitations will stand frm even if such personnel cause the breach willfully – this
is not the contractor’s “own fault”.61
The relevant criterion being whether the provisions limiting liability for breach
are “reasonable”, it will not sufce to consider whether the relevant person caused
the breach by an act or omission which itself must be considered as grossly negli-
gent or even willful. Rather, the perspective has to be widened to include elements
such as the type of loss created by the breach, its impact on the company, the
organizational position of the relevant actor, his insight into the alternatives and
consequences and other factors. However, it may be of help to try to analyze such
factors separately.
5.2 The contractor’s own gross negligence

In light of the discussion in section 5.1, the remaining question is whether an agreed
limitation of liability for breach may be invoked when the breach is caused by gross
negligence on the part of the contractor’s leading personnel. Based on the criteria
contained in the Contracts Act sect. 36, the question is whether invoking the provi-
sions in such circumstances “would be unreasonable or confict with generally-accepted
business practice”.
The only judgement from the Supreme Court of Norway on the censoring of agreed
limitation of liability after the entering into force of sect. 36 (Rt-1994–626, “quay-
inspector”) touches upon the question but does not resolve it. The inspector acting
on behalf of the liable party did indeed cause the damage by gross negligence, but
he could not be said to belong to the party’s leading personnel. The court notes that
the provision limiting the liability was part of an agreed standard contract developed
in negotiations between organizations representative of the parties. It should thus be
expected to balance the conficting interests, taking into consideration, among other
things, options for assuring the risk and which party would be most apt to take out
such insurance. The court also noted (p. 630) that the need for predictability and clear-
cut, simple solutions was an essential consideration in striking a balance and found
it unfortunate to disturb the balance established by the provision. This judgement
59 As an example, the contractor decides to use his limited personnel resources on a competing project,
thus inficting delay. This may amount to a basis for revising the agreed limitation or setting it aside, but
not necessarily.
60 See Rt-1994–626 (“quay-inspector”). However, in such situations there could also be gross negligence
on the part of leading personnel, e.g., in their organizing of the work or choice of personnel to perform or
supervise the work, see below in text.
61 See Rt-1948–370. However, modifcations are possible – depending, inter alia, on how the work has
been organized; see Brutaset (note 58) p. 116.
138
cannot be understood to indicate that the limitation would not have been accepted if
the inspector had been a part of the leading personnel; the question is held open.62
Older decisions from the Supreme Court63 – prior to the introduction of sect. 36 in
1983 – are still of interest. However, they do not provide clear answers. First, they do
not apply the same legal standard as provided by sect. 36 – the social norms in general
and accepted business standards specifcally may have changed since the older decisions
were handed down by the court. Second, none of the decisions concludes whether it is
acceptable to exclude liability for losses caused by gross negligence. Finally, the old case
law does not illustrate parallels to the mutually agreed liability limitation regime of the
present large construction contracts like NTK 15. Consequently, the guidance ofered by
this case law is vague. The issue of censoring such clauses must therefore be discussed
primarily based on the criteria presented by sect. 36 itself. In this context, the perspective
cannot be limited to the limitation provisions alone – it is necessary to consider the distinc-
tive character of the contractual relationship and the nature of the contractual breach.64
Legal theory appears to take as a starting point that there is no room for exclud-
ing oneself from liability for one’s own gross negligence.65 However, it is difcult to
identify any discussion of whether there is – or even should be – room for making
modifcations to this rule.66 The preparatory works of sect. 36 explicitly do not con-
clude on the question of whether the old provision of Norske Lov 5–1–2 (which was
not revoked upon the entry into force of sec. 36) excludes the right to limit liability
for one’s own gross negligence.67
When exclusion or limitation of liability follows from express provisions contained
in an agreed standard contract developed in the way NTK 15 is, there are strong
arguments that the parties’ agreement be honored. However, it can possibly not be
excluded that the contractor loses his right to limitation if the contractual breach is
caused by gross negligence on the part of leading personnel with the understanding
that the company is sufering a loss.
Both “gross negligence” and “leading personnel” are fexible criteria providing
little predictability if they were to be decisive for the outcome. Basing the result on
these criteria would therefore contradict a fundamental feature of the liability regime
of the contract, being that liability is to be allocated according to easily identifable
criteria without notable elements of discretion or equivocality. It would not be in
harmony with these considerations if the allocation of risk for liability due to con-
tractual breach were to be decided according to fexible criteria allowing for extensive
discretion. The urge to minimize transaction costs speaks against this solution.68 At
62 Hagstrøm, Ansvarsfraskrivelse p. 486 is of a diferent opinion.

63 A thorough exposé of these judgements is given by Hagstrøm, Ansvarsfraskrivelse p. 448.
64 See in this direction Krüger, Norsk kontraktsrett (Bergen 1989) p. 786. See also the summary of the
Swedish preparatory works related to the similar Swedish act (which in light of the prehistory of sect. 36
must be assumed to be of great interest when interpreting the Norwegian Act) in Hagstrøm, Ansvarsfra-
skrivelse pp. 462–465.
65 See the review in Hagstrøm, Ansvarsfraskrivelse pp. 424–430.
66 See Brautaset (note 58) at pp. 118–119 for a discussion of arguments that supports that the main rule
should be as stated.
67 NOU 1979:32 p. 19 (NL 5–1–2 contains the pacta sunt servanda principle). Hagstrøm, Ansvarsfra-
skrivelser claims that Norske Lov 5–1–2 prohibits such limitations.
68 Hagstrøm, Ansvarsfraskrivelse pp. 480–481 is more reserved.
139
knut kaasen
the same time the wish to prevent breach, which generally is a strong argument for
setting agreed exclusion clauses aside,69 has little bearing in this context. The risk
of losing professional reputation and esteem is likely to be a stronger incentive for
the leading personnel of the contractor not to commit willful or grossly negligent
breach than is the prospective of losing the right to limitation. Admittedly, this may
be diferent in situations where the contractor knowingly puts his interests (all things
considered) ahead of the company’s. However, such actions will result in the full
exposure for liability if the contractor’s knowledge includes the likely consequence
to the company provided that his action on this basis must be considered to imply
a reckless disregard of the company’s interests (see section 5.4).
On the other hand, willful misconduct is difcult to prove. The wish to avoid cases
of “hidden” (unprovable) willfulness may therefore be an argument for setting aside
limitation provisions also in case of gross negligence. Certainly, this would imply just
moving a difcult borderline. However, admittedly there is an important diference
between proving willful misconduct and drawing the line between simple and gross
negligence. Proving willful misconduct necessitates performing a challenging assess-
ment of the contractor’s consciousness, while distinguishing between simple and
gross negligence mostly requires examination and evaluation of just external facts.
This striking diference makes it questionable to equate willful misconduct and gross
negligence when deciding whether to censor agreed limitation clauses.
As noted, obtainable insurance programs may be relevant in the test of reasonable-
ness under sect. 36. If the contractor’s liability for breach is limited, the consequence
may be that loss exceeding the limit has to be carried by the company. Loss due to
physical damage may be – and is expected to be – covered by insurance; this is an
important element in the knock-for-knock regulation because there is a direct con-
nection between the liability and the insurance regimes of the contract (see section
6.2). To the extent that insurance is also an option for loss induced by breach, insur-
ance may be relevant in determining whether the limitation is unreasonable to the
company. However, there is no feasible insurance coverage available to cover losses
resulting exclusively from breach as opposed to losses caused by physical damage
(which may be caused by breach), and the contract does not assume there is. Origi-
nally, the company’s CAR insurance normally covered losses due to damage caused
by “faulty design/materials/workmanship”, i.e., cases of classical breach as opposed
to a fortuitous event. In practice, this implied that the company could claim against
the insurance if it sufered losses because the contractor’s liability for such breach was
limited. In recent years, the tendency seems to be that the CAR also covers physical
damage to objects included in the property insurance coverage of CAR if caused by
“a Defective part, faulty design, faulty materials, faulty or defective workmanship or
latent defect”, however, damages to the defective part itself are not covered unless
caused by another “insured peril” due to no fault or defect of the part itself.70 This
would seem to imply that the CAR policy no longer provides coverage for losses due
69 See, e.g., Hagstrøm, Ansvarsfraskrivelse pp. 479–480.

70 Sited from WELCAR, which normally is basis for CAR insurances on the Norwegian Continental Shelf.
WELCAR is developed by Joint Rig Committee (JRC) at Lloyd’s of London. WELCAR 2001 is available in its
original form at https://content.sbigeneral.in/uploads/c979f1604f6f4423b15ba81a545c8875.pdf.
140
to limitations to the contractor’s obligation to rectify defects, but possibly for losses
resulting from the contract limiting others of the contractor’s liabilities for breach.
5.3 Who is “the contractor”?

The contractor clearly loses his right to limit his liability for default if the default is
caused by his own willful misconduct (with knowledge of the consequences, see sec-
tion 5.4), and it cannot be ruled out that the situation will be the same if the default
is caused by his own gross negligence (see section 5.2). Because the contractor is
always a legal entity (as opposed to physical), it is necessary to determine which
physical persons and corporate bodies constitute “the contractor” in this context.
The conduct of these persons has the potential, in relation to the contractor’s default,
to result in the loss of its right to limitation – provided there is a relevant causal link
between their conduct and the contractor’s liability for breach.
The basic argument for setting aside agreed limitations if the default is caused
by willful misconduct is that behind every contractual obligation there should
be a contract party having a minimum liability for fulflling the obligation.71 If
the parties were allowed unlimited freedom to contract out of this fundamental
prerequisite for the contract being “binding”, the concept of pacta sunt servanda
would evaporate.
Consequently, a guiding observation may be that “the contractor” should be
identifed as the persons and corporate bodies who have been entrusted the task of
making decisions or taking actions which directly impact the contractor’s funda-
mental ability to perform his contractual obligations – as opposed to every single
element that would need to be in place in order to fulfll these obligations. By way
of illustration, the identity should possibly comprise of those who, on behalf of the
contractor, determine the qualifcations of the lead engineer, but should not include
the engineer who’s performing the calculations.
The contractor’s board of directors is clearly “the contractor” in this respect, as
is the CEO.72 This view must be extended beyond the concept of corporate bod-
ies. The technical manager must be identifed with the contractor in relation to
activities performed under his supervision,73 even though the “quay-inspector” case74
indicates that this cannot be taken without modifcations (the inspector could be
said to be the “technical manager” of the forwarding agent). On the other hand,
the agreed limitations clearly are valid if the default or damage is caused by gross
negligence (or even willful misconduct) on the part of subordinate employees or
71 Brautaset pp. 17 and 119 and note 58 above.

72 Possibly, there is a limit to this observation: if the contractor’s top management, or other persons that
he is liable for, substantially exceed what can be reasonably expected in their role, the agreed limitations to
the contractor’s liability for default may be upheld, in line with the views expressed in Rt-1959–849 (“the
apprentice”).
73 An example is an arbitral award 20 August 1973 (referred by Knudtzon in Nordisk Skibsrederforening
Medlemsblad Spesialnummer A 1984 s. 66) where a yard was held liable for decisions taken by its technical
manager, despite agreed limitations.
74 Rt-1994–626.
141
knut kaasen
subcontractors. For this purpose, neither of these groups may be identifed as the
contract party itself.
Beyond these observations, the border is hard to draw.75 The formal position of
the employee may serve as a good starting point, but it is not necessarily decisive
if it is not combined with real infuence on the matter in question. If a decision
by the board of directors does not impact the contractor’s fulfllment of his con-
tractual obligations (because other factors come in between), it would not be to
his detriment that the decision in itself would deprive him the right of limitation.
The other way around, the contractor would hardly lose his right to limit liability
due to the actions of an employee if the employee has not been given the task of
managing basic aspects of the contractor’s fulfllment in the relevant area – i.e., has
not been given a formal role on a higher organizational level. For this to happen,
there must be organizational disorderliness attributable to blameworthy actions by
the leadership of the contractor to such an extent that the agreed limitations may
not be invoked for that reason.
The hierarchical position is thus important, but often the individual circumstances
of the situation will be decisive. For example, it cannot be stated generally whether
the party’s project director is “the contractor itself ” in this context. It will depend,
inter alia, on how independently he is allowed to (and expected to) operate on behalf
of the party in managing the project.76
The parties’ “representatives”, appointed in accordance with NTK 15 art. 3.1, nor-
mally cannot be identifed with the party in relation to the provisions on limitation
of liability. Normally, the independent authority of the representatives is restricted,
and they do not belong to a sufciently high level of the party’s organization.
However, the contractor’s representative may well also be the project director. In
that case, the latter position may be relevant in identifying him with the contractor
in relation to limitation of liability. Nevertheless, a person’s hierarchical position
is not wholly decisive. The contractor may be identifed with persons having been
trusted with the supreme authority within defned areas of the contractor’s busi-
ness activities, no matter their formal position. Normally this would require that
the vested authority cover an independent function in fulflling the contractual
obligations in such a way that the person makes independent decisions substantially
efecting the fulfllment.77
75 An extensive analysis of the question of identifcation – specifcally in relation to limitations to the

shipowner’s liability according to the Maritime Act – is to be found in Ole Lund, Om erstatning for man-
gler, forsinkelse, utbedringsplikt og prisavslag ved skipsbygging, in AfS 8.313 f. The issue is also discussed
comprehensively in the arbitral award “Trans Tind” (ND 1984 p. 404). See also Jan Kaare Tapper, Fabrikas-
jonskontrakten (in Askheim et al., Kontrakter i petroleumsvirksomheten, Oslo 1983) p. 183 and Brautaset
(note 58) pp. 115–116.
76 An illustration in this respect is the arbitral award “Mørland 7” (ND 1988 p. 263). In this case the
contractor was identifed (in relation to the liability provisions of the contract) with his project director, who
had “the authority to make necessary decisions in relation to the performance of the project”, but who were
not authorized to sign (in company law meaning) on behalf of the contractor. It was of no relevance that
the manager was only coincidentally present when the decision was taken, nor that the decisions otherwise
would have been taken by an employee whom the contractor could not be identifed with.
77 In the same direction, Hagstrøm, Ansvarsfraskrivelse pp. 517–518.
142
5.4 The role of the concept of “willful misconduct”
As we have seen, if contractual breach is due to willful misconduct by the contractor,
the enforceability of provisions limiting his liability for the consequences stand at
risk. However, the unescapable condition for setting such provisions aside is whether
they are considered “unreasonable” under the circumstances (see section 4.1). The
question is therefore whether “willful misconduct” will necessarily satisfy this test.78
In my opinion the concept is not quite suitable for the role it has been given in this
context.
Traditionally, the evaluation of reprehensibility has moved along a continuum from
mere negligence79 via gross negligence80 to willful misconduct. Consequently, one
has coined sentences like “no-one can exclude liability for consequences of willful
misconduct”. This implies a need to decide whether the willfulness has to refer to
the efects that the action is likely or certain to have for the other party – in practice
inficting a loss – or whether it sufces that the default itself is caused willfully (or
by gross negligence, if that is the relevant criterion).81
The concept “willful misconduct” appears on these occasions to be understood in
line with the criminal law concept of intent. A key element is to which extent the
action or omission is conscious, and possibly whether the actor is conscious about
the consequences that the action or omission is likely to have to the other party. This
appears to be a sound starting point. A contractual party may breach his contrac-
tual obligations willfully without actually knowing or counting on the other party
thereby sufering a (substantial) loss – and without there being basis for criticizing
him for not considering this possibility. In ofshore contracts, this is not a theoretical
situation; frequently the contracts specify methods and actions to be implemented
by the contractor in constructing the contract object, but these requirements do not
necessarily need to be adhered to in order for the contract object to fulfl the con-
tractual requirements pertaining to the contract object at delivery. Nevertheless, not
fulflling the requirements amounts to breach, but not necessarily substantial losses
to the other party. The conditions for applying the criteria of sect. 36 for setting
agreed limitations aside (the reasonableness test) must at least be that the contractor
has breached his contractual obligations, that the company, for that reason, sufers
78 This is simpler in relation to the knock-for-knock regime (art. 30): For the limitations to be set aside,
the damage to the other contract party, his “group” (section 3.2 above) or third parties, as a minimum must
have been caused by the contractor’s leading personnel’s gross negligence or willful misconduct. This situa-
tion is hardly frequent but may occur if the management – typically for economic reasons – exposes the sur-
roundings to an unacceptable substantial risk for physical damage. It is more conceivable that such damage
is caused by an employee’s gross negligence. In such cases the provisions on limitation and indemnifcation
will be given full efect – this is the kind of situations they are designed for.
79 According to a widely used defnition: the failure to exercise the standard of care that a reasonably
prudent person would have exercised in a similar situation.
80 According to a widely used defnition: a high or serious degree of negligence; a very marked departure
from the standards by which responsible and competent people habitually govern themselves.
81 The proper answer has been disputed. Per Augdahl, Den norske obligasjonsretts almindelige del, 5.
ed. (Oslo 1989) p. 293 og Lund, Standardkontrakter, bilsalg og preseptoriske regler (in LoR 1964 pp. 66–81)
on p. 68 appear to put decisive weight on whether the breach itself is gross negligent or even willful, while
Per Brunsvig, AfS pp. 1.273, 3.172 and 13.354 argues that it is a prerequisite that the party, either grossly
negligently or willfully, has inficted a loss to the other party. The arbitral tribunal in “Trans Tind” (ND 1984
p. 404) endorses Brunsvig’s conclusion.
143
knut kaasen
a loss and that the contractor is aware of both of these facts – or at least that his
ignorance must be attributed to gross negligence at a sufciently high level of his
organization. Such consciousness must be assumed to be a condition for censoring
agreed limitations contained in the construction contracts (see section 5.2 on gross
negligence). However, it is not sufcient. The contractor may breach his duties with
open eyes, knowing that this will cause loss to the company, without necessarily losing
the right to limit his liability.82 The starting point is that the limitations are agreed
without any modifcations. If these provisions are to be censored, it must be because
the contractor, by markedly and fagrantly departing from the standard of conduct
of a reasonable person, sets his interests above those of the company.83 Under such
circumstances, applying agreed limitations would be “unreasonable”.
Neither “willful misconduct” (alternatively “intent”) nor “gross negligence” fts
well as an exhaustive criterion for evaluating the relevance of the contractor’s ideas
of the consequences of the breach. It may seem more proper to tie in directly to an
evaluation of the contractor’s balancing of his own interests against the company’s.
The crucial issue is not whether this act of balancing is “grossly negligent” or even
“willful” or “intentional” (but as noted, it is in practice a precondition that it must
fall into one of these categories). The decisive characterization is whether the bal-
ancing is “reckless” or “clearly disloyal” to the company’s interests, or something to
that efect.84
If the action or non-action chosen by contractor results in physical damage to
the company, as opposed to non-fulfllment of contractual obligations, the situation
may be diferent. The characteristics of the action/non-action would then be at the
forefront: was the damage caused by gross negligence or even willful misconduct (at a
sufciently high level of the contractor’s organization)? In making this determination,
it is often relevant to consider to which extent the contractor realized the possible
consequences of the action/non-action to the company – the duty of care is infuenced
by which consequences may result from lack of care. However, these consequences,
and specifcally whether the contractor evaluates them reasonably, are more central
in case of breach than if the contractor causes damages. Actions/non-actions leading
to breach are normally intentional in the sense that the contractor is well aware of
the fact that the action is or is not taken (e.g., delay resulting from not engaging a
night shift). The focus is therefore shifted to whether the decisions so made appear
to take the company’s interest into account to a sufcient degree.
The decisive criterion should be a general evaluation of whether the contractor’s
actions emerge as reckless or clearly disloyal towards the company, which is not
necessarily the case even if the breach is intended and the contractor knows that it
82 Krüger, pp. 784–785 om “forsettsbegrepet som kontraktsrettslig kvalifkasjonskriterium”.

83 This is illustrated in the “Trans Tind” arbitral award (ND 1984.404), where the tribunal underscores
that (my translation) “if the management has violated the rules that it is bound by, there is a willful breach on
the part of the management. However, the management may have assumed that this breach would not have
any detrimental efects on the fnal result. Only if this assumption must be categorized as grossly negligent
the provisions limiting the liability for breach should be censored. What entails such consequences is not the
formal breach, but actions that may damage the other party and that for this reason must be characterized
as grossly negligent.”
84 I am therefore not quite in line with Hagstrøm, Ansvarsfraskrivelse p. 491 on this point.
144
will cause loss to the company. It is only in such grave situations that it would appear
“unreasonable” to uphold the provisions limiting the liability (see the Contracts Act
sect. 36). For this reason, it does not appear adequate to base the discussion of these
questions solely on the traditional concept of willful misconduct or gross negligence –
other elements must be included.
6 Censoring the knock-for-knock regime?
6.1 The question

The risk of physical damage to persons and property in connection with the contract
work is allocated “regardless of any form of liability whether strict or by negligence,
in whatever form”; the decisive criterion is where (and sometimes when) the damage
hits, not who causes it and how it is caused. This implies an express exclusion of
liability, explicitly without any reservation. Can this arrangement be maintained in
light of the general rules on the enforceability of provisions excluding liability? The
relevant general rule, also in this context, is the Contracts Act sect. 36.
Several authors discuss this question. Bull85 assumes that the courts will not
censor such provisions where – as in the case of NTK 15 – the contract’s liability
regime is based on reciprocity, except in cases of the tortfeasor’s gross “own fault”
(egenfeil). He argues that the basic approach of the courts in relation to this type of
liability regimes should be that the regime is accepted as appropriate, adequate and
reasonable (p. 393). As will transpire from the following, I concur. Selvig86 discusses
the application of sect. 36 in professional relations. See also Sandvik and Ness.87
Wilhelmsen discusses the basis for censoring knock-for-knock clauses of ofshore
charter parties (maritime services contracts) and concludes that the legal situation
must be considered to be uncertain.88 Zak discusses the validity of liability clauses
in drilling contracts under Norwegian, English and American law. The discussion is
relevant also to complex liability and insurance regimes contained in other types of
contracts, including NTK 15. Referring to the complexity of the regimes and of the
factual context to which they apply, she voices objections against applying absolute
rules of invalidity in this area.89 The validity of the liability exemption clause of the
(then operational) standard ship building contract is discussed in the arbitral award
concerning “Trans Tind”.90
The issue has clear resemblance to the question of censoring provisions limiting
liability for default (see section 5), but there are also important diferences. The point
of departure is the same: the parties may exclude liability for loss and damage caused
by subordinate employees and subcontractors, regardless of any form of liability, in
whatever form. The parties may also exclude liability for loss and damage caused by
85 Bull (note 30) p. 394 (with further references).

86 Selvig, Kontraktsretten (in: Knophs oversikt over Norges rett, 10th ed. 1993 p. 553 f., particularly at p. 559.
87 Sandvik, xxx pp. 239–240, Ness, Ansvarsfordeling og forsikring i petroleumskontrakter, MarIus nr.
323 (2005) p. 59 f.
88 Wilhelmsen, supra note 4, at pp. 102–111, particularly at p. 110.
89 Zak, Ansvarsregulering i borekontrakter, MarIus nr. 415 (2013) p. 138.
90 ND 1984 p. 404.
145
knut kaasen
simple neglect by any person whom they are identifed with (see section 5.3), but not
for the consequences of their willful misconduct. The doubt relates to whether a party
may also exclude liability for gross negligence on the part of his leading personnel.91
In this discussion, there emerges a major diference between clauses excluding a
party from liability for physical damage that he has caused and those clauses limiting
his liability for breaching the contract. While the risk of sufering physical damage
or being held liable for having caused it can be insured, the availability of insurance
coverage for risk related to breach is much more limited. This puts a mark on the
discussion of reasonableness of the respective contractual regimes.
The combination of liability and insurance regimes may be viewed as an alterna-
tive to traditional tort law–based channeling of losses. The relevance of insurance
when determining whether agreed risk channeling is “reasonable” can be seen from
the “quay-inspector” case, where existing options for obtaining insurance and which
of the parties that were closest to buy insurance are elements given weight in the
test of reasonableness.92 This is also emphasized in the Swedish preparatory works
for the Swedish Contract Act, which contains a provision similar to the Norwegian
Contract Act sect. 36.93 Hagstrøm (p. 662) states that the Swedish preparatory
works allow for accepting exclusion of liability even for gross negligence if insur-
ance options make it unobtrusive. Røsæg94 fnds that the importance of insurance
in relation to the channeling of liability for cargo handling and storage can hardly
be overstated: it may be rational to waive claims for damages even in cases of
willful misconduct if that results in economic gain, e.g., by obtaining a rebate
exceeding the insurance premium. Ness95 provides further overview of the attitude
in Nordic jurisprudence regarding the relevance of insurance when determining the
reasonableness of clauses excluding liability. In all these examples, the insurance
arrangements contemplated are of a looser nature than the one NKT 15 makes an
integrated part of the liability regime.
91 This situation is not very theoretical. When the GBS structure (support structure held in place by
gravity) of the “Sleipner A” platform sank (August 1991) prior to delivery, it was argued by the client that
the loss was caused by the management of the contractor by gross negligence. The argument was that the
management did not organize the calculation of the strength of crucial parts of the structure in a proper way.
See also the arbitral award “Mørland 7” (see note 76 above; ND 1988 p. 263 f.): a barge was damaged when
four steel piles were lifted of from it under the supervision of the ofshore contractor who had chartered
the barge. The charterparty contained a knock-for-knock clause corresponding to the one found in NTK
15 art. 30.1. The contractor was acquitted from the claim concerning damages caused by the two frst piles
but held liable for the damage caused by the two last. In the latter case there was a willful ofering of the
owner’s interest for the beneft of the contractor, in that the contractor had decided that it would be better
inficting further damage to the barge than exposing more valuable interests to the risk of extensive dam-
age. (The piles were used to pile a steel jacket to the seafoor, and the jacket would possibly not withstand
the inclement weather conditions with just two piles in place.) This balancing of interests was defensible,
but the tribunal found that this willful causing of damage implied that the contractor could not rely on the
knock-for-knock clause.
92 Rt-1994–626 at p. 630. Insurance arrangements was also considered relevant in the “Chainco” judge-
ment (ND 1991 p. 180, Eidsivating court of appeals). See also Ness (note 85) p. 52.
93 SOU 1974:83. The Norwegian lawmaker explicitly states that Nordic uniformity of law is important
regarding sect. 36, se NOU 1979:32 p. 40 and OT.prp.nr. 5 (1982–83) p. 11 f.
94 Røsæg, Lastehåndterings- og forvaringstjenester, MarIus nr. 271 (2011) p. 4.
95 Ness (note 85) pp. 53–57.
146
6.2 “Organized fnancing of loss”
The link between insurance and the contractual liability regime implies that the latter
is not just a clear-cut exclusion of liability. The combined efect is that the system
establishes an “organized fnancing of loss”96 by means of insurance: the insurer shall,
according to the contract, in the last resort carry the risk that is channeled to a party.97
The implication of the liability regime is therefore not that the loss is shifted over to
the sufering party, leaving it to him to cover the risk by insurance as he sees ft, which
is normally the case. Rather, the liability provisions are part of a larger system where
a crucial and compulsory element is a rational organizing of the kind of insurance
coverage that the parties naturally need in light of the liability clauses. This has two
important consequences: frst, the reasonableness of the clauses excluding a party
from liability for loss and damage he has caused to the other party cannot be evalu-
ated as an isolated phenomenon; and second, insurance arrangements are brought
into the evaluation. The practical implication is that the contractual channeling of
the risk related to physical damage is supplemented by elements of insurance law,
where the fundamental question is what extent of protection a contractual party has
under the insurances he or the other party has acquired in fulfllment of their obliga-
tions under the contract’s insurance clauses. As far as this protection applies, the
distribution of risk under the contract can hardly be considered to be unreasonable.
The contract, expressly and in detail, links liability clauses to provisions requiring one
of the parties to acquire insurance coverage for the risk allocated by the liability rules.
This implies that it can, at least, not be easier to censor clauses excluding liability for
physical damage than censoring provisions limiting liability for defects. It is therefore
not conceivable that the reasonableness test of the knock-for-knock regime will be
more extensive than the one discussed in section 5. However, this presumes that the
required (and supposedly acquired) insurances actually cover the risk as intended. If
they do not, for reasons that are not attributable to the party losing the intended
protection, the general existence of insurance can no longer serve as an argument for
the liability clauses being reasonable. Consequently, even though we are dealing (like
in section 5) with an issue of censoring contracts on the basis of sect. 36, we have to
expand the perspective by investigating whether the insurer may refuse to cover the
loss and damage as contemplated in the contract. The point of departure is that the
insurer (assumedly) has undertaken to cover the risk allocated to a contractual party
by the liability regime of the contract, including the risk following from the other
party having the right under the contract to deny liability which would otherwise
follow from ordinary rules of tort or contract law. In other words, the liability provi-
sions of the contract defne the hazards at risk under the insurance.98 Further, as a
main rule, the insurer is bound by this liability regime – he has accepted that they
defne the hazards at risk if rules of insurance law do not relieve him from the obliga-
tion to cover the loss. In practice, this is a question of whether the insurer has waived
his right of subrogation against the party causing the damage in cases where the
96 Bull (note 30) p. 349.

97 Admittedly, there is normally a deduction for own risk.
98 Bull, Forsikringsrett (2008) p. 205 discusses the hazard at risk (farefeltet) of insurances.
147
knut kaasen
contract has channeled the loss to the party sufering the damage. According to the
contract, the terms of the insurance shall include such waiver.99
6.3 Can the waiver of subrogation be maintained?
6.3.1 Waiver of subrogation

Based on the above, the overall assessment of the validity of the knock-for-knock
regime partly rests on whether the insurer ultimately has to cover the loss resulting
from the damage caused by the contract party. The question is whether the insurance
terms on waiver of subrogation can be maintained no matter the party’s actions when
causing the damage.100
The starting point is that the insurer explicitly has waived any right of subroga-
tion against the tortfeasor.101 There is hardly any doubt that this is an element in a
rational allocation of risk and liability under the contract, established by professional
actors in the course of well-considered negotiations resulting in an agreed standard
contract which has been continuously in use since the mid-1980s. The issue is, how-
ever, not whether these elements of the contract are reasonable, but rather whether
the insurer’s waiver applies in all situations (see the Contracts Act sect. 36). This will
not be dealt with thoroughly, but some elements of the evaluation will be discussed.
6.3.2 The co-insured

In addition to the waiver of subrogation, the insurances required by the contract
shall also contain clauses naming the other party as co-insured. These two elements
may appear to constitute an overkill, as the co-insured party is protected from claims
99 NTK 15 art. 31.1 para. 3 and art. 31.2 para 3.

100 The background rules of law are basically the Norwegian Tort Act sect. 4–3, stating that an insurer
who has paid damages for property damages has a right of subrogation against the tortfeasor to the same
extent that the suferer could have claimed the tortfeasor based on the Act sect. 4–2. The main rule under
sect. 4–2 is that the suferer cannot claim the tortfeasor if the loss is recoverable under the suferer’s property
damage insurance. However, exempted from this are damages caused by the tortfeasor in the course of his
trade or economic activity. Most cases of damage caused by one party to the other in connection with the
contract work fall into this exemption. Consequently, the practical main rule according to background law
is that the insurer has a right of subrogation against the tortfeasor for compensation paid to the sufering
party. However, it follows from the main rule of the Act sect. 4–3 that the insurer is bound by clauses exclud-
ing the tortfeasor’s liability to the same extent as the clause can be maintained by the tortfeasor against the
sufering party. The rules therefore establish the subrogation principle: the insurer steps into the sufering
party’s legal position against the tortfeasor. Against this backdrop, the position of the tortfeasor according
to a property insurance should depend on whether his exclusion of liability as part of the knock-for-knock
regime can be upheld. So far, the situation resembles the issues of censorship discussed in section 5above.
However, since the contract requires the parties to obtain insurances where the insurer waives any right of
subrogation against the other party, we see that the “reasonableness” – and thus, validity – of the clauses
excluding liability for property damage depends on whether this waiver is binding in all situations. Admit-
tedly, the liability clauses are the point of departure, but the fnal allocation of the loss may depend not only
on the allocation contemplated in the contract, but also on the scope of the waiver of subrogation contained
in the insurance terms.
101 Two examples of such clauses: 1. “It is . . . agreed that all rights of subrogation hereunder shall
be waived as respects the Assureds hereon . . ., whenever contracts executed by the Assureds require such
waiver of subrogation, provided such waiver is agreed in writing prior to a loss as may be covered hereon.”
2. “Underwriters agree to waive rights of subrogation against any Assured and any person, company or
corporation whose interest are covered by this policy.”
148
of recourse to the same extent as the insured party has a claim under the policy.102
However, this overkill paves the way for the following notes on insurance law.
According to insurance law, the rights of a co-insured party are subject to some
distinct limitations, no matter the seemingly unreserved wording of the waiver and
co-insurance clauses. The insurer cannot waive his right of subrogation beyond the
restrictions imposed by rules of compulsory background law. According to the Act
relating to Insurance Contracts (No. 69/1989) sect. 4–9 frst para., the insurer is not
liable if the insured has intentionally brought about the insurance event. This implies
that the insurer is not prevented from holding liable a co-insured party who has caused
the damage intentionally, despite any agreed waiver of subrogation.
Further, it follows from sect. 4–9 second para. of the Act that the liability of the
insurer may be reduced or cease to exist if the insured has brought about the insurance
event by being grossly negligent. This applies, however, only to an insurance other
than liability insurance, that is to property damage insurance. The main rule of the
Act is therefore that gross negligence on the part of the insured does not relieve the
insurer from the obligation to cover the damage so caused. In the case of property
damage insurance, it depends on the degree of blame and the circumstances in general
whether the insurer is liable to cover the damage partly or in total.
6.3.3 Are the arrangements “reasonable”?

This normal system of the Act carries extensive weight when evaluating whether the
insurance system established by the contract is “reasonable”. However, this necessitates
deciding if the insurances required by the contract shall be labeled liability or property
damage insurances. Only if labeled liability insurance, the Act itself provides sufcient
basis for enforcing the clauses on waiver (and co-insurance).
The classifcation of the insurances is less obvious than one would expect. The
question may be illustrated by the situation where the contractor’s management by
acting grossly negligent has caused damage to the contract object prior to delivery. The
contractor shall carry out necessary measures to ensure that the work is completed
in accordance with the contract (NTK 15. 29.1), but his costs in doing so are limited
to an individually agreed maximum amount, inter alia, if the damage is covered by
the company’s Construction All Risks (CAR) insurance (art. 29.2). In his capacity as
co-insured under this policy, the contractor has an independent right of coverage.103
If the CAR, for this purpose, is considered a liability type of insurance, there can be
no room for rejecting the contractor’s claim for coverage referring to the Contracts
Act sect. 36. If the CAR is categorized diferently, the situation might be diferent.
Several approaches may be taken to this question. One may base the discussion
on the fact that the company owns the contract object as it comes into being (art.
22.1), implying that the contractor has to compensate the company’s loss caused to
102 Bull (note 30) pp. 387–388 with further references. It is, however, possible that the waiver of subroga-
tion clause in some situations will provide a wider protection than the one following from the position as
insured. See Ness (note 85) p. 86 f.
103 The issue in this situation is therefore not the waiver of subrogation – there is no claim form the
company that the contractor may step into. Hagstrøm, Ansvarsfraskrivelse appears to take a diferent view
at p. 476 f., especially pp. 87–491.
149
knut kaasen
its property – i.e., the insurance works as liability insurance. However, even under
this view, the underlying reality is that the contractor carries the risk of the contract
object until delivery; the right of ownership is not relevant in this context.104
6.3.4 The conclusion

The answer is not certain, and it is not necessarily decisive as to how the insurance
is categorized since the contract provisions on insurance do not have to be set aside
based on the Contracts Act sect. 4–9 (except in case of willful misconduct on the
part of the contractor’s management). The decisive question is whether upholding
the required waiver (and co-insurance) is unreasonable under the Contracts Act sect.
36 test in cases where the contractor has caused the insured event by his own man-
agement’s gross negligence.105
When evaluating reasonableness, it carries great weight that the contract’s provi-
sions on liability and insurance stand out as rational and appropriate – commercially
and legally. Coordinated insurance solutions imply that losses within the sphere of
the contract are covered irrespective of who has bought insurance, who has caused
the damage or loss and – at least apparently – whether the insurance in question is
categorized as liability or property damage insurance. Therefore, great weight should
be attached to the fact that the insurer has undertaken an unconditional obligation
to exclude the contractor from claims of subrogation and name him as co-insured. It
is also important that this arrangement is part of a well-known and well-established
contractual liability regime where the coverage at least in reality can be said to take the
form of liability insurance (having consequences under the Contracts Act mentioned
above). This system deserves acceptance on the additional basis that property damage
insurance containing a waiver of subrogation against a party who is also co-insured
under the policy, contains an element strongly resembling liability insurance.106
7 Do the NTK 15 provisions on exclusion and limitation of liability hold good?

The answer depends on whether the result emerges as “unreasonable” under the test
based on the Contracts Act sect. 36. This requires a complex evaluation of a number
of elements. The fundamental task of interpreting the contract provisions ofers few
challenges. There are no indications that the provisions are meant to imply less wide-
ranging exclusions and limitations than allowed under the background rules of law.
On the contrary, in relation to the knock-for-knock regime, the contract specifcally
provides that the exclusions and indemnifcations apply regardless of any form of
liability, whether strict or by negligence, in whatever form, and without any distinc-
tion regarding how, or by, or to whom the damage is caused.
104 The title is regulated for a diferent purpose, being a wish to pave the way for the company’s right to
take delivery of the contract object (in its present state) in competition with the contractor’s bankruptcy estate.
105 In this context, it may be of relevance that the “sliding scale” introduced by sect. 4–9 para. 2 opens
for a more fexible approach in the handling of grave instances than does the black or white approach result-
ing from categorizing the insurance as a liability insurance. It may seem ofensive to allow the contractor
full coverage in cases of qualifed negligence. On the other hand, all cases of damage except those caused by
management willful misconduct is covered under a liability insurance.
106 Ness (note 85) p. 109.
150
Important issues when deciding whether to censor these provisions have tra-
ditionally been who has caused the loss (management, subordinate employees
or others), what kind of negligence was involved (willful misconduct, gross or
ordinary negligence), the party’s insight into the consequences of his actions, and
what kind of losses are subject to limitation or exclusion (especially direct/indirect
losses). On this basis, it is possible to derive certain clear rules. The contractor
cannot rely on clauses limiting or excluding his liability for losses caused by his
management’s willful misconduct if the conduct can be categorized as a reckless
disregard for the company’s interests. On the other hand, he may normally limit
or exclude liability for losses caused by his management’s ordinary negligence or
even by gross negligence on the part of his other employees. Further, when decid-
ing whether the contractor may also limit or exclude liability for loss caused by
his management’s (i.e., his own) gross negligence, additional factors need to be
considered in order to determine the reasonableness. The help ofered by case law
is, in my opinion, limited in this respect. Nor does legal theory provide a clear
foothold, though it must be said that it, generally, is skeptical to limitation or
exclusion of liability in these situations. Both case law and legal theory identify
elements that must be considered relevant in evaluating the issue. However, this is
of limited help as long as neither of the sources discuss which implications follow
from these identifed elements operating in combination, in the way illustrated by
the complex construction contracts.
Four aspects appear central in this context – separately and especially combined.
First, the NTK 15 is an “agreed document”, created by highly professional parties
devoting huge resources to the task. Most of the contents – and all of the issues
discussed here – have remained unchanged through multiple revisions over a period
exceeding 30 years. We are far of from unbalanced, one-sided standards of shady
origins and based on doubtful legal understanding. And if there were to be a difer-
ence between the parties in terms of economic strength, it typically would imply that
the party that would have an interest in censoring the clauses limiting or excluding
liability would be the stronger party – the company – although market develop-
ments in recent years ofers examples of the opposite. Second, the clauses limiting
and excluding liability are, in principle, mutual, albeit the contractor is the one who
is – for practical reasons – the more likely benefciary of the system. We do not
look favorably upon the one-sided type of regimes that typically call for scrutiny by
the legal system. Third, the clauses limiting and excluding liability must be viewed
in connection with the insurance regime established (not just contemplated) by the
contract. The discussion of reasonableness must take the combined efect of these
regimes into account. This is markedly diferent from situations where there is just a
more or less relevant possibility of buying insurance coverage, which is not refected
in the contract’s risk regime and even less made an integral part of it. Finally, the
liability regime appears as a necessary part of a thoroughly prepared totality that seems
commercially sensible: an unlimited exposure to liability does not make commercial
sense in contractual relations of this type and size. Channeling the risk to insurance
appears to be a rational approach, and it is a beneft that the need for evaluating
causality and individual loss is reduced. The limitation and exclusion of liability are
parts of a complex and complete system, where it must be assumed – based on the
151
knut kaasen
history of the establishing of the system – that the parties internally have balanced
the pros and cons of each of the elements and of the totality.
Even though the starting point in more classical contractual relations is that a party
cannot rely on clauses limiting and excluding his liability for loss that he has caused
by gross negligence, the aforementioned characteristics of complex construction
contracts like NTK 15 are, in my mind, clearly convincing arguments for accepting
such clauses in such contracts. This view is applicable to both of the main forms
of limitation of liability; limiting and exclusion of liability for breach, and distribu-
tion of risk for losses caused by damage to the contract object or other property in
connection with the contract work. However, in the latter context the arguments are
even stronger: both the reciprocity and the direct link to the insurance come to the
forefront, and the knock-for-knock system stands out as a well-founded rational way
of handling this type of risks in this type of contracts.
Following this, my view is that the clauses of NTK 15 and similar contracts limit-
ing or excluding liability can also be maintained when the breach, damage or loss is
caused by the contractor’s own gross negligence. If the limitations or exclusions are not
to be respected, it must be because it is proven that the breach, damage or loss was
caused by willful misconduct on the part of the leading personnel of the contractor,
knowing that the consequence will be a reckless disregard of the company’s interests.
152
CHAPTER 7B
The obligation under Norwegian ofshore contracts to insure

the contract object as part of a knock-for-knock agreement
Erik Brannsten
1 Introduction and background

This chapter concerns the main features of the oil company’s obligation to insure
the contract object under an ofshore construction contract compared to the available
insurances in the market. Ofshore construction insurances are normally taken out
on a project basis to support a project overall knock-for-knock arrangement, such
as the knock-for-knock provisions found in the Norwegian ofshore contracts’ arts.
29 and 30, used for construction of production facilities in connection with an of-
shore development project on the Norwegian Continental Shelf (NCS).
The ofshore contracts most commonly used for this purpose are the Norwegian
Fabrication Contract 2015 (NF), the Norwegian Total Contract 2015 (NTK) and the
Norwegian Subsea Contract 2005 (NSC). Throughout this article, NF/NTK and NSC
will be referred to as the “ofshore contracts”. The parties to the contracts are the
“Company”1 and a “Contractor”, (see section NF/NTK art. 1.23/1.24 and NF/NTK
art. 1.15). In 2016, a new set of standard purchase conditions were also established,
referred to as “Norske Innkjøpsbetingelser/Norwegian conditions for purchase” NIB
16. These are often used for subcontracts for the purchase of materials and equipment.
Development of ofshore oil and gas production on the NCS is a complex and
hazardous task. Construction of production platforms and subsea installations takes
many years and involves large investments, technical challenges and a high risk for
loss and damage to personnel and property. When concluding an ofshore contract,
a Company’s main objective is to receive timely delivery of the contract object2
which is suitable for the purpose that the Company intends to use it for. Common
for the contracts is that the Company will become the owner of the contract object,
and that title will incrementally pass to the Company as the work progresses (see
art. 22). Should loss or damage to the contract object occur during construction, it
is crucial for the Company to be able to fnance, repair the damage and complete
the works within the delivery date. Such fnancing is one of the main purposes of
the project insurance.
1 The Company will normally be a group of companies which together holds a production license and
are thus also often referred to as a licensee. One of the licensees will be nominated as the operator and shall
take care of the operational activities of the license group.
2 The term ‘Contract Object’ is defned in NF/NTK 15 art. 1.12 and NSC 05 art. 1 (i) and is a reference
to the platform, modules, pipeline etc. which the contractor shall construct and deliver, and/or install. In
NTK 15 the contract object and the documentation are together defned as ‘Deliverables’; see art. 1.14.
DOI: 10.4324/9781003206798-9 153

erik brannsten
An ofshore project will normally be divided into several parts, each covered
by contracts with work executed in sequence and/or in parallel, with interfaces
between each other. Under each major project contract, there will be a value chain
forming a pyramid of subcontractors, sub-suppliers, sub-subcontractors and so on.
Subcontracts are often entered into on “back-to-back”3 terms, so that the rights
and obligations between company and contractor are refected in every subcontract
tier. As we will come back to, there is a strong incentive for the Company to take
out one project insurance where all contractors and suppliers are co-insured, which,
in turn, will from an insurance perspective remove the interfaces established by the
various project contracts.
1.1 Indemnifcation, liability and insurance

As pointed out above, construction activities and execution of a development project
will involve a risk for loss and or damage to property and/or personnel, as well as
pollution and damage to the environment. This risk stems from the complexity, value,
level of activity, and duration of such projects.
A loss or damage could be the result of negligence or another legal basis for liability,
such as the Company’s liability under § 10–9 of the Petroleum Act.4 In addition,
due to the material values and number of persons involved, any accident has the
potential to incur a loss greater than a company can manage fnancially, which risks
the completion of the project on time, should the contractor go bankrupt.
In addition to the incentives mentioned, due to the inherent risks, the Petroleum
Regulation5 § 73 also requires that the Company procures certain insurances related
to its activities:
Section 73. Insurance
The activities conducted by the licensee pursuant to the Act Chapters 3 and 4 shall be
insured at all times. The insurance must at least cover:
a) damage to facilities,
b) pollution damage and other liability towards third parties,
c) wreck removal and cleanup as a result of accidents,
d) insurance of the licensee’s own employees who are engaged in the activities.
The licensee shall ensure that contractors and subcontractors engaged in the activities
take out insurance for their employees to the same extent as the operator insures his own
employees.
When taking out insurance as mentioned in the frst paragraph literas a) to c), the
licensee shall provide reasonable insurance cover, taking into consideration risk exposure
and premium costs. Insurance as mentioned under litera d) shall be taken out as further
agreed with the organizations of the employees.
The Ministry may consent to the licensee using another form of security arrangement.
At the end of each calendar year, the licensee shall inform the Ministry about existing
insurance agreements, with an indication of the main terms. The Ministry may require
further insurance to be taken out.
3 See Knut Kaasen Tilvirkningskontrakter med kommentarer til NTK 15 og NF 15, Universitetsforlaget
2018, Kaasen 2018 (‘Kaasen 2018’) pp. 39 and 235 with further references.
4 Lov om Petroleumsvirksomhet av 29 November 1996 nr. 72 (Petroleum Act).
5 Forskrift til lov om petroleumsvirksomhet 27 June 1997 nr. 653 (Petroleum Regulation).
154
the obligation under norwegian offshore contracts
As indicated above, the Company’s legal insurance obligations are extensive, and the
Company has a strong incentive to procure such insurances at minimal cost while at
the same time fulflling its legal obligations. To achieve this, the Company should
avoid overlapping insurances within the project and reduce the amount of liability
insurance.
On this background the standard ofshore contracts contain reciprocal indemnities
and waivers of liability which in detail clarify how personnel and property losses
shall be distributed between the participants in the project, rather than containing
provisions distributing potential liability. The knock-for-knock arrangement intends to
displace ordinary tort liability and will thus deviate from the rules of liability found
in the Norwegian background law. This way each project participant will be able to
calculate its needs for insurance coverage and procure its insurances accordingly. Such
arrangement is often referred to as a “knock-for-knock” agreement.
1.2 Project all risk insurances

The insurances used in connection with a knock-for-knock agreement are commonly
known as a Project/Construction/Contractor/Builders All Risk (CAR/BAR) policy.
A main feature of such insurances is that they cover the whole project as such on
an all-risk basis. An all-risk CAR insurance represents the best cover that the insur-
ance market can ofer and covers all causes for damage, unless an exception is stated
in the policy.
There are several advantages of this approach. First, the total insurance costs over
the project period will be reduced. If all involved parties and participants insured
their liability, personnel, and property, then there would be extraneous overlapping
coverage. This is unnecessarily expensive and could increase the number of insur-
ance disputes in the project, since it would require discussion between insurers about
which insurance policy would cover a loss. Second, when determining where the loss
or damage has occurred, claim settlement becomes more efcient, as time-consuming
discussions and expensive examinations of liability and causation will be avoided.
Traditionally, project all-risk insurances are taken out through the London market
through use of international insurance brokers such as Aon, Marsh and Willis Tow-
ers Watson. A standard set of insurance terms which are often used as a starting
point is the WELCAR (Ofshore Construction Project Insurance Wording), which
was developed in 2001.6 WELCAR 2012 wording was fnalised and released for
consultation in the market, but work stopped and a new version has still not been
fnalised.7 WELCAR was originally developed under English law, however, when
used for projects on the NCS the choice of law is changed to Norwegian law, which
means that practices evolved under English law may be relevant when applying the
policy in Norway.
6 See Kaasen 2018, note 3 above, pp. 774–776.

7 Paul Wordley & Jonathan Bruce, ‘Revision to the ofshore construction policy’ (July 2012) Middle East
Insurance Review at www.hfw.com/downloads/HFW-MEIR%20Article%20-%20Ofshore%20Construc-
tion%20Policy%20[A4%204pp]%20August%202012.pdf.
155
erik brannsten
2 The knock-for-knock arrangement
2.1 Risk zones

A knock-for-knock agreement implies that each party must carry the loss or damage
which strikes his own so-called risk zone. In broad terms the four risk zones normally
used in the ofshore contracts are:
Art. 29 – the contract object and materials prior to being delivered;
Art. 30.1 – the property and personnel of Contractor group;8
Art. 30.2 – the property (except for the contract object prior to delivery) and
personnel of Company group;9
Art. 30.3 – genuine third parties which are not involved in the project.
It is a condition under the ofshore contracts that the legal entity or person of the
Contractor group is “participating in the Work” and that the legal entity or person
of the Company group is “involved in the [feld name] project”. This requirement
corresponds with the scope of the insurances, ref. below.
For the ofshore contracts, the Company group defnition is based on a so-called
extended family, where the “group” includes the Company, his afliates, other licence
group members, their employees and so on, and the Company’s other contractors, their
subcontractors, employees and so on. This as opposed to a so-called nuclear family solu-
tion, where the Company will only indemnify for losses sufered by himself, his afliates,
other licence group members, their ofcers, employees and so on. The efect of using
the extended group defnition is that the genuine third-party group is correspondingly
reduced to include only parties which are not part of the project, such as fshermen, ofcial
inspectors, guests and visitors. This will in turn reduce the need for liability insurance.
2.2 Waivers of liability, indemnifcation and waiver of subrogation

To make sure the knock-for-knock agreement will work, the arrangement must further
include:
1 That the parties waive their right to hold the responsible (or causing) party
liable for losses or damage sufered;
2 That they also waive their right to subrogation in case they have been held
liable by a third party within their own group;
3 That a party must indemnify the other (causing) party for loss or damage
sufered by anyone within his own group, in case such third party directs his
claim against the other (causing) party.10
2.3 Damage to the contract object

With the high focus on safety in the industry, losses or damage to property and
personnel rarely occur. It is more likely that the contract object is damaged as part
8 Defned in NF/NTK art. 1.16.

9 Defned in NF/NTK art. 1.21.
10 The system is illustrated in Kaasen 2018, note 3 above, pp. 770–772.
156
of the construction process or during installation, testing activities and so on. As
mentioned above, the contract object constitutes a separate risk zone, and the risk
for loss or damage passes from Contractor to Company upon delivery according to
the ofshore contracts (art. 19.1).11 Delivery occurs when the delivery protocol is
concluded, which may not necessarily correspond to the Company’s physical takeover
of the contract object.
As mentioned above, it is the Company’s main objective to receive timely delivery
of the contract object. Should a loss or damage to the contract object occur, it is
therefore agreed in art. 29.1 of the ofshore contracts that the Contractor will imme-
diately begin to repair, rebuild, and complete the contract object – and in the case
of a total loss, even start the construction of a new contract object. This obligation
applies regardless of the cause of the loss or damage.
A well-known total loss incident on the NCS happened when the substructure for
the Sleipner A platform collapsed during testing on 23 August 1991. The occurrence
was covered by insurance and eventually led to a massive insurance and subrogation
claim.12 Since the collapse happened before delivery, the occurrence was covered by
art. 29.1 of the contract and the Contractor had to build and complete a new platform
sub-structure. Completion of the Sleipner A platform was essential for Norwegian
gas sellers to commence export of gas to the European market in 1993 according to
the Troll gas sales agreements.13
Distribution of economic responsibilities between the Company and the Contrac-
tor is regulated by art. 29.2. As a starting point, the Contractor is economically
responsible for the repair or replacement costs, unless the loss or damage was caused
by someone in the Company group or by an event which it is not possible to insure
against under a CAR policy. Art. 29.2 specifcally mentions nuclear damage, war and
terrorism, which would require separate and more expensive insurances. Furthermore,
Contractor’s fnancial liability for repair and replacement costs will be limited to a
certain amount, often corresponding to the deductibles under the CAR insurance
or as agreed between the parties. A precondition for this limitation is that the CAR
covers the loss or damage in questions. If the policy does not provide cover, for
instance because the Contractor wilfully has caused the damage, the said limitation
will not apply. However, the limitation will provide protection if the policy does
not apply, and this is the “result of circumstances for which Company carries the
risk”, for instance if the Company has failed to pay the insurance premium. The
latter may also be the case if the policy provides sub-limits or other carve-outs and
limitations which the Company has chosen, unless the ofshore contract stipulates
that the Contractor will be liable for such costs which the policy does not cover. In
this respect, it is also important that the Contractor was able to review the policy
prior to calculating and submitting his tender (see the principle applicable for frame
11 See art. 29.1. For sale of goods the same follows from Section 13 frst paragraph of the Sale of Goods
Act.
12 Wikipedia, ‘Sleipner A’ (19 August 2021) at https://en.wikipedia.org/wiki/Sleipner_A.
13 See Christian Lund: “Sleipner Øst – Fra totalhavari til Nasjonal dugnad” in Industribygging og Rett-
sutvikling, Juridisk festskrift i anledning Norsk Hydros 100 års jubileum, pp. 91–103.
157
erik brannsten
agreements in art. 8.4). This is also often done by using a “dummy” policy at the
tender stage (see section 3.1).
2.4 Waivers and indemnities shall apply regardless of cause

The waivers and indemnities of the knock-for-knock arrangement shall in principle
apply regardless of how the loss or damage was caused. This is often expressed such
that the indemnity “applies regardless of any form of liability, whether strict or by
negligence, in whatever form, on the part of Company Group.”14
In general, Norwegian courts are likely to accept limitations of liability and waivers
which are part of “agreed documents” negotiated by equal parties, which are recip-
rocal, clear and predictable and which refect a deliberate allocation of risk backed
up by available insurances.15 Moreover, it is not likely that the agreed limitation
will be upheld in case of wilful misconduct (forsett). Nevertheless, it is still an open
question as to how the Norwegian Supreme Court will rule if a loss or damage has
been caused by gross negligence (grov uaktsomhet) by the responsible party himself
or anyone belonging to his higher management. An example from the courts is found
in a judgment by the Gulating Court of Appeal from 2014.16 The case concerned
the distribution of liability following a collision between the FSO17 vessel Njord B,
owned by the Njord licensees, and the ofoading tanker Navion Hispania, owned
by Teekay and chartered by Statoil ASA (which was not participating in the Njord
licence). The Navion Hispanie was chartered to lift and transport Njord oil which
Statoil had purchased. In this case the court found that the shipowner Teekay had
caused the damage to the Njord B vessel through gross negligence and therefore
concluded that the limitation of liability was suspended vis-à-vis Teekay, who became
responsible towards the Njord licensees for “any loss”.
3 The ofshore project insurances
3.1 Introduction
As discussed previously, the agreed indemnities (NTK 15 arts. 31.1 and 31.2) clarify
which insurances related to the construction project that the Company must procure
and maintain, and which insurances Contractor shall purchase and maintain, through-
out the project.18
According to the agreed distribution of risk and losses, the Company’s insurance
policy shall include
14 See NTK art. 30.1 para. 1 i.f. and the reciprocal provision in art. 30.2 for the indemnity provided by
Company.
15 See the Supreme Court decision in Rt. 1994 p. 626 which upheld a limitation of liability in relation to
a damage caused by gross negligence of an employee of the forwarding agent against whom the claim was
directed. The employee was not considered part of the management of the forwarding agent. See also the
discussion in Kaasen 2018, pp. 776–785.
16 Case 2013–06–25. LG-2012–077280.
17 Floating Storage and Ofoading vessel (lagerskip).
18 See also Section 73 of the Petroleum Regulations (F27.06.1997 nr 653).
158
1 Coverage for physical damage to the contract object, materials, Company’s
materials;
2 Transportation insurance (art. 31.1 (b));
3 Insurance for (so-called genuine) third-party liability up to NOK 500 million
(art. 31.1 (c)).
Accordingly, the Contractor is obliged to provide insurance cover for his own person-
nel, property and liability up to certain limits. The Company’s CAR insurance will
often provide an exclusion for the items or costs which, according to the ofshore
contract, fall within the Contractor’s obligation to insure.19 Contractor’s obligations
are as:
1 All risks, hull, and machinery insurance for vessels;
2 P&I insurance, including oil pollution insurance for vessels;
3 Liability insurance covering Contractor’s liability for damage to property
and personal injury under art. 30.3 for (so-called genuine) third-party liability,
to the extent such liability has not been covered by the insurance for vessels
(ref. above);
4 Personnel insurance which shall cover losses connected with illness, personal
injury, or accidental death in Contractor Group.
The insurance cover described above is closely linked to the various risk zones of
the ofshore contracts as mentioned in section 2.1.
Company’s CAR insurance refects the contract requirements in art. 31.1 and
usually has two sections:
1 Section I contains the construction all risk and transportation.
2 Section II contains third-party liability insurance.
It is common to submit a draft version of the CAR Project Insurance in the tender
documents as part of a competition so that the bidders can evaluate what the Com-
pany’s policy covers, what the deductibles are, the exclusions, limitations, carve-outs,
duration and so forth. Deductible levels can vary greatly depending on market condi-
tions, risks in the project, where the damage occurs and so on. Damage to a subsea
cable discovered before transportation and installation will often have a much lower
deductible compared to damage occurring during installation or testing of the cable
after it has been installed. It is important for the Contractor to be aware of these
limitations so that he can evaluate the need to buy additional insurance.
The same applies for the duration of the CAR. The intention is that the insurance
shall be valid from start of the work until the acceptance certifcate is issued, which
means that the CAR will be efective even during the guarantee period of two years
from delivery (in the policy this will often correspond to the maintenance period).
However, in the NF/NTK 15 versions the duration is linked to the actual duration
19 For instance, as follows: “Section I of this Policy will not respond in respect of: Loss of or damage
to contractor’s, sub-contractors’ and suppliers manufacturing facilities, also excluded hereunder are tools
or equipment intended for use in other projects’ works, unless the above-mentioned items are specifcally
identifed and the value is included in the declared Estimated Contract Price or specifed with a sub-limit
in the Schedule.”
159
erik brannsten
of the policy which can be shorter, for instance, if the Company decides to only
buy one year of discovery and maintenance. This will also typically be something a
tenderer should check when calculating his bid. Further, according to art. 31.1 (a),
the Company procures a Project Insurance which shall include the Contractors as
co-insured and a waiver of subrogation from the insurers (see art. 31.1 third para.).
If the insurance company could subrogate against the causing party, the knock-for-
knock arrangement would not be efective (see section 2.6).
In the CAR policy, Company will be nominated as the “Principal Assureds” and
the Contractors and/or Sub-Contractors and/or Manufacturers and/or Suppliers as
“Other Assureds”, thus implementing the requirement of the ofshore contracts for
the Contractor Group to be co-insured.
Welcar 2001 also contains a waiver of subrogation which ensures that a loss will
be covered by the insurers and cannot be claimed from the party causing the loss
or damage:
Underwriters agree to waive rights of subrogation against any Principal Assured(s) and/
or Other Assured(s). This waiver is not to apply to an Assured in the event that the said
Assured commits an act of gross negligence or wilful misconduct.
The assureds will be covered for their own negligent acts.20 However, to make the
waiver of subrogation compliant with the requirements of the ofshore contracts, the
last sentence of the waiver of subrogation clause should be deleted. However, as
pointed out above, it is not likely that the waiver will be upheld for any damage
caused by willful misconduct of the assured. As mentioned in section 2.7, the situ-
ation is not clear for damage caused by gross negligence by the assured.
3.2 Insured perils

Welcar section 1 provides an all-risk coverage, expressed as follows:
Subject to the terms, conditions and exclusions herein, this Policy insures against all risks
of physical loss of and/or physical damage to the property covered hereunder, provided
such loss or damage arises from an Occurrence within the Policy Period set out in Item
3 of the Declarations. [emphasis added]
“All risk” implies that any cause of loss or damage is covered unless the insurer can
prove that there is an exclusion stated in the policy. There are very few exclusions
provided in Welcar, normally limited to damage due to war, nuclear damage, or ter-
ror. This is refected in NTK 15 art. 29.2, frst para. However, there are buy-back
options available. In addition, there are limitations, sub-limits and exclusion clauses
to be aware of.21 Insurance is taken out for property located anywhere in the world,22
whether under construction or not. Further, such property must be part of Company’s
20 See also Skadeserstatningsloven (Skl) § 4–3 and Forsikringsavtaleloven (Fal) § 4–9.

21 Two types of exclusions for defective design or workmanship have been developed and are commonly
used in CAR policies: (1) the Design Exclusions (DE) written by the London CAR Group and (2) the Lon-
don Engineering Group (LEG) exclusions.
22 For certain infrastructure projects in Norway, it is common that the CAR policy is geographically
limited to Europe.
160
investment and included in the declared contract values stated in Section B of the
policy.
Property covered will include the contract object, goods, materials, equipment,
temporary works and so forth. Welcar expresses the cover as follows:
This insurance covers works executed anywhere in the world in the performance of all
contracts relating to the Project including (provided they are included in the contract
values declared to Underwriters and insured herein) materials, components, parts, machin-
ery, fxtures, equipment and any other property destined to become a part of the completed
project or used up or consumed in the completion of the project. This insurance shall
also cover (provided they are declared to and agreed by Underwriters) all temporary
works, plant, equipment, machinery, materials, outfts, and all property associated there-
with, whether such items are intended to form a permanent part of the works or not,
including site preparatory work and subsequent operational risks.
It is understood and agreed that any insured equipment and/or property that is not
for incorporation into the contract works shall be covered whilst it is being utilized in
the Project and whilst in transit from the Project site(s) until the earlier of the date of
arrival at its fnal destination or the 30th day after its removal from the Project site(s).
The policy provides coverage for “physical loss of and/or physical damage” only.
Such loss or damage could typically be caused by fres, collapses, breakage, frac-
turing, water leakages, electrical short circuits, pollution, spills, fallen objects and
so on.
A general requirement is, of course, that the damage must be unexpected, sudden
and accidental and not be the result of an inherent feature of the materials applied.
If, for instance, an environmentally friendly paint is used on the platform and the
paint starts to peel after one year, as expected according to the specifcations of the
paint, then the situation is neither a guaranteed matter nor a matter for insurance.
Design errors by themselves are not covered by the CAR insurance. But if a
physical loss or damage has resulted from a design error, then such damage will be
covered. If, for instance, a load-bearing structure has been under-dimensioned, and
this is discovered before any sign of collapse or physical alteration has materialised,
then any replacement or reinforcement will not be covered by the CAR. On the other
hand, if the beam breaks, insurance will cover the repair costs.
The same is the case for construction errors. If a beam has been installed wrongly,
for instance, it was not properly levelled or the contractor used carbon steel when he
should have used stainless, then any repair or replacement work will not be covered
by insurance. If, however, the steel beam was not dimensioned properly and has suf-
fered cracks, then this will be a damage covered by the insurance.
Excluded from the standard coverage is the defective part itself. If, for instance,
the defective wiring of an electrical transformer in a machine has started a fre, then
it could be argued that the cost of the transformer will not be covered by insurance,
but the machine and its damaged surroundings will be covered. On the other hand,
the cost of the faulty part may not be big compared to mob/demob cost, access costs
and installation costs which will be incurred.
In many instances it will not be clear what constitutes the defective part itself
despite it being a defned term of the CAR. The defnition reads: “any part of the
subject matter insured which is or becomes defective and/or unft or unsuitable for
161
erik brannsten
its actual or intended purpose, whether by reason of faulty design, faulty materials,
faulty workmanship, a combination of one or more thereof or any other reason what-
soever.” It should be possible to obtain a faulty part buy-back, but this may be costly.
3.3 Basis for recovery

The ofshore contracts provide that the CAR insurance shall provide coverage for
physical damage to the contract object and materials, which is not very precise.
Sometimes it will be stated that the Company will provide insurance coverage accord-
ing to the attached draft insurance policy. To investigate which costs the policy actu-
ally covers, the CAR policy draft attached to the tender documents must be studied
carefully.
As a starting point, no coverage exists for costs incurred if no physical damage
has occurred (see above concerning pure design faults or errors in workmanship).
This is sometime expresses such that “any portion of the Property Insured shall not
be regarded as damaged solely by virtue of the existence of any defect of material
workmanship design plan or specifcation.”
Likewise, the insurance does not provide cover for costs incurred by the assureds to
improve the original design, plan, specifcation or materials. For instance, the insurance
incident may have revealed that a load-bearing structure was insufciently designed
and that a much more robust structure should have been used instead. Insurance will
not cover the gap between the cost of the original design and the cost of the reinforced
design. This is often expressed as follows: “The cost of any alterations, additions or
improvements shall not be recoverable under this Policy unless specifcally agreed
hereunder.” On the other hand, the policy may contain a clause covering increased
costs of unbuilt works, where “the cost of the permanent or temporary works not
commenced at the date of the physical loss or damage shall exceed the cost which
would have been incurred but for the physical loss or damage.” This would typically
cover extra costs incurred if later work would need to be performed out of sequence,
during winter instead of summer and so forth. Further, the CAR insurance does not
cover maintenance costs or costs due to gradual deterioration.
When the CAR insurance responds to a physical loss or damage, the basic cost
recovery principle for items repaired or replaced is “new for old”, expressed such that
the assured will be compensated “the cost of repairing or replacing or re-erecting the
property damaged with new materials of like kind and quality.”
In case of a total loss the recoverable amount will be “the cost of reconstruction
or replacement of the Property Insured by property substantially the same as, but
not more extensive than, that lost or destroyed.” The replacement or reconstruction
cost will also apply where the cost of repair is higher.
The policy will also respond where damage is not repaired, and the recoverable
cost will be “the actual value of the property immediately before the occurrence.”
Certain costs related to the loss or damage will also be covered such as costs of
extra testing, leak search costs, frefghting expenses, wreck and debris removal and
clean up, cost of temporary repair, storage, access costs and so on. Nevertheless, for
any single physical damage, recovery will normally be capped at 125% of the latest
insured values which are given in Schedule B of the CAR policy.
162
4 Summary
As seen in this chapter, the Company has extensive insurance obligations which are
sought to be fulflled by the taking of project all-risk insurances. The insurance policy
is a complex document which is not clear on all aspects. In addition, some questions,
for instance, regarding limitation of liability have not been solved in background law.
Even though the Company shall insure the contract object, the content of the insur-
ance policy and the background law determines the actual cover and protection
extended to the Contractor. Thus, the Contractors will need to investigate thoroughly
what the limitations, sub-limits, deductibles, durations and exclusions of the CAR
policy are, to assess its own risk exposure and its potential need to procure additional
insurance.
163
CHAPTER 8
Liability and insurance clauses in contracts for

vessel services in the Norwegian ofshore sector
The knock-for-knock principle
1 Introduction
This chapter covers the Norwegian ofshore sector’s liability and insurance clauses
in contracts for vessel services. In this chapter, “contracts for vessel services” means
ofshore charter parties and contracts for drill and well services. The contracts referred
to in this context are the Supplytime 051 and the OLF2 Proposal “New Conditions
of Contract for Drilling and Well Services.”3
A main feature of such liability and insurance clauses is that they establish a systematic
liability and insurance system in all contracts that involve a particular project, whereby the
risk of damage is fnanced by insurance efected by the contractual party sustaining the
damage. The principle that damage should stay where it occurs is called the knock-for-
knock principle. Though the knock-for-knock system frst and foremost regulates damage
to the contractual parties, the liability system may also regulate damage to third parties.
The knock-for-knock principle is easy to establish in the contractual relationship
between two parties. The parties to a contract are free to regulate the risk of causing
damage to each other, including both limiting liability for damage caused to the other
party and waiving the right to claim damages from the other party. However, this free-
dom of contract only applies to agreements between the parties concerning damage to
economic interests held by the two parties, that is to say, damage to their goods, loss
of income and other losses sustained by those parties. The parties to the contract do
not have contractual freedom to regulate the tort position of an injured third party.
This position is regulated by tort law governing the relationship between the injurer and
the victim, and this tort law position cannot be departed from by a contract to which
the victim is not a party. If the knock-for-knock principle is to be extended to apply
to damage to personnel employed by the contractual parties, this must therefore be
achieved through indemnity and subrogation clauses. The same is true if the principle
is to be extended to apply to subcontractors or other cooperative parties on each side
1 Supplytime 2017 Time Charter for Ofshore Service Vessels (Supplytime).

2 Oljeindustriens Landsforening, today Norsk olje&gass.
3 OLF Proposal ‘New Conditions of Contract for Drilling & Well Services’ (Approved by the Develop-
ment & Operations Committee 26 February 2003) www.norskoljeoggass.no/contentassets/5cd867185b5c4
7fc85720d3d96a6f399/drilling-and-well-services.pdf. Some references will also be made to the Norwegian
Fabrication Contract 2015 (NF), which contain a similar regulation, cf. www.norskindustri.no/dokumenter/
leveringsbetingelser/nfntk-standardkontrakter/.
164 DOI: 10.4324/9781003206798-10

liability and insurance clauses in norwegian offshore sector
of the contract. The regulation of liability in the knock-for-knock principle is therefore
constructed partly as liability clauses and partly as indemnity and recourse clauses.
Much has been written about the knock-for-knock principle in Norwegian law.4
The main focus of this chapter is to discuss the validity of the clauses. In order to
do that, however, it is necessary to provide an overview of the main content and
structure of the clauses, as well as the rationale behind them. Furthermore, a brief
presentation of the relevant tort law and insurance law framework is also necessary.
2 Overview of the relevant tort law and insurance legislation
2.1 Tort law

According to Norwegian tort law, a person is liable for damage to another person
if three conditions are fulflled: there must be a legal basis for liability, there must
be an economic loss, and there must be legally relevant causation between the act or
omission of the injurer and the loss. These conditions are mainly developed in court
practice and are not stated in general legislation.
There are three main types of the basis for liability: liability for negligence, strict
liability, and vicarious liability. The main rule is that negligence is required to invoke
liability for damages. Strict liability will normally require legislative action. Vicarious
liability is regulated by the Compensation Act5 § 2–1, which states that an employer is
liable for any damage caused by negligence or a deliberate act from their employees.
This is therefore a combination of strict liability and negligence.
However, in several areas there is legislation providing for non-fault liability. Of
interest here is the regulation of petroleum and maritime legislation.
The Petroleum Act6 § 10–9 provides for extended liability of the licensees operating
on the Norwegian Continental Shelf:
If liability in respect of a third party is incurred by anyone undertaking tasks for a
licensee, the licensee shall be liable for damages to the same extent as, and jointly and
severally with, the perpetrator and, if applicable, his employer.
Thus, if a contractor performs drilling services for a licensee on the Norwegian

Continental Shelf and it causes damage while performing the services, the licensee
is jointly and severally liable with the contractor. The liability of the contractor is
regulated by ordinary tort law as outlined above.
Furthermore, both the Petroleum Act and the Maritime Code7 contain rules on strict
liability for pollution. A general feature of these rules is that the licensee or shipowner
has strict liability for pollution8 and that the liability is specifcally directed against these
persons so that claims may not be raised against other persons. Further the licensee or
4 Hans Jacob Bull, Tredjemannsdekninger i forsikringsforhold, Oslo 1988, Del IV, (Bull); Knut Kaasen,
Petroleumskontrakter, 2018, Del VIII (Kaasen); Monika Zak, Ansvarsregulering i borekontrakter – Gyldighets-
sensur i norsk, engelsk og amerikansk rett, Master Paper 19.05.2012 www.duo.uio.no/handle/10852/35075 (Zak).
5 Norwegian Compensation Act (Skadeserstatningsloven) 13 June 1969 no. 26.
6 Petroleum Act 1996 no. 72 (PA).
7 Maritime Code 1994 no. 39 (MC).
8 PA § 7–3, MC § 191.
165
trine-lise wilhelmsen
shipowner, having settled the claim, may not raise a subrogated claim against the persons
protected against claims, unless any such protected person acted deliberately or with gross
negligence.9 However, even if the charterer is protected against a pollution claim from a
third party, he is not protected against a claim for subrogation.10 The licensee, on the other
hand, may not claim recourse against anyone who by agreement with the licensee or his
contractors has performed tasks or work in connection with the petroleum activities.11
The starting point is therefore that the licensee and contractor are jointly liable
for damage caused during operations on the Norwegian Continental Shelf, and that
liability for pollution is specifcally channelled against the licensee and the shipowner.
2.2 Insurance law

Norwegian insurance contracts are governed by the Insurance Contract Act (ICA).12
ICA’s provisions are generally mandatory; however, there are several exceptions for
insurance relating to commercial activities.13 The relevant exclusions are for ships and
ofshore units and internationally transported goods, which includes transport to and
from the Norwegian Continental Shelf.14 As a starting point, therefore, the insurer
and parties to the charter or drilling contracts have full contractual freedom in rela-
tion to the content of the insurance contract.
Nordic marine insurance is regulated by the Nordic Marine Insurance Plan 2013
Version 2019 (NP).15 The NP is an agreed document which may be compared to
private legislation, and it is used for most marine insurance contracts in Norway. The
NP contains conditions for, inter alia, hull and loss of hire insurance for ocean-going
vessels and ofshore units. Hull insurance includes liability for collision, but ordinary
liability insurance is regulated by Protection and Indemnity insurance contracted in
the P&I Clubs. Of particular relevance here is that NP chapter 18, which regulates
insurance for mobile ofshore units, has special rules on subrogation and co-insurance
which are relevant for the knock-for-knock regulation.
3 The content and structure of the knock-for-knock principle
3.1 Type of loss and basis for liability

Both Supplytime and the OLF Proposal apply to the following groups of damage
and loss:
1 Damage to property, including the vessel;16
2 Personal injury or damage;17
9 PA § 7–4, MC § 193.
10 MC § 193 (c).
11 PA § 7–4 (a).
12 Act no. 69 of 16 June 1989 relating to insurance contracts.
13 ICA section 1–3 frst subparagraph.
14 ICA section 1–3 second subparagraphs (c), (d).
15 The Nordic Marine Insurance Plan of 2013, Version 2019, at www.nordicplan.org/The-Plan/.
16 Supplytime 14 (a) (i) and (ii), OFL proposal 8.1 (b) and 8 (2) b.
17 Supplytime 14 (a) (i) and (ii), OLF proposal 8.1 (a) and 8.2 (a).
166
3 Consequential damage;18
4 Pollution damage.19
In the OLF Proposal, the principle also applies to loss of or damage to in-hole
equipment, loss of hole, blowout, damage to reservoir and use of radioactive tools,
and infringement of patents/property rights.20
However, even if the core of the knock-for-knock principle is that each party
carries the damage that it has sustained, some losses will nonetheless be channelled
against one of the parties or divided between the two parties. This may be necessary
to comply with the mandatory pollution regulation.21
The knock-for-knock regulation applies regardless of the basis of liability that
may be invoked against the injurer, i.e., it applies to both strict liability and negli-
gence. This means that the regulation applies to liability which is based on either the
licensee’s contractor’s liability according to PA § 10–9, on strict liability for pollution
according to the petroleum or maritime regulation, or on negligence. Furthermore,
the knock-for-knock regulation applies to ordinary negligence, gross negligence, and
damage caused by intent. In addition, there is no distinction between faults made by
an employee and faults made by a company’s representative.
3.2 Who is included in the liability provisions – “the group concept”

The standard liability provisions frst and foremost apply to the parties to the con-
tract. In the charter party, this will be the “Owner”22 and the “Charterer”.23 In the
OLF proposal, the parties are the “Contractor”24 and the “Company”.25 However,
the standard liability provisions also apply to other parties.
Firstly, the provisions address both property damage and personal damage sufered
by persons employed by the parties to the contract. Secondly, the provisions apply not
only to the parties to the contract, but to others who are in a contractual relationship
with these parties. In the OLF Proposal, this follows directly from the clause’s wording,
where the Contractor and Company shall “indemnify” not each other, but instead the
“Company Group” and “Contractor Group”.26 In Supplytime, the result is more indirect
as the clause only states that the Owner and Charterer shall not be liable for damage to
the Charter Group/Owner Group, but where according to a Himalaya clause this freedom
from liability shall apply also to the beneft of other members of the respective Groups.27
18 Supplytime 14 (b) (ii), OLF proposal 8.10.

19 Supplytime 15 (a) and (b), OLF proposal 8.4 and 8.5.
20 OLF proposal 8.6–8.9.
21 According to OLF 8.4, the Company is liable for pollution from reservoir and property of the com-
pany, whereas the Owner according to Supplytime 15 is liable for all pollution from discharge, spills or leaks
from the vessel.
22 The Owner is the party stated in Box 2, Supplytime Part II Defnitions.
23 The Charterer is the party stated in Box 3, Supplytime Part II Defnitions.
24 The “Company” means X as operator on behalf of one or more Licence Groups as specifed; cf. OLF
proposal General conditions 1.1. Defnitions.
25 The Contractor is identifed by name, cf. OLF proposal General conditions 1.1. Defnitions. See
similarly NF Part 1 Art 1.14 (contractor) and 1.22 (Company).
26 OLF proposal 8.1 and 8.2, NF Art. 30.1 and 30.2.
27 Supplytime 14 (a) (i) and (ii); cf. 14 (d).
167
The “group concept” in this context is used to defne risk zones, which are zones for
which each party in the contract carry the tort risk. The point here is that the Owner/
Company or Charterer/Contractor not only agrees to be responsible for any damage
that befalls the company property or the property or persons of the employees, but
also assumes responsibility for such damage throughout the group.
The extent of the risk zones has varied over time,28 but both the Supplytime and
the OLF Proposal apply a broad group concept, or a “big family group” concept.29
Though the defnitions vary, the primary idea is that it consists of the contractual
party and the parties with whom they cooperate on a particular project, including
all the employees of such third parties.
The “Owner Group” means the “Owners”, “Owners’ Afliates”, “contractors and sub-
contractor” and “Employees of any of the foregoing”.30 The Company Group means the
Licensee Group, each of the participants herein, their afliated companies, the Company’s
other contractors and their contractors or subcontractors, the Company’s invitees, and
personnel employed in or engaged by the aforementioned corporate entities, and others
whose services are used by the Company.31 The Charterers Group is “the Charterers and
Charterers clients”. “Co-ventures of the foregoing”, “Afliates”, “contractors and sub-
contractors” and Employees of any of the foregoing”.32 The contractors Group is the
“Contractor, Contractor’s Afliated Companies participating in the Work, its Subcontrac-
tors and their contractors and subcontractors, participants in a joint venture or similar
partnership involved in the Work, Contractor’s invitees, and personnel employed in or
engaged by the aforementioned cooperate entities.33
In the contract between Owner/Company and Charterer/Contractor, the parties may

agree to waive the right to make a claim against a third party for liability. This could
be for damage to their property or person. This constitutes a promise not to make
any claim against the named third party. However, the parties may not, through their
contract, require other members of their respective groups to accept a similar waiver
of their rights. If matching waiver provisions are not included throughout the con-
tractual chain, a contractual party who has not agreed to a knock-for-knock provision
may refuse to accept responsibility for damage to his property and personnel and
may instead make a claim against the injurer. This will disrupt the system and can
easily lead to a chain of subrogated claims.34
To obtain an overall knock-for-knock governing provision in all the contracts relat-
ing to a particular project or work, the Company/Owner and Contractor/Charterer
must therefore include equivalent agreed liability terms in all their other contracts tied
to the same project or work and induce their contractual partners to do the same.
To the extent the same contractual terms are used throughout, this should secure
a consistent approach to liability. But such “back to back” regulation is not always
agreed. NF therefore includes a duty on the Company and Contractor to “ensure
28 Bull p. 333 f., Sofa Lazaridis, Maritime ofshore contracts Compendium, Sjørettsfondet 2011, p. 51.
29 Bull p. 347.
30 Supplytime Part II Defnitions.
31 OLF proposal 1.1 third paragraph. See similarly NF Art. 1.26.
32 Supplytime Part II Defnitions.
33 OLF proposal 1.1 sub-paragraph 10, see similarly NF Art. 1.16. Cf. further on the group principle in
drilling contracts Zak p. 21 f.
34 So-called loss carousel; cf. Bull p. 341.
168
as far as practicable that other companies” in the group agree to waive their right
in an equivalent way.35 Supplytime and the OLF Proposal do not include a duty to
secure equivalent provisions in other contracts.
The aim of the knock-for-knock principle is thus to allocate the risk for liability for all
the parties that are involved in the same project. But the project may also result in damage
or loss to a third party, that is to say a party outside the mentioned groups. This is not
regulated in Supplytime, but the OLF Proposal 8.3 contains the following provision:36
Subject to clause Article 8.4 – Pollution from reservoir and property of Company, Con-
tractor shall indemnify Company Group from and against any claim arising out of loss
or damage sufered by a Third Party in connection with the Work, to the extent that any
such loss or damage is caused by the negligence or breach of duty (whether statutory or
otherwise) of Contractor Group.
Subject to Article 8.5 – Pollution from Contractor’s property, Company shall indemnify
Contractor Group from and against any claim arising out of loss or damage sufered by a
Third Party in connection with the Work, to the extent that any such loss or damage is caused
by the negligence or breach of duty (whether statutory or otherwise) of Company Group.
These provisions mean that each party or group is liable for damage caused by their
own negligence. As a starting point this division of liability conforms to ordinary
contract law. However, as it follows from PA § 10–8 that the licensee (Company) is
severally and jointly liable with the contractor for all damage for which the contrac-
tor is liable, it means that this risk is transferred back to the contractor.
3.3 Freedom of liability, indemnity and subrogation

The starting point is that parties to the contract may only regulate their own duties
and rights in accordance with the contract. They are therefore free to limit their liabil-
ity against a contractual party, and waive their right to claim any liability in tort
from this party.37 As mentioned, they are also free to waive their right to claim any
liability from a third party and thus waive such right in regard to the whole group.
The parties to a contract may not, however, weaken the rights of a third party to a
contract. To the extent the third party is within the risk zones as defned in the con-
tracts, and that this party has included a similar knock-for-knock regulation in their
contract, this problem is solved through the matching contractual terms. However,
this restrictive approach also applies to third parties who do not have a contractual
relationship with the Company/Owner or the Contractor/Charterer. This will be the
case for all the employees in the groups, and for third parties outside the groups. Thus,
if the Owner/Company harms employee E of the Charterer/Contractor or causes dam-
age to E’s property, this damage shall – according to the contract – be compensated
by the Charterer/Contractor. However, E does not have to accept that the Charterer/
Contractor shall pay the claim. He may direct his claim to the Owner/Company
instead. If so, the knock-for-knock principle is obtained through a subrogated claim
from the Owner/Company against the Charterer/Contractor after E is compensated.38
35 NF Art 30.2 and 30.2 para. 2.

36 Third-party liability is also regulated in NF Art 30.3, but here with a diferent model.
37 Bull p. 346; Zak p. 23.
38 Bull p. 347; Zak p. 23.
169
If E accepts compensation from the Charterer/Contractor, this regulated approach
further implies that the Charterer/Contractor does not have a right to claim recourse
from the Owner/Company, even if the Owner/Company is at fault. This part of the
principle is particularly signifcant for work performed on the Norwegian Continental
Shelf. As mentioned, it follows from PA § 10–9 that the licensee is jointly liable for
any damage caused by a contractor. Thus, any party who sufers harm in Norwegian
petroleum activity may always make a claim against both the contractor and the
company. If the claim is raised against the party who is responsible according to the
knock-for-knock principle, subrogation according to PA § 10–9 is barred.39
The structure of the knock-for-knock principle is therefore a combination of free-
dom from liability/acceptance of not making a claim, a basis for recourse from the
party having paid the claim according to tort law, but who is not liable according to
the contract, and a bar to recourse from the party having paid the claim according
to the contract even if he was not liable according to ordinary tort law.
3.4 The insurance regulation

The regulation of liability is normally supplemented by a regulation of insurance. The
purpose of this is partly to secure that the liability risk of each party is fnanced by insur-
ance. Normally, it will be up to each contractual party to what extent he needs fnancial
security through insurance. However, if the knock-for-knock principle is put into efect
through indemnifcation after having frst paid the claim, it is important for the party
having paid the claim in the frst place that the other party is covered by liability insurance
which includes liability according to contract. Further, in the case of joint liability, in
particular according to PA § 10–9, if the Contractor cannot pay for the damages, the
Company will always be liable. Therefore, the contracts will provide the party with a duty
to take out proper insurance protection to cover its liability under the contract.40
Further, in order for the liability system to be carried through the contracts as out-
lined above, it is important that the division of risk is not disrupted by a subrogation
claim from the insurer.41 The starting point according to Norwegian law is that the
insurer, after having paid compensation for loss or damage, may claim subrogation
from the injurer that caused the loss.42 However, it is clear that the insurer does not
have a wider right to subrogation than the injured party could have claimed from the
injurer.43 If the injured party has agreed to waive his right to claim damage against
the injurer, a recourse claim from the insurer is similarly barred. However, this may
in turn mean that the insurer’s liability may be reduced by an amount equal to “that
which he is prevented from collecting because the assured has waived his right to claim
compensation from a third party, unless the waiver may be considered customary in
39 Bull p. 346; Zak pp. 23–24.

40 OLF proposal 8.12 para. 1–4, Supplytime 17 (a), but duty for the Owner only.
41 Kaasen p. 767; Zak p. 33.
42 Trine-Lise Wilhelmsen, Regress i skadeforsikring, Tidsskrift for erstatningsrett, forsikringsrett og
trygderett, 2019 p. 7 at pp. 12–13 with references.
43 Bull p. 489; Knut Selmer, Forsikringsrett, Oslo 1982, pp. 349 and 357.
170
the trade in question”.44 Therefore, in order to protect the position of the insurer and
the assured, it is necessary to secure against subrogation in the contract.
Two diferent contractual techniques are used in this context. The frst is the most
direct, and simply imposes a duty on the Contractor or Owner to require the insurers
to waive all rights of subrogation against the Company or Company group.45 Such
waiver of subrogation follows directly from NP § 18–9 for insurance of MOUs and
probably also from NP § 5–14 for hull insurance.46
The second technique is more indirect: the party not efecting the insurance shall
be named as co-insured under the policy.47 Such a right is automatically included
in NP clause 18–1 (i) subclause 2, and may be agreed in hull insurance.48 The main
content of co-insurance is that a third party with owner interest, security interest or
other economic interest in the insured property is insured for this interest under an
insurance efected by the “main owner” or “assured”.49 This is of lesser interest here.
But the co-insured also has indirect liability protection. This is not regulated directly
either in the NP or in the ICA, but it is presumed in the preparatory documents to
the ICA that the co-insured as injurer will have the same protection as the assured,
if they cause damage that constitutes an insured event under the casualty insurance
efected by the assured.50 This protection means that the co-insured has the same
protection against the insurer as he would have had as an assured if he had been
responsible for causing an insured event through a breach of the so-called duties
of due care.51 According to the NP clause 3–33, the insurer may, in cases where the
assured causes the damage through gross negligence, reduce his liability from 0%
to 100% depending on the degree of fault and circumstances generally. Among the
relevant circumstances taken into account in applying this rule will be the profes-
sionality of the assured, the risk involved in the activity and the injurer’s options
to avoid the risk. A co-insured will then have the same protection against recourse
from the insurer as he would have had if he had caused loss or damage under his
own casualty insurance. This is called the co-insured’s indirect liability insurance.
It follows from this that the protection for the injurer as a starting point is better
with a waiver of recourse clause than with a co-insurance clause in cases where the
injurer causes losses through gross negligence.
4 The rationale for the knock-for-knock principle
4.1 The need for contractual control of the liability risk

The regulation of liability through the knock-for-knock principle is unique for the
ofshore sector. Normally, contractual parties do not regulate the risk for tort law
44 NP Cl. 5–14; Bull p. 490; Selmer p. 357.

45 OLF proposal 8.12 para. 7, Supplytime 17 (a) (ii).
46 Bull p. 490. In the NP 2013 the provisions are Cl. 5–14 and Cl. 18–1 (i) sub-clause 1.
47 OLF proposal 8.12 paragraph 6, Supplytime 17 (a) (ii).
48 NP Cl. 8–1.
49 NP Cl. 8–1 and Cl. 7–1, and ICA § 7–1.
50 NOU 1987:24 p. 145; cf. pp. 151–152; Bull 2008 p. 539.
51 Bull p. 487.
171
liability and coverage under the other party’s insurance. However, there are three
main features in the ofshore sector that makes contractual control of the liability
risk important.
One feature is that the risk of causing damage during the operations is more sub-
stantial than in land-based projects or ordinary transport contracts.52 The ofshore
industry has always been a very risky and hazardous business, and this is more
accentuated as the upstream oil and gas sector has gradually moved into deeper and
more unsafe regions with adverse weather conditions.
A second feature is the huge capital sums invested, which easily results in enormous
losses if accidents happen. For example, the catastrophic Deepwater Horizon oil spill
is a primary illustration. Therefore, it is important that the allocation of this risk is
controlled by contractual regulation. A third feature is the involvement of many con-
tractors and subcontractors, which results in several potential injurers and victims.53
The substantial risk for damage and the number of people involved create a need
for foreseeability. Without regulation, liability claims will be handled according to
the ordinary tort law system. This requires investigation into which party was at fault
and could result in costly litigations, causing substantial economic uncertainty.54 For
example the Piper Alpha disaster led to claims against 24 diferent contractors. Among
those on board the platform who were killed, 134 were employed by contractors and
31 by the operator. Among those who survived, 55 were employed by contractors
and 31 by the operator.55 A clearer defnition of risk allocation between the parties
would transfer this uncertainty into a risk that may be calculated more accurately.
This will also help prevent difcult discussions between contractual parties who are
obliged to work together in long-term projects.56
4.2 Efcient insurance coverage

The principle may also be explained in terms of insurance coverage.57 Without risk
allocation, each party must purchase liability insurance to cover potential liability
for damage during the project. This will then be an addition to casualty insurance
covering damage to and loss of property and loss of income, and insurance covering
personal damage and death of employees. If an accident occurs, this could easily
result in overlapping insurances, where property damage loss is covered both by the
injured party’s casualty insurance, and also by the injurer’s liability insurance. Insur-
ance is a costly way to fnance risk; and, as with all ofshore sector’s costs, it is a
substantial amount. By channelling the risk for damage to the party where the dam-
age occurs, the need for liability insurance and thus the premium for this insurance
will be reduced.
52 Kaasen p. 765.
53 Ibid. p. 765.
54 Kaasen p. 766, Bull p. 353.
55 Caledonia North Sea Ltd v London Bridge Engineering Ltd [2002] UKHL 4; [2002] 1 Lloyd’s Rep 553, HL.
56 Bull pp. 353.
57 Cf. further Bull pp. 349–352; Kaasen p. 767.
172
The result of this is, of course, a higher risk exposure under the casualty insurance.
However, such insurance will normally be purchased in any case because damage may
easily occur without anybody being responsible through the tort law system. Typical
examples would be damage caused by natural disasters or mistakes made within the
insured’s own organisation. Anyone involved in the petroleum sector therefore needs
to make a risk assessment as to how best to handle and fnance these risks. The need
for casualty insurance is therefore not reduced even if a normal liability regime applies.
It is also generally considered cheaper to channel risk to the casualty insurance than
to divide the risk between casualty insurance and liability insurance, with a right for the
casualty insurer to claim recourse against the liability insurance. Recourse claims are
costly to invoke, and casualty insurance is generally cheaper than liability insurance.58
4.3 Loss prevention?
4.3.1 Loss prevention and efcient liability rules

A major argument against the knock-for-knock principle is that it is contrary to the
considerations of deterrence which are a key reason for regulation of tort liability. It
fails to adequately discourage parties to avoid causing damage in other parties’ risk
zones because negligence and faults have no consequences. In the Norwegian discus-
sion on this issue in relation to the knock-for-knock principle in the petroleum sector,
it is claimed that this consideration is exaggerated. Control over routines, the relation-
ship with public authorities and the parties’ desire to be qualifed to undertake further
projects on the Norwegian Continental Shelf are claimed to be more important.59
As a starting point, because there is difculty in measuring how a lack of deterrence
afects a specifc situation, it is difcult to have a meaningful opinion on the issue.
However, deterrence considerations can be analysed in terms of law and economics,
which can explain the signifcance of this consideration. The approach in this model
is to defne how diferent rules may infuence the optimal level of care, in order to
minimise the sum of costs for preventive measures and damage. This theory has
several main presumptions, which includes that the liable party or injurer is acting
rationally by seeking to minimise his own costs in relation to damage, that he has
full information regarding tort rules, and that the rules are enforced.60
As a starting point, the theory demonstrates three attributes: that no liability is
never optimal, that no-fault liability will induce the injurer to choose a level of care
that minimises the sum of prevention costs and expected damage, and that liability
for negligence will induce the injurer to choose the level of care necessary to avoid
liability so that he is only liable for loss prevention costs.61 This model conforms to the
traditional thinking on the signifcance of deterrence. However, the conclusions may be
diferent when there is a contractual relationship between the injurer and the victim.
58 Bull pp. 349–350.

59 Ibid. p. 355.
60 Erling Eide and Endre Stavang: Rettsøkonomi, 2. ed., Oslo 2018, (Eide and Stavang) pp. 243–244.
61 Eide and Stavang pp. 246–249 and Stephen Shavell, Economic Analysis of Accident Law, 1987, p. 8.
173
4.3.2 The efcient basis for liability in contractual relationships
In the event of a contractual relationship, for instance between an owner and a
charterer, the model must be supplemented by an assessment of the charterer’s will-
ingness to pay for the owner’s services. The charterer’s willingness to pay will depend
on how he assesses the risk involved in chartering the ship.62 The injurer’s/owner’s
inducement to prevent damage will therefore depend both on the risk of damage and
the possibility of charging more for the service.63
The model builds on a presumption that the owner maximises his own proft.
It is further presumed that he acts in a market with perfect competition, which
means that the price of the service provided is equal to the total costs necessary to
produce the service, including the costs connected to liability against the customers/
charterers.64 A third presumption is that the risk of damage depends on negligence
or non-negligence. With these presumptions, Table 8.1 – showing level of care, cost
of taking such precautions, percentage of probability of an accident and expected
accident losses – may be used as a basis for discussion.65
The owner’s costs in producing one unit of transport = 10. This amount does
not include costs in relation to care or liability against the charterer. Based on these
fgures, the optimal level of care may be discussed depending on the charterer’s
information about the risk.
In the frst case, the charterer has full information about the risk of damage. If
the owner does not face liability for damage, he will not have any costs connected to
duty of care or expected accident losses.66 In this case, an owner O-1 may choose to
sell one unit of transport for 10, which is the production cost. A charterer, who has
full information about the inherent risk in the transport, will know, however, that the
transport will in fact cost him 10 + 9 = 19 due to the accident costs. Another owner,
O-2, chooses to take due care. The cost per unit of transport will then be raised by
2 to 12. At the same time, the charterer’s risk will be reduced to 3. The charterer’s
total cost will therefore be 15. A cost-minimising charterer with full information
about the risk of damage will select O-2 with total costs of 15 instead of O-1 with
total costs of 19. The result is that the less careful O-1 will lose his customers to the
more careful O-2, even if no liability is imposed.
Table 8.1 The Optimal Level of Care
Level of care Cost of care Accident probability Expected accident

losses
None 0 9% 9
Care 2 3% 3
62 The model is based on Shavell p. 47 f.

63 Shavell p. 47 and pp. 51–52.
64 Shavell p. 47.
65 Taken from Shavell p. 49.
66 Shavell p. 52.
174
If the owner is facing strict liability, he will be liable for the cost of care and
expected accident costs. As 2 + 3 is less than 9, he will choose to take due care. If
we presume that the norm for negligence follows the optimal care in the ordinary tort
law model, the result is the same for liability for negligence. The result is therefore
that the owner will choose to take care regardless of liability for damage.67
In the second case, the charterer does not have sufcient information about the prob-
ability of damage to calculate the correct price for the service from diferent service
providers.68 Under this presumption, the owner cannot expect the charterer to pay extra
because the owner takes care, since the charterer does not know about the probability of
damage. Without any potential liability, the owner will therefore not take due care. But
if the owner is liable for negligence he will take optimal care in relation to the evalua-
tion of negligence.69 However, a negligence rule corresponding to the optimal level of
care presumes that the judge has sufcient knowledge to evaluate the carrier’s activity. If
this proves difcult, the evaluation of negligence may well lead to an over- or underesti-
mation.70 In case of no-fault liability, on the other hand, the owner will choose to take
due care as long as the costs of taking such care are lower than the potential liability.71
The model demonstrates that if services are provided in a market with perfect com-
petition and full information, the optimal level of care is not infuenced by liability
rules. This implies that if the parties to the contracts are professionals and have suf-
fcient volume of activity to establish their own statistics relating to the probability
of accidents and accident losses, liability for the damage caused by each party is not
necessary to obtain an optimal level of care. At the same time, a rule of liability for
negligence will result in transaction costs in relation to the settlements, which must
be calculated into the price of the transport.72 To the extent that the ofshore market
satisfes these presumptions, it may be argued that the knock-for-knock principle
conforms to economic efciency and that a liability regime is not necessary to obtain
the optimal level of care.
On the other hand, in regard to less professional or smaller market participants
with limited knowledge of the risk of damage, the knock-for-knock principle is not
defended by this model.
5 The validity of the regulation
5.1 Some starting points

Norwegian legislation does not contain a special rule prohibiting parties from freeing
themselves from liability for damage, or from waiving their right to claim for dam-
ages in tort. Therefore, the general starting point is that the parties are free to agree
67 Ibid. pp. 52–53.

68 Ibid. pp. 53–54.
69 Ibid. pp. 53–54.
70 Ibid. p. 56.
71 Ibid. p. 54.
72 Trine-Lise Wilhelmsen, Rett i havn, Oslo 2006, p. 337.
175
to such freedom from liability, regardless of the basis for this liability. Further, the
main rule is that contracts shall be fulflled as agreed.73
However, there are two general mandatory restrictions that are applicable to con-
tracts. The frst is NL 5–1–2,74 which prohibits contracts that are against law and
morality. The second is the Contract Act § 36,75 which states that contracts that are
unfair may be set aside partly or in full. Based on these rules, the legal theory assumes
that a contractual party cannot avoid liability for damage they cause deliberately.76
Further, it is a general view that liability limitation is accepted for acts or omis-
sions committed by ordinary employees, in contrast to those committed by company
leadership. This is also true if the act is made deliberately or with gross negligence.77
What is less certain is the extent to which freedom from liability for gross negligence
by the company itself may be valid. This must therefore be discussed based on the
two rules mentioned. It is also necessary to address the insurance clauses according
to these rules.
5.2 NL 5–1–2
NL 5–1–2 prohibits contracts that are contrary to the law or morality. The rule’s latter
portion regarding morality is relevant to the knock-for-knock principle. The expression
“contrary to morality” means where the contract is outside generally accepted moral
norms.78 The concept of “generally accepted moral norms” can be static or dynamic.
If the concept is static, previous Supreme Court judgments would be decisive regardless
of their occurrence. There are older Supreme Court judgments79 that implicate freedom
from liability for gross negligence by the company is void. Based on these judgments,
it is claimed in legal theory that limitation from liability for the company’s own gross
negligence is invalid as an absolute rule.80 With this interpretation, the knock-for-knock
principle cannot be applied in cases where damage is caused by gross negligence by
persons acting on behalf of the Owner/Company or Charterer/Contractor.
Expanding on the concept of “general accepted moral norms” also includes a
dynamic interpretation. This presumes a reference to the time of the evaluation and
that caution in the use of old judgments is required.81 There is some support for
this view in the preparatory documents to the Contract Act § 36, which claims that
the interpretation of NL 5–1–2 is uncertain in regard to limitation of liability.82 It
73 Kong Cristian Den Femtis Norske Lov av 25. april 1687 (NL) 5–1–1.
74 Ibid. 5–1–2.
75 Act 31. mai 1918 no. 4 on “avslutning av avtaler, om fuldmagt og om ugyldige viljeserklæringer
(Contracts Act).
76 Kai Krüger: Norsk Kontraktsrett, 1989 s. 784 with further references, Bull p. 394 and note 151 with ref-
erences, and Kaasen p. 777. An illustration of this principle may be found in ND 1988.263 “Mørland 7” NA.
77 Rt. 1994.626, Rt. 1948. 370, Rt. 1915.840, ND 1989.225 NA, ND 1991.180 Eidsivating, NOU 1979:32
Formuerettslig lempningsregel (NOU) p. 19, Bull p. 394 and note 152; Kaasen p. 777; Zak pp. 36–37 and
p. 41.
78 Zak p. 36 with further references in note 157.
79 Rt. 916.717 and Rt. 1926.712.
80 NOU p. 19 and references in note 1; Viggo Hagstrøm, “Om grensene for ansvarsfraskrivelse, særlig i
næringsforhold”, Tidsskrift for rettsvitenskap 4/1996, pp. 464 and 475.
81 Zak p. 36.
82 NOU p. 19.
176
is also argued that invalidity is merely a guiding principle, where there is room for
exceptions depending on the circumstances.83
If it is accepted that NL 5–1–2 only provides a guiding principle that gives room
for exclusions, the question is whether such exclusion is justifed for the knock-for-
knock principle in the ofshore sector. In this context, it is natural to analyse the
“morality” aspect of no liability in regard to the reasoning behind the liability rules.
Here, the most “immoral” aspect seems to be retribution, i.e., that a lack of retribu-
tion in cases of gross negligence by the company is in itself against “morality”. This
may be explained by a requirement for corrective justice in the relationship between
the parties. But a requirement for corrective justice is difcult to reconcile with the
development in the use of liability insurance. Further, it follows from the ICA84 that
the liability insurer will also cover liability caused by gross negligence by the assured,
which in this case will be the insured company. If the liability is insured, the retribu-
tion aspect is therefore reduced to an insurance premium payment. This payment is
made before the damage resulting in liability is caused. The only actual retribution
will therefore be that the liability insurer may raise the premium for the next insur-
ance period. And if they do, the assured may refuse renewal and enter into a contract
with another company. Therefore, there is not much left of corrective justice when
the liability is covered by liability insurance.
The knock-for-knock principle also means that the contractual partner as a victim
does not obtain repair of damage from the injurer. However, the knock-for-knock
principle assumes that repair of damage to the victim’s interests is fnanced through
casualty insurance covering loss of or damage to property and income, to the extent
that such fnancing is needed. Accidents to employees will be separately covered
through employment casualty insurance, which is mandatory in Norway.85 Repair of
damage is therefore secured through insurance.
The last consideration behind liability rules is deterrence. It may be argued that it
is immoral not to have rules that are aimed at preventing damage. However, deter-
rence’s efect on liability must be seen in conjunction with developments in public
safety regulations and requirements, and the attitude towards safety issues within
companies. This may imply that deterrence through liability is less needed. Further-
more, if it can be demonstrated that liability is not needed because the market will
secure optimal care, it is difcult to see why a notion of morality should prevent an
efcient development of risk sharing.
It may therefore be argued that the earlier view of morality in court practice and
legal theory is outdated by recent developments in insurance and security legislation,
as well as by the economic models on the deterrent efect of liability where there is
a contractual relationship between the parties. However, since there are no decisions
from the Norwegian Supreme Court, the conclusion is uncertain.
A possible underlying reason for disagreements on this issue is a more general
confict between corrective justice as the raison d’être for liability rules, and the legal
83 Lasse Brautaset, “Kontraktsreguleringen ved salg av gass”, Norsk Gassavsetning. Rettslige hovedele-
menter, Sjørettsfondet 1998, pp. 117–118; Zak pp. 37–38.
84 ICA § 4–9 para. 2.
85 Yrkesskadeforsikringsloven 1989 no. 65 § 3 cf. § 1.
177
and economics-based approach where liability rules are analysed in terms of efciency
and optimal care.86 The goal of maximising wealth in the legal and economics-based
approach is considered “immoral” because it does not consider distributive and
corrective justice.87 But even if such more fundamental considerations of justice are
relevant in relation to liability regimes, they appear less useful in a professional and
well-organised contractual setting.
5.3 The Contract Act § 36
5.3.1 Overview and some starting points

The Norwegian Contract Act § 36 provides:
An agreement may be wholly or partially set aside or amended if it would be unreason-
able or confict with generally accepted business practice to invoke it. The same applies
to a unilaterally binding disposition.
When making a decision, account will be taken not only of the contents of the agree-
ment, the position of the parties and the circumstances prevailing at the time of conclu-
sion of the agreement, but also of subsequent events and circumstances in general.
The assessment according to this provision is therefore diferent from that according
to NL 5–1–2. A contract may be invalid due to immorality even if it is totally fair
between the parties. This may, for example, be the case if the party claiming freedom
from liability has paid the other party a fair price for obtaining this right, and this
payment is used to buy alternative fnancing for costs of damage through insurance.
On the other hand, a contract may be unfair without being immoral, for instance, if
a change of circumstances has resulted in an altered level of risks for damage being
incurred, making the knock-for-knock agreement unfair for one of the parties.
The Contract Act § 36 primarily applies to consumer contracts, but it may also be
applied to professional contracts.88 However, the threshold for applying this rule is
higher in such contracts than in consumer contracts.89 In the oil and gas industry, it is
generally recognised that adjusting and rewriting contracts according to the Contract
Act § 36 can only occur under special circumstances.90 However, it follows from the
Contract Act § 36’s preparatory documents that a primary reason for establishing
this rule was clauses used in the oil and gas sector; namely, the clauses giving the
company a right to the beneft of technological developments made by the contrac-
tors during building projects.91
Even if contractual revision is not out of the question, it seems that the attitude
towards contractual revision has hardened over the last 20 years. Generally, the change
86 Cf. for instance Mårten Schultz, Kausalitet. Studier i skadestandsrettslig argumentation, Stockholm
1992 p. 101 f., and Trine-Lise Wilhelmsen, Årsakssammenheng i erstatningsretten, Oslo 2011, p. 15–19.
87 Schultz p. 127, p. 131 f. and p. 141 f., Wilhelmsen 2011 p. 18.
88 NOU pp. 47 and 61, Ot. Prp. no 5 (1982–1983) Om lov om endringer i avtaleloven 31. mai 1918 nr. 4
m.m. (Generell formuerettslig lempningsregel) (Ot. Prp.) p. 33, Generalklausul i förmögenhetsrätten, SOU
1974:82 p. 111, Regjeringens propositioner 247/1981 p. 14, Kai Krüger, Kontraktsrett 1989, p. 422, Viggo
Hagstrøm, Obligasjonsrett, Oslo, 2. ed. 2011, p. 290.
89 Rt 1999.922 at p. 932; Zak p. 42.
90 ND 1990.204 NA Ula and ND 2000.240 NA Troll.
91 NOU pp. 47 and 61.
178
of attitude is because courts have favored predictability at the cost of fairness, par-
ticularly in relation to later events resulting in more extensive losses than expected.
There are very few cases where an agreement is set aside according to the Contract
Act § 36 over the last 20 years, and in those cases the reasoning is inapplicable in our
context.92 This process conforms to the Supreme Court’s development and practice
of giving more weight to the objective understanding of the wording in the contract
where the parties are professional.93
The evaluation according to the Contract Act § 36 is a broad evaluation of the cir-
cumstances listed in the provision in relation to the individual contract. However, not all
the circumstances are relevant for this paper. The argument “the position of the parties”
refers to the situation where one of the parties lacks the competence or ability to enter
the agreement, or lacks the knowledge and experience to understand it, or there is a
clear inequality between the parties in regard to the contract. As a general consideration
this seems less relevant for the types of contracts discussed here. The same is true for
the argument “the circumstances prevailing at the time of conclusion”, which refers to
duress, misuse of negotiation power, exploitation and information failure. Therefore,
the argument that is more generally relevant is that of the content of the agreement.
5.3.2 The content of the agreement

The content of the agreement as a reason for setting aside this kind of liability provision
means that the knock-for-knock principle applied in cases of gross negligence by the
company itself would be unfair. The preparatory documents to the Norwegian Contract
Act § 36 provision sheds little light on limitation of liability clauses in general. However,
the equivalent Swedish rule and their preparatory documents contains relevant remarks.94
In those preparatory documents, it is stated that the application of limitation of liability
clauses shall not be limited to a specifc degree of fault, but instead it depends on a
total evaluation of the specifcs of the actual contract. In cases where freedom from
liability is tied to fnancing through insurance, the main purpose will be to limit recourse
from the insurer, and a convenient liability and insurance regime limiting the costs of
recourse processes should not be denied through strict principles of fairness.95
Based on these statements, Swedish legal theory has assumed that limitation of
liability clauses in professional contracts should be treated diferently from other
contracts, in particular if they are combined with insurance.96
92 Rt. 2014.351 (energy company freed from a contract where they would have to pay revenue to the
landowner for 50 years against no performance from the landowner), Rt. 2013.769 (revision of insur-
ance settlement with no cover for income disability after such disability proved to be 50%), Rt. 2013.388
(consumer investment in structured saving products set aside because misleading information of central
parameters for the investment and the risk for loss gave a wrong picture of the possibility of revenue on the
investment), Rt 2008.969 (Fraud or misleading information), Rt 2001.603 (a continuation of an agreement
seemed meaningless).
93 Cf. for instance Rt. 2010.1345, Rt. 2002.1155, Rt. 2000.806.
94 Relevant for the Norwegian Contract Act § 36 also because this paragraph is a result of Nordic legis-
lative cooperation with identical rules in all the Nordic countries, cf. Wilhelmsen, Avtaleloven § 36 og øko-
nomisk efektivitet, Tidsskrift for rettsvitenskap 1995 no. 1, pp. 13–14; Hagstrøm (2011) p. 287 f.; Zak p. 43.
95 SOU 1974:83 pp. 180–181.
96 Claes-Robert von Post, Studier kring 36 § avtalslagen med inriktning på rent kommersiella förhål-
landen, Stockholm 1999, p. 207; Jan Ramberg og Christina Ramberg, Allmän avtalsrätt, Åttonde upplagan,
179
A similar focus on professionalism and insurance is found in court practice concern-
ing the Nordic Freight Forwarder Agreement (NSAB), which is an agreed standard
contract with a long tradition. NSAB states that the freight forwarder’s liability for
damage is limited regardless of fault.97 In U 1993.851 this clause was set aside by
the Danish Supreme Court when the freight forwarder negligently failed to follow
its own established practise for delivery of goods.98 However, the limitation was
accepted in U 2005.243 and U 2006.632. In U 2005.243 the Danish Supreme Court
simply stated that the clause must equally be accepted as written even in cases of
gross negligence. In U 2006.632, the situation was that the company failed in the
planning and performing of the service. It is not directly stated that the failure was
grossly negligent, but this seems to be presumed in the lower court, which set the
limitation aside. The Danish Supreme Court referred to U 2005.243, and stated that
the limitation could not be set aside, according to the Contract Act § 36, in cases
of gross negligence. The main arguments were that the freight forwarder contract is
an agreed standard contract where the limitation is part of a total liability regime,
which presumably rests on a total evaluation where considerations of efcient insur-
ance play a central role.
This result conforms to Norwegian practice, but here the negligence is tied to the
employee, and not to the company. In Rt 1994.626, the same clause was accepted
with similar reasoning found in U 2006.632. A similar view is found in previous
arbitration and appeal cases with towing contracts.99
The Norwegian theoretical discussion on this issue is divided into two factions. One
faction argues that such clauses should be set aside due to traditional considerations
of fairness similar to those following from the discussion on NL 5–1–2.100 Arguments
in relation to the knock-for-knock principle in the Norwegian Fabrication Contract
are that the activity constitutes a risk for personal safety and pollution, and that
transaction costs are presumably small compared to the importance of these interests.
But even if they are not, efciency considerations must be given less weight than
the need for a liability regime to protect personal safety and environment.101 This
argument seems to overlook the fact that the risk for pollution damage is regulated
by mandatory regulation in the MC and the PA, and that this regulation is adhered
to in the contracts. Furthermore, under certain market conditions considerations of
deterrence may not necessitate a liability regime.
The other faction argues that the principle should be accepted as it is described.102
Main arguments here are that the knock-for-knock system is supported by both
Stockholm, 2010, p. 215; Thorsten Lundmark, Friskrivingsklausuler giltighet og räckvidd, Uppsala, 1996,
p. 133; Zak p. 44.
97 NSAB 1975 § 25; NSAB 2000 § 22 cf. § 5 (no limitation in case of deliberate acts); NSAB 2015 § 23.
98 Cf. further Hagstrøm (1996) p. 435.
99 ND 1989.225 NA and ND 1991.180 Eidsivating.
100 Hagstrøm (1996) p. 422, p. 478 f.; Jo Hov and Alf Petter Høgberg, Alminnelig avtalerett, Oslo 2009,
p. 402; Lasse Simonsen, ”Kreditors mangelsbeføyelser – særlig for tilvirkningskontraktene”, Jussens Venner
1999 p. 305 at p. 380; Zak p. 41.
101 Hagstrøm (1996) p. 481.
102 Kaasen p. 776 f., Bull p. 391 f., Erik Røsæg, “Lastehåndterings og forvaringstjenester”, MarIus
no. 271, p. 41.
180
parties to the contracts103 and that freedom of liability is closely tied to an insurance
regime securing the interests of the victim.104
What can be concluded here is that freedom from liability in cases where the dam-
age is caused by gross negligence by the company itself can only be achieved if some
minimum requirements are fulflled: the contract should be agreed to secure involve-
ment and acceptance by both parties, the freedom from liability should be tied to a
systematic insurance regulation to secure that all potential victims are compensated,
the liability and insurance system should refect a thorough analysis of what combina-
tion of liability insurance and casualty insurance is most convenient for the parties,
and the system should reduce transaction costs. But even when these conditions are
fulflled, acceptance by the court is still uncertain.
5.3.3 The insurance clauses

The knock-for-knock principle is, as mentioned, combined with waiver of subroga-
tion clauses and co-insurance clauses in the insurance policies. The waiver of subro-
gation clause is similar to the indemnity clause because it is not tied to any degree
of fault. However, the position of the co-insured will mean that the co-insured party
obtains an indirect liability cover for ordinary negligence, but in the case of gross
negligence the result may be a reduction in the compensation.
It must be presumed that these insurance clauses follow the same mandatory regime
as the indemnity clauses. If the indemnity clause is deemed invalid according to NL
5–1–2 or unfair according to the Contract Act § 36, but the waiver of subrogation or
co-insured’s protection is upheld, the position of the party claiming to be indemnifed
will difer according to whether the injured party makes the claim against the injurer,
or prefers to claim coverage from his own insurer. In the frst case, the injurer must
compensate the injured party. In the second case, he will be free from liability. Such
an arbitrary result is inconsistent and contrary to considerations of fairness.105 The
result must therefore be that the mandatory rules apply similarly to the waiver of
subrogation clauses and the co-insured’s indirect liability protection.
103 Bull p. 393; Kaasen p. 778.

104 Kaasen pp. 778–780; Bull pp. 393–394.
105 Hagstrøm (1996) pp. 485–486. See also Selmer p. 130; Bull p. 319.
181
CHAPTER 9
Statutory liability regulation versus contractual

risk allocation in upstream oil and gas
The Norwegian case
Kristofer Svendsen
1 Introduction
This chapter looks at the interaction between the statutory regulation of liability for
damage or loss caused by petroleum spills from ofshore installations, as set forth in
Chapter 7 of the Norwegian Petroleum Act,1 and allocation of risk in oil and gas
contracts achieved through the use of knock-for-knock clauses. The chapter frst sets
forth the liability regime under Chapter 7, then examines the contractual allocation
of risk in model oil and gas contracts used on the Norwegian Continental Shelf
(NCS), before determining the validity of said clauses and its efects when interacting
with Chapter 7 liability. It is important to remember, however, that knock-for-knock
clauses cover many more situations than just possible Chapter 7 pollution damage.
In other words, situations arising under Chapter 7 of the Petroleum Act are only a
part of what these risk allocation clauses cover.
2 Statutory regulation of risk under Chapter 7 of the

Norwegian Petroleum Act
2.1 Introduction
The Norwegian authorities started examining delict liability for pollution damage
caused by oil spills from ofshore installations as early as 1970, which resulted in a
committee report in 1973.2 The Norwegian government’s work on the issue was tem-
porarily halted due to a belief that the work leading up to and the fnalized Conven-
tion on Civil Liability for Oil Pollution Damage Resulting from Exploration for and
Exploitation of Seabed Mineral Resources3 would take efect. When the convention
did not receive necessary backing from important countries, such as the United King-
dom, the Norwegian government appointed another committee to suggest regulation
for compensation of pollution damage from ofshore installations. The committee
1 Lov om petroleumsvirksomhet 29 November 1996 Nr. 72.

2 NOU: Erstatningsansvar for forurensingsskader. Om erstatningsansvar m.v. for forurensingsskader i
forbindelse med undersøkelse etter og utvining av undersjøiske naturforekomster (1973:8).
3 Convention on Civil Liability for Oil Pollution Damage Resulting from Exploration for and Exploita-
tion of Seabed Mineral Resources, London (1 May 1977).
182 DOI: 10.4324/9781003206798-11

regulation versus risk allocation in upstream oil and gas
gave its report in 1981,4 which was afrmed, with minor changes, in Chapter 7 of the
Petroleum Act.5 The frst edition of the Petroleum Act came into force in 19856 and
received an overhaul in 1996.7 The Petroleum Act of 1996 is currently in force and
only minor changes took place in Chapter 7 of the 1996 edition.
The Petroleum Act contains four liability regimes exclusively dealing with petro-
leum activities. Chapter 7 of the Petroleum Act is the main legislation regulat-
ing liability for petroleum pollution damage.8 Chapter 8 regulates compensation
of fnancial losses sufered by Norwegian fshermen as a result of the petroleum
activities occupying fshing, causing pollution and waste, or inficting damage by a
facility or actions in connection with the placing of a facility.9 The regulation of
Chapter 8 damages does not apply to petroleum pollution damage under Chapter
7. Section 5–4 of the Petroleum Act regulates liability in the abandonment phase,
while section 10–9 of the Petroleum Act regulates liability of the licensee for dam-
age caused by a legal entity of physical person who performs work for the licensee.
Section 10–9 does not, however, apply to losses resulting from petroleum pollution
damage under Chapter 7 or losses sufered by Norwegian fshermen as a result of
petroleum activities under Chapter 8.
The chapter focuses on the efect of Chapter 7 on contractual risk allocation.
The Maritime Code10 defnes the outer limits of Chapter 7, and more generally
the Petroleum Act. The Code mainly regulates various aspects of vessels and
shipping and includes provisions on liability of oil pollution from vessels, among
others. Importantly, the Maritime Code regulates ‘drilling platforms and similar
mobile constructions’ when these units are in motion and therefore categorized
as vessels.11
The Pollution Control Act12 also assists in defning the outer limits of Chapter 7,
and the Petroleum Act. The Pollution Control Act regulates compensation for pol-
lution damage, unless regulated by other legislation, contract, or other things.13 The
Pollution Control Act does not apply to situations specifcally regulated by Chapter 7
of the Petroleum Act.14 However, Chapter 7 is not exhaustive regarding the defnitions
4 NOU: Erstatningsansvar for forurensningsskade som følge av petroleumsvirksomhet på norsk konti-

nentalsokkel (1981:33).
5 Ot. prp.nr. 72. Lov om petroleumsvirksomhet (1982–1983).
6 Lov om petroleumsvirksomhet (repealed) 22 March 1985 Nr. 11.
7 Lov om petroleumsvirksomhet 29 November 1996 Nr. 72.
8 Id. at § 7–1.
9 Id. at § 8–1.
10 Lov om sjøfarten The Norwegian Maritime Code is translated into English in MarIus No. 393 (24
June 1994 No. 39).
11 Id. at § 507.
12 Lov om vern mot forurensninger og om avfall (Forurensningsloven) 13 March 1981 Nr. 06.
13 Id. at § 53.
14 The preparatory works states that the Pollution Act supplements or complements the special rules ‘as
far as the rules ft’. Ot.prp.nr.33 (1988–1989) Om lov om endringer i lov 13 mars 1981 Nr. 6 om vern mot foru-
rensninger og om avfall (forurensningsloven) m.v (Erstatningsansvar ved forurensningsskade) (1988–1989)
pp. 99–100 and 104. See also a discussion in this topic in Hans Christian Bugge, Forurensningsansvaret
(1999) (Tano Aschehoug, Oslo) para. 6.3.2.2.
183
kristoffer svendsen
of pollution damage, which is seen below.15 Finally, the general rules, lex generalis,
of delict law apply when the special rules, lex specialis, of delict law come short.16
2.2 Liability for petroleum pollution damage
2.2.1 Unlimited no-fault liability

The Petroleum Act places unlimited no-fault liability on the licensee for pollution
damage caused by petroleum spills from ofshore installations.17 Usually there are
several licensees under a license, of which one is the operator. In these situations,
claims for compensation should be directed to the operator. If the operator does not
pay claims when due, the licensees pay the claims according to their participating
interest in the license.18 If a licensee fails to pay its share, that share is allocated
proportionally between the remaining licensees. Interestingly, the Act continues with
a very practical approach to the procedural aspect of mass claims requiring the
operator without undue delay, by public announcement, to provide information
regarding the party to whom claims for compensation for pollution damage should
be directed and of the period of limitation.19 The Ministry of Petroleum and Energy
decides where actions should be brought when questions of venue arise.20
The Act places unlimited no-fault liability on the licensee because the licensee has
the authorization to conduct the activity, the fnal say, obtains the profts from the
activity (high fnancial reward), holds bargaining power (insurance, contracts, etc.),
and therefore should hold the risk and the consequences of oil pollution damage.21
The equitable policy consideration ‘prevention’ is also used to justify strict liability,
based on which the licensee will take better precautions to hinder an accident as
he has a direct fnancial interest in the matter.22 Furthermore, the licensee must
obtain liability insurance, of which the Ministry of Petroleum and Energy receives
summaries every year and ensure that appropriate insurances are maintained.23 The
15 The Pollution Act covers all non-petroleum pollution damage such as other chemicals and waste-
related petroleum pollution damage. Bugge states that it is reasonable to apply the Pollution Control Act
to supplement and complement the Petroleum Act (a) for damages sufered by other groups of people then
Norwegian fshermen as a result of other types of pollution then for petroleum pollution, and (b) in situa-
tions where other types of pollution than petroleum pollution infict personal harm on Norwegian fshermen
or a reduction in fshing abilities. Bugge, Forurensningsansvaret (1999), para. 6.3.2.2, pp. 263–264.
16 The general rules of the law of compensation are compiled in the Damages Compensation Act (Lov
om skadeserstatning 13 June 1969 Nr. 26) as well as in case law, preparatory works, legal theory, and by
applying and balancing the equitable policy considerations (reelle hensyn).
17 Lov om petroleumsvirksomhet (29 November 1996 Nr. 72). § 7–3.
18 Id. at § 7–3.
19 Id. at § 7–7.
20 Id. at § 7–8. The Ministry decides where the action shall be brought if (a) the efuence or discharge
has taken place or the damage has been caused outside the area of any court district; (b) it cannot be dem-
onstrated within which court district the efuence or discharge has taken place or damage has been caused;
(c) the efuence or discharge has taken place in one court district and the damage is caused in another court
district (d) damage has been caused in more than one court district.
21 NOU: Erstatningsansvar for forurensningsskade som følge av petroleumsvirksomhet på norsk kon-
tinentalsokkel. 1981:33, p. 21.
22 Id. at p. 25.
23 Erik Røsæg, The Norwegian perspective with regard to liability regimes concerning oil rigs and instal-
lations, in Ofshore contracts and liabilities (Baris Soyer & Andrew Tettenborn eds., 2015), p. 281.
184
average insured amount appears to be 250 million USD for each licensee covering
its pro rata share of the liability.24
The geographical scope of Chapter 7 is subject to a special regulation in section
7–2, which is diferent from all other liability regimes. The section applies lex loci
damni when pollution damage ‘occurs in Norway or inside the outer limits of the
Norwegian continental shelf or afects a Norwegian vessel, Norwegian hunting or
catching equipment or Norwegian facility in adjacent sea areas.’25 The consequence of
the scope of Chapter 7 is a unilateral extension of protection in delict law to Norwe-
gian interests harmed outside of Norway. This privilege has resulted in discrimination
against only Russian interests, which do not receive any judicial remedy on pollution
damage caused by a Norwegian operator sufering an oil spill on the Norwegian side
of the sea border in the Barents Sea, and that oil spill inficts pollution damage on
the Russian side of the sea border in the Barents Sea.26 A Russian injured party
forced to pursue a legal claim against a Norwegian licensee without assets in Russia
may receive no compensation, because there is no agreement about recognition and
enforcement of foreign court judgments between Norway and Russia.27
2.2.2 Pollution damage

The term ‘pollution damage’ is defned as ‘damage or loss caused by pollution as a
consequence of efuence or discharge of petroleum from a facility, including a well,
and costs of reasonable measures to avert or limit such damage or such loss, as well as
damage or loss as a consequence of such measures.28 Damage includes personal injury,29
and damage to fshermen.30 Chapter 7 does not cover discharge of petroleum from ves-
sels transporting petroleum, which is covered by Chapter 10 of the Maritime Code.
The Act defnes ‘pollution damage’ in relation to events for which liability under
Chapter 7 of the Act may arise, but does not assist in defning compensable damage. The
main delict statute, the Compensation for Damages Act,31 and other delict legislation
do not notably assist much in defning compensable damage.32 The preparatory works
24 Id. at p. 281.
25 Lov om petroleumsvirksomhet 29 November 1996 Nr. 72. § 7–2.
26 Kristofer Svendsen, Compensable damage ex delicto as a result of harm in the Barents Sea caused by
petroleum spills from ofshore installations. A Norwegian and Russian comparative legal analysis of confict
of laws, the concept of harm, losses sufered by third parties, and environmental damage and its valuation
and calculation, caused by petroleum spills from ofshore oil rigs and installations in the Barents Sea (2015)
Tromsø Ph.D. in law, Faculty of Law, UiT – Arctic University of Norway).
27 No other sea bordering countries to Norway are afected due to the Convention on jurisdiction and
the enforcement of judgments in civil and commercial matters, Lugano (30 October 2007).
28 Lov om petroleumsvirksomhet 29 November 1996 Nr. 72. § 7–1. Petroleum is defned in § 1–6a, and
facility is defned in § 1–6d.
29 Ot. prp.nr. 72. Lov om petroleumsvirksomhet. 1982–1983, p. 70.
30 Chapter 8 damage (diferent from Chapter 7 damage) is specifc damage to Norwegian fshermen
caused by petroleum activities (and maybe non-petroleum substances). Norwegian fshermen may be com-
pensated for their economic loss caused by pollution damage or waste from petroleum activities. Compen-
sation includes loss of fshing opportunities and equipment due to the aforementioned pollution. Lov om
petroleumsvirksomhet 29 November 1996 Nr. 72, § 8–1.
31 Lov om skadeserstatning 13 June 1969 Nr. 26.
32 Sören Koch, Det erstatningsrettslige skadebegrepet – en sammenligning mellom tysk og norsk rett,
Tidsskrift for erstatningsrett, forsikringsrett og velferdsrett pp. 250–281 (2010), p. 255. Koch summarises the
understanding of damage in legal literature on pages 256–261.
185
kristoffer svendsen
give examples of certain types of damage that would qualify as pollution damage, such
as harm inficted to wildlife in the sea and on land, the soiling of beaches and fshing
gear, the closure of a water area as an obstacle for fshing and shipping, the soiling
of real estate and objects, and costs incurred by the public or others for cleaning up
soiled beaches.33 The preparatory works confrm that compensable pollution damage
must fulfll the prerequisites for every type of damage compensable in delict law.34
Section 6 of the Pollution Control Act defnes pollution through four sentences.
The frst sentence qualifying for pollution is ‘the introduction of solids, liquids or
gases to air, water or ground.35 Efuence or discharge of petroleum would be within
the wording of section 6. The fnal part of each of the four sentences is ‘which cause
or may cause damage or nuisance to the environment. 36 The wording ‘damage or
loss caused by pollution’ in section 7–1 fulflls the fnal part of the frst sentence of
section 6. Thus, the defnition of pollution in the Pollution Control Act defnes ‘pol-
lution’ in section 7–1,37 while the Petroleum Act defnes petroleum,38 efuence and
discharge,39 and facility.40
2.2.3 Channelling liability to the licensee and licensee’s recourse

The Petroleum Act channels unlimited liability for pollution damage to the licensee.
The courts have discretionary power to reduce this unlimited liability partly or com-
pletely upon (1) an inevitable event of nature, (2) an act of war, (3) the exercise of
public authority, or (4) a similar force majeure event that has contributed to a con-
siderable degree to the damage or its extent under circumstances which are beyond
the control of the liable party.41 Liability can only be reduced to the extent it is rea-
sonable, with particular consideration to the scope of the activity, the situation of the
party that has sustained damage and the opportunity for taking out insurance on
both sides.42
The channeling provisions in section 7–4 bar claims for petroleum pollution damage
to be directed towards licensee’s contractors, manufacturers or suppliers of equip-
ment, anyone who undertakes measures to avert or limit pollution damage,43 and

34 Id. at p. 35 agreeing with NOU: Erstatningsansvar for forurensingsskader. Om erstatningsansvar m.v.
for forurensingsskader i forbindelse med undersøkelse etter og utvining av undersjøiske naturforekomster.
1973:8.
35 Lov om vern mot forurensninger og om avfall (Forurensningsloven) (13 March 1981 Nr. 06). § 6 1.
36 Id. at § 6.
37 U. Hammer, et al., Petroleumsloven (Universitetsforlaget, Oslo. 2009). p. 532.
38 § 1–6 (a) of the Petroleum Act.
39 The preparatory works to the Petroleum Act do not expressly distinguish between “efuence”
(utstrømming) and “discharge” (utslipp). The legislator uses the word “efuence” when describing a leak from
an oil pipeline. Ot. prp.nr. 72. Lov om petroleumsvirksomhet. 1982–1983, p. 70.
40 Lov om petroleumsvirksomhet 29 November 1996 Nr. 72, § 1–6d.
41 Id. at § 7–3.
42 Id. at § 7–3.
43 The full text reads ‘anyone who undertakes measures to avert or limit pollution damage, or to save
life or rescue values which have been endangered in connection with the petroleum activities, unless the
measures are performed in confict with prohibitions imposed by public authorities or are performed by
someone other than public authorities in spite of express prohibition by the operator or the owner of the
values threatened.’ Id. at § 7–5(c).
186
employees of the licensee or employees of any of the above groups.44 The same parties
shielded from liability in section 7–4 are shielded from indemnity in section 7–5. The
licensee can only seek recourse from the parties listed in section 7–4 if ‘the person
in question or someone in his service has acted willfully or by gross negligence.’45
Similarly, a shipowner is also strictly liable for oil pollution damage, with a similar
group46 protected through channeling of liability except when a person caused damage
with intent or through gross negligence, and with the knowledge that such damage
would probably result.47 In summary, injured parties cannot direct claims towards
these parties, and the licensee cannot demand that these parties accept liability for
pollution damage, unless intent or gross negligence. This begs the question whether
licensees on the NCS are solid enough to carry the liability of a potential oil spill.
Importantly for our current discussion, section 7–5 states:
Any agreement on further recourse in respect of those against whom liability cannot be
claimed pursuant to Section 7–4, second paragraph, shall be invalid.
The parties can, however, further limit access to recourse,48 but not extend recourse
in contravention with section 7–5. This would indicate that the shielded parties could
contractually be protected against pollution damage inficted by them through, for
example, gross negligence. However, it is doubtful in Norwegian legal theory whether
a party can contractually agree to avoid liability for gross negligence.49
3 Contractual allocation of risk on oil and gas contracts in Norway
3.1 Introduction
The contractual allocation of risk in oil and gas contracts is often achieved through
the careful wording of clauses describing which parties should pay which parties for
damage or harm arising out of diferent situations as a result of physical risks mate-
rializing. Gordon describes three vehicles used to regulate and manage these physical
risks in the oil and gas industry: (1) indemnity and hold harmless clauses, (2) clauses
which exclude or limit liability for ‘consequential losses’, and (3) overall limitations
of liability.50 Gordon takes a further dive into indemnity and hold harmless clauses
in Chapter 2 of this book. This part of the chapter looks at the main risk allocation
44 Id. at § 7–4.
45 Id. at § 7–5.
46 (a) The servants or agents of the owner or the members of the crew; (b) the pilot or any other person
who performs services for the ship; (c) the ship operator (reder) or manager or operator of the ship, the
charterer, the consignor, shipper, owner of cargo or consignee; (d) any person performing salvage operations
with the consent of the owner or on the instructions of the public authority; (e) any person taking preventive
measures; (f) all servants or agents of persons mentioned in subparagraphs (c), (d) and (e).
47 Lov om sjøfarten The Norwegian Maritime Code is translated into English in MarIus No 393 24
June 1994 No. 39, § 193.
49 Viggo Hagstrøm, Om grensene for ansvarsfraskrivelse, særlig i næringsforhold, 1996 Tidsskrift for
rettsvitenskap pp. 421–518 (1996); see Wilhelmsen in Chapter 8 and Kaasen in Chapter 7a of this book.
50 Greg Gordon, Chapter 14, Risk Allocation in Oil and Gas Contracts, in Oil and Gas Law: Current
Practice & Emerging Trends (Greg Gordon, et al. eds., 2011), p. 14.2 f.
187
kristoffer svendsen
clause used in model oil and gas contracts in Norway, and whether they make refer-
ence to or take into account the above-mentioned section 7–5 of the Petroleum Act.
3.2 Industry-negotiated model contracts

The NCS is well known for the wide use of industry-negotiated model contracts (also
called agreed documents) setting forth a set of standard conditions developed for
contracting on the NCS. These model contracts are not mandatory to use but are
widely used in engineering, procurement, and construction on the NCS. The follow-
ing contracts are the main industry-negotiated model contracts:
• The Norwegian Total Contract of 201551 (NTC 15);
• The Norwegian Total Contract of 2015 Modifcation52 (NTC 15 Mod);
• The Norwegian Total Contract of 2015 for module and modifcation (modi-
fcation with single delivery of the entire contract object);53
• The Norwegian Total Contract of 2015 for module and modifcation (modi-
fcations with separate delivery of module, prefabricated items, and ofshore
permanent works);54
• The Norwegian Fabrication Contract of 201555 (NFC 15);
• The Norwegian Subsea Contract of 200556 (NSC 05).
The NTC 15 is recommended to regulate contracts for the delivery of larger compo-
nents for production of hydrocarbons on the NCS where the contractor is responsible
for engineering, procurement, construction, and possibly installation (EPC(I)). NTC
15 Mod is recommended to regulate contracts for larger modifcations of platforms
on the NCS where the contractor is responsible for EPC(I). The NTC 15 for module
and modifcations is recommended when the delivery also includes a new module, and
the contractor is responsible for EPC(I). This standard contract is issued in two ver-
sions, the frst version for which the contract object is delivered collectively, while the
second version caters for separate delivery of the module, prefabricated items, and
ofshore permanent works. The NFC 15 is recommended for larger contracts for delivery
of fabrication, such as machinery and mechanical equipment, to the NCS. Finally, the
NSC 05 is intended to regulate contracts covering ‘marine operations such as installa-
tion of pipelines, cables, umbilicals and other subsea structures and related subsea
51 The contract can be retrieved at www.norskindustri.no/contentassets/69b36f82c6f341a68c1aa7691e4

f5e06/versjoner-lagt-inn-des.-2017/norsk-totalkontrakt-2015.pdf
f5e06/versjoner-lagt-inn-des.-2017/norsk-totalkontrakt-2015-modifkasjon.pdf.
53 The contract can be retrieved at www.norskindustri.no/contentassets/69b36f82c6f341a68c1aa7691e
4f5e06/versjoner-lagt-inn-des.-2017/norsk-totalkontrakt-2015-modul-og-modifkasjon-samlet-levering.pdf.
f5e06/versjoner-lagt-inn-des.-2017/norsk-totalkontrakt-2015-modul-og-modifkasjon-separat-levering.pdf.
f5e06/versjoner-lagt-inn-des.-2017/norsk-fabrikasjonskontrakt-2015.pdf.
56 The contract can be retrieved at www.norskoljeoggass.no/drift/publikasjoner/hms-og-drift/
norwegian-subsea-contract/.
188
construction work where the use of vessels is involved.57 The NSC 05 captures both
installation-only contracts as well as full EPC(I)-type contracts and addresses specifc
risks in connection with subsea work and the operation of vessels.58
Large industrial actors, namely the predecessor of the Federation of Norwegian
Industries (Mechanical Industry Association) and the Norwegian Union of Iron
and Metalworkers (now part of the Norwegian United Federation of Trade Unions)
started the negotiations of these model contracts, the NTC and the NFC, in the
1970s, with the frst draft presented in 1983. Other large industrial parties, such as
Hydro, Statoil, and Saga, joined in and the frst edition of the model contract was
presented in 1987.59 The current editions are a result of eforts between the Federation
of Norwegian Industries, the Norwegian Oil and Gas Association, and companies
in the industry. The predecessor of the Norwegian Oil and Gas Association, OLF,
initiated the preparation of the model contract NSC 05 with participation of Statoil,
Stolt Ofshore, Subsea 7, and Technip Ofshore Norge.
3.3 The risk allocation clause

These model contracts were negotiated to implement a standard for a developing industry
to rely on, even the commercial playing feld through balancing commercial contracting
provisions, efectively train staf, and minimize transactional costs. In all the model
contracts above, article 30 contains the provision titled ‘Exclusion of liability. Indemni-
fcation’, which is the preferred method of the industry to allocate physical risk.
All the above-mentioned model contracts, except for a couple of words in the NSC
05, contain the exact same wording of their exclusion of liability and indemnifcation
clause. This is not particularly strange as exclusion of liability and indemnifcation
work best in the contractual chain when all parties use the same clause, ‘back-to-back’.
Article 30 of the above-mentioned contracts sets forth a mutual indemnity clause60
for loss or damage to Contractor and Company Group, Contractor’s indemnifcation
obligation for third-party claims and its limitation, indemnifcation for industrial
property infringements, and notifcation procedures. In other words, article 30 consists
of a combination of Gordon’s three vehicles.
An mutual indemnity clause means ‘a contractual device where the parties with
the one hand give and with the other hand take an indemnity in respect of a spe-
cies of loss which, if the indemnity is to avoid circularity, must not be identical to
each other, but which are usually closely related61 Articles 30.1 and 30.2 set forth
a mutual indemnity clause for personal injury, death, and loss of or damage to
property in relation to the subject matter of the contract applying when sufered by
the parties to the contract, the Contractor and the Company Group. The identical
57 See OLF, Norwegian Subsea Contract, NSC 05, Conditions of Contract (2005). ‘Introduction to
NSC 05.’
58 Id.
59 Knut Kaasen (2006) Petroleumskontrakter med kommentar til NF 05 og NTK 05 (Universitetsfor-
laget, Oslo), pp. 23 f.
60 A mutual indemnity is also called ‘reciprocal indemnity’, ‘cross indemnity’, or a ‘knock-for-knock’
indemnity. Gordon, Chapter 14, Risk Allocation, in Oil and Gas Contracts, 2011. p. 14.5.
61 Id.
189
kristoffer svendsen
mutual indemnity clause between the Contractor and Company Group reads as fol-
lows (identical in the NSC 05):
30.1 Contractor shall indemnify Company Group from and against any claim concerning:
a) personal injury to or loss of life of any employee of Contractor Group, and
b) loss of or damage to any property of Contractor Group,
and arising out of or in connection with the Work or caused by the Contract Object
in its lifetime. This applies regardless of any form of liability, whether strict or by neg-
ligence, in whatever form, on the part of Company Group.
Contractor shall, as far as practicable, ensure that other companies in Contractor Group
waive their right to make any claim against Company Group when such claims are covered
by Contractor’s obligation to indemnify under the provisions of this Art. 30.1.62
30.2 Company shall indemnify Contractor Group from and against any claim
concerning:
a) personal injury to or loss of life of any employee of Company Group, and
b) loss of or damage to any property of Company Group, except as stated in Art. 29,
and arising out of or in connection with the Work or caused by the Contract Object
in its lifetime. This applies regardless of any form of liability whether strict or by negli-
gence, in whatever form, on the part of Contractor Group.
Company shall, as far as practicable, ensure that other companies in Company
Group waive their right to make any claim against Contractor Group when such claims
are covered by Company’s obligation to indemnify under the provisions of this art.
30.2.63
The parties, the company group and the contractor, agree to indemnify each other
from and against any claim concerning their own employees’ personal injury or death
and loss of and damage to any of their own property. The wording ‘from and against
any claim’ would mean a loss or damage independent of amount and injured party.64
Articles 30.1 and 30.2 apply to the internal relationship among the contracting par-
ties. In article 30.3, however, the contracting parties regulate their interaction with
public authorities and third parties. Article 30.3 reads:
Until the issue of the Acceptance Certifcate, Contractor shall indemnify Company Group
from:
a) costs resulting from the requirements of public authorities in connection with the
removal of wrecks, or pollution from vessels or other foating devices provided
by Contractor Group for use in connection with the Work, and
b) claims arising out of loss or damage sufered by anyone other than Contractor
Group and Company Group in connection with the Work or caused by the
Contract Object,
even if the loss or damage is the result of any form of liability, whether strict or by
negligence in whatever form by Company Group.
The Acceptance Certifcate is a document issued by Company when the work has
been completed according to the contract.65 Article 23 discusses, amongst other
things, the Acceptance Certifcate and requires the Company to the Acceptance
62 The wording of the NFC 15.

63 Id.
64 Hans Jacob Bull (1988) Tredjemannsdekninger i forsikringsforhold: en studie av dekningsmodeller,
med basis i sjøforsikringsretten og i petroleumskontraktenes ansvars- og forsikringsregulering (Sjørettsfon-
det, Oslo), p. 394.
65 See Norwegian Fabrication Contract 2007 (NF 07), art. 1.10.
190
Certifcate when the work has been completed under the contract, deeming the cer-
tifcate to be issued 30 days after either the guarantee period has ended.66
This part of article 30.3 does move a possible heavy fnancial burden over on the
Contractor as the Company Group requires unilateral indemnity from the Contractor
for removal and pollution expenses from vessels or other foating devices provided by
the Contractor Group for use in connection with the work. More importantly is the
sub-paragraph b) of article 30.3 requiring the Contractor to unilaterally indemnify
the Company Croup from third-party claims for loss or damage sufered in connec-
tion with the Contractor’s work or caused by the object of the contract, even if the
loss or damage is the result of any form of liability, whether strict or by negligence
in whatever form by Company Group. Sub-paragraph (b) is in direct breach of
articles 7–4 and 7–5 of the Petroleum Act, which expressly shields the contractor
from liability for petroleum pollution damage and bans the licensee from seeking any
indemnity from the contractor.
Continuing in article 30.3, a separate limitation of contractor’s liability follows
this paragraph:
Contractor’s liability for loss or damage arising out of each accident shall be limited to
NOK . . . million. This does not apply to Contractor’s liability for loss or damage for
each accident covered by insurances provided in accordance with Art. 31.2.a) and b),
where Contractor’s liability extends to the sum recovered under the insurance for the loss
or damage.
Company shall indemnify Contractor Group from and against claims mentioned in
the frst paragraph above, to the extent that they exceed the limitations of liability men-
tioned above, regardless of any form of liability, whether strict or by negligence, in
whatever form, on the part of Contractor Group.
After issue of the Acceptance Certifcate, Company shall indemnify Contractor Group
from and against any claims of the kind mentioned in the frst paragraph above, regard-
less of any form of liability, whether strict or by negligence, in whatever form, on the
part of Contractor Group.
This limitation of contractor’s ‘third-party liability’ to a set amount does not negate
article 30.3’s infringement on articles 7–4 and 7–5 of the Petroleum Act. Previous
editions of this part of article 30.3 contained a specifc amount of million NOK,
which have increased through negotiation rounds throughout the years, the latest
being 5 million NOK (approx. 530,000 EUR) in the 2007 edition.67 The current
edition’s ability to set your own limitation might increase the licensee’s risk of having
to cover a certain amount of money for petroleum pollution damage without the
amount being insured. That said, many companies self-insure so the insurance aspect
might be a non-issue.68
It is interesting to note that the Contractor in article 30.3, and the contracting par-
ties in articles 30.1 and 30.2, is obligated to indemnity the Company Group ‘regard-
less of any form of liability, whether strict or by negligence, in whatever form’ on
66 Id. art. 23.5.

67 See Norwegian Fabrication Contract 2007 (NF 07).
68 Neil A. Doherty & Cliford W. Smith Jr., Corporate Insurance Strategy: The Case of British Petro-
leum, 6 (3) Journal of Applied Corporate Finance pp. 4–15 (1993).
191
kristoffer svendsen
the part of any of these two parties. This means that article 30 applies to, amongst
other, no-fault liability for petroleum pollution damage.
According to Norwegian contract law, a party cannot contractually avoid liability
for intent, but can arguably avoid liability for gross negligence (see more in Kaasen,
Chapter 7a, section 5.2).
This quick look at the model contracts shows no reference to section 7–5 of the
Petroleum Act, or any indication that the model contracts incorporated the wording
of section 7–5.
4 Contractual allocation of risk breaches with Chapter 7

The contracts consist of three main elements: (1) the parties to the contract waiver
the ability to claim compensation from each other even if the elements to claim
compensation is fulflled; (2) the parties to the contract waiver the ability to claim
recourse from each other when a third-party damage or a damage to own employees
are covered according to the agreement, even though a recourse action could take
place; and (3) the parties commit to hold the other party harmless, even though one
party might have to pay a third-party claim or an employee claim that should not
have been paid by the party in another legal sense.69
The Norwegian model contract contains a mutual indemnity clause for damage
or loss among the contracting parties, a unilateral indemnity clause for third-party
damage to the beneft of the Company Group, and a fxed-amount limitation of this
unilateral indemnity obligation left to the contracting parties to decide. However, the
model contracts do not consider section 7–5 of the Petroleum Act, which states that
there cannot be any changes to the channeling of liability for pollution damage set
forth in Chapter 7 of the Petroleum Act. As such, Norwegian law does not allow
for an indemnity clause shifting the liability for pollution damage away from the
licensee or creating a risk allocation model diferent from the statutory regulation
of liability for pollution damage in Chapter 7. Thus, mutual indemnity clauses
would be invalid for pollution damage under Chapter 7.
69 Knut Kaasen, Ansvarsbegrensning i fabrikasjonskontrakter, in Industribygging og rettsutvikling.

Juridisk festskrift i anledning Hydros 100-årsjubileum 2005 (Odd Ivar Biller, et al. eds., 2005), p. 241.
192
CHAPTER 10
Applying knock-for-knock in Germany

1 Introduction
Conducting constructions, operations and maintenance ofshore poses a large number
of challenges for developers, operators and contractors. Typically, there are many
parties involved, very large sums of money at stake, and constantly evolving and
therefore potentially dangerous work environments to contend with. Consequently,
the allocation of risk between the parties is a fundamental issue for consideration at
the outset of contract negotiations. In the ofshore oil and gas sector, the so-called
knock-for-knock regime has been practiced for decades. In the German ofshore wind
farm sector, although many parties are keen to contract on a knock-for-knock basis,
there has been and remains resistance from some. This is because the knock-for-knock
regime is fundamentally diferent from the liability regime and fault-based concept
of German law. Consequently, German companies have been initially hesitant to
accept these clauses, which were developed in the context of English law. The clauses,
and their validity, are still not fully tested by German courts, so there are several
legal uncertainties in this regard. In any case, the standard wording of knock-for-
knock clauses needs to be adapted to be valid under German law.
This chapter means to illustrate the common legal issues regarding the knock-for-
knock clause under German law, because, in contrast to English law, German law
signifcantly limits the parties’ right to exclude and/or limits their liability, especially
in circumstances where the respective clause will most likely be treated as general
terms and conditions (Allgemeine Geschäftsbedingungen).
First, we will give an overview on the use and prevalence of knock-for-knock
clauses in the German market (section 2) and then continue to illustrate the Ger-
man approach to clauses that are deemed general terms and conditions and whether
knock-for-knock clauses qualify as such (sections 3.1 and 3.2). Next, we will focus
on the fairness standards a knock-for-knock clause has to meet and the prohibition
to exclude liability for wilful/intentional misconduct (sections 3.3 and 3.4). Finally,
we will shortly review of available case precedent (section 4) and then give a general
recommendation as to how to conduct contract negotiations and amend the respec-
tive clauses in order to avoid disputes and minimise legal risks (section 5).
2 Use of knock-for-knock clauses in the German market

As already mentioned in the introduction, knock-for-knock clauses originated in the
English insurance market. German developers and contractors only encountered this
DOI: 10.4324/9781003206798-12 193

eckehard volz and anna-sophie waldmann
legal concept recently, once they became more heavily involved in the ofshore sector,
particularly the ofshore wind sector. The frst German ofshore wind farm Alpha
Ventus was commissioned in 2010, many of the frst projects still had very few inter-
national connections, and most if not all the contracting parties involved were German
companies operating under the familiar German law and liability regime. Consequently,
knock-for-knock clauses have only become more frequent and common during about
the last fve years. While there were initially many companies in the market that
refused to accept this liability regime, it became rapidly clear that German developers,
wanting to attract international contractors, and German contractors, tendering for
international projects, would need to adapt. Many quickly became used to the concept,
which, admittedly, has many advantages as it makes liability risks more foreseeable,
bankable, and insurable. Today, many German companies operating in the ofshore
wind sector will accept or ask for knock-for-knock clauses to be part of their contracts.
Outside of this industry, however, the concept remains largely unknown.
3 Validity issues under German law

Due to some elements of German law and since knock-for-knock clauses have their
origin in a diferent jurisdiction, there are elements of the liability regime that are
invalid and some that are potentially invalid under German law. Therefore, the knock-
for-knock clause needs to be adapted to work under a German contract. Carve-outs
from the general concept that each party shall cover its own losses need to be made.
In the following, we will explain the background on why these adaptations are neces-
sary and what exactly needs to be amended.
3.1 Court review of general terms and conditions

Under German law, all general terms and conditions are subject to court review.
While in principle there is freedom of contract, court review serves to protect the
party that has not drafted the general terms and conditions – particularly consumers,
but also experienced commercial parties – from unfair treatment and unexcepted
provisions. In the following, we will frst illustrate the general principle of court
review of general terms and conditions and then go on to apply this principle to
knock-for-knock clauses.
3.1.1 General principle

In respect to general terms and conditions, the German Civil Code (BGB) contains
rules in sec. 305 et seq. BGB to ensure the fairness and appropriateness between the
parties, which must be complied with to avoid those particular clauses that are
considered to be invalid by German courts. Accordingly, general terms and condi-
tions are subject to supervision by the German courts as regards their enforceability.
Such supervision predominantly intends to protect consumers from unreasonable
and surprising clauses. However, most of these rules also apply to business
contracts.
German courts will therefore examine whether any of these clauses are unfair to
the detriment of one of the parties. As far as particular clauses of the general terms
194
applying knock-for-knock in germany
and conditions do not comply with the regulations set out in sec. 305 et seq. BGB,
the relevant clause will be deemed invalid and consequently as non-binding between
the parties of the contract.
If a particular clause is considered invalid, this afects the whole clause, not only
the invalid part. Thus, special attention must be paid during the drafting process and
in case of a liability regime, even a minor drafting error can undermine the whole
concept of the liability in a contract.
3.1.1.1 Definition of general terms and conditions

Sec. 305 (1) sent. 1 BGB reads:
Standard business terms are all contract terms pre-formulated for a multitude of contracts,
which one party to the contract (the user) presents to the other party upon the entering
into of the contract.
From a legal perspective any standard form contract, such as the BIMCO SUPPLY-
TIME 2005, WINDTIME, FIDIC, LOGIC or any other standard business terms,
regardless of the actual frequency of its use by your company or others, will be deemed
general terms and conditions under German law pursuant to sec. 305 (1) BGB, unless
(1) the terms of the contract were not pre-formulated for a multitude of contracts –
which will be very difcult to prove as generally standard forms are used as the con-
tractual basis – or (2) the clause has been individually negotiated (see below) between
the parties.
3.1.1.2 Content test

In addition, any clause from standard business terms, if not individually negotiated,
is subject to the so-called content test set out in sec. 307 (1) and (2) BGB, which read:
(1) Provisions in standard business terms are inefective, if they unreasonably disad-
vantage the other party to the contract with the user contrary to the requirements
of good faith. An unreasonable disadvantage may also arise from the provision
not being clear and comprehensible.
(2) An unreasonable disadvantage is, in case of doubt, to be assumed to exist, if a
provision
1. is not compatible with the essential basic principles of the statutory provi-
sion from which it deviates, or
2. limits essential rights or duties inherent in the nature of the contract to such
an extent that attainment of the purpose of the contract is jeopardised.
Whether or not a standard knock-for-knock clause is to the unreasonable disadvantage

of a party will be determined by the relevant court or arbitral tribunal within the scope
of a test of the reasonableness of content (Inhaltskontrolle) by looking at the principles
set out in secs. 307, 308 and 309 BGB (Content Test). This Content Test therefore has
two legs: frst, general principles of fairness must be observed; second, secs. 308 and 309
contain a list of terms that are prohibited in any case and of terms that can either be
valid or invalid subject to interpretation of the specifc wording of the term and whether
it allows for an interpretation in a way that is not unfair or grossly detrimental.
Strictly speaking secs. 308 and 309 BGB are only directly applicable for consumer
contracts. Nevertheless, the German courts take these provisions into account as
195
guidelines to business contracts, however certainly also looking at the common busi-
ness practice in a particular industry.1
3.1.1.3 What is “individually negotiated”?

Thus, for business-to-business contracts it is advisable to prevent the courts or an
arbitral tribunal from conducting the Content Test on any clause, i.e., to prevent clauses
of the contract to be regarded as general terms and conditions at all. Under German
law pursuant to sec. 305 para. 1 sent. 3 BGB any clauses, which are individually negoti-
ated – this may include clauses that were originally general terms and conditions provi-
sions – are not regarded as being general terms and conditions. However, the notion
of what is individually negotiated (im Einzelnen ausgehandelt) has once more been
subject to several court decisions by the German Federal Supreme Court (BGH):
NEGOTIATION OF TERMS DEMANDS MORE THAN DEBATING

According to the BGH negotiating (aushandeln), in the sense of sec. 305 para. 1 sent.
3 BGB, demands more than simply debating or disputing (verhandeln) the terms of a
contract. In fact, the user of general terms and conditions, which frst need to contain
provisions that amend or add essential content to statutory provisions, must addition-
ally put the terms to serious disposition and allow the contractual partner to protect
his own interests by being able to infuence and modify the content of the contract.2
Therefore, the user must clearly and seriously agree to amend respective clauses,
if requested by the contractual partner. However, a general reference that all terms
of the contract were open to discussion is not considered to be sufcient to establish
the necessary negotiation in respect of the essential content of individual clauses.
The same applies for protocols of the negotiations, which contain wordings like “all
terms of the contract have been thoroughly and seriously negotiated”; this is equally
considered to be insufcient for proving that individual negotiations in the sense of
sec. 305 (1) sent. 3 BGB were conducted.
ADDITIONAL SECURITY BY A “CAVEAT” CLAUSE?

Until recently, as a means of additional security, it was possible and advisable to
insert a rider clause aiming to change the character of the contract for the knock-
for-knock clause into an ordinary contract clause. This was to prevent the above-
mentioned strict limitation rules for German general terms and conditions by
suggesting that the draft contract was individually negotiated between the parties.
Such a “caveat” clause would have had to state that both parties considered the
entire agreement as in line with the common industry standard and as a reasonable
and fair allocation of rights and duties, highlighting that each clause was known and
1 BeckOK BGB/H. Schmidt, 58. Ed. 1.5.2021 Rn. 6, BGB § 307 Rn. 6; BeckOK BGB/H. Schmidt, 58.
Ed. 1.5.2021, BGB § 307 Rn. 96; MüKoBGB/Basedow, 8. Auf. 2019, BGB § 310 Rn. 11.
2 BGH, judgment dated 20.3.2014 – VII ZR 248/13 = NJW 2014, 1725; BGH order dated 20.11.2012 –
VIII ZR 137/12, BeckRS 2013, 5597 Rn. 7; BGH, order dated 19.3.2019 – XI ZR 9/18, NJW 2019, 2080;
MüKoBGB/Basedow, 8. Auf. 2019, BGB § 305 Rn. 35; Wolf/Lindacher/Pfeifer/Pfeifer, 7. Auf. 2020 Rn.
36, BGB § 305 Rn. 36.
196
part of the individual negotiations. The following additional wording was added at
the end of the contract, if made subject to German law:
This contract was in its entirety freely negotiated by both parties and both parties have
duly noted all its contents and acknowledge that all the terms of this contract constitute
a fair and balanced allocation of rights and duties which properly refect the . . . industry’s
customs and practices.
Of course, parties never entirely discounted the risk that a court or arbitral tribunal
might overrule this clause and still qualify such clause and with it any other clause
of the contract, although negotiated and changed, as general terms and conditions.
But it at least supported and may in future still support the argument of the parties
that the contract is not constituted of general terms and conditions, including the
implications this may have, but was indeed negotiated openly.
Nevertheless, as mentioned above, the BGH has made clear that such clauses do
not automatically prevent a contract from still being regarded to be general terms and
conditions.3 Whether a contract is agreed based on general terms and conditions or
individually is a question of fact, which cannot be contractually regulated by the par-
ties, because this corresponds to the attempt to waive the general terms and conditions
provisions. Being able to waive these provisions individually would exceed the scope of
sec. 305 para. 1 sent. 3 BGB. The BGH is of the opinion that an economically superior
party of the contract could then force such an “individual clause” into the contract to
annul the protective purpose of German general terms and conditions provisions for the
entire contract. For this reason, even in the commercial business area, a “caveat” clause
will be regarded as legally non-binding, because sec. 305 et seq. BGB do not underlie
the disposition of the parties to the contract, but instead have legally binding force.
3.2 Are knock-for-knock clauses general terms and conditions?

Most knock-for-knock clauses used in practice are part of standard contract forms, such
as BIMCO SUPPLYTIME, WINDTIME, LOGIC or other pre-formulated terms used
in the industry. Under German law, these standard forms will be considered general terms
and conditions, which then by extension applies to the knock-for-knock clause contained
therein. If a tailor-made contract is drafted, copying the wording from one of these stan-
dard forms into the tailor-made contract without any amendments or negotiations, it is
very likely that a court would also consider the knock-for-knock clause part of general
terms and conditions. As illustrated above, individual and earnest negotiations about the
content and wording of the knock-for-knock clause can make it an individually agreed
term, but ultimately this will be decided by German courts or arbitral tribunals on a
case-by-case basis, taking into account the circumstances of the individual matter.
3.3 Applying the reasonableness test

Having established the basic principles in section 3.1 and whether knock-for-knock
clauses are considered as such, we will now look at the standards for validity that
3 BGH, judgment dated 20.3.2014 – VII ZR 248/13 = NJW 2014, 1725.
197
are applied by German court. As explained above, the Content Test has two legs,
the general fairness and reasonableness test specifed in sec. 307 BGB and the more
specifc list of terms not allowed as per secs. 308 and 309.
3.3.1 Commercial contracts versus consumer contracts

As mentioned previously, the second leg of the Content Test only applies, in principle,
to consumer contracts. Since knock-for-knock clauses are not generally used vis-á-vis
consumers, this might therefore appear irrelevant. However, it is important to know
that German courts use the list of terms prohibited in secs. 308 and 309 as indicators
for what could be considered unreasonable or unfair also in a strictly commercial
contract between businesspeople. Therefore, case precedent and legal commentary
on the list of prohibited terms are signifcant also in a strictly commercial context
and need to be considered during the drafting process for knock-for-knock clauses.
3.3.2 Fairness test – what is market practice?

Sec. 310 para. 1 sent. 2 BGB states that when interpreting and applying sec. 307
BGB in connection with business-to-business contracts “reasonable account must
be taken of the practices and customs that apply in business dealings” by the courts.
Accordingly, a court would have to consider the ofshore wind industry’s customs
and practices. Given that the wind industry is young in the context of established
legal practices, it is barely foreseeable to what extent a court or arbitral tribunal
would respect the ofshore wind industry’s practices and customs in this context.
That said, German courts and arbitral tribunals will consider that the industry is
dominated not only by German companies but companies with an international
background, which use knock-for-knock clauses frequently. Should a knock-for-
knock clause ever be tested in front of a German court, it is to be expected that
the court will not only look towards the use of knock-for-knock clauses in Germany,
but to the industry, if the parties of that hypothetical case have an international
background.
Finally, as we will elaborate on in more detail below, it is important to keep in
mind that there is as yet no court decision on the validity of knock-for-knock clauses.
Therefore, the question whether German courts will accept them as common industry
practice remains unanswered and it is quite difcult to predict how German courts
will resolve the confict between the ofshore wind industry’s customs and practices
on the one hand and the guidance taken from the principles contained in secs. 308
and 309 BGB on the other hand, when ruling on whether a knock-for-knock clause
is considered to be unreasonably disadvantageous pursuant to sec. 307 BGB and thus
invalid or not. It will largely depend on the individual circumstances of the matter.
How those circumstances may be infuenced in favour of the validity of the clause
is discussed below.
3.3.3 The “risky” terms

As mentioned above, sec. 309 BGB, inter alia, is not directly applicable to com-
mercial agreements between professional businesses but the German courts consider
them as a guideline for business contracts. With regard to knock-for-knock clauses
198
(e.g., in Clause 14 BIMCO SUPPLYTIME 2005) sec. 309 No. 7 BGB is most
relevant:
Even to the extent that a deviation from the statutory provisions is permissible, the fol-
lowing are inefective in standard business terms: . . .
a) any exclusion or limitation of liability for damage from injury to life, body or health
due to negligent breach of duty by the user or intentional or negligent breach of
duty by a legal representative or a person used to perform an obligation of the user;
b) any exclusion or limitation of liability for other damage arising from a grossly
negligent breach of duty by the user or from an intentional or grossly negligent
breach of duty by a legal representative of the user or a person used to perform
an obligation of the user.
It follows that a German court or arbitral tribunal may render a knock-for-knock

clause invalid, since it attempts to exclude liability for damages arising from (1)
personal injury to life and body or health, (2) wilful/intentional misconduct (not only
by representatives and employees, but the parties themselves), and (3) gross negligence.
There is a considerable risk that this could happen because German courts tend to
view these liability exclusions as highly detrimental and thus potentially unfair or
unreasonable even in a purely commercial contract. Consequently, in such circum-
stances a court may consider the entire clause to be invalid, even if only a part of
such clause is afected. This means – in case of any damage or loss – that the German
statutory liability regime of the German Civil Code, which is generally a fault-based
liability regime, would apply and the party claiming damages would be entitled to
damages without any limitation, provided the other party was at fault.
It follows, that the following elements of any knock-for-knock clause and the
overall liability regime of a contract containing such knock-for-knock clause need
to be carefully considered and potentially adapted in a German law context:
• Wording that attempts to limit or exclude liability for bodily injury or death;
• Wording that attempts to limit or exclude liability for gross negligence;
• Wording that attempts to limit or exclude liability for intentional misconduct
of employees or legal representatives of the parties;
• Wording that attempts to put in place an overall liability cap for the above-
named types of damages or the overall damages payable under a contract.
It needs to be emphasised that the above limitations are not generally invalid or
prohibited. However, they are terms listed under secs. 308 and 309 that will be under
scrutiny of a German court. Even commercial contracts need to be fair and reason-
able to some extent and a court will look at the overall liability provisions to establish
whether the whole of the contract – on balance – could be to the clear detriment of
one party. Including all the above-mentioned liability limitations or even exclusions
and low overall caps could tip the balance towards unreasonableness and make a
knock-for-knock clause fail the Content Test.
3.4 Invalid terms under German civil law – no exclusion of liability for intent
Aside from the Content Test applied to general terms and conditions, there are two
elements of a typical knock-for-knock clause that are invalid in any case, irrespective
199
of whether it is individually negotiated or not and whether the overall balance of the
contract is achieved or not. Sec. 276 para. 3 BGB does not allow the exclusion or
limitation of liability for wilful or intentional misconduct (Vorsatz) of a party to a
contract. And German law also does not allow contractual agreements to the detriment
of third parties that are not signatories to the contract at hand. To illustrate this, we
will briefy explain the general concept of liability for fault under German law.
3.4.1 General principle of liability under sec. 276 BGB

The central civil law norm in Germany for fault-based liability for breaches of both
contractual and statutory obligations is sec. 276 BGB:
Sec. 276 Liability of the debtor
(1) The debtor is responsible for intent and negligence if a stricter or milder liability is
neither stipulated nor can be inferred from the other content of the obligation, in
particular from the assumption of a guarantee or a procurement risk. . . .
(2) Any person who fails to exercise the due care required in the course of ordinary
business shall be guilty of negligence.
(3) Liability for intent may not be waived in advance for the debtor.
Sec. 276 thus contains an attribution rule, which determines under which circumstances
non-performance, bad performance, delay or breaches of statutory duties (comparable to
tort) are attributed and claims are established. However, it is not a specifc basis for a
claim.4 Sec. 276 para. 1 sent. 1 BGB contains the legislative decision that the principle of
fault requires (simple) fault,5 which results in a fundamentally fault-based liability regime
of German statutory law. Fault-based liability means the obligation to pay damages based
on conduct that is not only objectively unlawful but also personally culpable.
The principle of fault, which is the basis of sec. 276 (1) sent. 1 BGB, binds liability to
the prerequisite that the damage or unlawful condition has been caused by a reproach-
able conduct of the person against whom a claim is made.6 The BGB itself does not
explicitly defne the term “fault”. According to the history and system of the law, it
is the generic term for the two forms of fault mentioned in para. 1 sent. 1, namely
intent and negligence. In both forms, fault is an expression of subjective reproachability.
Intent means the knowledge and intention of the realisation of the facts, i.e.,
the deliberate breach of duty. The law does not provide a defnition of intent. The
so-called intent theory defnes it for private law by an intellectual and a voluntary
element, each of which must extend to the unlawfulness of the result: intent is the
knowledge and intention of the unlawful result.7
Negligence, on the other hand, according to the legal defnition of sec. 276 para.
2 BGB, means disregarding the care required in the ordinary course of business. In
civil law, a distinction is made between two diferent forms of negligence: simple
negligence and gross negligence. To determine the degree of care required in the
4 BGH, judgement dated 13. 11. 1953 – I ZR 140/52 _ NJW 1954, 229; MüKoBGB/Grundmann, 8. Auf.
2019, BGB § 276 Rn. 1–2; Jauernig/Stadler, 18. Auf. 2021, BGB § 276 Rn. 1.
5 BeckOGK/Schaub, 1.3.2021, BGB § 276 Rn. 4; MüKoBGB/Grundmann, 8. Auf. 2019, BGB § 276 Rn. 1–2.
6 BGH, judgement dated 09.07.1992 – VII ZR 7/92 = NJW 1992, 3158.
7 BGH, judgement dated 20. 11. 2012 – VI ZR 268/11 = NJW-RR 2013, 550; HK-BGB/Reiner Schulze,
10. Auf. 2019, BGB § 276 Rn. 6; MüKoBGB/Wagner, 8. Auf. 2020, BGB § 823 Rn. 47.
200
ordinary course of business, the respective circles/groups of the public must be con-
sidered, considering the special features of the type of transaction and the groups
of persons typically involved in it.8 Gross negligence requires that the care required
in the ordinary course of business is violated to a particularly serious degree. This
is to be assumed if something was not observed that should have been obvious to
anyone in the given situation,9 obvious and simple considerations were not considered
or concerns that should have been obvious due to typical indications of the danger
were recklessly disregarded.
According to sec. 276 para. 1 sent. 1 BGB, liability for intent and negligence only
arises if a stricter or milder liability is neither stipulated by law or contract nor can
be inferred from the other content of the obligation. This means that the law itself
exceptionally provides for strict liability or allows it by contract. Contractual exten-
sions of liability are permissible by individual agreement within the limits of secs. 138,
242 BGB. These prohibit immoral agreements and prescribe conduct that is guided
by the principle of good faith. Examples of this are the guaranteed agreement or
the assumption of strict liability.
Finally, as indicated already above, sec. 276 para. 3 BGB prohibits the waiver
or exclusion of liability for intentional misconduct in prior to the damage having
occurred.
3.4.2 No agreements to the detriment of third parties

As a general concept of German civil law, contractual agreements to the detriment
of third parties that are not part of the agreement itself, are invalid.10 This afects
knock-for-knock clauses where they establish a “Contractor’s Group” and “Employer’s
Group” concept, where subcontractors and afliates companies are included in the
knock-for-knock liability regime. The intention of the knock-for-knock regime and
the concept of establishing “groups” of companies that fall on either side of the liabil-
ity spheres is easy to understand: it means to be a catch-all clause to establish a
back-to-back liability in a multi-party project contract, with several subcontractors
and supplies and service providers involved during all stages of the project.
From a German law point of view, however, including third parties in such an
agreement is highly problematic, because it attempts to take away title to sue or claim
damages from such third parties. Therefore, a German court would consider this an
agreement to the detriment of a third party, which is invalid.
3.4.3 Consequences for knock-for-knock clauses under German law

Because of sec. 276 para. 3 BGB, even if the clause has been individually negotiated
prior to the signing of the contract, a knock-for-knock clause will be deemed invalid
under German law if it aims to exclude liability for wilful or intentional misconduct.
8 BGH, judgement dated 02.10.1985 – IV a ZR 18/84 = NJW 1986, 1100.

9 BAG, judgement dated 12. 11. 1998–8 AZR 221–97 = NJW 1999, 966; BGH, judgement dated
18.12.1996 – IV ZR 321/95 = NJW 1997, 1012.
10 BGH, judgement dated 12.11.1980 – VIII ZR 293/79 = NJW 1981, 275; BVerfG, judgement dated
23. 04. 1986–2 BvR 487/80 = NJW 1987, 827; BeckOGK/Mäsch, 1.1.2021, BGB § 328 Rn. 123; BeckOK
BGB/Janoschek, 58. Ed. 1.5.2021 Rn. 5, BGB § 328 Rn. 5; MüKoBGB/Gottwald, 8. Auf. 2019, BGB § 328
Rn. 261.
201
Hence a valid knock-for-knock clause under German law must always contain a
“carve-out” for this damages.
In contrast, the attempt to govern rights and claims of third parties that are part
of the Contractor’s Group or Employer’s Group would likely not as such render
the knock-for-knock clause invalid, because it does not directly afect either of the
parties to the contract. But it is important to note, that neither of the parties to the
contract would be able to refer to the knock-for-knock clause between them, if a
third party were to make claims against either of them.
4 Available case law

Knock-for-knock clauses have not yet been the subject of any published German court
case or arbitral award, so it remains to be seen how a German court would treat the
above illustrated legal issues. There are many uncertainties in this regard, starting with
the question of whether the clause will be considered part of general terms and condi-
tions, to the question to what extent the court will accept the clause as common industry
practice and fnally what limits will be placed on the efects of the clause.
5 Practical consequences and advice for use of knock-for-knock

clauses under German law
Considering the above and the uncertainty given that the above addressed issues have
not been decided on in the context of a knock-for-knock clause subject to German
law by German courts or tribunals, it is very difcult to give fully reliable advice and
predict a precise outcome to its validity under German law.
Nevertheless, if the contract shall be governed by German law, the best approach
seems to be to avoid agreeing on knock-for-knock clauses by way of standard terms,
but rather to enter into an individual agreement.
To do that, it would be advisable to:
1 Never refer to standard contracts only, but to always customise a contract
for the specifc purposes.
2 Before entering into a contract under German law, allow the other side to
propose and negotiate the wording of the knock-for-knock clause. By doing
so, it is more likely that an individual agreement as described above will be
assumed. When negotiating a contract the drafting party should always
express its willingness to really negotiate, i.e., no clause of the contract, in
particular the clauses on liability, should be “set in stone” from the beginning
and throughout the proceedings of negotiations.
3 Document that negotiations took place, ideally by way of clear written cor-
respondence between the parties. One could refer to “common custom” and
the advantages of a knock-for-knock clause to demonstrate that both parties
really wanted to implement this particular liability regime and be a part of
the contract.
4 If necessary, change the knock-for-knock clause to be compatible with Ger-
man law or, alternatively, choose English law to govern the contract as a
whole or parts of the contract, the liability regime (if so, preferably as a
202
whole) in particular. A partial choice of law clause is permissible under
German law, if the division is clear and does not result in contradicting
provisions within the contract as a whole. As far as the diferent sub-questions
for the chosen law regimes, however, cannot be consolidated with each other
in an appropriate manner, i.e., are deemed contradictable, the applicable law
will be determined by reverting to objective connecting factors.
Please note that these recommendations will not automatically make a knock-for-
knock clause contained in standard terms valid under German law, and legal counsel
should always be sought. However, it will strengthen the position to argue that an
individual contract was entered into, and that for this reason the strict rules of the
German Civil Code do not apply.
203
CHAP TER 11
Knock for knock under Brazilian law

1 Introduction
Historically, knock-for-knock clauses have not deserved a lot of attention from Bra-
zilian scholars. This is not due to the negligence of Brazilian scholars, but rather a
consequence of the very limited (not to say rare) historical use of knock-for-knock
clauses in Brazil, which also leads to very little case law on this matter. The adver-
sarial culture consolidated in contractual relations in Brazil also plays a relevant role
in the resistance to the type of risk allocation inherent to the nature of such clause.
Even though Brazil is a relevant global player in the ofshore oil and gas explora-
tion and production industry, where the knock-for-knock liability regime is more
commonly adopted worldwide, it has not been a practice to adopt such kind of risk
allocation regime in Brazil.
Apparently, the main reason for this is that Petrobras has maintained a de facto
monopoly over the Brazilian oil and gas exploration and production industry and
has historically opted to implement fault-based liability in its services contracts with
all kinds of suppliers, including drilling services agreements and O&M agreements.
To better illustrate the prominent position of Petrobras in the Brazilian upstream
industry, it operated 93% and owned 73% of all Brazilian oil and gas production in
November 2021,1 even after 20 years from the opening of the market.
More recently, though, this scenario started to change as Petrobras decided to
implement knock-for-knock provisions in some of its services contracts, which has
been a long-time request from services providers. In this sense, the debate around
knock-for-knock clauses has gained momentum and tends to be further developed in
the coming years. The fact that Petrobras has been conducting a relevant divestment
program, which includes several upstream assets, is also opening the market for new
operators that tend to drive an increase in the use of knock-for- knock provisions.
Although Petrobras’ decision to adopt the knock-for-knock regime has been well
received by services providers in Brazil, the risk allocation refected in the contracts
ofered by Petrobras in its bid processes has received signifcant critics from the market.
In part, such critics are related to the fact that knock-for-knock provisions appear
as the centrepiece of a wider range of specifc indemnities, exclusions and limited
1 See Ministry of Mining and Energy, Brazilian oil and gas production report, prepared yearly (Agên-
cia Nacional de Petróleo, Gás Natural e Biocombustíveis – ANP), available at www.gov.br/anp/pt-br/
centrais-de-conteudo/publicacoes/boletins-anp/boletim-mensal-da-producao-de-petroleo-e-gas-natural.
204 DOI: 10.4324/9781003206798-13

knock for knock under brazilian law
liability clauses which, depending on the wording, may undermine the cornerstone
of knock-for-knock clause.
Therefore, the purpose of this chapter is to contribute to a better understanding on
the nature of knock-for-knock clauses and how it fts in the Brazilian legal framework
with special focus on the Brazilian oil and gas industry.
Section 2 of this chapter presents the civil liability regime in Brazil to get an overview
of the diferent types of liability recognized by Brazilian law, to better understand the
issues surrounding knock-for-knock provisions such as (1) contractual liability and
torts liability, (2) fault-based liability and strict liability and (3) the concept of damages.
Section 3 describes how limitation of liability clauses are regulated in Brazil
with special focus on issues around its validity and efectiveness. For purposes of
this chapter, we adopted a wide concept of limitation of liability clauses, including
clauses that restrict or fully release a party from a liability it would bear should the
contract was silent.
Section 4 departs from the classifcation of knock-for-knock provisions as a limi-
tation of liability clause to discuss some relevant issues arising from its application
under Brazilian law. We also analyse the link between knock-for-knock provisions
and parties’ potential insurance strategy, forming the so-called knock-for-knock
system.
Section 5 is dedicated to two relevant conclusions. Firstly, Brazilian law, in general,
provides a supportive backdrop to knock-for-knock clauses. Secondly, knock-for-
knock clauses are well suited for some operations, especially in the oil and gas indus-
try, which may lead to their increasing use by Petrobras as well as by the increasing
number of independent operators in the Brazilian upstream industry.
2 Overview of civil liability regime in Brazil

Knock-for-knock provisions are deemed as limitation of liability clauses in Brazil,
which aim to contractually limit or release the liability of a party that would other-
wise bear such liability. Before entering a more detailed analysis of knock-for-knock
provisions, it is relevant to set some important considerations regarding key aspects
of civil liability under Brazilian law.
2.1 Contractual liability and tort liability

The obligation to indemnify arises when a duty is breached. The basis for the duty
may either be a pre-existing and voluntary legal relationship (a contract) or a general
legal duty (the law).2
As many other jurisdictions of both civil law and common law traditions, Brazilian
law recognizes two diferent categories of civil liability in connection with the basis of
such liability: (1) contractual liability, when the liability arises from the non-fulflment
of a contractual obligation and (2) torts liability or non-contractual liability, when
the liability arises from the breach of duty provided by law.
2 Sergio Cavalieri, Programa de Responsabilidade Civil (2021) p. 16.
205
felipe t. boechem and márcio opromolla
Contractual obligations are categorized between (1) obligation of means and (2)
obligation of result.
Where there is an obligation of a result, liability arises if a party fails to achieve or
deliver the result as contractually agreed. For example, a seller is liable for delivering
the goods which was sold; a constructor must deliver the construction according to
the schematics and perfectly usable conditions and so forth.
Where there is an obligation of means, no liability arises for lack of delivery of the
result. There is only liability for failure to take the steps reasonably required to achieve
the expected result. Some examples of obligations of means are typically those of
physicians (to cure a patient) and of lawyers (defending their clients’ interests) – both
must employ their best eforts and technical expertise to achieve the intended result
but are not liable (except for a failure to employ best eforts and technical expertise)
for the outcome.
Non-contractual liability is divided between (1) subjective liability and (2) strict
liability, each of those containing its own subdivisions3 (which will be described in
further detail below).
Brazilian scholars usually highlight two key diferences between contractual and torts
liability. First, in contractual liability, one party breaches a “positive” duty (one which
requires action or omission), while in tort liability the ofender breaches a “negative”
duty of not causing harm (e.g., a duty of care). Second, in contractual liability, as a
general rule, (1) guilt is presumed for obligations of result, so the defaulting party
need to prove that another rule exempted them from performance to avoid or mitigate
its indemnifcation obligation; and (2) guilt must be proven by the creditor in case of
obligations of means. In torts liability, as a general rule, guilt is not presumed and
must be proven by the creditor/victim of non-contractual obligation.4 The distinction
above, however, has nuances, as there are several rules that apply to both contractual
and non-contractual liability (arts. 393,5 4026 and 4037 of the BCC). Moreover, there
are peculiarities concerning consumer relations under Brazilian Consumer Law (which
has its own set of rules and diferent principles), which is not within the scope of this
chapter.
2.2 Fault-based liability and strict liability

There are certain requirements that must be met for liability to arise whether under
contractual liability or non-contractual liability. These requirements are slightly
3 Cavalieri, p. 21.
4 Important exceptions pertaining to strict liability are applicable, as will be addressed in in this chapter.
5 “Art. 393. The debtor is not liable for losses resulting from a fortuitous event or force majeure, unless
he has expressly made himself liable therefor.”
6 “Art. 402. Except where otherwise expressly provided for by law, the losses and damages owed to the
creditor cover, in addition to what he efectively lost, that which he reasonably failed to proft.”
7 “Art. 403. Even where non-performance results from the debtor’s wrongful conduct, losses and dam-
ages only include efective losses and lost proft that are the direct and immediate efect of non-performance,
without prejudice to the provisions of the legislation governing procedure.”
206
diferent for each but follow essentially the same structure: liability arises when one
proves that (1) a damage occurred and (2) a conduct breached either law or contract
(give or take additional requirements).
Notwithstanding, liability can also be divided between (1) fault-based liability or
subjective liability (responsabilidade subjetiva), which has traditionally been the general
liability regime under BCC; and (2) strict liability (responsabilidade objetiva), which,
although exceptional, has been gaining more applicability since the new BCC was
enacted in 2002.
Fault-based liability takes place when the following criteria are identifed: (1) a
“guilty” conduct (intention or negligence), (2) causation, and (3) a damage. The bur-
den of proof of these elements is generally placed on the plaintif (creditor/injured
party) (art. 373, I8 of the Brazilian Code of Civil Procedure9).
As to strict liability, Brazilian law adopts the “theory of risk”, whereby agents are
deemed to have assumed the risk embedded in their activity and are therefore liable even
when there is no “guilt” (intention or negligence; e.g., art. 927, sole paragraph,10 and 93111
BCC). Therefore, one need only to prove the damage and causation to entail liability.
In this sense, strict liability is triggered by (1) evidence of the act or omission (i.e.,
evidence of its occurrence, and not of the fault or intention motivating it); (2) the
damage allegedly sufered by the party; and (3) the link of causation between the
conduct and the damage. Strict liability applies to cases specifed in the law such as
consumer rights under Law No. 8,078/1990 and environmental matters pursuant to
Law No. 6,938/81) or when the ofender’s activity, by its nature, implies inherent risks
to third parties (e.g., a carrier of hazardous or fammable substances).
2.3 Indemnifable losses

The BCC sets out the general legal regime regarding indemnifcation of damages in con-
nection with both contractual liability and torts liability. Accordingly, only actual damages
(danos emergentes) and loss of profts (lucros cessantes) supported by the aggrieved party
which are considered a direct and immediate efect (por efeito dela direto e imediato) of
the breach of a legal or contractual obligation by the other party are indemnifable.
Therefore, hypothetical damages (meaning those that are not a direct consequence of the
conduct) are not indemnifable under Brazilian law. Allowing indemnifcation for hypotheti-
cal damages would ultimately create a situation of unjust enrichment (art. 884 BCC12).
8 “Art. 373. The burden of proof is on: I – the plaintif, as to the constitutive facts of his or her right.”
9 Federal Law no. 13.105 of 16 March 2015.
10 “Art. 927. Anyone who, through an illicit act (arts. 186 and 187), causes damage to another is obligated
to repair it. Sole paragraph. The obligation to repair the damage will exist, regardless of fault, in the cases
specifed by law or when the activity normally carried out by the person who caused the damage entails, by
its nature, risk to the rights of others.”
11 “Art. 931. Save for other cases provided for in special legislation, sole proprietors and companies are
liable for the damage caused by the products put into the market, regardless of fault.”
12 “Art. 884. Anyone who, without just cause, enriches themselves at the expense of others, will be
obliged to repay the unduly earned, after updating the monetary values.”
207
Actual damages or positive damages consist of the efective reduction or loss of
assets sustained by the aggrieved party. A typical example is the damage sustained
by a car in the case of a trafc accident (the damages payable being the amount
needed to repair the vehicle).
Loss of profts (lucros cessantes) or negative damages consist of values that an
injured party fails to receive/earn due to the illicit act. For example, in the case of
the trafc accident, the taxi driver involved in the accident may claim loss of profts
for the revenue lost due to inactive days. Loss of profts may also be present when
a company loses revenues due to a breach of contract, for example.
As one can note, the concepts of “direct damages”, “indirect damages” and “loss
of profts” are relatively broad and are not further detailed in the law. Therefore, the
defnition of which damages are indemnifable will need to be assessed on a case-by-
case basis considering the facts and evidence of each specifc case.
It is worth highlighting that the expression “consequential loss” is not used by Brazil-
ian law. In this sense, if the expression “consequential loss” is used in a contract – which
is often the case in the Brazilian oil and gas industry – it is recommended to include
a defnition of its content in the contract to mitigate disputes around its meaning.
Indemnifable damages under Brazilian law can also be classifed in the following
broad categories: material, moral, aesthetic, loss of a chance, and social or difuse.
Other categories exist but are seldomly applied.
Material damages (danos materiais) are losses that afect the physical property of
a person, legal entity, or depersonalized entity.
Moral damages (danos morais) are due when there is a breach of personal rights,
such as the right to privacy, freedom, sexual or religious preferences and so forth. The
monetary indemnifcation for this kind of damage seeks to partially mitigate the distress
and emotional consequences of an immaterial harm which is caused to an injured party.
Aesthetic damages (danos estéticos) are treated as a separate form of moral damage.
As a rule, these damages occur when a person sufers wounds, scars, cuts, injury or
loss of organs, amputations, or other anomalies that afect human dignity. It concerns
an existential scope of the aggrieved party and aims at repairing the dread stemming
from this sort of damage. Such damages are presumed (in re ipsa) and need no proof
other than the demonstration of the damage.
The loss of a chance (perda de um chance) occurs when an aggrieved party sufers
a frustration of a reasonable expectation or of a future opportunity (that would take
place to a degree of certainty) due to the illicit act. Had circumstances followed their
normal course, chances would not have been lost and the aggrieved party would
reasonably and to a degree of certainty be in a diferent fnancial situation. It is a
case law construction and is not frequently relied upon by plaintifs.
Therefore, for this sort of damages to be considered indemnifable, the relevant
chance must be serious and real13 (proof of which must be made by the plaintif).
Therefore, liability for loss of a chance is contingent on the degree of certainty of the
13 “Civil liability for loss of chance is not limited to the category of moral damages, since, depending
on the circumstances of the specifc case, the lost chance may also have the legal nature of material damage.
The chance must be serious and real, not being restricted to a priori percentages” (Pronouncement No. 444,
V Jornada de Direito Civil).
208
event that was frustrated due to the illicit act. For this reason, the lost chance must
represent much more than a simple subjective hope, and proof is usually complex.
Social damages are difuse/collective, involving rights in which the victims are inde-
terminate or indeterminable. These damages are usually related to consumer issues,
environmental and other aspects of collective nature. Difused rights are those that
deserve special protection, as they do not afect anyone in particular and, at the same
time, everyone. Examples of difused rights are the rights to a healthy environment,
the prohibition against misleading advertising and the right to public safety.
3 Limitation of liability clause

The limitation of liability clause changes the ordinary consequences of a contractual
default and contractually implement between the parties only an agreed risk alloca-
tion that difers from the risk allocation that would apply should the parties remain
silent in the contract.
It enables the parties to make an efcient management of the project risks and
reallocate them to enable a project that otherwise would not be feasible or would be
feasible in a less efcient arrangement. In this sense, the non-liability clause is aligned
with the main social and economic function of contracts, which is the promotion of
the efcient fow of market relations.
In Brazil, there is no piece of legislation that sets forth a general prohibition or
imposes conditions to a limitation of liability clause. Consequently, pursuant to the
freedom to contract principle,14 the parties are entitled to agree on the non-liability
clause. In specifc and justifed situations, the Brazilian law forbids such type of provision,
such as in consumer and transportation contracts. In consumer contracts, the natural
unbalance between the consumer and the supplier can justify such legal prohibition.
Most Brazilian scholars and judicial precedents recognize the validity of the non-
liability clause as a legitimate manifestation of contractual freedom, but they subject
such validity to the following conditions:
1 Both parties shall agree with the limitation of liability clause (bilateral con-
sent), not being admitted unilateral declarations of a party with the purpose
of limiting its own liability.
2 There must be some contractual balance among the parties, which would
prevent such provisions, for example, in most adhesion contracts and employ-
ment agreements.
3 It cannot be agreed with the purpose of excluding or transferring essential obliga-
tions of the parties as otherwise it would signifcantly modify the main contractual
obligation without which the contract would not have been executed.
4 It does not limit the liability in case of wilful misconduct or gross negligence
under the basis that the duty to indemnify for losses arising from acts per-
formed with wilful misconduct or gross negligence is a principle of civil liabil-
ity with status of public order.
5 It does not result in violation of the public order.
14 Article 5, II of the Brazilian Federal Constitution.
209
For the purposes of this chapter, we will provide a few considerations on the fourth
and ffth conditions, which trigger more debates. The majority of Brazilian scholars
generally argue that the provision limiting the indemnifcation obligation in case of
wilful misconduct or gross negligence (1) would correspond to a contractual exemp-
tion for the party to default its contractual obligation as if, in practice, the contractual
obligation was not contracted and, consequently would be a violation of the parties’
good faith duty which is expressly recognized under the Brazilian Civil Code; and
(2) could lead to a situation in which signifcant damages caused by such acts would
remain unpunished, so such provision would promote undesirable social impacts and
be against the public order.
The frst argument is refuted by some scholars who support that the indemnifcation
obligation is not the only available remedy to enforce a party to fulfl its contractual
obligations. There are other means to enforce a contractual obligation, such as the
possibility of the non-defaulting party seeking an injunction, claiming contractual
fees, terminating the contract or not fulflling its own obligations due to such default.
According to these scholars, the other remedies cannot be ignored when discussing
the validity of the limitation of liability clause.
Regarding the second argument, it shall be pointed out that the concept of public
order is abstract and vague. Unfortunately, Brazilian scholars do not clearly defne
the content of public order. Notwithstanding, two basic ideas are usually involved
in the discussions of this matter.
The frst idea is that the concept of public order mandatorily involves a public
interest that cannot be waived. Under this perspective, the argument to support
the invalidity of the provision excluding the indemnifcation obligation in case of
wilful misconduct or gross negligence would be that it creates incentives for the
parties to adopt negligent behaviours in the fulflment of their contractual obliga-
tions that could have material adverse impacts not only on the counterparty but
also on third parties (externalities). Although one may say that this argument is
reasonable, it cannot be applied to all and any situation without analysing the
specifcities of each case.
The second idea is that, in the context of contract interpretation, the public
order shall be construed as an exception to the principle of contractual freedom
which is a fundamental pillar of the theory of contracts. Following the general
interpretation principles, the interpretation of an exception shall be restrictive,
which means that it should avoid expanding the situations covered by such excep-
tion. It is necessary to demonstrate that the situation under analysis would be
detrimental to the public interest because ultimately it is on behalf of the public
interest (and not the private interests) that it will be allowed a potential judicial
intervention to invalidate the non-liability clause. Expanding the concept of public
order to situations in which there is no clear public interest to be protected could
result in unfair decisions.
More recently, the arguments for the validity of limitation of liability provisions
were strengthened by Law 13,874/2019, which included a new provision in the Bra-
zilian Civil Code to state that civil and commercial contracts shall be presumed as
balanced, unless there are actual elements to the contrary, and the risk allocation
defned in such contracts shall be respected and observed.
210
4 Knock-for-knock clauses in Brazil
Knock-for-knock clauses are contractual provisions used worldwide, which basically
defne that each party shall bear the damages caused to its people, goods, equipment,
and properties regardless of whether such damages were caused by the other party.
To quote a simple example, see the following language used by Petrobras in a drilling
contract:15
The CHARTERER undertakes to indemnify PETROBRAS Group for any and all Dam-
ages involving:
i. The CHARTERER GROUP personnel.
ii. goods or facilities owned by the CHARTERER Group
PETROBRAS undertakes to indemnify the CHARTER Group from any and all Dam-
ages involving:
i. PETROBRAS Group personnel.
ii. goods or facilities owned by PETROBRAS Group.
From the analysis of the knock-for-knock provisions, it is possible to extract some

relevant considerations about them.
First, the knock-for-knock provision has the nature of a limitation of liability
clause under Brazilian law, as it releases the party that caused the damages from its
duty to indemnify the party that sufered such damages. Please note that the general
liability regime set forth by Brazilian legal system imposes on the ofender the duty to
indemnify the damages caused by its acts or omissions, so the knock-for-knock clause
changes the ordinary liability allocation provided by law. As previously mentioned, it
shall be emphasized that, as a general principle of contractual law, knock-for-knock
provisions are binding among the parties to the contract only.
Second, knock-for-knock clauses in the oil and gas industry are usually structured
with an initial provision covering damages to people and property as the one men-
tioned above and subsequent clauses covering other specifc issues like pollution.
In drilling services agreements, for example, it is common to have specifc provisions
regulating (1) pollution arising from the equipment above the rotatory table, which is
usually allocated to the contractor; (2) pollution arising from equipment owned by the
contracting party, equipment below under the rotatory table, wells and reservoir, which
are usually allocated to the contracting party; and (3) big risks such as damages to wells,
reservoirs and other geological formations, as well as damages arising from blowout
and the use of radioactive substances, which are usually allocated to contracting party.
Some Brazilian scholars argue that the principle of the human dignity, which is
defned as a fundamental principal of the Brazilian Constitution, prevents the pos-
sibility of limitation of liabilities for damages in connection with the life and physical
integrity of an individual and, consequently, the validity of knock-for-knock clauses
could be challenged specifcally regarding the release from damages caused to people.
We understand, though, that the release, operated by a knock-for-knock provision
in connection with damages caused to people, should be valid and enforceable under
Brazilian law. The knock-for-knock provision inserted in commercial contracts does
15 “Look before you leap: are your oil patch liability clauses enforceable? (An analysis under civil law
jurisdictions with emphasis on Brazil)” (2021) Journal of World Energy Law and Business 14, 49, at 65.
211
not restrict any legal recourse that an individual may have against any of the parties
to the contract as such commercial contract is binding among its parties only (does
not bound such individual). For example, if “Party A” causes damage to an employee
of “Party B”, such employee will be entitled to seek indemnifcation from Party A
(torts liability) and, if Party A pays for such indemnifcation, Party A will be entitled
to seek reimbursement from Party B. Having said that, we understand that the human
dignity principle is not, as general rule, undermined by knock-for-knock provisions.
Third, knock-for-knock clauses usually cover both contractual liability and torts
liability. Therefore, “Party A” would contractually bear the damages to its personnel
and equipment regardless of whether such damages were caused by Party A, another
party to the contract or a third party.
Fourth, although the knock-for-knock provision is agreed in a contract between
its parties (i.e., contractor and contracting), it usually covers not only damages to
equipment and personnel of such parties but also damages to the equipment and
personnel of their respective group (controlling companies, controlled companies,
and companies under common control). In this sense, contractor will bear all the
damages not only to its own equipment and personnel but also to the equipment
and personnel of other companies of the “contractor group”.
Fifth, companies usually opt to contractually exclude the applicability of a knock-
for-knock provision in case of wilful misconduct of the party that caused the
damage. However, there is often negotiation around limiting such cases to wilful
misconduct of the senior supervisory personnel (SSP) of the parties and the defni-
tion of the SSP. In such cases, contractors tend to push for higher thresholds (if
possible excluding people working in loco), while the contracting parties tend to
push for lower thresholds.
Finally, it is important to give some special consideration to the interconnection
between knock-for-knock provisions and the insurance strategy of the parties. By
agreeing to traditional knock-for-knock liability, the parties somehow recognize that
they will bear the risk of their own equipment and people and, therefore, defne their
risk and insurance strategy.
The allocation of liability promoted by knock-for-knock provisions simplifes the
distribution of risks by, for example, avoiding several discussions on who caused the
damages and the extent of its liability. Such risks are then more easily identifable
and channelled to one of the parties, which enables an efcient management of risk,
including by means of insurance. That is why experts on this topic often refer to
knock-for-knock provisions and insurance coverage as the “knock-for-knock system”.
5 Conclusion
Based on the analysis above, it is possible to extract two relevant conclusions.
First, Brazilian law provides a supportive backdrop to knock-for-knock clauses,
even though it contains certain broad and vague concepts that may be somehow
used to restrict the applicability of such provisions. As mentioned during this paper,
Brazilian law supports the freedom to contract and, consequently, the allocation of
risks contracted by sophisticated parties in balanced contractual position. Having
said that, the Brazilian oil and gas industry, especially in ofshore exploration and
212
production, tends to be a feasible feld for the development of the knock-for-knock
provision. Notwithstanding, special and careful attention should always be given to
the specifcities of each case.
Second, although the knock-for-knock provisions have only started to be more
frequently used in the Brazilian oil and gas industry in the past few years, we pre-
sume its use will increase considering (1) its benefts in terms of simplifcation of
risk allocation and the possibility of developing a more rational and efcient insur-
ance coverage, (2) Petrobras’ openness to use the clauses in its contracts and (3)
more independent companies are starting to operate in Brazilian upstream industry
demanding the use of these clauses.
213
CHAP TER 12
The implications of litigation post Deepwater

Horizon on knock-for-knock clauses in US law
Cindy Matherne Muller, Jeanne Amy, and Jennifer David
1 Introduction
For many decades, operations in the oil and gas industry have been conducted under
agreements with indemnity obligations that, with some variations, require a party to
indemnify the counterparty for certain claims when caused by the counterparty’s
negligence. This is known in the industry as “knock-for-knock” indemnity obligations,
which are often agreed to with respect to personal injury and property damage claims
that could be brought by each party, their respective afliates, contractors, and
employees against the counterparty. This indemnity scheme is designed to create a
“no-fault” system for litigation and places the risk of the loss on the party who car-
ries frst-party insurance coverage for the respective losses. The indemnity scheme
also avoids having two potential tortfeasors litigate the fault of each other to the
beneft of the injured party and his attorney.
For ofshore operations of the coast of the United States, the threshold issue in
determining the enforceability and scope of any indemnity provision is what law will
apply. The choice of law issue is not as simple as honoring the choice of the par-
ties. Instead, a practitioner has to consider the potential application of federal law,
maritime law, the Outer Continental Shelf Lands Act (OCLSA),1 or state law. The
applicable law could be outcome determinative, as various states have enacted “anti-
indemnity” statutes, which, in varying degrees, render indemnity obligations which
purport to indemnify a party against its own negligence unenforceable as a matter
of public policy. These statutes vary in breadth with respect to the unenforceability
of the indemnity and related insurance obligations. Care must be taken in drafting
indemnity and insurance provisions to take advantage of any statutorily permissive
structures to avoid unenforceability. Key to any drafting or enforcement is an under-
standing of what law could, or is most likely to, apply.
2 Applicable law
A state or federal court in the United States will have to determine the applicable
law to any contract containing indemnity provisions. Courts may consider the parties’
choice of law in the contract, but certain limitations on the parties’ freedom of
contract are imposed by the OCSLA. The conficts of law issue is complicated by
1 43 U.S.C. § 1331 et seq.
214 DOI: 10.4324/9781003206798-14

the implications of litigation post deepwater horizon
dual state and federal legal systems under which federal law can pre-empt state law.
Moreover, a federal court could have jurisdiction either under federal question, in
which a federal statute is applied, or over a controversy on the basis of diversity of
citizenship of the parties, or under admiralty jurisdiction. The basis of jurisdiction
could afect the confict of law analysis employed by the court.
2.1 Federal court jurisdiction

Article III of the US Constitution is the basis for the formation and jurisdiction of
US federal courts. Article III states:
The judicial Power shall extend to all Cases, in Law and Equity, arising under this Con-
stitution, the Laws of the United States, and Treaties made, or which shall be made,
under their Authority; . . . to all Cases of admiralty and maritime Jurisdiction; – to
Controversies . . . between Citizens of diferent States.2
This jurisdiction is commonly broken into diferent categories, including federal

question jurisdiction, admiralty jurisdiction, and diversity jurisdiction.
2.1.1 Federal question jurisdiction

Lower federal courts have limited jurisdiction and are empowered to hear only those
cases authorized by a federal statute, the US Constitution, or a US treaty.3 This is
referred to as “federal question” jurisdiction. A district court (trial court) has federal
question jurisdiction when an action arises under the Constitution, laws, or treatises
of the United States.4 A case arises under federal law if a well-pleaded complaint
establishes that either (1) federal law creates a cause of action or (2) the plaintif’s
right to relief necessarily depends of the resolution of a substantial question of
federal law.5 For most cases brought under federal question jurisdiction, the suit
arises under the federal law that created the cause of action.6 As a general rule, state
courts have concurrent jurisdiction over federal claims with the federal courts unless
(1) Congress provides that the federal courts have exclusive jurisdiction or (2) there
is a clear incompatibility between the state court jurisdiction and the federal interest.7
When jurisdiction is concurrent, a cause of action can be pursued in either state or
federal court. Admiralty and maritime cases8 and Jones Act maritime personal injury
cases9 are subject to the concurrent jurisdiction of state and federal courts.
2.1.2 Diversity jurisdiction

Federal courts have diversity jurisdiction when the suit involves a controversy between
parties of diverse citizenship and the amount in controversy exceeds $75,000.10 When
2 U.S. Const. art. 3, § 2, cl. 1.

3 Kokkonen v Guardian Life Ins. Co. of Am., 511 U.S. 375, 377 (1994).
4 U.S. Const. art. 3, § 2; 28 U.S.C. § 1331.
5 Empire HealthChoice Assur., Inc. v McVeigh, 547 U.S. 677, 689–690 (2006).
6 Merrell Dow Pharmaceuticals, Inc. v Thompson, 478 U.S. 804, 808 (1986).
7 Gulf Ofshore Co., Div. of Pool Co. v Mobil Oil Corp., 453 U.S. 473, 478 (1981).
8 28 U.S.C. § 1333; see Ofshore Logistics, Inc. v Tallentire, 477 U.S. 207, 222–223 (1986) (“savings to
suitors” clause allows state courts to entertain in personam maritime causes of action).
9 Engel v Davenport, 271 U.S. 33, 37 (1926); see 46 U.S.C. §§ 30104, 30105.
10 28 U.S.C. § 1332(a).
215
cindy matherne muller, jeanne amy, and jennifer david
jurisdiction is based on diversity, there must be complete diversity of citizenship between
the parties, which requires that no plaintif can be a citizen of the same state as any
defendant.11 Determining diversity of citizenship is factually based and diferent rules
apply to determine the citizenship of corporations and limited liability companies.
Moreover, when a court is sitting in diversity jurisdiction, the court must apply the
conficts of laws rules under the law of the forum state, or the state in which the federal
court is located.12 The choice of law rules of a forum state are varied, but many follow
an interest analysis under which the court applies the law of the state with the most
interest in the controversy. Such a determination is also highly fact dependent.
2.1.3 Admiralty jurisdiction

Under 28 U.S.C. § 1333(1), US federal district courts are granted “original jurisdic-
tion, exclusive of the courts of the United States, of any civil case of admiralty or
maritime jurisdiction.” This grant includes jurisdiction “over all contracts . . . which
relate to the navigation, business, or commerce of the sea.”13 A maritime contract is
“one that relates to the ship in its use as such or to commerce or to navigation on
navigable waters or to transportation by sea or to maritime employment.”14 Types
of contracts which have been held to invoke maritime jurisdiction include, but are
not limited to: contracts to furnish services, supplies, or accessories to a particular
vessel; charter parties; contracts for carriage of goods and afreightment; stevedoring
contracts; contracts for the carriage of passengers by water; maritime insurance
contracts; salvage contracts; towage contracts; and mortgages covered by the Ship
Mortgage Act of 1920,15 cargo claims subject to the Carriage of Goods by Sea Act,16
and the Harter Act of 1893.17
Maritime tort jurisdiction is based either upon statute or by law, a judicial test
based upon situs and status which includes relationships to and efect upon maritime
commerce. The US Supreme Court noted that the boundaries of admiralty jurisdic-
tion over contracts – as opposed to torts or crimes – are “conceptual rather than
spatial” and have always been difcult to draw.18 Contracts containing both maritime
and non-maritime elements (“mixed contracts”) are maritime contracts for the pur-
poses of federal admiralty jurisdiction where the focus of the contract is maritime
regardless of the locale.
In recent cases, the Supreme Court noted that the fundamental interest giving rise
to maritime jurisdiction is the protection of maritime commerce.19
11 Owen Equipment & Erection Co. v Kroger, 437 U.S. 365, 374 (1978).
12 Klaxon Co. v Stentor Elec. Mfg. Co., Inc., 313 U.S. 487 (1941); Erie R.R. Co. v Tompkins, 304 U.S.
64 (1938).
13 De Lovio v Boit, 7 Fed. Cas. 418 (C.C.D. Mass. 1815) (No. 3776).
14 See J.A.R., Inc. v M/V LADY LUCILLE, 963 F.2d 96 (5th Cir. 1992).
15 Now known as the Commercial Instruments and Maritime Lien Act, 46 U.S.C. § 31301 et seq.
16 46 U.S.C.S. § 30701.
17 46 U.S.C. §§ 30701–30707.
18 Kossick v United Fruit Co., 365 U.S. 731, 735 (1961).
19 Norfolk Southern Ry. Co. v James N. Kirby, Pty Ltd., 543 U.S. 14, 25 (2004) (citing Exxon Corp. v
Cent. Gulf Lines, Inc., 500 U.S. 603 (1991)).
216
2.2 Applicable law
Work on the Outer Continental Shelf is performed on a wide variety of structures,
including vessels, drill ships, jack-up rigs, lift boats, semi-submersible platforms,
single point anchor reservoir (SPAR) and tension leg structures, foating production
platforms (FPP), and foating production and storage and ofoading (FPSO).
Whether these structures are considered vessels is a key component to any choice-
of-law analysis, as the use of a vessel is a critical factor in determining whether a
contract is maritime and whether maritime law, the OCSLA, or state law might
apply.
2.2.1 Outer Continental Shelf Lands Act

The OCSLA makes it clear that federal law applies ofshore:
The Constitution and laws and civil and political jurisdiction of the United States are
extended to the subsoil and seabed of the outer Continental Shelf and to all artifcial
islands, and all installations and other devices permanently or temporarily attached to
the seabed, which may be erected thereon for the purposes of exploring for, developing,
or producing resources therefrom or any such installation or other device (other than a
ship or vessel) for the purposes of transporting such resources, to the same extent as if
the outer Continental Shelf were an area of exclusive Federal jurisdiction located within
a State.20
In the Energy Policy Act of 2005, Congress amended the OCSLA to authorize the
regulation of the development of ofshore energy from sources other than oil and
gas. Under 30 C.F.R. § 585.100, this authority was delegated to the Bureau of Ocean
Energy Management, which exercises its authority in multiple phases for leasing,
siting, and permitting. Recently, in the National Defense Authorization Act for 2021,
Congress amended the OCSLA to extend its application in certain instances to “non-
mineral energy resources,” which removed any question that the OCSLA would apply
to ofshore wind development in the United States.
Congress recognized that federal law is incomplete – e.g., there is no federal law of
contract, tort, liens, etc. Therefore, Congress adopted the law of the adjacent state
as surrogate federal law where necessary to fll in gaps in federal law:
To the extent that they are applicable and not inconsistent with this subchapter or
with other Federal laws and regulations of the Secretary now in efect or hereafter
adopted, the civil and criminal laws of each adjacent State, now in efect or hereafter
adopted, amended, or repealed are hereby declared to be the law of the United States
for that portion of the subsoil and seabed of the outer Continental Shelf, and artifcial
islands and fxed structures erected thereon, which would be within the area of the
State if its boundaries were extended seaward to the outer margin of the outer Con-
tinental Shelf, and the President shall determine and publish in the Federal Register
such projected lines extending seaward and defning each such area. All of such
applicable laws shall be administered and enforced by the appropriate ofcers and
courts of the United States. State taxation laws shall not apply to the outer Conti-
nental Shelf.21
20 43 U.S.C. § 1333 (a)(1).

21 43 U.S.C. § 1333(a)(2)(A).
217
The Supreme Court confrmed that Congress intended to have adjacent state law
apply rather than have federal courts create a new federal common law.22 In Chevron,
the Supreme Court explained:
Congress specifed that a comprehensive body of state law should be adopted by the
federal courts in the absence of existing federal law. . . . Congress made clear provision
for flling in the “gaps” in federal law; it did not intend that federal courts fll in those
“gaps” themselves by creating new federal common law.23
Importantly, courts have consistently held that Congress’s choice of the substantive
laws of the adjacent state as federal law is a mandatory choice of law, which applies
irrespective of the parties’ choice of law in a contractual provision or the confict of
laws rules of the adjacent state. In Matte v Zapata Ofshore Co.,24 the Fifth Circuit
did not enforce a choice-of-law provision contained in a non-maritime contract opt-
ing for the application of general maritime law. The court held that enforcement of
the parties’ choice of law would be in violation of the federal policy expressed in the
OCSLA: that the law of the adjacent state be applied as surrogate federal law.
Similarly, in Union Texas Petroleum Corp. v PLT Engineering, Inc., the court stated:
UTP argues that even if OCSLA applies, the parties have chosen admiralty law through
the choice of law provisions in their contracts. This argument can not prevail.
Although Louisiana’s choice of law rules might enforce this choice of law provision,
OCSLA will not. We fnd it beyond any doubt that OCSLA is itself a Congressionally
mandated choice of law provision requiring that the substantive law of the adjacent state
is to apply even in the presence of a choice of law in the contract to the contrary.25
Courts have rejected the argument that application of an adjacent state’s law requires
applying that state’s choice of law principles.
Having decided that state law should apply, it becomes necessary to determine which
state’s substantive law governs. Rodrigue mandates that the law of the adjacent state,
in this case Texas, applies. The Defendants have made an argument that in applying
Texas law, this Court should apply Texas’ conficts of law rules for contracts. Defendants
argue that Texas’ conficts rules would require that Kentucky’s substantive law should
apply because the contract has little connection with Texas. However, while enticing
on common sense grounds, this argument must legally fail. In Chevron Oil Company v
Huson, supra, the Supreme Court held that a State’s confict of law rules have no rel-
evance in an OCSLA case when a Federal court is applying adjacent state law as sur-
rogate Federal law.26
Indeed, the Supreme Court in Chevron stated:

It was the intent of Congress, expressed in the Senate Committee Report, in the Confer-
ence Report, and on the foor of the Senate, that state laws be “adopted” or “enacted”
as federal law. Thus a federal court applying Louisiana law under § 1333 (a)(2) of the
Lands Act is applying it as federal law – as the law of the federal forum. Since the federal
22 Chevron Oil Co. v Huson, 404 U.S. 97 (1971).

23 Id. at 104–105.
24 784 F.2d 628 (5th Cir. 1986).
25 Union Texas Petroleum Corp. v PLT Engineering, Inc., 895 F.2d 1043, 1050 (5th Cir. 1990) (footnote
and citations omitted).
26 Walter Oil & Gas Corp. v NS Group, Inc., 867 F. Supp. 549, 553 (S.D. Tex. 1994) (footnote omitted).
218
court is not, then, applying the law of another forum in the usual sense, ordinary confict
of laws principles have no relevance.27
Thus, when a federal court has federal question jurisdiction under the OCSLA, it
will apply applicable federal law. Where federal law is silent, however, the law of the
adjacent state will apply.
2.2.2 State waters

If a contract is performed on a platform in state territorial waters rather than on
the Outer Continental Shelf, and the work has no maritime connection, the OCSLA
will not operate as a choice-of-law trump card. Instead, when the OCSLA does not
apply, a federal court sitting in diversity jurisdiction must follow the choice of law
rules of the forum state.28
2.2.3 Maritime law

Whether maritime law applies requires the establishment of maritime jurisdiction,
an inquiry which usually depends on the nature of the claim:
[I]n the context of oil and gas exploration on the Outer Continental Shelf, admiralty
jurisdiction and maritime law will only apply if the case has a sufcient maritime nexus
wholly apart from the situs of the relevant structure in navigable waters.29
To establish maritime jurisdiction for a tort, a party must show maritime situs and
a connection to traditional maritime activity.30 If the dispute arises out of a maritime
contract, i.e., the nature of the contract is maritime, then maritime law applies rather
than state law. A maritime contract has been described as one “relating to a ship in
its use as such, or to commerce or navigation on navigable waters, or to transporta-
tion by sea or to maritime employment.”31
2.2.3.1 Maritime contracts

Although ofshore oil and gas exploration on the Outer Continental Shelf has been
ongoing for more than 50 years, the question of whether a contract is maritime in
the oil and gas context has been murky and often conficting. Recent jurisprudence
has simplifed the test for determining a maritime contract. In 2004, the Supreme
Court held that courts must examine the primary objective of a contract when deter-
mining whether the contract is maritime.32 In part based on this pronouncement, the
Fifth Circuit announces a new test to determine whether an oil and gas service
contract is maritime in nature. The Fifth Circuit issued a unanimous en banc decision
in In re Larry Doiron, Inc.,33 which provided that to determine whether an oil and
27 Chevron, 404 U.S. at 102.

28 Roberts v Energy Dev. Corp., 104 F.3d 782 (5th Cir. 1997); Klaxon, 313 U.S. at 496 (holding that a
federal court sitting in diversity must apply the confict of law rules of the state within which it sits).
29 Laredo Ofshore Constructors, Inc. v Hunt Oil Co., 754 F.2d 1223, 1230 (5th Cir. 1985).
30 Champagne v Tetra Applied Technologies, Inc., No. 05–299, 2006 U.S. Dist. LEXIS 37457, *13–14
(S.D. Tex. Feb. 6, 2006).
31 M/V Lady Lucille, 963 F.2d at 98.
32 Norfolk Southern Ry. Co. v James N. Kirby, Pty Ltd., 543 U.S. 14 (2004).
33 879 F.3d 568 (5th Cir. 2018) (en banc), cert. denied, 138 S. Ct. 2033 (2018).
219
gas service contract is maritime, a court must ask (1) whether the contract provides
services to facilitate the drilling or production of oil and gas on navigable waters
and (2) whether the parties expect that vessel will play a substantial role in the comple-
tion of the contract.34 There, the decision dealt with determining the maritime nature
of contracts dealing with exploration and production, but the court stated that the
test could apply to non–oil and gas activity that involved maritime commerce.35
Thereafter, the Fifth Circuit expanded the Doiron test and announced that it applies
to all mixed-services contracts, not just those arising in the oil and gas context.36 In
Barrios, the plaintif was injured while ofoading a generator from a crew boat to a
barge. The plaintif’s employer had executed a master service contract with recipro-
cal indemnity provisions with the owner of a dock for repair work to be performed
at the dock facility, including the installation of a concrete containment rail. To do
the work, the plaintif’s employer chartered a barge, which moved up and down the
river using a tugboat and a winch. The plaintif’s employer contracted with the barge
owner to transport crew and equipment to the barge via a crew boat.
The issue before the court in Barrios was whether the master service agreement
was a maritime contract. If the master service contract was a maritime contract, then
the indemnity provision would be enforceable under the general maritime law. But
if the underlying contract was non-maritime in nature, then Louisiana law would
invalidate the indemnity provision. The district court determined that the contract
was a “land-based construction contract,” and therefore governed by Louisiana law.
On appeal, the court reviewed whether the dock contract was maritime in nature.
To make that determination, the Fifth Circuit grappled with the appropriate test to
apply to this non–oil and gas contract. The court noted that only one district court
had applied the Doiron test to a mixed-services contract.37 The Fifth Circuit found
that Doiron should apply as written because it streamlines the inquiry and is consis-
tent with the Supreme Court’s decision in Norfolk Southern Railway Co. v Kirby, 543
U.S. 14 (2004), which requires courts to look to the primary objective of a contract
when determining whether the contract is maritime. Therefore, the court employed
a broader test to determine whether the dock construction contract in Barrios was
maritime in nature.
The test enunciated by the Barrios court is as follows: “To be maritime, a con-
tract (1) must be for services to facilitate activity on navigable waters and (2) must
provide, or the parties must expect, that a vessel will play a substantial role in the
completion of the contract.” Applying that test to the facts of the case, the Fifth
Circuit determined that the frst prong was met because the construction of a concrete
containment rail was to facilitate activities on navigable waters. The dock extended
into the Mississippi River and the containment rail was designed to prevent materi-
als from spilling into the river. On the second prong, the court determined that the
parties contemplated that the substantial use of a vessel, namely a barge that would
be shifted using a tugboat and winch, would be required to complete the contract.
34 Id. at n. 576.
35 Id. at n. 52.
36 Barrios v Centaur, LLC, 942 F.3d 670 (5th Cir. 2019).
37 See Lightering LLC v Teichman Group, LLC, 328 F. Supp. 3d 625, 627–629 (S.D. Tex. 2018).
220
Accordingly, the Fifth Circuit reversed the district court and remanded the case,
holding that the master service contract was maritime in nature.
Accordingly, under Fifth Circuit jurisprudence, “[t]o be maritime, a contract (1)
must be for services to facilitate activity on navigable waters and (2) must provide,
or the parties must expect, that a vessel will play a substantial role in the completion
of the contract.”38
2.2.3.2 What is a vessel?

The second prong of the test enunciated in Doiron, required a determination of
whether equipment used ofshore for energy exploration and production or construc-
tion of any ofshore facility is a vessel. Under maritime law, a “vessel” is defned as
every “water-craft or other artifcial contrivance used, or capable of being used, as
a means of transportation on water.”39 When determining vessel status, the relevant
inquiry is “whether the watercraft’s use ‘as a means of transportation on water’ is a
practical possibility or merely a theoretical one.”40 “[A] ‘vessel’ is any watercraft
practically capable of maritime transportation regardless of its primary purpose or
state of transit at a particular moment.”41 “[A] watercraft is not ‘capable of being
used’ for maritime transport in any meaningful sense if it has been permanently
moored or otherwise rendered ‘practically incapable’ of transportation or movement.”42
Accordingly, the vessel status inquiry hinges on whether the vessel is practically
capable of its maritime operations and “transport[ing] persons or things over water.”43
In Lozman, the Supreme Court held that a foating houseboat capable of being
towed was not a vessel and emphasized that “[n]ot every foating structure is a ves-
sel,” and that the statutory defnition codifed at 1 U.S.C. § 3 “applies to an ‘artifcial
contrivance . . . capable of being used . . . as a means of transportation on water.’”44
The Supreme Court cited the meaning of “transportation” as involving the “convey-
ance (of things or persons) from one place to another” and noted that this defnition
must be applied “in a practical, not theoretical, way.”45 The Supreme Court held that
a structure does not fall within the scope of this statutory phrase unless a reasonable
observer, looking at the foating structure’s physical characteristics and activities,
would consider it capable to a practical degree of serving as a special purpose vessel.46
In Lozman, the Supreme Court held that a structure is not a vessel merely because
it is “capable of foating, moving under tow, and incidentally carrying even a fair-sized
item or two when they do so.”47 Instead, the structure must be “practically capable”
of being used in maritime commerce as a means of transportation for persons or
goods on water. The court focused on the physical features of the “vessel,” including
38 Barrios, 942 F.3d at 680.

39 Stewart v Dutra Constr. Co., 543 U.S. 481, 489 (2005) (citing 1 U.S.C. § 3).
40 Id. at 496.
41 Id. at 497.
42 Id. at 493; see also Riley v Alexander/Ryan Marine Servs. Co., 983 F. Supp. 2d 884, 887 (S.D. Tex. 2013).
43 Lozman v City of Riviera Beach, 568 U.S. 115, 121 (2013).
44 Id. at 121.
45 Id.
46 Id.
47 Id. at 121.
221
whether it had a rudder, steering mechanism, unraked hull, electrical connection to
land, portholes or regular doors and windows, and any means of self- propulsion – all
things that a “reasonable observer” would conclude that it was not designed to trans-
port persons or things over water. The “reasonable observer” test in Lozman makes
any determination of what is a vessel a fact-intensive inquiry, leading to uncertainty.
Post-Lozman decisions are instructive on how to apply the Supreme Court’s test for
evaluating vessel status. In Martin v Fab-Con, Inc.,48 the court addressed whether a
quarters barge used as a foating hotel for workmen at a job site qualifes as a “ves-
sel.” The quarters barge did not have a rudder or steering mechanism, was incapable
of self-propulsion, was spudded in place, and had remained stationary for the past
several years.49 The court reasoned that a reasonable observer would not believe that
the quarters barge was designed or capable to a practical degree for transporting
people or cargo over water.50 Accordingly, the court found that the quarters barge
was not a “vessel” for purposes of general maritime law.51
The Fifth Circuit’s post-Lozman decision in Baker v Director, Ofce of Workers’
Compensation Programs52 is also instructive as to the application of Lozman principles
to ofshore drilling structures. In Baker, the plaintif was injured while building a
housing module for use on a tension leg ofshore oil platform named BIG FOOT.53
The injury occurred while the BIG FOOT was berthed at a waterside marine fabrica-
tion yard in Houma, Louisiana.54 Ultimately, the Fifth Circuit concluded that BIG
FOOT was not a vessel. In support of its reasoning, the Fifth Circuit noted that,
like the foating house in Lozman, BIG FOOT had no means of self-propulsion,
had no steering mechanism or rudder, and had an unraked bow.55 Furthermore,
BIG FOOT could be moved only when under tow, and after being towed to the
Outer Continental Shelf, it was not anticipated that BIG FOOT would relocate for
approximately 20 years.56 Although BIG FOOT was required to carry a captain and
crew when under tow, those individuals were present only to ensure BIG FOOT
was safely towed to its permanent location on the OCS.57 In sum, the Fifth Circuit
found that “a reasonable observer, looking to BIG FOOT’s physical characteristics
and activities, would not consider it designed or capable to a practical degree for
carrying people or things over water.”58
Like in Baker, most determinations of vessel status are not made at the Supreme
Court. Instead, Fifth Circuit jurisprudence provides guidance on what devices used
in ofshore oil and gas exploration have been found to be vessels.59
48 9 F. Supp. 3d 642, 649 (E.D. La. 2014).

49 Id.
50 Id.
51 Id. at 651.
52 834 F.3d 542 (5th Cir. 2016).
53 Id. at 544.
54 Id.
55 Id. at 547.
56 Id.
57 Id.
58 Baker, 834 F.3d at 547.
59 Single Point Anchor Reservoir (SPAR) platforms are not vessels. Fields v Pool Ofshore, 182 F.3d
353 (5th Cir. 1999); Mendez v Anadarko Pet. Corp. 466 F. App’x 316 (5th Cir. 2012); Tension Leg Platforms
222
2.2.4 Maritime law incomplete, supplemented with state law
Courts have recognized the need for national uniformity in the substantive law govern-
ing maritime commerce. Where federal maritime law is incomplete and the issue does
not require national uniformity, courts may apply state law.60 In Wilburn Boat v Fire-
man’s Fund Ins.,61 the Supreme Court held that where there is no established general
maritime law on a particular issue, and the subject does not require national uniformity,
admiralty courts may look to the rule of the state in which the court is sitting. There,
the court held that state law will control the interpretation of a marine insurance policy
only in the absence of a federal statute, judicially fashioned remedy, or need for national
uniformity. As discussed below, state law typically governs the interpretation of insur-
ance policies that are often used “hand in glove” with indemnity provisions.
2.2.5 Tension between OCSLA and maritime law

In addition to the complications of whether state or federal courts could hear a
controversy, and the diferent bases of federal court jurisdiction, given the nature of
ofshore operations, an unavoidable tension exists between the application of the
OCSLA and maritime law. Claims and contractual disputes arising out of work
performed ofshore in the United States often involve determining the tension between
the application of admiralty and maritime law and the OCSLA.
In Union Texas Petroleum Corp. v PLT Engineering, Inc.,62 the Fifth Circuit set
forth a three-part test which must be met for adjacent state law to apply rather than
maritime law: (1) the controversy must arise on a “situs covered by the OCSLA” (i.e.,
on the subsoil, seabed, or artifcial structures permanently or temporarily attached
thereto); (2) federal maritime law must not apply of its own force; and (3) the
adjacent state’s law must not be inconsistent with federal law. With the recent Fifth
Circuit decision in Doiron, it appears that the Doiron test will be applied to answer
the second question of the Union Texas Petroleum test. If a vessel plays a substantial
role in the completion of the contract, it is likely that maritime law will apply of
its own force, which would prevent application of the law of the adjacent state as
surrogate federal law.
As should be apparent, the determination of applicable law, whether maritime
law or the law of the adjacent state as surrogate federal law under the OCSLA, is a
fact-intensive inquiry that is not always predictable. The applicable law matters and
may well determine whether a knock-for-knock indemnity provision is enforceable.
are not vessels. Warrior Energy Servs. Corp. v ATP Titan, 551 F. App’x 749 (5th Cir. 2014); Baker v Dir.,
OWCP, 834 F.3d 542 (5th Cir. 2016); Mooney v W&T Ofshore, Inc., 2013 AMC 1480 (E.D. La. 2013). Semi-
submersible drilling units or mobile ofshore drilling units (MODU) are vessels. Ofshore Co. v Robinson, 266
F.2d 769 (5th Cir. 1959); In re Oil Spill, Deepwater Horizon, 745 F.3d 157 (5th Cir. 2014). Drill ships are ves-
sels, and Floating Production Storage and Ofoading Vessels (FPSOs) are vessels. BW Ofshore USA, LLC
v TVT Ofshore AS, Case No. 14–1052, 2015 U.S. Dist. LEXIS 153840 (E.D. La. Nov. 13, 2015). Jack-up
rigs are generally considered vessels. Texas Co. v Gianfala, 222 F.2d 382 (5th Cir.), reversed on other grounds,
350 U.S. 879 (1955). Although jack-ups have been held not to be vessels in navigation when attached to the
sea bottom within OCSLA jurisdiction. Barker v Hercules Ofshore, Inc., 713 F.3d 208, 2013 AMC 946 (5th
Cir. 2013); Demette v Falcon Drilling Co., 280 F.3d 492, 2002 AMC 686 (5th Cir. 2002).
60 The Tungus v Skovgaard, 358 U.S. 588, 1959 AMC 813 (1959).
61 348 U.S. 310, 1955 AMC 467 (1955).
62 895 F.2d 1043 (5th Cir. 1990).
223
3 Indemnity in maritime
Indemnity clauses in maritime contracts are usually enforceable in accordance with
their clear and unequivocal terms. A contract to indemnify another for his own
negligence imposes an extraordinary obligation.63 Long-established principles of
general maritime law require that indemnifcation for an indemnitee’s own negligence
be clear and unequivocally expressed.64 The clear terms will be enforced even when
the negligence of the party seeking indemnity contributed to or caused the injury if
the terms are clear that the indemnity was intended to apply without regard to the
negligence of the indemnifed party.65
3.1 Enforceability of indemnity provisions in a maritime contract

As a matter of frst impression, the court in In re Oil Spill by the Oil Rig “Deepwater
Horizon” confronted, among other things, whether a contractual indemnity agreement
that provides indemnity for a party even if caused by that party’s gross negligence
in a drilling contract is enforceable under maritime law.66 That case arose out of the
20 April 2010 explosion and fre onboard the DEEPWATER HORIZON, which
resulted in a signifcant loss of life and millions of gallons of oil that spilled into
the Gulf of Mexico.
Transocean, as the owner of the DEEPWATER HORIZON, and BP as the operator
entered into a drilling contract that purportedly required BP to defend and indemnify
Transocean for pollution emanating below the surface of the water, even in situations
where Transocean is found to be strictly liable for the discharge or the pollution was
caused by Transocean’s negligence or gross negligence.67 Both Transocean and BP fled
cross-motions for partial summary judgment to determine whether such a provision
was enforceable under maritime law.68
Transocean argued that the intent of the parties must be honored and the provi-
sion should be enforced as written, so BP’s indemnity obligation should extend to
compensatory damages, punitive damages, and statutory penalties owed under the
Clean Water Act (CWA).69 BP admitted that it owed indemnity for Transocean’s fault
or negligence, but denied that it owed indemnity for claims arising under a strict
liability theory (like that of a statutory requirement under the CWA) or for Trans-
ocean’s gross negligence.70 BP also argued that public policy invalidates the indemnity
provision because a party cannot obtain indemnity for its own gross negligence or
damages assessed for punitive damages or strict liability, i.e., civil penalties assessed
pursuant to the CWA.71
63 Stewart v Grand Isle Shipyard, Inc., Case No. 10–4016, 2011 U.S. Dist. Lexis 148603 (E.D. La. Decem-
ber 23, 2011).
64 Id.
65 Theriot v Bay Drilling Corp., 783 F.2d 527 (5th Cir. 1986).
66 841 F. Supp. 2d 988, 1000 (E.D. La. 2012).
67 Id. at 992.
68 Id. The parties did not dispute that maritime law applied to the drilling contract. Id. at 994.
69 Id. at 992; 33 U.S.C. §§ 1251 et seq.
70 Id. at 992–993.
71 Id.
224
The court addressed whether indemnifcation for gross negligence and punitive
damages was against public policy.72 First, the court tackled the issue of gross negli-
gence, which raised an inherent confict between “freedom of contract, which weighs
in favor of enforcing the indemnity, and a reluctance to encourage grossly negligent
behavior, which weighs against enforcing the indemnity.”73 On the public policy front,
the court reasoned that the drilling contract allocated risk to both parties such that
Transocean would have been incentivized not to act in a grossly negligence manner.74
When considering freedom of contract, the court determined that the parties had
equal bargaining power in negotiating the drilling contract.75 Therefore, the court held
that if Transocean is determined to be grossly negligent in causing the pollution, it
could seek contractual indemnity from BP.76 Accordingly, a contractual indemnity
provision that indemnifes for one’s own gross negligence is likely to be valid and
enforceable as it relates to compensatory damages.
The court next examined whether the contractual indemnity provision was enforce-
able as it relates to punitive damages.77 The court noted that the purpose of punitive
damages is “to punish the defendant for egregious conduct, teaching him not to do it
again, and to deter others from engaging in similar behavior.”78 Yet, “[i]n situations
where the defendant’s conduct is so extreme as to merit an award of punitive damages,
the cost of such must be placed upon the party responsible, and not transferred to
a party innocent of any wrongdoing.”79 Based on this reasoning, the court held that
Transocean could not shift the burden of punitive damages to BP, thereby rendering
the indemnity provision as it relates to punitive damages void and unenforceable.80
Finally, the court considered whether BP owned indemnity for civil penalties that
could be assessed under the CWA against Transocean.81 The CWA imposes a civil
penalty on “[a]ny person who is the owner, operator, or person in charge of any
vessel, onshore facility, or ofshore facility from which oil or a hazardous substance
is discharged” where the discharge is “harmful to the public health or welfare or the
environment of the United States.”82 The court reasoned that similar to the aim of
punitive damages, the primary goals of the CWA “are to punish and deter future
pollution.”83 In the same way that public policy prohibits indemnifcation for puni-
tive damages, a provision that seeks to indemnify for civil penalties assessed under
the CWA is void and against public policy.84
72 In re Oil Spill by the Oil Rig “Deepwater Horizon”, 841 F. Supp. 2d at 997.
73 Id. at 1000.
74 Id. at 1000–1001.
75 Id. at 1001.
76 Id. at 1003.
77 Id.
78 In re Oil Spill by the Oil Rig “Deepwater Horizon”, 841 F. Supp. 2d at 1003 (quoting Daughdrill v Ocean
Drilling & Expl., 665 F. Supp. 477, 481 (E.D. La. 1987)) (internal quotation marks omitted).
79 Id. (quoting Daughdrill, 665 F. Supp. at 482) (internal quotation marks omitted).
80 Id.
81 Id.
82 33 U.S.C. § 1321(b)(3), (4), and (7).
83 In re Oil Spill by the Oil Rig “Deepwater Horizon”, 841 F. Supp. 2d at 1004.
84 Id. at 1006.
225
In sum, in contracts where maritime law applies and the parties are on equal foot-
ing in negotiation, an indemnity provision that includes indemnifcation for gross
negligence will likely be enforceable. If the parties seek to indemnify for punitive
damages or strict liability under a statutory scheme designed to punish or deter
certain behavior, then it is unlikely to be enforceable.
3.2 Indemnity under the Longshore and Harbor Workers’ Compensation Act
The Longshore and Harbor Workers’ Compensation Act (LHWCA)85 invalidates
indemnity agreements where an employer is bound to indemnify a vessel owner
for the vessel’s negligence.86 The LHWCA is a workers’ compensation scheme
whereby an employer is strictly liable to an injured employee if a qualifed employee
is injured on the job. The statute also provides a means for an employee to recover
against a vessel owner directly for an injury “caused by the negligence of a vessel”
under § 905(b) of the LHWCA. That same provision provides that “the employer
shall not be liable to the vessel for such damages directly or indirectly and any
agreements or warranties to the contrary shall be void.”87 The aim of the statute
is to hold the negligent vessel owner liable for its own negligence instead of shift-
ing liability via contractual indemnity to the employer, who is subject to strict
liability for its injured employee. While this type of risk allocation may be rendered
unenforceable under the LHWCA, courts have held that insurance protection is
available for vessel owners named as additional insureds on the employer’s liability
insurance policies.
For example, in Voisin v O.D.E.C.O. Drilling Co., the Fifth Circuit held that an
additional insured provision did not contravene § 905(b) of the LHWCA.88 There, an
employee of a contractor was injured aboard an ofshore drilling vessel of the coast
of Texas.89 The injured employee sued the employer-contractor and the owner of the
drilling vessel, Odeco Drilling Inc. (Odeco), for vessel negligence and unseaworthi-
ness.90 Odeco fled a third-party suit against the employer for breach of a Master
Service Contract between Odeco and the employer, asserting that the employer failed
to name Odeco as an additional assured as required by the agreement.91 Indeed, the
Master Service Contract required that the employer name Odeco as an additional
assured on its liability policies with a waiver of subrogation.92 The agreement also
had a “broadly worded indemnity” provision in favor of Odeco “to the maximum
extent permitted by law and to support such indemnity by liability insurance cover-
age to be furnished by [the employer].”93
85 33 U.S.C. § 901 et seq.

86 33 U.S.C. § 905(b).
87 Id.
88 744 F.2d 1174 (5th Cir. 1984).
89 Id. at 1175.
90 Id.
91 Id.
92 Id.
93 Id. at 1176 & n. 1.
226
Odeco later settled with the personal injury plaintif, but maintained its claim
against the employer.94 The trial court found that the indemnity and additional
insured provisions under the Master Service Contract were void under the LHWCA
and held that Odeco was not entitled to recover against the employer for the sums
paid to the personal injury plaintif.95 On appeal, Odeco conceded that § 905(b) of the
LHWCA invalidates an indemnity that requires an employer to indemnify for vessel
negligence, but urged that the LHWCA does not invalidate an additional assured
provision.96 The Fifth Circuit agreed with Odeco and held that § 905(b) does not
render an additional assured provision unenforceable.97
The Fifth Circuit looked to the legislative history of § 905(b) of the LHWCA.98
The statute was amended in 1972 to include § 905(b), but prior to the amendment,
the Supreme Court had rendered decisions that permitted vessel owners to recover
against employers for vessel negligence.99 Congress enacted § 905(b) to preclude this
indirect liability of employers for vessel negligence resulting from an injury to an
employee.100 The court reasoned that an additional assured provision is not a similar
form of indirect liability prohibited by § 905(b).101 The Fifth Circuit held that the
employer breached the Master Service Contract by failing to name Odeco as an
additional assured as required by the agreement.102
Section 905(c) of the LHWCA provides a remedy for longshoremen to pursue a
cause of action for vessel negligence against a vessel owner for injuries sustained on
the Outer Continental Shelf and for which an injured employee can recover under §
4 of OCSLA.103 The LHWCA does not preclude, however, the enforcement of “any
reciprocal indemnity provision whereby the employer of a person entitled to receive
benefts under this Act . . . and the vessel agree to defend and indemnify the other
for cost of defense and loss or liability for damages arising out of or resulting from
death or bodily injury to their employees.”104
3.3 Insurance obligations and additional assured status

Generally, under federal law in the Fifth Circuit, insurance obligations take precedence
over indemnity obligations. In Ogea v Lofand Bros.,105 Ogea, an employee of Inter-
national Hammer, was injured working on a drilling rig operated by Lofand Brothers
and located on a fxed platform owned by Phillips. Lofand fled a third-party demand
94 Voisin v O.D.E.C.O. Drilling Co., 744 F.2d at 1176.

95 Id.
96 Id.
97 Id. at 1179.
98 Id. at 1177.
99 Id. (citing and quoting Longmire v Sea Drilling Corp., 610 F.2d 1342 (5th Cir. 1980) (discussing Seas
Shipping Co. v Sieracki, 328 U.S. 85 (1946) and Ryan Stevedoring Co. v Pan Atlantic Steamship Corp., 350
U.S. 124 (1956)).
100 Voisin, 744 F.2d at 1177.
101 Id.
102 Id. at 1179.
103 33 U.S.C. § 905(c).
104 Id.
105 622 F.2d 186 (5th Cir. 1980).
227
against Phillips seeking indemnity. Phillips asserted a counterclaim against Lofand
alleging that Lofand was obligated to obtain liability insurance covering Phillips,
and that its failure to do so constituted a breach of contract.
The specifc indemnity provisions in the contract required Phillips to defend and
indemnify Lofand for all claims for personal injury brought by Phillips and its con-
tractors’ personnel. However, Phillips pointed to an additional assured provision in the
contract, which required Lofand to obtain insurance of $500,000 under which Phil-
lips was named as an additional assured. The Fifth Circuit concluded that the parties
clearly intended that Phillips would not be held liable for any injuries incurred on its
ofshore platform up to $500,000; therefore, the $500,000 additional assured cover-
age provided by Lofand to Phillips must be exhausted frst. However, the indemnity
obligations owed by Phillips to Lofand did not come into play because the claim
was settled for $60,000. Consequently, Ogea left open the possibility that if a claim in
excess of $500,000 had been brought, exhausting the insurance provided to Phillips,
then the indemnity obligation that Phillips owed to Lofand would be implicated.
Later, in Tullier v Halliburton Geophysical Services, Inc.,106 the parties to a time
charter for a vessel used in the ofshore oil and gas industry agreed to indemnify
each other for job-related liability and to back up the cross-indemnity provisions with
insurance. The charterer agreed to obtain CGL insurance with appropriate maritime
endorsements. The vessel owner agreed to provide P&I insurance under the SP23
form with the “as owner” clause deleted. Last, the contract required that all insur-
ance obtained by the vessel owner shall name the charterer as an additional assured
and shall provide that it shall be primary insurance with respect to the charterer,
irrespective of any excess or other insurance clauses.
From these provisions, the Fifth Circuit concluded in Tullier that the vessel’s insur-
ance policy was intended to specifcally cover the charterer as an additional assured
with the “as owner” limitations deleted. The policy was also intended to constitute
primary coverage for the charterer. The Fifth Circuit relied on Ogea and other Fifth
Circuit jurisprudence for the proposition that a party who has entered into a con-
tractual indemnity provision, but who also names the indemnitor as an additional
assured under its own liability policies, must frst exhaust the liability insurance it
agreed to obtain before seeking contractual indemnity.
After these cases, the oil and gas and maritime industries operating within the
reach of the Fifth Circuit adapted. Most contracts now include language that limit
obligations to name as additional assureds and waive subrogation “to the extent of
the indemnity obligations assumed by Contractor under this agreement.” Market par-
ticipants in the ofshore industry understand the necessity of such limiting language
and do not normally provide resistance to the addition of such language.
The efectiveness of this language was tested in litigation between BP and Trans-
ocean and its underwriters related to the Macondo well blowout. Following the spill,
BP sought coverage as an additional assured from Transocean for both the liabilities
assumed in the drilling contract by Transocean and for BP’s sub-surface well pollution –
liability that was expressly assumed by BP under the drilling contract. Transocean’s
106 81 F.3d 552, 1996 AMC 1561 (5th Cir. 1996).
228
insurers fled a declaratory judgment action to limit BP’s additional assured coverage
to only the obligations assumed by Transocean under the drilling contract, which
expressly excluded sub-surface pollution. The Fifth Circuit certifed a question to the
Texas Supreme Court on BP’s additional assured status.107
The Supreme Court of Texas answered that BP was an additional assured under
Transocean’s liability policies, but only to the extent of the liabilities that Transocean
assumed in the drilling contract. The court found that Transocean’s contractually assumed
liability did not include the sub-surface well pollution that resulted from the oil spill.
Specifcally, the Texas Supreme Court determined that the insurance policies incorporated
the drilling contract by reference by including an additional assured endorsement that
names other parties as additional assureds “as required by an Insured Contract.” By
incorporating the drilling contract by reference, the insurance policies required that the
court consult the drilling contract to determine BP’s additional assured status. Upon
consulting the drilling contracts, the court held that BP’s additional assured status was
inexorably linked, at least in some respect, to the extent of Transocean’s indemnity obli-
gations. Therefore, the court held that coverage aforded to BP, as an additional assured,
was limited to the obligations assumed by Transocean under the drilling contract alone,
which did not include the additional pollution and environmental risk.
4 Application of state law

The enforceability of a contract’s indemnity provision ultimately turns on the law
governing the contract and whether a state’s anti-indemnity statute applies. Most US
states have anti-indemnity statutes, so when contracting for work to be performed in
a particular state, great care should be taken to understand the parameters of each
state’s particular provisions.108 Pertinent to the oil and gas industry, Louisiana,109
Texas,110 New Mexico,111 and Wyoming112 each have oilfeld anti-indemnity statutes
that, under certain circumstances, restrict indemnity between companies and contrac-
tors. When it comes to ofshore oilfeld disputes, however, the oilfeld anti-indemnity
statutes of Texas and/or Louisiana are naturally more often implicated, considering
those states border the US Gulf of Mexico. Accordingly, the body of case law inter-
preting the application of anti-indemnity statutes in the ofshore context largely stems
from Texas, Louisiana, and the Fifth Circuit. For that reason, the oilfeld anti-
indemnity statutes of Texas and Louisiana are more fully summarized herein.
4.1 The Louisiana Oilfeld Indemnity Act

The Louisiana Oilfeld Indemnity Act (LOIA)113 was passed to address the concerns of
oilfeld contractors over the “widespread inclusion” of indemnity obligations in master
107 In re Horizon, 470 S.W.3d 452 (Tex. 2015).

108 See Appendix A for survey of anti-indemnity statutes across the United States.
109 The Louisiana Oilfeld Indemnity Act, La. R.S. § 9:2780.
111 New Mexico Oilfeld Anti-Indemnity Act (NMOAIA), N.M. Stat. Ann. § 56–7–2.
112 Wyoming Oilfeld Anti-Indemnity Act (WOAIA), Wyo. Stat. § 30–1–131.
113 La. R.S. § 9:2780.
229
service agreements, which oil companies require contractors to sign.114 The indemnity
provisions exposed oilfeld contractors to worker’s compensation and tort liability, gener-
ally without an equal bargaining position with the oil companies requiring the master
service agreement.115 To that end, the Louisiana legislature enacted LOIA to restrict
defense and indemnity obligations in certain contracts relating to the drilling industry.
Specifcally, LOIA declares null and void and against public policy any provision
in a contract “pertaining to a well for oil, gas, or water, or drilling for minerals” to
the extent that it provides for defense and indemnity “for damages arising out of
or resulting from death or bodily injury to persons, which is caused by or results
from the sole or concurrent negligence or fault (strict liability) of the indemnitee.”116
In other words, LOIA only applies to contracts pertaining to a well and for loss or
claims arising out of personal injury or death, not property damage. Louisiana and
Fifth Circuit courts broadly interpret LOIA when evaluating indemnity claims.117
When determining whether LOIA applies to a contract, courts apply a two-step
process: (1) the contract must pertain to a well and (2) the contract must be related
to exploration, development, production, or transportation of oil, gas, or water.118
With respect to the frst prong, whether the contract pertains to a well is a fact-
intensive inquiry analyzed on a case-by-case basis.119 While the focus of the inquiry
is the location of the work required under the contract, a court will examine the
following factors:
1 Whether the structures or facilities to which the contract applies or with
which it is associated are part of an in-feld production system;
2 The geographical location of the structure or facility relative to a well or wells;
3 The purpose or function of the facility or structure in question;
4 The owner and operator of the relevant facility or structure, and the owner
and operator of the well or wells that produce the gas in question;
5 Any number of other details afecting the functional and geographic nexus
between a well and the structure or facility that is the object of the
agreement.120
To clarify, the services required under the contract need not be rendered on site or
in connection with drilling or operation of a specifc well to fall within the purview
of LOIA.121 Further, the fact that a contract also may pertain to numerous other
wells or to objects other than wells is of no consequence for purposes of applying
LOIA, so long as the services rendered relate to the “exploration development, pro-
duction, or transportation of oil, gas, or water.”122
114 Meloy v Conoco, Inc., 504 So. 2d 833 (La. 1987).

115 Id.
116 La. R.S. § 9:2780(B).
117 Roberts v Energy Development Corp., 104 F.3d 782, 784 (5th Cir. 1997).
118 Lloyds of London v Transcontinental Gas Pipe Line Corp., 38 F.3d 193, 196 (5th Cir. 1994).
119 Id.; see also Verdin v ENSCO Ofshore Co., 104 F. Supp. 2d 682 (W.D. La. 2000), af’d, 255 F.3d 246
(5th Cir. 2001).
120 Verdin, 104 F. Supp. 2d at 690 (W.D. La. 2000) (internal citations omitted).
121 Livings v Service Truck Lines of Texas, Inc., 467 So. 2d 595, 599 (La. App. 3 Cir. 1985).
122 Id.
230
In Marcel v Placid Oil Co.,123 the Fifth Circuit set forth a judicially created excep-
tion to LOIA, which provides that an indemnity provision in an otherwise qualifying
contract is enforceable when the indemnitor purchases insurance for the indemnitee
and the indemnitee reimburses the indemnitor for its portion of the premium. The
exception is known as the Marcel exception. Stated diferently, the Marcel exception
to LOIA requires that the indemnitee pay its portion of the premium for additional
assured status to the indemnitor, separate and apart from the contract price. In creat-
ing the exception, the Marcel court explained that the purpose of LOIA is served by
shifting the responsibility of paying for liability coverage to the indemnitee:
The LOIA is aimed at preventing the shifting of the economic burden of insurance cover-
age or liability onto an independent contractor. If the principal pays for its own liability
coverage, however, no shifting occurs. We see no need to prevent such an arrangement in
order to give efect to the LOIA.124
It is worth noting that the Marcel court stressed that the exception “does not apply
if any material part of the cost of insuring the indemnitee is borne by the indepen-
dent contractor procuring the insurance coverage.”125 Some courts have strictly con-
strued this requirement when considering the cost and amount of the premium paid
by the indemnitee.126
In addition to its oilfeld anti-indemnity statute, Louisiana also has a separate
anti-indemnity statute applicate to other certain types of contracts, including con-
struction contracts127 and certain motor carrier transportation contracts relative to
unloading, loading, transporting, or services incidental thereto (i.e., storage).128 Under
123 11 F.3d 563, 569–570 (5th Cir. 1994).

124 Id. at 569–570.
125 Id. at 570.
126 See, e.g., Amoco Prod. Co. v Lexington Ins. Co., 98–1676 (La. App. 1 Cir. 9/24/99); 745 So. 2d 676,
writ denied, 99–3553 (La. 2/25/00); 755 So. 2d 253 (holding that oil company’s payment of $2,000 for $11 mil-
lion in primary and excess liability coverage as an additional insured was “clearly inadequate consideration”
that failed to satisfy the Marcel exception to LOIA).
127 Under La. R.S. § 9:2780.1(A)(2), a “construction contract” is defned as “any agreement for the
design, construction, alteration, renovation, repair, or maintenance of a building, structure, highway, road,
bridge, water line, sewer line, oil line, gas line, appurtenance, or other improvement to real property, or repair
or maintenance of a highway, road, or bridge, including any moving, demolition, or excavation, except that
no deed, lease, easement, license, or other instrument granting an interest in or the right to possess property
will be deemed to be a construction contract even if the instrument includes the right to design, construct,
alter, renovate, repair, or maintain improvements on such real property.” The statute expressly excludes con-
tracts related to the design, construction, alteration, renovation, repair, or maintenance of “any dirt or gravel
road used to access oil and gas wells and associated facilities” as well as “[o]il fow lines or gas gathering lines
used in association with the transportation of production from oil and gas wells from the point that oil and
gas becomes co-mingled for transportation to oil storage facilities or gas transmission lines.”
128 La. R.S. § 9:2780.1(A)(1) defnes a “motor carrier transportation contract” as “any contract, agreement,
or understanding covering the transportation of property, other than agricultural products as defned in R.S.
9:3306 and timber without limitation, for compensation or hire by a motor carrier, entrance upon property
by the motor carrier for the purpose of loading, unloading, or transporting property, other than agricultural
products as defned in R.S. 9:3306 and timber without limitation, for compensation or hire, or a service inci-
dental to any such activity, including but not limited to storage of property, other than agricultural products as
defned in R.S. 9:3306 and timber without limitation, except the Uniform Intermodal Interchange and Facilities
Access Agreement administered by the Intermodal Association of North America or other agreements provid-
ing for the interchange, use, or possession of intermodal chassis, containers, or other intermodal equipment.”
231
this Louisiana Anti-Indemnity Statute,129 indemnity provisions and provisions requir-
ing an indemnitor to procure liability insurance to cover the acts of an indemnitee
contained in a construction contract or certain motor carrier transportation contracts
are null, void, and unenforceable as against public policy. In particular, La. Rev. Stat.
§ 9:2780.1 provides, in pertinent part:
Notwithstanding any provision of law to the contrary and except as otherwise provided
in this Section, any provision, clause, covenant, or agreement contained in, collateral
to, or afecting a motor carrier transportation contract or construction contract which
purports to indemnify, defend, or hold harmless, or has the efect of indemnifying,
defending, or holding harmless, the indemnitee from or against any liability for loss or
damage resulting from the negligence or intentional acts or omissions of the indemnitee,
an agent or employee of the indemnitee, or a third party over which the indemnitor
has no control is contrary to the public policy of this state and is null, void, and
unenforceable.
Notwithstanding any provision of law to the contrary and except as otherwise pro-
vided in this Section, any provision, clause, covenant, or agreement contained in, col-
lateral to, or afecting a motor carrier transportation contract or construction contract
which purports to require an indemnitor to procure liability insurance covering the acts
or omissions or both of the indemnitee, its employees or agents, or the acts or omis-
sions of a third party over whom the indemnitor has no control is null, void, and
unenforceable. However, nothing in this Section shall be construed to prevent the
indemnitee from requiring the indemnitor to provide proof of insurance for obligations
covered by the contract.
To the extent ofshore activities involve the construction, design, maintenance, or

repair of platforms, other ofshore structures, or oil and gas lines, Louisiana’s anti-
indemnity statute applicable to construction contracts may still apply. The Louisiana
anti-indemnity statute for certain construction and motor carrier transportation
contracts does contain a statutory exception under La. Rev. Stat. § 9:2780.1(I). The
statutory exception provides that an indemnity or additional assured provision may
still be enforceable where the indemnitor obtains insurance coverage only to the extent
of the indemnity obligation and there is evidence that the indemnitor recover the
cost of the required insurance in the contract price.130
129 La. R.S. § 9:2780.1.

130 The statutory exception under La. R.S. § 9:2780.1(I) specifcally states that:
Nothing in this Section shall invalidate or prohibit the enforcement of the following:
(1) Any clause in a construction contract containing the indemnitor’s promise to indemnify, defend,
or hold harmless the indemnitee or an agent or employee of the indemnitee if the contract also
requires the indemnitor to obtain insurance to insure the obligation to indemnify, defend, or
hold harmless and there is evidence that the indemnitor recovered the cost of the required
insurance in the contract price. However, the indemnitor’s liability under such clause shall be
limited to the amount of the proceeds that were payable under the insurance policy or policies
that the indemnitor was required to obtain.
(2) Any clause in a construction contract that requires the indemnitor to procure insurance or
name the indemnitee as an additional insured on the indemnitor’s policy of insurance, but only
to the extent that such additional insurance coverage provides coverage for liability due to an
obligation to indemnify, defend, or hold harmless authorized pursuant to Paragraph (1) of this
Subsection, provided that such insurance coverage is provided only when the indemnitor is at
least partially at fault or otherwise liable for damages ex delicto or quasi ex delicto.
232
4.2 Texas Oilfeld Anti-Indemnity Act
Under Texas law, indemnity obligations in oilfeld contracts that indemnify a party
for its own negligence are likewise generally void as against public policy. Specifcally,
the Texas Oilfeld Anti-Indemnity Act (TOAIA)131 states:
Except as otherwise provided by this chapter, a covenant, promise, agreement, or under-
standing contained in, collateral to, or afecting an agreement pertaining to a well for
oil, gas, or water or to a mine for a mineral is void if it purports to indemnify a person
against loss or liability for damage that:
(1) is caused by or results from the sole or concurrent negligence of the indemnitee,
his agent or employee, or an individual contractor directly responsible to the
indemnitee; and
(2) arises from:
(A) personal injury or death;
(B) property injury; or
(C) any other loss, damage, or expense that arises from personal injury, death,
or property injury.
Certain exceptions exist to the TOAIA, however, including an exception for insurance-
supported indemnities.132 To clarify, Texas law provides an exception for where the
parties “agree in writing that the indemnity obligation will be supported by liability
insurance coverage to be furnished by the indemnitor.”133 With respect to a mutual
indemnity obligation, “the indemnity obligation is limited to the extent of the cover-
age and dollar limits of insurance or qualifed self-insurance each party as indemnitor
has agreed to obtain for the beneft of the other party as indemnitee.”134 As for a
unilateral indemnity obligation, “the amount of insurance required may not exceed
$500,000.”135
Under Texas law, any indemnity clause must further comply with “fair notice”
requirements – i.e., the indemnifcation clause must be “express” and “sufciently
conspicuous.”136 The “express” requirement requires that the intent to indemnify for
a party’s negligence must be expressly stated. Practitioners have developed extensive
language stating the express causes for which indemnity will apply. As for the conspicu-
ous requirement, Texas courts look to whether the indemnity provision, on its face,
“attract[s] the attention of a reasonable person when he looks at it.”137 “Language
may satisfy the conspicuousness requirement by appearing in larger type, contrasting
colors, or otherwise calling attention to itself.”138 Generally, indemnity provisions in
Texas contracts should appear in bolded, contrasting, and/or all capitalized font in
order to ensure the provision is sufciently conspicuous for purposes of meeting the
“fair notice” test.

133 Id.; see also Ken Petr. Corp. v Questor Drilling Corp., 24 S.W.3d 344, 346 (Tex. 2000).
134 Tex. Civ. Prac. & Rem. Code § 127.005(b).
135 Tex. Civ. Prac. & Rem. Code § 127.005(c).
136 The fair notice requirements are not required if it can be shown that the party had actual knowl-
edge of the indemnity provision, such as where the provision was revised by the parties during the contract
negotiation. See, e.g., Cleere Drilling Co. v Dominion Expl. & Prod., Inc., 351 F.3d 642, 647 (5th Cir. 2003).
137 Storage & Processors, Inc. v Reyes, 134 S.W.3d 190, 192 (Tex. 2004) (internal quotations omitted).
138 Id.
233
Like Louisiana, Texas also has a Texas Construction Anti-Indemnity Act139 that
renders void certain indemnity agreements in construction contracts, as well as
related provisions that require procurement of additional-assured coverage.140 There
are several exceptions to the Texas Construction Anti-Indemnity Act. For example,
the anti-indemnity statute does not apply to a claim “for the bodily injury or death
of an employee of the indemnitor, its agent, or its subcontractor of any tier.”141 The
act also does not apply for certain other contracts, including, inter alia, contracts for
public works for a municipality.142
5 Conclusion
In the United States, the enforcement of indemnity obligations designed to
indemnify a party for its own negligence is not simple. In an ofshore setting, the
enforcement is complicated by the tension between application of federal and
state law. A substantial amount of litigation involves the fact-intensive inquiries
to determine applicable law to a particular indemnity obligation. As many agree-
ments used in other industries ofshore take the form of master service agreements,
care should be taken to ensure that indemnity and insurance provisions are drafted
to maximize enforceability under the laws most likely to apply. Under any given
work order, the facts could change, as could the location and thus, the law of
the adjacent state. Practitioners have developed provisions to ensure enforceability
under all of the likely legal regimes, through reciprocal indemnities, additional
assured provisions, and premium allocations. These provisions should be revised
or updated often to adapt to or take advantage of new considerations as juris-
prudence continues to develop. Companies should review their existing master
service agreements and examine long standing agreements to consider whether
the indemnity language drafted at the time of execution remains enforceable under
current jurisprudence.
139 Tex. Ins. Code § 151.102 provides:

Except as provided by Section 151.103, a provision in a construction contract, or in an agreement collateral
to or afecting a construction contract, is void and unenforceable as against public policy to the extent that
it requires an indemnitor to indemnify, hold harmless, or defend a party, including a third party, against a
claim caused by the negligence or fault, the breach or violation of a statute, ordinance, governmental regula-
tion, standard, or rule, or the breach of contract of the indemnitee, its agent or employee, or any third party
under the control or supervision of the indemnitee, other than the indemnitor or its agent, employee, or
subcontractor of any tier.
140 Tex. Ins. Code § 151.104.

141 Tex. Ins. Code § 151.103.
142 Tex. Ins. Code § 151.105.
234
APPENDIX A
State Anti-Indemnity Statute Is an Indemnitee Permitted to Indemnify

for its Sole Negligence?
Alabama None Permissible if the language is clear

and unequivocal. Mobil Oil Corp. v
Schlumberger, 598 So. 2d 1341 (Ala.
1992).
Alaska Alaska Stat. § 45.45.900 No. Prohibited by anti-indemnity statute.
Arizona Ariz. Rev. Stat. §§ 32– No. Prohibited by anti-indemnity statute.
1159, 34–226, 41–2586 Ariz. Rev. Stat. § 32–1159.
Arkansas Ark. Code § 4–56–104, No. Prohibited by anti-indemnity statute.
§ 22–9–214 Ark. Code §§ 4–56–104(b), 22–9–
214(b).
California Cal. Civ. Code § 2782 No. Prohibited by anti-indemnity statute.
Cal. Civ. Code § 2782.
Colorado Colo. Rev. Stat. §§ 13–50. No. Prohibited by anti-indemnity statute.
5–102, 13–21–111.5 Colo. Rev. Stat. §§ 13–50.5–102(8)(a),
13–21–111.5(6)(b).
Connecticut Conn. Gen. Stat. § 52–572k No. Prohibited by anti-indemnity statute.
Conn. Gen. Stat. § 52–572k(a).
Delaware Del. Code. Ann. 6 § 2704 No. Prohibited by anti-indemnity statute.
Del. Code Ann. 6 § 2704(a).
Florida Fla. Stat. Ann. § 725.06 Permissible only if the agreement caps
the indemnity obligation at an amount
“that bears a reasonable commercial
relationship to the contract.” Fla. Stat.
Ann. § 725.06(1).
Georgia Ga. Code Ann. § 13–8–2(b) No. Prohibited by anti-indemnity statute.
Ga. Code Ann § 13–8–2(b).
Hawaii Haw. Rev. Stat. § 431:10– No. Prohibited by anti-indemnity statute.
222 Haw. Rev. Stat. § 431:10–222.
Idaho Idaho Code Ann. § 29–114 No. Prohibited by anti-indemnity statute.
Idaho Code Ann. § 29–114.
Illinois 740 ILCS § 35/1 No. Prohibited by anti-indemnity statute.
740 ILCS § 35/1.
Indiana Ind. Code Ann. § 26–2–5 No. Prohibited by anti-indemnity statute.
Ind. Code Ann. § 26–2–5.
(Continued)
235
(Continued)
State Anti-Indemnity Statute Is an Indemnitee Permitted to

Indemnify for its Sole Negligence?
Iowa Iowa Code Ann. § 537A.5 No. Prohibited by anti-indemnity statute.

Iowa Code Ann. § 537A.5.
Kansas Kan. Stat. § 16–121 No. Prohibited by anti-indemnity statute.
Kan. Stat. § 16–121(b).
Kentucky Ky. Rev. Stat. § 371.180 No. Prohibited by anti-indemnity statute.
Ky. Rev. Stat. § 371.180.
Louisiana La. R. S. §§ 38:2216(G) Permissible if the language is clear and
(public contracts), 9:2780 unequivocal. Barnett v Am. Constr.
(oil and gas), 9:2780.1 Hoist, Inc., 2011 1261 (La. App. 1 Cir.
(construction) 02/10/12); 91 So. 3d 345.
In oil and gas contracts, see Marcel v
Placid Oil Co., 11 F.3d 563 (5th Cir. 1994).
In construction contracts, see La. R. S. §
9:2780.1(I).
Maine None Permissible if contract clearly and
unequivocally reflects a mutual intention
on behalf of the parties to indemnify for
the indemnitee’s sole negligence. Emery
Waterhouse Co. v Lea, 467 A.2d 986
(Me. 1983).
Maryland Md. Code Ann. § 5–401 No. Prohibited by anti-indemnity statute.
Md. Code Ann. § 5–401(a)(1)–(2).
Massachusetts Mass. Gen. Laws Ann. ch. No. Prohibited by anti-indemnity statute.
149 § 29C Mass. Gen. Laws Ann. ch. 149 § 29C.
Michigan M.C.L.A. § 691.991 No. Prohibited by anti-indemnity statute.
M.C.L.A. § 691.991.
Minnesota Minn. Stat. §§ 337.01, No. Prohibited by anti-indemnity statute.
337.02 Minn. Stat. § 337.02.
Mississippi Miss. Code. Ann. § 31– No. Prohibited by anti-indemnity statute.
5–41 Miss. Code. Ann. § 31–5–41.
Missouri Mo. Rev. Stat. § 434.100 No. Prohibited by anti-indemnity statute.
Mo. Rev. Stat. § 434.100.
Montana MCA § 28–2–211 No. Prohibited by anti-indemnity statute.
MCA § 28–2–211.
Nebraska Neb. Rev. Stat. § 25–21, 187 No. Prohibited by anti-indemnity statute.
Neb. Rev. Stat. § 25–21, 187(1).
Nevada N.R.S. § AB 125, § 2 Prohibited in residential construction
(applies to residential contracts. N.R.S. § AB 125, § 2.
construction contracts)
New Hampshire N.H. Rev. Stat. § 338-A:1, No. Prohibited by anti-indemnity statute.
338-A:2 N.H. Rev. Stat. § 338-A:1, 338-A:2.
New Jersey N.J. Stat. Ann. § 2A:40A-1 No. Prohibited by anti-indemnity statute.
N.J. Stat. Ann. § 2A:40A-1.
New Mexico N.M. Stat. Ann. § 56–7–1 No. Prohibited by anti-indemnity statute.
N.M. Stat. Ann. § 56–7–1(1).
236
State Anti-Indemnity Statute Is an Indemnitee Permitted to

Indemnify for its Sole Negligence?
New York N.Y. Gen. Oblig. Law § No. Prohibited by anti-indemnity statute.
5–322.1 N.Y. Gen. Oblig. Law § 5–322.1(1)–(2).
North Carolina N.C. Gen. Stat. Ann. No. Prohibited by anti-indemnity statute.
§ 22B-1 N.C. Gen. Stat. Ann. § 22B-1.
North Dakota None Permissible if contract clearly conveys the
parties’ intent. Rupp v Am. Crystal Sugar
Co., 465 N.W.2d 614 (N.D. 1991).
Ohio Ohio Rev. Code Ann. No. Prohibited by anti-indemnity statute.
§ 2305.31 Ohio Rev. Code Ann. § 2305.31.
Oklahoma Okla. Stat. title 15 § 221 No. Prohibited by anti-indemnity statute.
Okla. Stat. title 15 § 221.
Oregon Or. Rev. Stat. § 30.140 No. Prohibited by anti-indemnity statute.
Or. Rev. Stat. § 30.140(1)–(2).
Pennsylvania None No case law found.
Rhode Island R.I. Gen. Laws § 6–34–1 No. Prohibited by anti-indemnity statute.
R.I. Gen. Laws § 6–34–1.
South Carolina S.C. Code Ann. § 32–2–10 No. Prohibited by anti-indemnity statute.
S.C. Code Ann. § 32–2–10.
South Dakota S.D. Codified Laws § No. Prohibited by anti-indemnity statute.
56–3–18 S.D. Codified Laws § 56–3–18.
Tennessee Tenn. Code Ann. § 62–6– No. Prohibited by anti-indemnity statute.
123 Tenn. Code Ann. § 62–6–123.
Texas Tex. Ins. Code Ann. §§ No. Prohibited by anti-indemnity statute.
151.102, 151.103 Tex. Ins. Code Ann. § 151.102.
Utah Utah Code Ann. § 13–8–1 No. Prohibited by anti-indemnity statute.
Utah Code Ann. § 13–8–1.
Vermont None Permissible if contractual language
clearly expresses the intent to indemnify
for sole negligence of indemnitee.
Tateosian v Vermont, 945 A.2d 833
(Vt. 2007).
Virginia Va. Code Ann. § 11–4.1 No. Prohibited by anti-indemnity statute.
Va. Code Ann. § 11–4.1.
Washington Wash. Rev. Code. § No. Prohibited by anti-indemnity statute.
4.24.115 Wash. Rev. Code. § 4.24.115(1)(b).
West Virginia W. Va. Code Ann. § No. Prohibited by anti-indemnity statute.
55–8–14 W. Va. Code Ann. § 55–8–14.
Wisconsin None Permissible if provision is clear and
unambiguous. Gunka v Consolidated
Papers, Inc., 508 N.W.2d 426 (Wis. Ct.
App. 1993).
Wyoming Wyo. Stat. § 30–1–131 Permissible if clearly stated. Union Pac.
Resources Co. v Dolenc, 86 P.3d 1287
(Wyo. 2004).
237
CHAP TER 13
Mutual hold harmless clauses in France and

francophone civil law systems
1 Introduction
Although there is no general theory of risk in French law, the concept of risk occupies
a central place in the French civil law tradition.1 According to its Latin etymology,
the word ‘risk’ fnds its origins in the idea of “shared risk between contracting
parties”,2 suggesting the idea was conceived in a contractual setting.
French law has also gradually introduced the concept of risk in matters of civil
liability. Although long subordinated to the existence of fault, the principles of civil
liability have expanded to the point of recognising ‘no fault’ liability, the idea being
that where someone’s actions create a risk for third parties, the author of those
actions should be liable for the harm caused.3
The abundance of topics linked to risk highlights the importance of its place in French
civil law. In addition, technological development and the associated risks it has created
have given rise to new challenges that need to be refected in contractual relations.
In this chapter we will focus on the analysis of the principle of contractual allocation
of risk from a French civil law perspective and by extension other francophone civil law
jurisdictions. More specifcally, we will look at the acceptance within such legal systems
of so-called mutual hold harmless clauses as a technique for contractually allocating risk.
1.1 Francophone civil law systems

French law has long served as a point of reference for legislators in other jurisdic-
tions and has made its contribution to the creation of the international legal order.
In Europe, the Napoleonic campaigns resulted in the export of a codifed legal
system, as well as the institutions underpinning it. In North America, the laws of
Quebec and Louisiana fnd their origins in French civil law (as well as English com-
mon law), and in Latin America many of the States created in the 19th century drew
inspiration from the French legal system.
In Africa, French colonisation meant that French law became an essential source
of legislation at the time of independence. To this day, French law is intricately woven
into the legal fabric of most francophone African jurisdictions.
1 A. Downe, La gestion des risques contractuels par le contrat: Étude du droit français, 2021, p. 29.
2 Ibid.
3 P. Le Tourneau, Droit de la responsabilité et des contrats, Dalloz Action, 12e éd., 2020, p. 83.
238 DOI: 10.4324/9781003206798-15

mutual hold harmless clauses in francophone civil law systems
The analysis in this chapter will be limited to French law and other francophone
civil law systems in Africa.
1.1.1 The French Civil Code of 1804
1.1.1.1 French law of obligations: contract law and civil liability

Promulgated in France in 1804, the French Civil Code (or Napoleonic Code) com-
bines all of the rules governing the legal status of persons, of property and relations
between private individuals, the latter dominated by what has become known as the
‘law of obligations’.
The French Civil Code divides the law of obligations into two broad categories,
depending on whether the obligations arise from an agreement between the parties
or by operation of the law. The detailed provisions of Title III of the Code, on
contracts and contractual obligations, refected the individualistic conception of the
1804 Code, based on the autonomous will of the individual.
Title V covers “commitments formed without agreement”, which are further distin-
guished between commitments resulting “from the sole authority of the law” (such as
family relations or the protection of those designated as incapable) and those which
arise as a result of a characteristic personal to the debtor. The latter includes the
French notions of ‘quasi-contracts’ and tortious and quasi-tortious acts.
In the context of the above, parties to a contract are free to agree in advance to
limit, exclude or re-allocate their liability for failure to perform their obligations.
Despite their frequent use in day-to-day contractual relations, such agreements are
not subject to any dedicated regulation under French law, although they are subject to
certain limitations when used in the context of certain types of contracts or sectors.
1.1.1.2 Francophone African legal systems

The Civil Code of 1804 was exported to North Africa, specifcally to Algeria, Morocco,
Tunisia and Egypt, where the legislation is strongly inspired by the code. It was then
exported to sub-Saharan Africa; to the former territories of French West Africa,
namely Benin, Burkina Faso, Côte d’Ivoire, Guinea, Mali, Mauritania, Niger and
Senegal; in French Equatorial Africa, now Congo-Brazzaville, Gabon, the Central
African Republic and Chad;4 as well as Cameroon and Togo, both formerly under
French mandate. Finally, the Civil Code of 1804 was exported to the Great Lakes
region, namely Rwanda, the Democratic Republic of Congo and Burundi (former
colonies of Belgium), as well as to southern Africa where it was implanted in the
Comoros, Mauritius, Madagascar and the Seychelles.5
Following independence, a number of States opted to retain a number of principles
of the French Civil Code. This was the case in Tunisia and Morocco in 1906 and
1913, for instance, in the form of a Code on obligations and contracts.6
4 B. Bokolombe, L’infuence du modèle juridique français en Afrique, 2016, p. 75.

5 Ibid.
6 M. Grimaldi, “L’exportation du code civil” in Pouvoirs, 2003, p. 91.
239
bertrand montembault and paul morton
Generally speaking, francophone African States retain a close link to the Civil
Code of 1804, to the point where it remains in force in certain jurisdictions to this
day (e.g., in Chad, the Republic of Congo and Gabon).
1.1.2 Recent developments
1.1.2.1 French law

With few exceptions, the French Civil Code remained largely unchanged until ordi-
nance no. 2016–131 of 10 February 2016 came into efect on 1 October 2016,7
bringing with it a general reform of the law of contracts and obligations. In addition
to removing certain provisions that had become increasingly criticised as obsolete,8
the 2016 reform served above all to codify a number of concepts and jurisprudential
principles developed to respond to the shortcomings of the 1804 code.
The rules relating to contractual non-performance, which were addressed in a
scattered and incomplete way in the Civil Code, have been grouped together in the
same place in article 1217 et seq. of the Civil Code.
This reform was also intended to be followed by a reform of the law of civil liability,
a frst draft of which was published in 2017. Similarly to the 2016 reforms, the reform of
the law of civil liability is intended to modernise rules originally written more than two
centuries ago, as well as to codify a signifcant body of jurisprudence in a number of areas.
1.1.2.2 Francophone African legal systems

As noted above, with the exception of certain jurisdictions, such as the Democratic
Republic of Congo where the law of obligations is inspired by the Belgian model,
most of the legislation in this area in francophone African countries is based on
French civil law.9
Generally speaking, the law of contracts and more generally of obligations in these
jurisdictions has changed little since independence, with very few having adopted new
legislation on obligations. Notable exceptions include the Senegalese Code of Civil
and Commercial Obligations, based on the amended law no. 63–62 of 10 July 1963,
the Guinean Civil Code based on law no. 2019/035/AN of 4 July 2019 (replacing the
1983 Civil Code) and Malian law no. 87–31/AN-RM of 29 August 1987 establishing
the general regime on obligations. Although these texts have their specifcities,10 they
remain fundamentally inspired by the French civil law tradition.
Despite the relatively few developments across individual jurisdictions in franco-
phone Africa, there is the potential for a harmonisation of the law of obligations at
7 Ordinance no. 2016–131 on the reform of the law of contracts, the general regime and proof of obliga-
tions was promulgated on 10 February 2016 and entered into force on 1 October 2016. This ordinance was
ratifed by the law no. 2018–287 of 20 April 2018. The ordinance modifed Book III of the French Civil Code
and added a Title III relating to the sources of obligations, a Title IV relating to the general regime covering
obligations and a Title IV bis on the proof of obligations.
8 For instance, the ordinance removed any reference to ‘cause’, a central concept under French contract
law and condition for validity of a contract.
9 Although contract law in Cameroon notably combines elements of both French civil law and common
law (B. Bokolombe, L’infuence du modèle juridique français en Afrique, 2016, p. 75).
10 These are notably in areas such as family law, where Islamic law has had an important infuence.
240
a regional level through the Organisation for the Harmonisation of Business Law
in Africa (OHADA), which published a draft uniform act on contract law (in 2006)
and a draft uniform act on the general law of obligations (in 2015).11
In practice, academic and jurisprudential developments in French law continue to
play a persuasive role in francophone African countries.
1.2 Mutual hold harmless clauses: a contractual re-allocation of liability
1.2.1 Origins and growing prevalence in French law

The remainder of this chapter will focus on the articulation of mutual hold harmless
clauses with the notions of contractual and non-contractual liability12 under French law.
‘Mutual hold harmless’ or ‘knock-for-knock’ clauses, which allow the parties to a
contract to give mutual undertakings to assume any losses they may sufer, irrespec-
tive of who caused the loss, have their origins in the English language, common law
world, but are increasingly found in French law contracts. Such clauses are particu-
larly common in high-risk industries, such as the petroleum industry, as well as the
nuclear and space13 sectors. By avoiding the need to establish fault in the context of
complex industrial operations with multiple actors, the mutual hold harmless mecha-
nism allows participants to reduce the overall cost of operations by avoiding lengthy
dispute resolution processes and multiple layers of insurance (see Gordon, Chapter 2,
for more information about the general understanding of mutual indemnity, mutual
indemnity and hold harmless, etc.).
In addition to addressing losses sufered by the parties themselves, mutual hold
harmless clauses frequently also deal with the way in which third-party liability is
addressed between the contracting parties. These third parties may be related to the
contracting parties (such as afliates or subcontractors) or may be ‘true’ third parties
that are otherwise unconnected to the parties or the subject matter of the contract.
This distinction between liability between the parties and liability towards third par-
ties will play an important role in the interpretation of mutual hold harmless clauses
under French law (as will be seen in the remainder of this chapter).
As in other jurisdictions, such a contractual re-allocation of liability between
parties derogates from the principles of French law on liability, which is based on
the notion of fault. However, these clauses have received relatively little attention or
analysis from a French law perspective.
1.2.2 A note on English terminology

As noted above, mutual hold harmless clauses have their origins in contracts drafted
in English and subject to the laws of common law jurisdictions (with a particularly
heavy infuence from the United States and England). The specifc language and
patterns of drafting used in such clauses have also developed from practices in
11 The latter was developed following the lack of progress made on the draft uniform act on the law of
contracts. Finally, it is also worth noting that a Uniform Act on General Commercial Law exists already.
12 Which we will refer to here as ‘civil liability’.
13 See, for instance, A. Rakibi, “Les clauses réciproques d’abandon de recours et de garanties dans les
contrats de l’industrie spatiale”, 2014.
241
diferent industries (notably the oil and gas industry). The result is a collection of
operative terminology that includes, among others, ‘to hold harmless’, ‘to indemnify’,
‘to defend’, ‘to release’ and ‘to be responsible for’. While some of these terms will
have specifc legal meanings in many jurisdictions, others are less well defned and
appear to be included simply as a matter of industry custom.
Although a number of these concepts will equate to similar or equivalent concepts
under French law, this chapter will focus on the re-allocation of liability based on
the principles of civil liability under French law and the construction of mutual hold
harmless clauses from those principles.
2 The legal nature of mutual hold harmless clauses under French law
2.1 The legal framework for mutual hold harmless clauses
2.1.1 Principles of the law on civil liability under French law

We saw above in relation to the law of obligations that a person may owe an obliga-
tion either because they agreed to such an obligation or because the law imposes it
on them. Mutual hold harmless clauses fall under both regimes as they permit the
parties to address both contractual liability and non-contractual (or tortious or
delictual) liability of the parties.
As in many jurisdictions, contractual and non-contractual liability are governed by
diferent legal regimes. To begin with, while contractual liability may only be triggered
by a breach of contract,14 non-contractual liability may be triggered by diferent
operative events, such as an act of negligence or as a result of vicarious liability of
diferent kinds.15 Reparations are also based on diferent principles. Whereas a breach
of contract will entitle the other party to recover all losses foreseen or foreseeable at
the date of execution of the contract, compensation in respect of non-contractual
liability under French law is based on the principle of full reparation.
The starting point is that contractual and non-contractual liability are non-cumu-
lative, meaning that a creditor of a contractual obligation may not rely on the rules
governing non-contractual liability.16 However, this principle is subject to certain
exceptions. First, the Supreme Court (Cour de Cassation) has held that the ‘non-
cumulative’ principle should not prevent a party to a contract from relying on
subsidiary legal grounds for a claim. In other words, where there is doubt over the
extent of the contractual obligations, the claim may, although principally based
on contractual grounds, also rely on non-contractual grounds.17 A claim based on
non-contractual liability will also be allowed where there is no link between the loss
sufered and the obligations under the contract.18 As a result, there is now a much
14 Article 1231–1 of the French Civil Code.

15 Articles 1240 to 1244 of the French Civil Code.
16 Cour de Cassation, 11 January 1989, 86–17.323. The proposed reform of the law of obligations tabled
in 2016 retained this ‘non-cumulative’ principle.
17 Cass. com. 24 October 2018, no. 17–25.672.
18 Civ. 2e, 19 November 1964.
242
more nuanced position under French law in relation to potential non-contractual
liability arising as between contracting parties.
There are also important diferences in relation to contractual limitation or exclu-
sion of contractual and non-contractual liability. Limitation and exclusion of liability
clauses are valid in principle when applied to contractual liability (subject to certain
exceptions), but this principle is more limited when applied to non-contractual liability
(see section 3.1.1.1.1).
2.1.2 Contractual mechanisms for allocating risk between the parties

The parties to a contract governed by French law generally have signifcant freedom
to allocate between them risks arising out of a party’s civil liability. The legal mecha-
nisms available for allocating such risks will depend on whether the risk relates to
liability of the contracting parties towards each other or liability towards third parties
(in each case whether tortious or contractual).
Where the objective is to allocate liability of the parties towards each other, a number
of options are available. First, the parties can choose to act either at the level of the
obligation itself, by defning the conditions for the existence of the obligation and
preventing the liability from arising in the frst place.19 Alternatively, the objective
can be achieved by intervening at the level of the consequences for non-performance,
by restricting the consequences of a failure to perform the obligation.
Addressing the consequences of a failure to perform, the second option above,
can be done either by limiting the ability of the other party to act through a waiver
of recourse (clauses de non-recours), or by limiting the right of the other party to
compensation. Such a limitation on the right to compensation can take several forms,
either an outright exclusion of a right to damages or compensation (clause de non-
responsabilité), a limitation of that right or a pre-agreed compensation in the event of
non-performance (e.g., through a penalty clause, whereby the parties agree in advance
on the compensation to be paid in the event of specifc instances of non-performance20).
Although conceptually distinct, the same jurisprudential principles apply to both exclu-
sions of liability and waivers of recourse, both in relation to their validity and their efects.21
2.1.3 Contractual mechanisms for allocating third-party liability

Parties to a contract may also seek to re-allocate third-party liability between them.
Such third parties may either be related to the contracting parties, including for
instance their subcontractors or afliates, or ‘genuine’ third parties without any con-
nection to the parties or the contract. Although the contractual approach may vary
19 Ph. Delebecque, Fasc. 115: “RÉGIME DE LA RÉPARATION – Modalités de la réparation. – Règles

particulières à la responsabilité contractuelle. – Conventions relatives à la responsabilité”, JurisClasseur Con-
trats, 2018, p. 3.
20 A penalty clause under French law is a clause “by which the parties determine, on a comprehensive
basis in advance, the compensation due upon a breach of the contractual obligation”. See J. Carbonnier,
Droit civil: les biens, les obligations, 2004, p. 2222.
21 There are, however, some distinctions. For instance, in the case of an exclusion of liability, a breach
by the debtor may be remedied otherwise than through engaging the contractual liability of the debtor. The
creditor may seek the forced performance or judicial termination of the contract W. Dross, Dictionnaire des
clauses ordinaires et extraordinaires des contrats de droit privé, 2016, p. 542.
243
depending on the type of third parties involved, the task is more straightforward
insofar as such a contractual arrangement does not afect the rights of the third-party
having suferance the loss.
2.2 Legal characterisation of mutual hold harmless clauses under French law
2.2.1 Allocation of risk between the parties

The general approach in French law contracts is to rely on a combination of an
exclusion of liability and a waiver of recourse, as can be seen in the example below:
Each party shall bear any physical losses . . . sufered by it and its afliates in connection
with the performance of this contract. Accordingly, each party irrevocably undertakes
not to bring any claim or to initiate any proceedings against the other party or its afli-
ates for any reason whatsoever.22
Although exclusions of liability and waivers of recourse are well established under
French law, there is no general legal defnition of, or codifed framework for, clauses
allocating liability between the parties, including on a mutual basis.23 Nevertheless, such
mechanisms are increasingly found in long-term, high-value contracts, notably in the
oil and gas, mining and space sectors, as well as certain infrastructure contracts.
Such clauses will clearly cover contractual liability between the parties. However, the
intention of the parties is generally that they should extend to capture non-contractual
liability as well. As a result of the contra proferentem interpretation of such clauses in
many common law jurisdictions, the drafting in English mutual hold harmless clauses
will frequently explicitly refer to non-contractual liability.24 The drafting in French law
contracts is often less explicit, but reference to losses incurred in connection with the
performance of the contract would generally be interpreted as including contractual
and non-contractual liability. However, as noted above, the position on limiting non-
contractual liability between contracting parties is more complex.
2.2.2 Allocation of third-party liability

2.2.2.1 Related third parties
The parties will often look to expand the scope of a mutual hold harmless regime
to include a wider group of subcontractors, afliates and joint venture partners linked
to the parties (referred to here as members of a party’s ‘group’). Such related third
parties will generally be involved in two ways. First, the beneft of the waiver of
recourse granted by each of the parties may be extended to members of the parties’
groups, meaning that each of the parties undertakes not to bring claims against the
members of the other party’s wider group.
22 Quoted in L. Ravillon, Espace extra-atmosphérique – aspects contractuels, 2015, p. 11.

23 Neither the 2016 reform of the law of contracts and obligations nor the proposed reform of the law
on civil liability included any proposals to defne any mechanisms for the allocation of liability between
parties. See N. Dowlatshahi, S. Salem, La réforme du droit des contrats, une réforme à parfaire en matière de
responsabilité, 2019, p. 4.
24 For instance, “howsoever arising, whether in contract or tort, irrespective of negligence or breach of
statutory duty”.
244
Second, the parties will generally seek to extend the burden of the waiver of
recourse to the members of each party’s group, such that each party is protected
from claims brought by members of the other party’s group as well. However,
French law, like other jurisdictions, is clear that any clause purporting to limit or
exclude the tortious liability of the parties towards third parties is of no efect:
“articles 1382 and 1383 of the Civil Code are a matter of public policy (ordre public),
from which it follows that their application may not be infringed in advance by
contract”.25 Accordingly, there are two options available for extending the cross-
waiver of claims. The frst option is that the parties undertake to procure that
members of their group will not bring claims against the other party. Although
this will have no impact on the third party having sufered the loss, the related
party to the contract will nevertheless be in breach of its contractual obligation
towards the other party as a result of any claim brought by a member of its group
against the other party.
The second option is for each party to indemnify the other for any losses sufered
as a result of a claim brought by another member of the indemnifying party’s group,
as in the example below:
Where one or more afliates of a party pursues the other party and/or its afliates . . .
the frst party guarantees the other party and/or its afliates in respect of the outcome of
such a claim, and shall advance all necessary funds for the defence of their interests and
reimburse all amounts that they may be required to pay as direct or indirect result of such
a claim.26
2.2.2.2 Unrelated third parties

Occasionally, mutual hold harmless clauses may also extend to other third-party
liability, beyond the members of each party’s extended group. In this case, related
third parties need to be distinguished from ‘genuine’ third parties that do not
have any link to the parties or the subject of the contract between them. The
objective with respect to this type of third-party liability is not to extend the
beneft and the burden of the cross-waiver of claims, but simply to allocate a
defned risk (i.e., of a third-party claim against a party) between the parties. In
this case, the parties may agree to indemnify the other party for any losses arising
out of such a claim (as seen above in relation to claims from related third
parties).
Unlike the exclusion of liability and waiver of recourse between parties, an indem-
nity in respect of third-party claims (whether from related or unrelated third parties)
simply transfers the burden of compensating the third party from one contracting
party to the other. Under French law such an undertaking is qualifed as an ‘agree-
ment analogous to insurance’ (pacte analogue à l’assurance), which is addressed in
further detail in section 3.2.
25 Voir Cass. 1e civ, 2 août 1951, JCP 1951, II, 6592.

26 Quoted in L. Ravillon, Espace extra-atmosphérique – aspects contractuels, 2015, p. 11.
245
3 The legal regime governing mutual hold harmless clauses under French law
3.1 Exclusions of liability and waivers of recourse
3.1.1 Validity and exceptions

It is well established in French law that exclusions of liability and waivers of recourse
are valid in principle,27 at least insofar as these relate to contractual liability. However,
French law provides for a number of conditions for the validity of clauses dealing
with liability and limitations on their efect.
3.1.1.1 Exceptions linked to the law of obligations
3.1.1.1.1 LIMITING NON-CONTRACTUAL LIABILITY

Although clauses limiting or excluding liability are, on the basis of the principle of
freedom of contract, generally permitted, French jurisprudence has consistently held
that such limitations or exclusions are not valid when applied to non-contractual (i.e.,
tortious or delictual) liability,28 on the basis that tortious or delictual liability is a
matter of public policy (ordre public). However, the clear distinction between contrac-
tual and non-contractual liability in the context of limitation or exclusion of liability
has been brought into question on a number of fronts and the courts have taken a
more nuanced position.29 The growing consensus is that clauses limiting or excluding
delictual or tortious liability will be upheld to the extent that the loss sufered is not
the result of the negligence (faute) of the party having caused the loss. This is also
the position adopted in the proposed reforms of the French law of civil liability.30
It is worth noting that the restrictions above relate to non-contractual liability
between contracting parties and therefore necessarily apply to a small range of
circumstances, rather than the usual circumstances giving rise to delictual liability
where the parties are otherwise unconnected contractually. Contractual provisions
dealing with delictual or tortious liability towards third parties is less problematic
and is dealt with in section 3.2.
3.1.1.1.2 ESSENTIAL OBLIGATIONS

In practice, an exclusion of liability could apply to any contractual obligation. How-
ever, under French law certain obligations are characterised as ‘essential’, in respect
of which an exclusion of liability or waiver of recourse (for enforcement of the
obligation) will not be enforceable.31 Since the reform of the law of contracts in
2016, article 1170 of the French Civil Code provides that “any clause which deprives
the essential obligation of the debtor of its substance is deemed unwritten”, refecting
a well-established jurisprudential position.32
27 Cass 1re civ., 19 January 1982, no. 80–15745; more generally, see G. Viney, P. Jourdain and S. Carval,
Les efets de la responsabilité, 4th edition, LGDJ, 2017, nos. 185–187.
28 C. Bloch et al., Droit de la responsabilité et des contrats, 12th ed, Dalloz, para. 2335.22.
29 G. Viney, Introduction à la responsabilité. 4th ed, LGDJ, 2019, p. 433.
30 See, for instance, articles 1284 and 1286 of the draft law of 29 July 2020.
31 F. Buy, M. Lamoureux, J. Mestre, J.C. Roda, Les principales clauses des contrats d’afaires, 2019 p. 575.
32 For instance, Cass. Req. 19 janv. 1863, DP1863, 1, p. 248.
246
Such an essential obligation can be characterised as the principal undertaking in
respect of which the parties specifcally contracted. The exclusion of liability in rela-
tion to such an undertaking “would amount to removing the legally binding nature
of the commitment, and would thus have the efect of transforming it into a moral
commitment only”.33
Although a mutual hold harmless clause under which the parties waive any recourse
against each other would be subject to the requirements of article 1170 of the Civil
Code, we are not aware of any jurisprudence invalidating such a clause on this
basis. In any event, the ‘essential’ nature of the relevant obligations would need
to be considered in the context of a particular mutual hold harmless regime. It is
certainly conceivable that in many cases, the mutual hold harmless provisions would
not generally extend to the performance of the core obligations of the contract (e.g.,
the provision of the transportation or other services), but rather to the liability with
respect to, for instance, property and personnel, as a result of the performance or
failure to perform the core obligations.
3.1.1.2 Specific legislative exceptions34

Clauses excluding or limiting liability are often restricted or regulated by specifc
legislation that applies either to particular sector or to certain types of contracts. It
is therefore important to consider the application of such legislative provisions to
any contract that includes a mutual hold harmless regime.
For instance, exclusions of liability are prohibited in the context of transportation
of merchandise by land. Article 133–1 of the French Code of Commerce provides
that the transporter remains responsible for the loss of all objects being transported,
except in the case of force majeure, and any clause to the contrary is of no efect.
Similarly, in relation to maritime transportation, article L. 5422–15 of the French
Transportation Code provides that “any clause having as its purpose or efect, directly
or indirectly, to reduce the transporter’s liability defned in the provisions of article
L. 5422–12 is null and of no efect”.
3.1.2 Efects of waiver of recourse and exclusions of liability clauses under

French law
3.1.2.1 Strict interpretation
Similarly to the position in many common law jurisdictions, clauses related to the
liability of the parties will be subject to particular scrutiny by the courts, insofar as
they derogate from the ordinary law.35 More specifcally, mutual hold harmless clauses
will be interpreted strictly, on the basis that they restrict the right to act.36
33 Ibid.
34 There are a number of other legislative restrictions that would cover exclusion of liability and waiver
of recourse clauses but that are not relevant in the context of hold harmless clauses. These include, for
instance, the restriction on abusive clauses (clauses abusives) in standard form contracts (contrats d’adhésion),
as well as consumer law restrictions on clauses limiting the rights of recourse of consumers. These legislative
restrictions are not further addressed in this chapter.
36 See Cass. 3e civ., 2 mars 2010, no. 09–12.272, in relation to a general waiver of recourse in relation
to insurance.
247
Furthermore, such clauses will be at higher risk of being inefective to the extent
that they appear contrary to the interests of one of the parties. In other words, they
will be interpreted contra proferentem, the objective of the court in such cases being
“less to defne the common intention of the parties than to neutralise a clause that
arouses mistrust”.37 Often this contra proferentem principle is applied in situations
where the provisions of a contract are unbalanced, which should not be the case for
mutual hold harmless clauses given that these are by defnition drafted on a recipro-
cal basis. However, a reciprocal clause will, in the event of ambiguity, be interpreted
against the party seeking to enforce it.
3.1.2.2 Exceptions linked to the nature of the fault

The conduct of the party seeking to enforce the clause can also have an impact on
the efectiveness of a mutual hold harmless clause, notably where such conduct con-
stitutes intentional or wilful misconduct (faute dolosive)38 or gross misconduct (faute
lourde).39
Generally speaking, the French courts will refuse to apply exclusion or limitation
of liability clauses in the face of faute lourde or faute dolosive by the debtor in respect
of its contractual obligations.
It is commonplace to expressly provide that any waiver of recourse is dis-applied
in the event of dol or faute lourde. However, even in the absence of any express refer-
ence in the contract, the French courts have repeatedly refused to apply waivers of
recourse in the presence of faute lourde.40
This limitation on the application of an exclusion of liability or waiver or recourse
is a key diferentiating element when comparing the efect of such clauses under
French law to that under other laws. Although such clauses will frequently include
a carve-out in the event of wilful misconduct, under most common law governed
contracts the mutual hold harmless principle is generally drafted to apply in instances
of negligence falling short of wilful misconduct. Depending on the defnitions applied,
the scope for achieving this result under French law is more limited.
3.1.2.3 Exceptions linked to the nature of the loss

The efectiveness of a waiver of recourse or exclusion of liability under French law
in the context of mutual hold harmless clauses also needs to take into account
restrictions on liability clauses that relate to personal injury.
Most authors contend that, on the basis of the principles in relation to the human
body in article 16–1 of the French Civil Code, that any clauses dealing with liability
38 In French law, the notion of ‘dol’ in the context of the performance of a contract is defned as a con-
scious or deliberate breach of one’s obligations, but without the debtor necessarily having had the intention
to cause harm (Cass. civ, 2, 10 novembre 2021, no. 19–12.660; Cass. civ, 2, 28 février 2013, no. 12–12.813;
Cass. civ, 3, 27 juin 2001, no. 99–21.017 99–21.284).
39 ‘Faute lourde’ is defned as a breach that “cannot result from a mere breach of a contractual obligation,
even an essential one, but must be deduced from the seriousness of the debtor’s behaviour” (Cass. com, 29
juin 2010, no. 09–11841); it is “characterised by extremely serious behaviour, bordering on dol, demonstrat-
ing the inability of the obligor to perform the contractual obligation he had accepted” (Cass. ch. mixte, 22
avr. 2005, no. 03–14112).
40 Voir Cass. 1e civ., 31 mars 1987, no. 79–17.129.
248
are inefective insofar as they relate to bodily harm. Although this is not a universally
accepted position,41 it is likely that any clause seeking to re-allocate the liability for
personal injury will not be enforceable.
Given that mutual hold harmless clauses frequently include personal injury within
their scope, this is a key limitation on such clauses where they are governed by French
law, as compared to similar clauses governed by other laws.
3.2 Indemnifcation of third-party losses

An indemnity in respect of third-party liability typically included in mutual hold
harmless clauses constitutes what is termed an ‘agreement analogous to insurance’
(pacte analogue à l’assurance), which refers to a provision pursuant to which a person
not acting in a professional capacity as an insurer undertakes contractually to indem-
nify the counterparty in respect of the liability of the latter towards third parties in
the context of the performance of their contract.42
3.2.1 Validity of clauses indemnifying third-party losses

The validity of clauses indemnifying third-party losses is well established under French
law.43 Such clauses can apply to both the contractual and non-contractual (tortious
or delictual) civil liability of the beneftting party. Although French law generally
restricts the ability to contractually restrict one’s non-contractual liability (see section
3.1.1(A)), nothing prevents a party from agreeing that another person will bear the
fnancial consequences of such liability, on the basis that the person having sufered
the loss is unafected by the agreement.
3.2.2 Limitations
Clauses indemnifying against third-party claims, such as those found in mutual hold
harmless clauses, will be subject to certain principles and limitations applicable to
insurance contracts. One of the most important principles in the context of an
indemnity covering third-party liability is that French law prohibits a person from
insuring against their own intentional misconduct, but does allow insurance against
all other forms of misconduct.44 This means that such a clause will not be enforce-
able in the event of the wilful misconduct (faute intentiontionelle) of the benefciary
indemnity, but will be in the face of other misconduct, even gross misconduct (faute
lourde) of the benefciary. The position in relation to indemnifcation for third-party
liability is therefore more permissive than the position in relation to exclusions of
liability and waivers of recourse, where such a clause will also fail in the face of gross
misconduct (faute lourde),45 as set out in section 3.1.2.2. Accordingly, the scope for
41 W. Dross, Dictionnaire des clauses ordinaires et extraordinaires des contrats de droit privé, 2016, p. 303.
42 Ph. Delebecque, Fasc. 115: “RÉGIME DE LA RÉPARATION – Modalités de la réparation. – Règles
particulières à la responsabilité contractuelle. – Conventions relatives à la responsabilité”, JurisClasseur Con-
trats, 2018, p. 7.
43 Civ. 3e, 10 January 1978, no. 76–11.111.
44 Y. Aubin, T. Portwood, Les clauses récriproques d’abandon de recours et de garanties contre les recours
des tiers, RDAI/IBLJ, no. 6, 2001, p. 684.
45 Cass. 1e civ. 15 avril 1961, JurisData no. 1961–000198.
249
re-allocating third-party liability is broader than it is for the liability of the parties
towards each other.
Acknowledgement
The authors wish to express their gratitude to Sumerya Dinç and Majda Hashemi
for their invaluable assistance on this chapter.
250
CHAPTER14
Lessons from the application of

knock-for-knock clauses under Malaysian law
1 Introduction
This chapter discusses the application of knock-for-knock in countries unaccustomed
with the clause. The use of legal provisions without a proper understanding of them
will cause the indemnity and hold harmless clauses to be interpreted outside their mean-
ing, potentially opening the door to abuse by a party in a dominant position. This
chapter provides a case study of how these concerns might become mixed up in Malaysia.
This chapter is divided into two major sections. The frst section provides an overview
of Malaysian legal framework including the statutory and judicial approach on indemnity
clauses in Malaysia. The second section investigates the issues and problems in Malaysia
with regard to risk allocation provisions and indemnity clauses under the oilfeld service
contracts. This section also presents the research fndings of the empirical studies that
have been conducted in Malaysia and the analyses on indemnity clauses under oilfeld
service contracts which were used and drafted by three operators in Malaysia.
1.1 General overview of the Malaysian legal system

In order to fully understand the underlying legal processes and the policies from
which they derive, it is crucial to describe the Malaysian legal system. This section
will explain the Malaysian legal structure, including its historical basis, the Constitu-
tion and Malaysia’s fundamental laws and institutions.1
Malaysian law is settled on the English legal system, and it is a common law juris-
diction.2 This is due to Britain’s occupation of Malaya, Sabah and Sarawak from the
early 19th century until the 1960s.3 Malaysian law is a hybrid of English law, which
governs most aspects of daily life, and Sharia law, which governs family and personal
matters for Muslims.4 Malaysian legislation is based on the Federal Constitution of
Malaysia and Acts of Parliament. The British introduced the fnal version of the
Civil Law Ordinance, which was frst enacted in the Straits Settlements in 1878, a
1 Arumugam Rajenthran, Malaysia: An Overview of the Legal Framework for Foreign Direct Investment
(Citeseer 2002).
2 Sharifah Suhana Ahmad, Malaysian Legal System (MLJ 1999).
3 Ibid.
4 Geofrey Wilson Bartholomew, The Commercial Law of Malaysia: A Study in the Reception of English
Law (MLJ 1965).
DOI: 10.4324/9781003206798-16 251

wan mohd zulhafiz wan zahari
year before Malaya gained independence in 1957.5 The formal importation of the
English common law and the rules of equity into the national legal system were done
through the Ordinance. The Ordinance that remains until today was revised in 1972
and renamed as the Civil Law Act 1956.6
The Federal Constitution is the supreme law of the land.7 The constitution provides
for a unique dual justice system: secular laws on one side (e.g., criminal and civil) and
Sharia law on the other side.8 The constitution also provides the legal framework for the
laws, legislation, courts and other administrative aspects of the law. Further, it defnes
the government, monarch and its powers and the rights of the citizens.9 Federal laws
enacted by the Parliament of Malaysia apply throughout the country.10 There are also
state laws enacted by the State Legislative Assemblies which apply in the particular
state.11 In respect of the judiciary, the Malaysian Court system is infuenced by and
in fact, is quite similar to the English Court system which split the Subordinate and
Superior Courts.12 The Penghulu Court is the lowest level of the Subordinate Courts,
and each of the specifc districts is presided by a headman that is appointed by the
state government. The next level in the court hierarchy is the Magistrate’s Courts which
deals with minor civil and criminal cases, and at the highest level of the Subordinate
Courts are the Sessions Courts. The Superior Courts are made up of the High Court,
Court of Appeal and Federal Court (which is the supreme court in the land).13
The laws of Malaysia can be divided into two types of laws: written law and
unwritten law.14 Written laws are laws which have been enacted or legislated by the
lawmakers. In contrast, unwritten laws are laws which are not contained in any statutes
and can be found in case decisions. This is also known as common law or case law.
In situations where there is no law governing a particular situation, Malaysian case
law may apply. If there is no Malaysian case law, English case law can be applied.15
The principle of stare decisis also applies in Malaysian law.16 Thus, any decisions by
a court higher in the hierarchy will be binding upon the lower courts.
1.2 The law of contract in Malaysia

In Malaysia, parties typically have freedom of contract.17 The Privy Council in Ooi
Boon Leong v Citibank N.A.18 confrmed that “parties to an agreement have much
5 JN Matson, ‘The Confict of Legal Systems in the Federation of Malaya and Singapore’ (1957) 6 Int.
Comp.L.Quart. 243.
6 The CLA 1956 (Revised 1972) is in fact a consolidation of the Civil Law Ordinance 1956, Sabah’s
Application of Laws Ordinance 1951 and Sarawak’s Application of Law Ordinance 1949.
7 Malaysian Constitution. Art 4.
9 Abdul Aziz Bari, Malaysian Constitution: A Critical Introduction (Other Press 2003).
11 Ibid.
12 Arfah Hamzah and Ramy Bulan, An Introduction to the Malaysian Legal System (Fajar Bakti 2003).
13 Ibid.
14 Wu Min Aun, An Introduction to the Malaysian Legal System (Kuala Lumpur: Heinemann Educa-
tional Book (ASIA) Ltd 1975).
15 Subject to sections 3 and 5 of Civil Law Act, which will be discussed later.
16 Hamzah (n12).
17 Ahmad SA Alsagof, Principles of the Law of Contract in Malaysia (MLJ 1996).
18 [1984] 1 MLJ 222.
252
lessons from the application of knock-for-knock
scope to negotiate and incorporate terms acceptable to them”. Currently, there are
no statutory restrictions in Malaysia on contractual provisions purporting to exclude
or limit the liability of one or both of the parties in relation to indemnity and
liability.19
The issue of fairness and unfairness of contractual terms is seen in two aspects:
procedural unfairness and substantive unfairness. Procedural fairness refers to
fairness during the contract formation process.20 A contract can be regarded as
procedurally unfair when it is entered by one party through an unfair means, such
as coercion, undue infuence, or fraud.21 The court may fnd that such contracts
are voidable under Part III of Malaysian Contracts Act 1950.22 In contrast, sub-
stantive fairness focuses on the fairness in the allocation of substantive rights
and obligations under the contract itself.23 The contract is considered to be sub-
stantively unfair if the contract terms are unfair or biased or “one-sided”.24 That
said, usually, “courts have been less willing to interfere in cases of substantive
fairness bearing in mind the concept of freedom of contract”.25 However, it is
argued that “the development of monopolistic business, in the public and private
sector alike, has made it impossible for the weaker party actually to exercise the
freedoms in many cases”.26
The Malaysian judicial position on the doctrines of unconscionability and
inequality of bargaining power remains unclear and ambiguous.27 In Fui Lian
Credit & Leasing Sdn Bhd v Kim Leong Timber Sdn Bhd,28 Datuk Cheong Siew
Fai J appeared to recognise the doctrine of unconscionability, where the court,
referring to the English cases of Multiservice Bookbinding Ltd v Marden29 and
Hart v O’Connor,30 stated that:
In order that a party may free himself from complying with an agreement he had entered
into, he must show that, in the eyes of the court, it was unreasonable. A bargain cannot
be unfair and unconscionable it is shown that one of the parties to it has imposed an
objectionable term in a morally reprehensible manner, that is to say, in a way which
afects his conscience or has procured the bargain by some unfair means.
19 Toby Hewitt, ‘An Asian Perspective on Model Oil and Gas Services Contracts’ (2010) 28 J.Energy
Nat.Resour.L. 331, 333.
20 Hart v O’Connor [1985] A.C. 1000.
21 Part III of Malaysian Contracts Act 1950.
22 Maryam Rafei and Nazura Abdul Manap, ‘Legal Position of Click Wrap Agreement’, Proceedings of
International Conference on Computer Communication and Management (ICCCM 2011) (2011).
23 Naemah Amin, ‘Protecting Consumers against Unfair Contract Terms in Malaysia: The Consumer
Protection (Amendment) Act 2010’ (2013) 1 MLJ 1.
24 Norliza Abdul Hamid and Hariati Mansor, ‘The Legal Implications of the Consumer Protection
(Amendment) Act 2010 on Contract Terms in Malaysia’, Annual Summit on Business and Entrepreneurial
Studies (ASBES 2011) Proceeding (UiTM 2011).
25 Ibid.
26 John N Adams and Roger Brownsword, ‘The Ideologies of Contract’ (1987) 7 Leg.Stud. 205, 208.
27 Andrew Phang Boon Leong, Cheshire, Fifoot and Furmston’s Law of Contract, vol. 2 (Butterworths
Asia, Singapore 1998).
28 [1991] 1 CLJ 522.
29 [1979] Ch. 84.
30 [1985] A.C. 1000.
253
However, the judge did not further clarify the status of Hart31 in the Malaysian
context, specifcally he did not clarify whether or not the doctrine of unconscionabil-
ity could be adopted as an independent ground for relief.32
Moreover, the principle of inequality of bargaining power formulated by
Lord Denning in the case of Lloyds Bank v Bundy33 has also been rejected by
the local courts because there is no established precedent in Malaysia.34 This
could be seen in the judgment of Visu Sinnadurai J. in Polygram Records Sdn
Bhd v The Search.35
In another case, Gopal Sri Ram JCA explained in his judgment at the Court of
Appeal in Saad Marwi v Chan Hwan Hua & Anor36 that the doctrine of uncon-
scionability, which is also referred to as the doctrine of inequality of bargaining
power, should be explicitly recognised in Malaysian law. Gopal Sri Ram JCA
stated that
the time has arrived when we should recognize the wider doctrine of inequality of bar-
gaining power. . . . What is therefore called for is a fairly fexible approach aimed at doing
justice according to the particular facts of a case. . . . That brings me to the third alter-
native. This is to adopt the English doctrine [of unconscionability] but apply it in a broad
and liberal way as in Canada.37
Unfortunately, the subsequent case at the Court of Appeal did not follow the prin-
ciples in Saad Marwi. In the case of American International Assurance Co Ltd v Koh
Yen Bee (f),38 on the issue of the applicability of the doctrine of unconscionability,
Abdul Hamid Mohamad JCA chose not to follow Gopal Sri Ram JCA’s view. He
contended that the two cases were completely contradictory and that section 14 of
the Contracts Act of 1950 only recognised coercion, undue infuence, fraud, misrep-
resentation and mistake as elements that infuence free consent.39 It has been argued
that this ruling fails to acknowledge that, in reality, it is necessary to strike a balance
between the need for fair contracts and the competing interest of legal certainty.40
This situation had left Malaysian law in a state of ambiguity and uncertainty regard-
ing its recognition and acceptance of the doctrines of unconscionability and inequality
of bargaining power.41 As a result, any party who is subjected to procedurally unfair-
ness has no recourse to seek justice and fle a cause of action in court for uncon-
scionability and unequal bargaining power.
31 Ibid
32 Ibid.
33 [1975] QB 326.
34 Cheong May Fong, Contract Law in Malaysia (Sweet & Maxwell Asia 2010).
35 [1994] 3 MLJ 127.
36 [2001] 3 CLJ 98.
37 Ibid.
38 [2002] 4 MLJ 301.
39 Ibid., 318.
40 Cheong May Fong, ‘A Malaysian Doctrine of Inequality of Bargaining Power and Unconscionability
After Saad Marwi’ [2005] MLJ 4.
41 Siti Aliza Alias and Zuhairah Arif Abdul Ghadas, ‘Inequality of Bargaining Power and the Doctrine
of Unconscionability: Towards Substantive Fairness in Commercial Contracts’ (2012) 11 Austl.J.Basic&App.
Sci. 331.
254
1.2.1 Statutory and judicial approach on indemnity clauses in Malaysia
This section demonstrates the statutory and judicial approach to indemnity clauses
in Malaysia. Sections 77 and 78 of the Malaysian Contract Act 1950 govern the law
relating to indemnity. However, there are no legal constraints on contractual clauses
that seek to exclude or limit the liability of one or both parties with respect to
indemnifcation and liability.42 The High Court in Amman Singh v Vasudevan43 held
that section 77 of the Malaysian Contracts Act 1950 defnes what a contract of
indemnity is,44 whereas section 78 provides the rights of the indemnity holder when
he is sued by the promisor.45
In the context of section 77, the Supreme Court in South East Asia Insurance Bhd
v Nasir Ibrahim46 held that a contract of indemnity is “a contract by which one party
promises to save the other from loss caused to him by the conduct of the promisor
himself, or by the conduct of any other person”. Such consideration provided under
contract of indemnity is regarded as valuable and legally acceptable. In Macon Works
& Trading Sdn Bhd v Phang Hon Chin & Anor,47 the court applied the common law
meaning of consideration and followed the classic statement contained in Currie v
Misa,48 where it was observed that “[a] valuable consideration may consist either in
some right, interest, proft or beneft accruing to one party or some forbearance,
detriment, loss or responsibility given, sufered or undertaken by the other”. Gunn
Chit Tuan SCJ in delivering the judgment in South East Asia Insurance Bhd v Nasir
Ibrahim49 said that the essence of consideration was that the promisee had imposed
some kind of burden or detriment, for example, a legal responsibility.50 In this case,
42 Hewitt (n.19) 331; Wan Zulhafz, ‘Unfair Contract Terms Act 1977: Does It Provide a Good Model
in Regulating Risk Allocation Provisions in Oilfeld Contracts in Malaysia?’ (2015) 8(1) Int.J.Trade Glob.
Mark. (IJTGM) 3.
43 [1972] 1 LNS 7.
44 Section 77 of Malaysian Contract Act reads:
Contract of Indemnity
A contract by which one party promises to save the other from loss caused to him by the conduct of the
promisor himself, or by the conduct of any other person, is called a ‘contract of indemnity’.
Illustration A contracts to indemnify B against the consequences of any proceedings which C may take
against B in respect of a certain sum of RM200. This is a contract of indemnity.
45 Section 78 of Malaysian Contract Act reads:
Right of Indemnity holder when sued
The promisee in the contract of indemnity, acting within the scope of his authority, is entitled to recover
from the promisor –
(a) all damages which he may be compelled to pay in any suit in respect of any matter to which the
promise to indemnify applies; (b) all costs which he may be compelled to pay in any such suit if,
in bringing or defending it, he did not contravene the orders of the promisor, and acted as it would
have been prudent for him to act in the absence of any contract of indemnity, or if the promisor
authorized him to bring or defend the suit; and (c) all sums which he may have paid under the
terms of any compromise of any such suit, if the compromise was not contrary to the orders of
the promisor, and was one which it would have been prudent for the promisee to make in the
absence of any contract of indemnity, or if the promisor authorized him to compromise the suit.
46 [1992] 2 MLJ 355; [1993] 1 SCR 89, SC.
47 [1976] 2 MLJ 177, HC.
48 (1875) LR 10 Exch 153.
49 [1992] 2 MLJ 355; [1993] 1 SCR 89, SC.
50 Ibid., 356.
255
the Supreme Court held that the promisee had to assume legal responsibility to its
detriment by providing a third-party indemnity.51 In sum, it could be said that the law
in Malaysia recognises the concept of indemnity. However, the law has not provided
explanation with regards to contractual indemnity.52
An example of the use of an indemnity clause in the oilfeld service contracts in
Malaysia can be seen in the case of Sabah Shell Petroleum Co Ltd & Anor v The
Owners of and/or any other Persons Interested in The Ship or Vessel The “Borcos
Takdir”.53 In this case, the plaintif (“the operator”) was engaged in ofshore oil
and gas exploration and production activities in the ofshore areas of Sabah and
Sarawak.54 The plaintif operated ofshore platform structures including underwater
oil pipeline and had deployed the defendant’s vessel, the Borcos Takdir, to deliver
such cargoes for food, equipment, tools and machinery to the plaintif’s ofshore
platform structures. However, the vessel damaged the pipeline.55 The plaintif thus
commenced the action against the defendant for the substantial loss and damage they
had sufered as a consequence of the defendant’s negligence and/or breach of duty or
breach.56 The defendant contended that by virtue of Clause 21 of the Contract, there
was an agreement that any claim for physical loss to the property of Operator would
be subject to a contractual limit of RM5,000,000.57 Clause 21 provides as follows:
Clause 21 – responsibilities and indemnities
(1) The Contractor shall be absolutely liable for and shall indemnify the Company
(i.e., the Operator), its Associated Companies, Co-Venturers and PETRONAS
from and against any and all physical loss of or damage to:
(a) WORK or any part thereof, howsoever arising, and shall, at its own costs
and expense, restore the WORK lost or damaged, replace or repair any
unfxed MATERIALS or consumables and the like which have been lost or
damaged, remove and dispose of any debris and proceed with the perfor-
mance of the WORK; and
(b) Property of the Company, its Associated Companies, Co-Ventures and
PETRONAS caused by or contributed to by or resulting from any act,
omission, fault, lack of due diligence or negligence of the Contractor, its
Sub-contractors or any of their respective employees, servants and agents
unless such loss or damage was caused by the sole negligence of the
Company
The Contractor’s liability under this Clause 21(1)(b) herein shall be limited to Malaysian
Ringgit Five Million (RM5.000.000) for any one accident or series of accidents arising
out of any one event. The liability in excess of such limit shall be governed by applicable
law.58
51 Ibid.
52 Wan M Zulhafz and Nasarudin Abdul Rahman, ‘Unfair Risk Allocation in Oil and Gas Upstream
Service Contracts in Malaysia: The Necessity for Oilfeld Anti-Indemnity Act’ (2020) Int.J.Bus.Soc.
(IJBS) 21.
53 [2012] 5 MLJ 515.
54 Ibid.
55 Ibid.
56 Ibid., 516.
57 Ibid., 558.
58 Ibid.
256
The High Court held that the above clause is an indemnity clause that provides
coverage in two tranches. The frst tranche is the amount of RM5,000,000 and is
payable by the contractor regardless of whether liability is not wholly attributable to
it, as long as the incident in question was not caused by the sole negligence of the
operator.59 The second tranche is only available if liability has been determined and
adjudicated in accordance with the law.60
In relation to the interpretation and construction of contracts, the court in Malaysia
will generally apply an objective or reasonable man test in construing an indemnity
clause. In Lee Kam Wah v Associated Asian Securities (Pte) Ltd,61 the plaintif
agreed “to indemnify the company (the defendants) of any losses and damages that
may arise on any contracts written by (him)”.62 At the foot of the document, the
plaintif signed and “(agreed) to the above terms and conditions and hereby (indem-
nifed) the company for any losses or damages arising from any contracts written or
any transactions entered into by (him)”.63 The defendants sought indemnity from
the plaintif.64 The High Court held that the wording of the indemnity clauses was
inelegant and implied that indemnities had to cover every conceivable loss or dam-
age.65 Further, the court stated:
It would be wrong to give such clauses such an open-ended construction without trying
to ascertain what the parties really intended from the terms of the document in question
and the surrounding circumstances at the time the parties entered into the agreement
unless the evidence clearly established that the plaintif was negligent or had been a party
to the fraudulent transaction.66
At the Supreme Court level, in Ayer Hitam Tin Dredging Malaysia Bhd v YC Chin
Enterprises Sdn Bhd,67 the Court in ascertaining the meaning of “the appropriate
indemnity clauses” held that the existence of an agreement depends upon the inten-
tion of the parties, who must be ad idem, which may be inferred from (1) the language
used, (2) the parties’ conduct having regard to the surrounding circumstances and
(3) the object of the contract.68
In another case, the Federal Court in Malaysia construed the indemnity clause
as exemption and exclusion clauses. This can be seen in the case of CIMB Bank
Bhd v Maybank Trustees Bhd,69 where the Federal Court agreed with the Court of
Appeal’s approach and interpreted the indemnity clause under Clause 14.1 of the
Trust Deed as an exemption clause where the Court of Appeal referred to the prin-
ciple of exemption clause under English case of Karsales (Harrow) Ltd v Wallis,70
59 Ibid., 565.
60 Ibid.
61 [1991] 3 MLJ 286.
62 Ibid., 287.
63 Ibid.
64 Ibid.
65 Ibid.
66 Ibid.
67 [1994] 2 MLJ 754.
68 Ibid., 755.
69 [2014] MLJU 117.
70 [1956] 2 All ER 866.
257
where Lord Denning ruled that “it is now settled that exempting clauses of this kind,
no matter how widely they are expressed, only avail the party when he is carrying
out his contract in its essential respects”.71 He further stated that if the party “has
been guilty of a breach of those obligations in a respect which goes to the very root
of the contract, he cannot rely on the exempting clauses”.72 On the other hand, the
Court of Appeal held that the indemnity must be resolved by the proper construc-
tion of the exclusion clauses. The Court of Appeal had then cited few cases relating
to exclusion clause such as Hotel Anika Sdn Bhd v Majlis Daerah Kluang Utara,73
Anderson v Fitzgerald,74 and Guardian Assurance Co Ltd v Condogianis.75
When there is ambiguity between the indemnifcation clause and the exemption
clause, the court will apply the contra proferentem rule to settle the issue. In Syari-
kat Uniweld Trading v The Asia Insurance Co Ltd,76 the appellant ran a motor
vehicle repair and welding shop. A fre broke out as a result of an unintentional
leak in a gasoline gas pipe, resulting in the destruction of a third-party vehicle. The
appellant compensated the owner of the motor vehicle and sought indemnity from
the respondent. The appellant sought indemnifcation because the insurance policy
expressly stated that the respondent would indemnify the appellant for any sums for
which the appellant became legally obligated to pay. However, the insurance policy
excluded the appellant from being held liable for accidental damage to third-party
property caused by the appellant’s negligence. That exemption clause was invoked
by the respondent. The High Court granted the appeal and held that, when there is
an ambiguity between the indemnity clause and the exemption clause, the insurance
policy should be construed in favour of the appellant under the contra proferentem
rule. As a result, the insurance company (respondent) was ordered to indemnify
the appellant. In Malaysia, contracting parties typically use extensive and broad
phrases such as “will not be liable for any damage however caused”, “will not in
any circumstances be responsible” and “arising from any cause whatsoever” in order
to be exempted from liability for negligence.77 These expressions are commonly used
to refer to liability for negligence.78 However, it is important to note that the party
making the claim must be able to demonstrate that the damage was not caused by
his negligence and that he took reasonable steps to fulfl his duty of due diligence.79
In Chin Hooi Nan v Comprehensive Auto Restoration Service Sdn Bhd,80 Siti Norma
Yaakob J found that it is well established that an exemption clause, no matter how
broad and general, does not relieve the respondents of the burden of proving that the
71 Ibid., 868.
72 Ibid., 869.
73 [2007] 1 MLJ 248.
74 (1853) 4 HLC 484; (1853) 10 ER 551.
75 (1919) 26 CLR 231.
76 [1996] 2 MLJ 160.
77 Alsagof (n17).
78 Ibid.
79 Ibid.
80 [1995] 2 MLJ 100.
258
damage was not caused by their negligence and misconduct.81 They must demonstrate
that they fulflled the duty and responsibility with reasonable diligence and care.82
In Sekawan Guards Sdn Bhd v Thong Guan Sdn Bhd,83 Wan Adnan J held that
the appellant carry the burden of demonstrating that the exemption clause applied
and that the clause protected it from liability unless the loss was caused solely by
his negligence.84 To succeed, the appellant must demonstrate that other people were
negligent.85 In the above case, the appellant had contracted to provide security ser-
vices at the premises of the respondent.86 At the premises, a theft occurred, result-
ing in the loss of the respondent’s goods, which included 169 cartons of cigarettes
worth RM158,237.15.87 Under the contract, the appellant’s liability was limited to
RM100,000.88 The respondent fled a breach of contract lawsuit against the appel-
lant. In addition, the respondent sued for negligence.89 The appellant relied on an
exemption clause in the contract.90 However, the court found that the appellant was
solely negligent and held that the exemption clause did not protect the appellant.91
In Malaysia, a party cannot use an exemption clause to protect himself from
the consequences of a fundamental breach of contractual obligations.92 In Sze Hai
Tong Bank Ltd v Rambler Cycle Co Ltd,93 the carrier’s agent wrongfully released the
respondents’ goods to the consignee, who produced a written indemnity by the con-
signee’s bank (the appellant) in favour of the carrier, without the production of the
bill of lading.94 The carrier’s agent was aware that this act was an illegal act.95 The
respondent sued the carrier because the consignee never paid for the goods.96 The
carrier admitted liability to indemnify the carrier if it was found liable.97 The carrier
was argued to be immune from liability under Clause 2 of the bill of lading, which
stated that the liability of the carrier shall be deemed to cease absolutely after the
goods were discharged from the ship.98 The Privy Council held that the carrier was in
breach of contract and was not protected by the exemption clause.99 The clause must
be limited and modifed to the extent necessary to give efect to the main object and
intent of the contract.100 The court ruled that the carrier should not be allowed to
deliberately disregard its obligation to deliver, that is, against production of the bill
81 Ibid.
82 Ibid.
83 [1995] 1 MLJ 811.
84 Ibid., 816.
85 Ibid.
86 Ibid., 813.
87 Ibid., 814.
88 Ibid.
89 Ibid.
90 Ibid.
91 Ibid., 816.
92 Alsagof (n17) 278.
93 [1959] MLJ 200.
94 Ibid., 201.
95 Ibid.
96 Ibid.
97 Ibid., 200.
98 Ibid., 201.
99 Ibid., 200.
100 Ibid., 202.
259
of lading.101 A fundamental breach of obligations of a contract cannot be allowed
to pass unnoticed under the cloak of a general exemption clause.102
It is important to note that, procedurally, a party cannot sue for a breach of
contract while also claiming indemnifcation from the same party who committed
the purported breach. The High Court ruled in Siong Electronic Industries (1981)
Sdn Bhd v Sanyo Sales & Service Sdn Bhd,103 that a party cannot sue for a breach
while also seeking indemnifcation from the party who has committed the alleged
breach.104 The application of indemnity clauses was limited by the High Court.105 The
court held that indemnity clauses were only applicable in situation where it saved the
promisee from loss caused by third-party claims. In law, indemnity clauses denote a
contract in which the promisor undertakes an original and independent obligation
to indemnify the promise against third-party claims.106
In terms of evidence, the indemnitee must be able to demonstrate to the court that
he sufered loss prior to fling a claim for such indemnity. In Azman bin Mahmood
& Anor v SJ Securities Sdn Bhd,107 there was a question about whether a customer
could be held liable on his indemnity for losses incurred as a result of a transaction
where the loss claimed could not be proven.108 The majority view of the Federal
Court was negative on this question.109
Based on the discussion above, it could be argued that Malaysian law applicable
to indemnity clauses is too brief, unclear and incomprehensive. The Contract Act
under sections 77 and 78 only lays down the defnition of contract of indemnity
and provides the right to the indemnity holder to sue against the indemnitor. The
courts use the principle of the “reasonable man” in construing the indemnity clause,
i.e., by looking at (1) the intention of the parties, (2) the language used and (3) the
surrounding circumstances when the time the parties entered into contract. It could
be argued that the courts have taken the correct approach to interpret indemnity
clauses. However, it could also be argued that it is inappropriate for the court to go
beyond that and treat indemnity clauses in the same way as exemption and exclusion
clauses. This is because, indemnity clauses are used by the parties in oilfeld service
contracts to allocate risk.
Conventionally, indemnity clauses distinguish between negligence and gross negli-
gence. Treating indemnity clauses like exemption and exclusion clauses may defeat the
function of indemnity clauses. For example, the indemnitee may seek indemnifcation
even if the damage was solely caused by his own negligence. However, under the
principle of exemption clauses from the case of Sekawan Guards Sdn Bhd v Thong
101 Ibid.
102 Ibid., 200, 202.
103 [1997] MLJU 162.
104 Ibid.
105 Ibid.
106 See Halsbury’s Laws of England, vol. 20 (4th edn, Butterworths 1983), para. 345.
107 [2012] MLJU 660.
108 Ibid., 6.
109 Ibid., 16.
260
Guan Sdn Bhd,110 the indemnitee was not allowed to claim for indemnity since the
damage arose solely from his own negligence.
1.3 Empirical study of oilfeld service contracts in Malaysia

An empirical study was conducted in Malaysia during the period of February to
April 2014. Seven case studies were done and the purpose was to elicit the opinion
of key players in the oil and gas industry on the issue of risk allocation. The empiri-
cal study also sought to ascertain their attitudes towards the current trends of
indemnity clauses under oilfeld service contracts in Malaysia as well as their percep-
tion on how to address this problem.
It is important to study actual contracting practice because, as Korobkin sug-
gests, it is useful “to describe in-depth the contracting patterns and norms generally
followed by a particular type or group of contracting party”.111 In addition, Eigen
suggests that empirical are a signifcant tradition in legal scholarship because they
help scholars “understand the diversity of disciplinary approaches and framings of
questions about contracts raised in modern empirical explorations”.112 Eigen also
argues that it is valuable to concisely articulate the interrelation between contract
doctrine, theory and empirics.113 The data source for these studies was primarily semi-
structured interviews of ten respondents from seven organisations representing the
contractor and operator. It is suggested that the low response rate could be indicative
of a limited awareness of the subject matter of this research among the professionals
in the oil and gas industry in Malaysia. Secondary data consisted of documentary
evidence including samples of indemnity clauses of oilfeld service contracts drafted
by three operators which were provided by the contractors and feld notes written
by the researcher during the interviews.
The present empirical study was conducted to support a previous empirical study
that was done in 2008. In the previous study, a total of three case studies were con-
ducted in diferent companies in Malaysia in respect of their experiences and percep-
tions about specifc issues on procurement in the oil and gas industry in general.114
The fndings indicate that there is a strong objection in the Malaysian oil and gas
industry to the one-way adversarial style of the operator-contractor relationship and
the way the contractor is being forced to be on the receiving end when it comes to
price fuctuation, overall risks and increased workloads.115 Even though the previous
study is not designed specifcally to deal with the issue of contractual risk allo-
cation provision, there was clear assertion by the contractors in expressing their
110 [1995] 1 MLJ 811.

111 Russell Korobkin, ‘Empirical Scholarship in Contract Law: Possibilities and Pitfalls’ [2002]
U.Ill.L.Rev. 1033, 1040.
112 Zev J Eigen, ‘Empirical Studies of Contract’ [2012] Ann.Rev.L.Soc.Sci. 2.
113 Ibid.
114 Mohammad Fadhil Mohammad, ‘Procurement Strategies for the Oil and Gas Industry: To Cap-
ture Changing Values and Dealing with Multi Cultural Complexity’, The Proceedings of the International
Conference on Construction and Building Technology (ICCBT2008), Universiti Teknologi MARA (UiTM),
Malaysia (UiTM, Malaysia 2008) 33.
115 Ibid.
261
dissatisfaction pertaining to unequal bargaining and the dominant position held by
the operator in negotiating the contractual terms.
There are three main questions to be addressed in this study. The frst question is
that, considering the relationship between operators and contractors, to what extent
do both parties exercise free will and have equality of bargaining power in negotiat-
ing the contractual terms? The second question is, are the risk allocation provisions
and the indemnity clauses fair to both parties? The third question is, to what extent
does the insurer provide coverage in respect of the risks which have been indemni-
fed by the parties?
In addressing these questions, the discussion will be based on the legal theories
pertaining to the concept of contractual risk allocation and indemnity clause, the
theory of freedom of contract, the doctrine of inequality of bargaining power and
fairness.116 These legal theories informed the construction of the interview questions.
The author also used the theories as a tool for data analysis.
1.4 Research design

The research method chosen for this study was semi-structured interview with oilfeld
contractors and operators. Lists of predetermined open-ended questions were posed
to the respondents in person. On some occasions, questions developed naturally dur-
ing the course of the interview. This research method was chosen because it has some
level of predetermined order but still allows fexibility in the way issues are addressed
by interviewees.117 It also contains high validity as it scrutinises the ideas of the
interviewees about the subject matter. Hence it encourages in-depth knowledge shar-
ing by the respondents. Therefore, the researcher has a chance to pose complex
questions and issues. Moreover, new ideas could also be discussed with the respon-
dents, which is advantageous for the research.
The questions posed during the interviews can be divided into two parts. The
frst part is connected to the background of both the company and the interviewees.
Meanwhile, the second part relates to the research investigation. This part deals with
the process of (1) contract formation; (2) contractual negotiation; (3) preparation of
standard form contracts; (4) interviewees’ experience with regards to risk allocation
provisions, indemnity clauses and insurance requirements; and (5) interviewees’ views
on the law and practice of risk allocation provisions and indemnity clauses in Malaysia.
1.4.1 Research process

The semi-structured interviews were conducted with fve companies representing the
contractor, as well as one company and one legal frm representing the operator.
The respondents were (1) the heads or managers of the participating companies’
legal departments, (2) contract managers, (3) procurement managers, (4) principal
technical managers, (5) head project managers of the companies and (6) a lawyer
116 Wan M Zulhafz, ‘On the Contractual Risk Allocation in Oil and Gas Projects’ (2017) Law Rev.
(LR) 168.
117 K Louise Barriball and Alison While, ‘Collecting Data Using a Semi-structured Interview: A Discus-
sion Paper’ (1994) 19 J.Adv.Nursing 328.
262
who used to litigate for operators in court. The contractors were selected from large,
medium and small companies. The respondents were chosen due to their prominence
and experience in contractual matters. The large companies are usually equipped
with complete legal and contract departments, whereas the medium size company
might have a legal department but not necessarily a contract department, since
contractual matters are normally handled by the engineer who is also the principal
technical manager. On the other hand, most of the small companies do not have a
legal department or a contract department. Therefore, a procurement manager who
comes from a technical background would handle contractual matters.
The author experienced some difculties with getting appointments with the respon-
dents, especially the operators, who are quite reluctant to be interviewed and not
easily accessible. The respondents were given the questions together with a consent
form before the interview took place to allow them to have a general idea of what
was expected from them during the interview. The interviews were conducted in
their ofces and the respondents were required to return the signed consent forms
before the interview took place. This was done to ensure that the respondents fully
understood the subject matter of the research and to confrm that they were vol-
untarily participating in the research. Both the recording of the interview and the
notes taken during the interview have been kept confdential. All information about
the respondents is made anonymous.
2 Description of the case studies

This section will report the evidence found from the case study conducted amongst
oil and gas companies in Malaysia.
• Operator A
Operator A is the largest oil company in Malaysia. It engages in locating, exploring
and producing oil and gas reservoirs. It also engages in oil and gas well drilling, as
well as engineering, constructing and commissioning natural gas pipelines and associ-
ated facilities. The company is based in Kuala Lumpur, Malaysia.
• Operator B
Operator B is a multinational company engaging in exploration activities. It entered
Malaysia in 1999 and currently holds the majority interest in seven separate produc-
tion-sharing contracts and three gas-holding agreements. The company is based in
Kuala Lumpur, Malaysia.
• Operator C
Contractor C principally engages in exploration and production activities, as well as
operation and maintenance of foating, production storage and an ofoading tanker
facility.
• Contractor A
Contractor A’s involvement in the oil and gas sector focuses on pipeline services such
as pre-commissioning; commissioning and de-commissioning; ofshore transportation
263
and installation; operation and maintenance; fabrication and construction; topside
major maintenance and hook-up commissioning; Engineering, Procurement, Con-
struction and Commissioning (EPCC); onshore pipeline and construction; underwater
services; and ship management and catering. Contractor A has entered into a few
contracts with Operators A, B and C.
• Contractor B
Contractor B is one of the world’s largest integrated oil and gas services and solu-
tions providers. This company provides end-to-end solutions and services to the
upstream petroleum industry and covers activities such as installation of ofshore
pipelines and structures, fabrication of ofshore structures, accommodation and sup-
port vessels, drilling vessels, hook-up and commissioning, topside maintenance ser-
vices, underwater services, ofshore geotechnical and geophysical services, project
management, diving services, ofshore support services, infrastructure and specialised
steel fabrication works. Contractor B has a number of ongoing jobs with Operators
A, B and C.
• Contractor C
Contractor C has been active in the oil and gas industry for 35 years. It holds a
business license to trade with Operator A. It is a pioneer in supplying metering and
regulating skids in Malaysia for natural gas transmission and distribution. The com-
pany provides EPCC for custody transfer and non-custody transfer metering and
system integration for onshore and ofshore applications. It also represents and is
business partners with global leaders in the areas of control valves and instrumenta-
tion, measurement and fow management, safety, electrical and environmental work.
Contractor C has signed a few contracts with Operators A, B and C.
• Contractor D
Contractor D provides quality and value-added products and services, especially
valve-related products in the oil and gas industry. Contractor D has experience in
dealing with and providing its product and services to Operators A and B.
• Contractor E
Contractor E engages in trading and supply of protective coating systems in the oil
and gas industry. Apart from that, it also specialises in the design, manufacturing
and installation of pipeline strengthening, repair and protection systems for oil and
gas installations. Operator A is one of the main clients of Contractor E.
• Legal Firm Z
Legal Firm Z specialises in handling cases that are related to maritime, vessels and
disputes in oil and gas contracts, and most of its clients are operators. The frm is
located in Kuala Lumpur.
Table 14.1 provides a summary for better understanding about the number of
respondents involved in this study.
The next section will present the fndings on how operators and contractors perceive
the distribution of contractual risk during contract formation.
264
Table 14.1 Categories of Respondents
Categories of Respondents
units of analysis
No. Units of analysis No. of Designation
respondents
Contractor 1 Contractor A 2 Legal head + contract
manager
2 Contractor B 2 Legal manager +
contract manager
3 Contractor C 2 Legal manager +
principal technical
4 Contractor D 1 Procurement manager
5 Contractor E 1 Procurement manager
Operator 6 Operator A 1 Head project
7 Legal Firm Z 1 Legal practitioner
Total number of respondents 10
2.1 Findings
Two fndings of empirical study will be discussed in this section: an analysis on the
perception of the contractual formation process between operators and contractors,
and an analysis on the issues of risk allocation and indemnity clauses which are
commonly used by the parties. The analysis of indemnity clauses will be divided into
four parts: (1) liability with regard to personal injury or death of employees and loss
of property of the parties; (2) third-party claims; (3) liability with regard to pollu-
tion; and (4)insurance coverage.
2.2 Perception of the contractual formation process

This section presents the evidence on how the respondents perceive contractual
negotiation during the contract formation process and also whether there is equal
bargaining between operators and contractors.118 The case study was undertaken with
the aim to primarily investigate the extent of contractual formation process that leads
to the distribution of risk allocation provision in oilfeld service contracts.
Competitive tendering is a standard practice for operators to obtain information
on the optimal price in the contracting market.119 Tendering is part of the contract
formation process. “Formation” in this context means the process by which a contract
comes into being.120 Usually, the operator would invite several contractors to bid for
118 Wan M Zulhafz, ‘Perception of Contractual Risk Allocation in the Oil and Gas Contracts in Malay-
sia’ (2018) 11 Int.J.Trade Glob.Mark. (IJTGM) 127.
119 Kees Berends, ‘Engineering and Construction Projects for Oil and Gas Processing Facilities: Con-
tracting, Uncertainty and the Economics of Information’ (2007) 35 Energy Pol’y. 4260; MJ Greaves, ‘Under-
standing and Managing Risk in Real Estate Investments’ (1983) 1 Surveyor Valuer (SISV) 21.
120 Robinson M Nigel and Lavers P Anthony, Construction Law in Singapore and Malaysia (Butter-
worths & Co (Asia) Pte Ltd 1988) 68.
265
the oilfeld service job in accordance with detailed technical specifcations outlined in
the bid invitation121 and subsequently issue a package of documentation including a
blank form tender and the instruction to tenderers. This package is called Invitation
to Tender (ITT) or sometimes Invitation to Bid (ITB).
Negotiation of the contractual terms is crucial before the operator accepts one of
the bids and concludes the contract, as it enables the parties to identify one or more
incompatibilities between them and work to fnd a mutually acceptable solution.122
Generally, standard forms of contract are not used for major oil and gas projects; in
almost all cases, stand-alone bespoke contracts are used by the major oil companies
which bear little resemblance to the well-established standard forms except perhaps
the international FIDIC123 contract general conditions, from which a lot of bespoke
construction contracts are derived.124 Most major oil companies, large independents
and national oil companies have developed their own company of contract forms.125
One of the respondent representing contractor remarked:
The operators or the clients will usually request a quote by issuing RFQ or sometimes,
they may issue more formal request for proposals by way of RFP, ITT or ITB to the
contractors. So, if the contractors wish to perform the described work, then they will
respond with the price that they would charge. The acceptance of the bid by the client
is considered as an “award” of the contract. These documents i.e. the RFQ, RFP, ITT
and ITB contain COC (Condition of Contract) that has already been prepared by the
operators. The operators, they prefer to use their own customized contract or the COC.
This COC would be given to us (i.e. the contractors) as a base to initiate contractual
negotiation.126
Most of the oilfeld service contracts are based on one standard model form or
another and they tend to be similar in basic content with few signifcant modifca-
tions depending on the particular commercial activities involved.127 Standard forms
of contract are rarely utilised for signifcant oil and gas projects; instead, the major
oil frms almost always use stand-alone customised contracts that bear little rela-
tion to the well-established standard forms.128 The operators will often keep their
own database of model form of contracts and use them as a foundation in contract
negotiations.129 This practice apparently saves the parties time and transaction
costs and is more convenient, as they are working on the model form of contracts
121 Owen L Anderson, ‘The Anatomy of an Oil and Gas Drilling Contract’ (1989) 25 Tulsa L.J. 359.
122 David A Hensher and John Stanley, ‘Transacting under a Performance-Based Contract: The Role of
Negotiation and Competitive Tendering’ [2008] Transportation Research Part A: Policy and Practice 1143.
123 FIDIC stands for Federation International Des Ingenieurs-Conseils.
124 Vincent Hooker, ‘Major Oil and Gas Projects - the Real Risks to EPC Contractors and Owners’
(2010) 26 Const.L.J. 98.
125 Cary A Moomjian, ‘Drilling Contract Historical Development and Future Trends Post-Macondo:
Refections on a 35 Year Industry Career’, IADC/SPE Drilling Conference and Exhibition on 7th March
2012 in San Diego, California, USA (Society of Petroleum Engineers 2012) 9 http://www.drillingcontrac-
tor.org/wp-content/uploads/2012/04/Drilling-Contract-Historical-Development-and-Future-Trends-Post-
Macondo.pdf.
126 Respondent 2 from Contractor A.
127 Ibid.
128 Hooker (n126).
129 Hewitt (n19) 331.
266
with which they are already familiar.130 One of the respondents confrmed this
practice:
Operators usually prepare the contracts. Many oil and gas contracts from my experience
tend to be “standard bespoke” i.e. they are “standard” with respect to the individual
principal contracting entity – but on an industry basis – bespoke as they are not written
by e.g. standards committees or seek industry participation in their drafting of the stan-
dard terms.131
The contracts proposed by operators are subject to qualifcation by contractors that

tender for the work, resulting in negotiated contracts.132 The most important single
factor in any contract negotiation is the balance of bargaining power.133 The inequal-
ity of bargaining power during contractual negotiation can create contractual unfair-
ness between the parties. The dominant position of oil companies or the operators
relative to many contractors in the oil and gas industry is such that they may assume
that they can impose whatever conditions they wish.134 During the contract negotia-
tion process, there is a high possibility of inequality in bargaining power to take
place between the operators and contractors, especially when there is “imposition of
will by a dominant party on a party with inferior economic bargaining power, who
is being unfairly coerced into indemnifying the dominant party, stating that the result
will rest on the relative bargaining power of the parties”.135 Based on the fndings of
the case studies, all respondents agreed that the operators hold a dominant position
during contractual negotiation process. For example, one respondent said:
Of course, the operator would have a better and dominant position in negotiating the
contract. The operator is the one who create the job opportunity for the contractors. As
contractors, we provide them with services so as to make some proft. Sometimes, whether
we like or not, we are going to agree with most of the conditions set out by the
operator.136
Another respondent also confrmed this scenario:

The operators are our clients as well as the project owners; of course, they have better
position in negotiating the contracts.137
One of the respondents claimed that this scenario happened because of contractors’
desperation for jobs:
Basically, yes, the operators have greater bargaining power over the contractors. In order
to secure a job, the contractors have lesser rights to make bargaining. We can say that
the contractors actually desperately need job from the operators, while the operators have
130 Ibid.
131 Respondent 5 from Contractor C.
132 Leslie Edwards, Practical Risk Management in the Construction Industry (Thomas Telford 1995).
133 Chris Thorpe, Fundamentals of Upstream Petroleum Agreements (C P Thorpe Ltd 2008) 282.
134 Peter Cameron, ‘Liability for Catastrophic Risk in the Oil and Gas Industry’ [2012] IELR 207, 207;
also see ibid.
135 N Stephen Kinsella, ‘Oilfeld Indemnity and “Seperate Insurance” Provisions in the Wake of Getty
Oil’ [1994] Tex.Oil Gas L.J. 29.
136 Respondent 3 from Contractor B.
267
number of players (candidates) to do the job. For example, clause like indemnity, the
contractor has to take it as it is. If the operators see there are a lot of qualifcations
made by one contractor during the negotiation process, the operator might as well go
for another contractor.138
Historically in the upstream industry, the party holding bargaining power has used
their power to impose its preferred terms on the weaker party.139 The party with a
better bargaining position, such as the operator’s, will be able to use their position
to gain more favourable contract terms.140 However, the contractors might still sign
an imbalanced agreement in order to secure a job even though they will be subject
to unacceptable risk. In fact, some of the contractors risk double jeopardy, as any
extraordinary risks will not be accepted by subcontractors and will be passed down
in the usual way from the operator to the contractor but then not passed on in the
subcontract.141 In this situation, the contractor is exposed to both his own contractual
risk and the subcontractor’s risk.142 Generally, operators’ statutory duties pertaining
to the award of contracts give theoretical power to the contracting parties in order
to ensure fair, open and competitive tendering practices; however, only a brave con-
tractor would challenge an operator with whom he aspires to work in the future.143
Here, the inferior party may well fnd that he is given no opportunity to negotiate
the one-sided terms prior to the conclusion of the contract but that he should “take
it or leave it”.144 This is a problem “in so-called negotiated contracts if the stronger
party chooses to dictate to the other”.145 The stronger party may use his power by
declining to enter any contract at all.146
2.3 Perception of risk allocation and indemnity clauses

Semi-structured interviews were conducted to fnd the perception of risk allocation
from the people in the industry and to discover their views on its current practice in
Malaysia. Interviewees’ perceptions will be compared to samples of indemnity clauses
which have been obtained from Operator A, Operator B and Operator C.
In general, ranges of risk allocation clauses are commonly seen in contracts used
in the oil and gas industry including indemnity and hold harmless clauses, clauses
excluding liability for “consequential losses” and clauses limiting overall liability.147
The actual sharing of risk, indemnities and provisions for supporting insurances will

139 Thorpe (n135) 282.
140 Ibid.
141 Helen Franklin, ‘Irretrievable Breakdown? A Review of Operator/Contractor Relationships in the
Ofshore Oil and Gas Industry’ (2005) 23 J.E.R.L. 1, 7.
142 Ibid.
143 Ibid., p. 6.
144 Hugh Beale, ‘Unfair Contract Terms Act 1977’ [1978] Brit.J.Law Soc. 114, 115.
145 Ibid.
146 Stewart Macaulay, International Encyclopedia of Comparative Law (Mohr 1974) 18.
147 Greg Gordon, ‘Risk Allocation in Oil and Gas Contracts’ in Greg Gordon, John Paterson and Emre
Usenmez (eds), Oil and Gas Law: Current Practice & Emerging Trends, vol 2nd (Dundee University Press
2011) 443.
268
be determined by the wording of the relevant contract documents.148 In fact, many
contracts are non-standard or they are standard form contracts which are made
non-standard by additional clauses.149 Non-standard contracts transfer more risk to
the tenderer.150
It is important to note that subsequent to the Macondo oil spill in the Gulf of
Mexico in 2010, there was an attempt by operators to depart from the established
standard form contracts and shift greater risk to the contractor in the event of
catastrophe.151 Post Macondo, operators are now less willing to deviate from their
pro-forma contracts, and this creates a “take it or leave it” situation.152 Even before
the Macondo incident, some contractors claimed that they were “required to take on
onerous responsibilities without appropriate margins to absorb the consequences of
the associated risks”.153 Many contractors consider such non-negotiable contracts to
be problematic, primarily because they often contain onerous provisions in important
areas such as allocation of risk, and this can create signifcant risk exposure.154 As a
result, at the conclusion of the contracts the contractors will end up swallowing greater
risk and liabilities. As one of the respondents representing contractor remarked:
Usually, the risks are allocated through certain provisions such as indemnity clause, limi-
tation clause and exclusion clause. From what I can see, these clauses are usually one-
sided. . . . Well, the thing is that, the operators would not entertain if we put so much
qualifcation in the contracts. Sometimes, the operator is just going to say, “If you can’t
comply and you have a lot of exceptions to the clauses, then you will be disqualifed”.
As a result, we are not going to get the job. . . . We always have problem when it comes
to indemnity clause. We are expected to bear most of the liabilities. The worst part is
that, sometimes they (operators) expect us to be liable for something which is due to
their faults . . . we have no choice but to agree with such clause. It is always be the case,
either that we take or leave it.155
In distributing the risk between the operator and contractor, one of the respondents
claimed that contractors were usually at the losing end and expressed their dissatis-
faction at being made to indemnify operators’ negligence. This can be seen in the
following remark made by one of the respondents:
Supposedly, anything that are risky to us, then we need to take steps to mitigate such
risks or deviate from such terms and conditions. However, most of the time, the contrac-
tor always be at the losing end, this is because in order to secure a big job, whether the
contractors like it or not, the contractors have to meet the operator’s demand and must
get ready to take those risks. . . . The problem with the indemnity clause is that, when
148 Edwards (n134).

149 Ibid.
150 Ibid.
151 Cameron (n17) 204. For further details, see ‘In Re: Oil Spill by the Oil Rig “Deepwater Horizon” in
the Gulf of Mexico, on April 20, 2010: Memorandum in Support of Transocean’s Motion for Partial Sum-
mary Judgment against BP to Enforce BP’s Contractual Obligations, including BP’s Obligation to Defend,
Indemnify and Hold Transocean Harmless against Pollution Claims’ (US District Court, Eastern District
of Louisiana, 1 November 2011).
152 Moomjian, ‘Drilling Contract Historical Development and Future Trends Post-Macondo: Refec-
tions on a 35 Year Industry Career’ (n127).
153 Franklin (n143).
154 Ibid.
269
the operator transfers their liabilities to us by asking us indemnify it and even though it
was happened due to their negligence.156
The respondent also claimed that indemnity clauses were one-sided and that almost all
of the liability in respect of indemnity clauses was placed on contractors. It could be
argued that the best commercial policy would place responsibility for risk on the party
best able to manage it, e.g., the party with the relevant insurance coverage.157 Insuring
or contractually transferring risk to the insurer and allowing the premium to settle any
charges to the other party could mitigate risk exposure and in fact, this is the most
economically benefcial and practical way for risk to be dealt with.158 Insurance is used
by the indemnitor as a risk cushion in a situation when he is responsible for his own
employees and equipment. The insurance in fact is the underlying driver in this case
rather than an ancillary tool for risk management.159 This is particularly natural for
super-majors, who attempt to self-insure and minimise transaction costs.160 But this
reason is not applicable to some contractors, as they cannot aford self-insurance:
The indemnity clauses mostly are one sided. The contractors would be made to be liable
for most of the liabilities, for example the indemnity with regards to property and equip-
ment of the operators, the third-party liability, not to mention pollution. Usually we try
to keep it and make it consistent with the insurance coverage, for example per occurrence
how much we’ll be liable. Usually it is always unlimited liability and most of the time
the clients refuse to negotiate on that as well.161
It is argued that the practicalities of risk allocation limited by certain basic require-
ments for those to whom risk is being transferred such as ability to undertake a
hazardous task, willingness to take the risk, fnancial capability if the risk event
occurs, continued existence and adequate fnance during the period of liability.162 It
is also argued that, ideally, the responsibility for indemnifying the consequences of
a risk event resulting from the activities of one of the contracting parties should rest
with the party who has control over that risk.163 The operator is always in the best
position to control the risk, and this was confrmed by the operator himself:
As operator, it is mandatory to perform a project risk assessment and front-end study.
If the risks are uncertain, big or economically not meeting the target proft margin, the
project will not proceed.164
The operator also confrmed that both quantifed risk and qualitative risk, including
indemnity, would be transferred to contractors. He also contended that it is the
responsibility of contractors to understand and convert the risk into monitories:
156 Respondent 5, Contractor C.

157 Edwards (n134).
158 Max Abrahamson, ‘Risk Management’ (1984) 2 ICLR 241.
159 Caledonia North Sea Ltd. v British Telecommunications Plc same v Kelvin International Services Ltd.
same v London Bridge Engineering Ltd. same v Norton (No. 2) Ltd (In Liquidation) same v Pickup No. 7
Ltd. same v Stena Ofshore Ltd. same v Wood Group Engineering Contractors Ltd. – [2002] 1 Lloyd’s Rep 553.
160 ‘BP Annual Report and Accounts’ (2007) 39 https://www.bp.com/content/dam/bp/business-sites/en/
global/corporate/pdfs/investors/bp-annual-report-accounts-2007.pdf
161 Respondent 3, Contractor B.
162 Edwards (n134).
163 Ibid.
164 Respondent 10 from Operator A.
270
As operator, all the quantifed risk will be transferred to contractor and stated in contract.
The contractors are to put the price of each risk identifed in contract. It is responsibility
of contractor to understand and convert the risk into monitories. . . . The risks are made
clear to contractor. Contractor will put the prices for those scope specifed in the contract.
Any risk is to be priced by contractor. . . . Uncounted quantity will use reimbursable
cost plus. . . . Indemnity scope is to be taken by contractor; cost of premium for indem-
nity will be claim to operator.
Therefore, the contractor is inserted into a position where he has to absorb the risk
and cost that risk into the price. Two respondents confrmed this scenario:
Let say, there are some conditions that we could not aford to accept them in the event
the operators attempt to shift greater risks to the us – now the operators go for competi-
tive bids so our chances to be awarded is lesser if we stick to our qualifcation.165
Another respondent commented:

However, we are not in the position to change the conditions, so what we normally do
is, we will take note on that and accordingly advise to our technical people, “please
calculate this risk into your cost”. Well, the thing is that, the operators would not entertain
if we put so much qualifcation in the contracts. Sometimes, the operator is just going
say, “if you can’t comply and you have a lot of exceptions to the clauses, then you will be
disqualifed”. As a result, we are not going to get the job.166
However, it is difcult for contractors to calculate and set an ideal price for their
services after absorbing the risk. Due to the highly competitive nature of the bidding
process, the contractor faces a dilemma when setting the price. The contractor is
either afraid of not getting the job if the price he sets is too high after converting
the risk into monitories. Or he fears exposure to fnancial problems if the price he
sets is too low. Two respondents commented on this:
We need to set the price. The price should be an ideal one. Not too high as there is pos-
sibility our submission would be rejected, but not too low to the extent that it might
jeopardies our proft.167
The second respondent said:

Usually, in order to mitigate the risks, we have to cost in the impact into pricing. Whatever
the risks involved; the cost has to be refected in the pricing. Sometimes, this would be
a problem, when we were trying to cost in everything, the cost will be too high and we
afraid that we are not going to get the job. But, if we neglect the risks now, then if
anything happens in future, the risks would be at our own cost. So, it is real challenge
for us to draw a middle line between these two.168
It is true that a contract is an equalisation of accepting between controllable and

uncontrollable risks with the price deemed appropriate to undertake the work.169
However, it is said that the main reason for the increase in overall costs is due to the

169 Jur Tunay KÖKSAL, ‘FIDIC Conditions of Contract as a Model for an International Construction
Contract’ (2011) 1 BMR 32.
271
usage of disclaimer clauses in allocating risk such as indemnity clauses.170 This is
because, once the risk is transferred to the contractor and “the contractor has no
means by which to control the occurrence or outcome of the risk, the contractor
must either insure against it or add a contingency to the bid price”.171 The cost of
transferring risk to the contractor through such clauses presents a few hidden costs
such as “restricted bid competition, increased potential for claims and disputes and
above all, more adversarial owner-contractor relationships.”172
One of the respondents, who is a practicing lawyer and used to litigate on behalf
of the operator, shared her perspective on this issue:
I assume when the contractors signed the contract, the contractors were fully aware of
those liabilities that they will be carrying on. So, if anything happens in future, I think
it is their obligation to get protection for those liabilities by way of insurance. Plus, the
parties are able to practice contractual freedom. I don’t see any problem with regard to
this matter, unless, there is issue with regard to fairness, but such allegation should be
proven by separate cause of action and proper hearing in court.173
It is true that as a matter of contractual freedom, the parties may freely decide the
terms of the contract as they wish including the risk allocation provisions.174 However,
this particular respondent might not be aware of the actual situation in the industry
during the contract formation process. For example, if there is inequality of bargain-
ing power between the parties, then contractual freedom has not been exercised
properly. As a result, the contract becomes one-sided and unfair risk allocation has
taken place.
The contractors are actually aware of the situation, but they are not in a position
to change or qualify any of the terms. Nor are they given the opportunity to discuss
or negotiate on the allocation of risk. One of the respondents remarked:
The most we can do is to voice our dissatisfaction to the clients. Sometimes they may
listen to us, unfortunately most of the time they are not. How would we mitigate? Basi-
cally, it is good to have both parties to sit down and explain and discuss about the risks
involved in each project. But I suppose they are going to say that, it is them who invest
money; thus, they will not accept any of our qualifcation with regard to the risks.175
There are circumstances, however – possibly owing to the self-insurance of certain

risks – where one party will attempt to pass additional risks on to the other.176 This
may require the contractor to indemnify the operator for the operator’s negligence
170 F Hartman, ‘Construction Dispute Resolution through an Improved Contracting Process in the
Canadian Context’ (Loughborough University of Technology, UK 1993); G Jergeas and F Hartman, ‘A
Contract Clause for Allocating Risk’ [1996] AACE Int. Trans.; JG Zack Jr., ‘“Risk-Sharing” – Good Con-
cept, Bad Name’ (1996) 38 Cost Engin. (Morgantown, West Virginia) 26.
171 GF Jergeas and FT Hartman, ‘Contractors’ Protection Against Construction Claims’, Annual
Meeting-American Association of Cost Engineers (AACE 1994) 8.
172 C Robert, ‘Managing Change Orders and Claims’ (1997) 2 J. Manage. Eng. 27; D Becker, ‘The
Cost of General Conditions’ [1993] Am. Assoc. Cost. Eng. Trans. 7; Kwaku A Tenah, ‘The Design-Build
Approach: An Overview’ (2000) 42 Cost Engin. 31.
173 Respondent 10 from Legal Firm Z.
174 Privy Council in Ooi Boon Leong v Citibank N.A. [1984] 1 MLJ 222, confrmed that “parties to an
agreement have much scope to negotiate and incorporate terms acceptable to them”.
175 Respondent 10 from Legal Firm Z.
272
to the operator’s employee and property, or vice versa. This would represent an
uninsurable risk to the contractor and the adjustment severely undermines the knock-
for-knock indemnity regime and may also give rise to the possibility of increased
costs as the contractor attempts to insure a risk in which it does not have a true
insurable interest.177 One of the respondents commented on this scenario:
We always have problem when it comes to indemnity clause. We are expected to bear
most of the liabilities. The worst part is that, sometimes they (operators) expect us to be
liable for something which is due to their faults. Could you imagine that? But, what can
we say; we have no choice but to agree with such clause? It is always be the case, either
that we take or leave it. . . . The risks have to be covered by insurance. We do not aford
to take the risks without any fnancial back up from the insurance company. . . . This is
the problem. Sometimes, we have to absorb all risk regardless whether such risk is being
covered by the insurance company.178
Moreover, from the contractor’s point of view, the risk is not one which can be passed
down to subcontractors. Again, parties who seek protection from the other party’s
negligence in an adjusted knock-for-knock regime similarly undermine the benefts
of the concept since they, the indemnifying party, will in any event be obligated to
secure additional insurance to cover the consequences of their negligence to the other
party’s employees or property. Moreover, it is likely that the existence of negligence
may frst have to be proved in the courts for an indemnity to operate, which defeats
one of the primary objectives of the knock-for-knock regime.
In the same vein, a contractor will often seek to limit its liability for third-party
risks arising out of the agreement to the extent that this exceeds the limit of its third-
party insurance.179 This is consistent with an overall desire to cap risks over which
it has little or no control, including the catastrophic risks associated with ofshore
exploitation; e.g., fre, explosion, blowout (underground and surface) and pollution
emanating from the reservoir. Without liability caps, the contractor will have to bear
what are normally uninsured risks which may well be disproportionate to the size,
nature and fnancial return of the contract.
However, the quid pro quo for appropriate protection by the operator is that the
contractor should be prepared to insure the operator.180 It is not reasonable to
expect limitation of, and/or indemnity from, catastrophic and third-party risks that
the contractor’s own insurance covers. Contractors should be prepared to be open
about the scope and terms of their insurance cover as well as other material infor-
mation about the size and resources of the company and, where applicable, parent
and group company structure. In this regard, one of the respondents commented:
Again, the indemnity clauses mostly are one sided. The contractors would be made to
be liable for most of the liabilities, for example the indemnity with regards to property
and equipment of the operators, the third-party liability, not to mention pollution.
177 Ibid.
180 Wan M Zulhafz, ‘An Empirical Study on the Contractual Risk Allocation Provisions and Indemnity
and Hold Harmless Clauses in the Oilfeld Service Contracts in Malaysia’, Paper Proceedings on ‘Second
International Conference on Interdisciplinary Legal Studies (ICILS) 2015’ on 9th–10th June 2015 in Toronto,
Canada (Unique Conference Canada, June 2015).
273
Usually, we try to keep it and make it consistent with the insurance coverage, for example
per occurrence how much we’ll be liable. Usually it is always unlimited liability and most
of the time the clients refuse to negotiate on that as well.181
In order to combat this problem, some of the respondents suggested that fairness
could be achieved by legal intervention where the legislator passes a law to protect
contractors. Another respondent suggested that it should be an anti-indemnity
law:
I think we need the government interference to come out with a regulation in order to
achieve fairness.182
Another respondent commented:

I think it is quite ideal to have some sort of protection by the legislator on this matter.183
A third respondent stated:

I know that in US, they have oilfeld anti-indemnity law, but I am not quite sure how
far does the law really efcient to address this issue. What I can confrm, we have nothing
yet in Malaysia to that efect. To certain extent, I think yes, we need rules by the govern-
ment to solve the problem with regards to indemnity clauses.184
One of the respondents shared his views, saying that the relevant authority should
issue guidelines in order to address this problem:
I think it is good if the authority come out with guidelines to monitor this problem but
it has to be tightened up with our Petroleum Act.185
Therefore, based on the fndings from the semi-structured interviews with the respon-
dents, it could be argued that operators has dominant position over contractors and
in a better position during contractual negotiation in Malaysia. Due to the lesser
bargaining power, contractors are given fewer opportunities to modify the terms and
conditions in order to make them fairer. However, as discussed above, despite the
unfair terms and conditions, contractors end up agreeing to them in order to secure
present and future jobs. Moreover, contractors are exposed to the risk without covered
by insurance. This situation is regarded as a serious problem because it may lead to
great fnancial losses to contractors.
3 Analysis of the indemnity and hold harmless clauses of

Operators A, B and C
In order to assess the claims made by the contractors, the indemnity clauses in three samples
of oilfeld service contracts which were drafted by three diferent operators: Operator A,
Operator B and Operator C are each analysed in four diferent contexts, namely:

183 Respondent 8 from Contractor D.
185 Respondent 7 from Contractor E.
274
1 Liability with regard to personal injury or death of employees and loss of
property of the parties;
2 Third-party claims;
3 Liability with regard to pollution;
4 Insurance coverage.
These indemnity clauses are compared with two model form contracts which are
produced by LOGIC186 and FIDIC.187 LOGIC is a not-for-proft subsidiary of Of-
shore Energies UK (OEUK; formerly Oil and Gas UK).188 LOGIC operates as the
custodian for cross-industry projects that aim to increase the efciency of working
practice in the UK Continental Shelf (UKCS). The standard contracts for the UK
ofshore oil and gas industry (formerly CRINE contracts) were developed by the
Standard Contracts Committee and are issued by LOGIC for use within the industry
between companies and their contractors.189
In South and South East Asia (India, Indonesia, Malaysia, Thailand and Vietnam),
LOGIC is widely used as a starting point in drafting contracts due to its international
character.190 The General Conditions of Contract for Construction are prepared for
the scopes of work that are related to major fabrication, topsides installation and
hook-up, signifcant topsides modifcations, construction services contracts for top-
sides works and also other contracting arrangements such as EPCC Contracts or
Engineering, Procurement, Fabrication and Installation (EPFI) Contracts.191
The International Federation of Consulting Engineers (FIDIC) publishes interna-
tional standard form contracts for works and also provides related materials such as
standard pre-qualifcation forms.192 The FIDIC (Silver Book) model contract is usually
used by the industry as a foundation in drafting turnkey construction contracts for
lump-sum projects.193 The standard forms of contracts by LOGIC and FIDIC are
the two examples that are widely used by the oil and gas industry players around the
world as a foundation to draft oilfeld service contracts. These two contracts, par-
ticularly their indemnity clause provisions, will be compared to the contracts drafted
by Operator A, Operator B and Operator C later in this chapter.
The analysis will also make reference to the General Contracting Principles (FAIR
Principles) which are provided by the International Marine Contractors Association
(IMCA).194 These principles serve the long-term interests of all players in the oil
and gas industry by encouraging fair contractual provisions according to the parties’
186 LOGIC Construction Contract, 2nd Edition, 2003.

187 FIDIC Conditions of Contract for EPC Turnkey Projects (Silver Book), 1999.
188 www.oeuk.co.uk/.
189 See www.logic-oil.com/.
190 Hewitt (n19) 332.
191 ‘General Conditions of Contract (including Guidance Notes) for Construction’ in Standard Contract
for U.K. Ofshore Oil and Gas Industry, vol October (2nd edn, LOGIC 2003) 2.
192 See more at http://fdic.org/.
193 Huse; Axel-Volkmar Jaeger and Götz-Sebastian Hök, ‘FIDIC Contract Documents’, FIDIC-A
Guide for Practitioners (Springer 2010) 125.
194 IMCA is an international trade association representing ofshore, marine and underwater engineer-
ing companies. One of its roles is to communicate and promote dialogue amongst the various companies,
both clients and contractors, for the beneft of the whole industry.
275
respective risks and rewards.195 It is vital to note that the FAIR Principles are not
law but merely provide guidelines in drafting oil and gas contracts, which promotes
well-established industry custom and practice.196
3.1 Liability regarding personal injury or death of employees and loss

of property of the parties
According to the FAIR Principles, both parties should grant reciprocal, unlimited
indemnities for the costs of loss of or damage to both parties’ property and person-
nel of whatsoever nature or the property and personnel of the parties’ group, irre-
spective of cause (including negligence).
This principle ensures that each party to the contract has clear and easily under-
stood liability during the project for both people and property. Further, this alloca-
tion of risk could aim to avoid increasing either party’s liability levels as a result of
involvement in the project which would require additional insurance cover or the
implementation of an alternative risk management regime for the project.197
It is important to note that the FAIR principles are not law. Hence they are not
legally binding. The FAIR principles operate as a guideline to the parties on how
to draft contracts with fair contractual terms and conditions. For this reason, the
FAIR principles will be referred to in the analysis below in setting out a benchmark
for what would be a fair indemnity clause.
3.1.1 Under LOGIC

In the oil and gas industry, contractual indemnity clauses often take the form of a
specifc type of broad form indemnity known as the “knock-for-knock” provision,
where an individually party agrees to indemnify the other for injuries or damage to
its personnel and property regardless of fault.198 In other words, each party agrees
to take full responsibility for all bodily injury or property damage claims made by
its own employees, regardless of which party may actually be responsible for the
injury.199 For example, under the LOGIC Construction Contract Ed, 2 October 2003,
the reciprocal indemnities between the Company and the Contractor relate to the
parties included in the Company and Contractor Groups as follows:
22.1 The CONTRACTOR shall be responsible for and shall save, indemnify, defend and
hold harmless the COMPANY GROUP from and against all claims, losses damages,
costs (including legal cost) expenses and liabilities in respect of:
(a) Loss or damage to property of the CONTRACTOR GROUP whether owned,
hired, leased or otherwise provided by the CONTRACTOR GROUP arising
from, relating to or in connection with the performance or non-performance of
the CONTRACT; and
(b) personal injury including death or disease to any person employed by the CON-
TRACTOR GROUP arising from, relating to or in connection with the perfor-
mance or non-performance of the CONTRACT; and . . .
195 Louys (n289).

196 Culligan and de Roo (n21).
197 ‘IMCA General Contracting Principles (Rev.1)’ (2011), www.imca-int.com, accessed 6 January 2014.
198 Egbochue (n166).
199 Evans and Butler (n334).
276
22.1 The COMPANY shall be responsible for and shall save, indemnify, defend and
hold harmless the CONTRACTOR GROUP from and against all claims, losses damages,
costs (including legal cost) expenses and liabilities in respect of:
(a) loss of or damage to property of the COMPANY GROUP whether
(i) owned by the COMPANY GROUP, or
(ii) leased or otherwise obtained under arrangements with fnancial institution by the
COMPANY GROUP
which is located at the WORKSITE arising from, relating to or in connection with
the performance or non-performance of the CONTRACT, but excluding the PERMA-
NENT WORK; and . . .
(b) personal injury including death or disease to any person employed by the COM-
PANY GROUP arising from, relating to or in connection with the performance
or non-performance of the CONTRACT; and . . .
22.6 All exclusions and indemnities given under this Clause 22 (save for those under
Clause 22.1(c), 22.2(c) and 22.5(b)) and Clause 25 shall apply irrespective of cause and
notwithstanding the negligence or breach of duty (whether statutory or otherwise) of
the indemnifed party or any other entity or party and shall apply irrespective of any
claim in tort, under contract or otherwise at law.
In respect of personal injury including death or disease of the contractor’s employees,

the contractor is required to indemnify the company who, in the context of this research,
is the operator. On the other hand, the company will indemnify the contractor for the
company’s employees with regard to personal injury including death or disease. In respect
of loss or damage to property, both are required to be responsible and indemnify each
other with regard to their own property. Clause 22.6 has made clear that both of the
contractors and company will be liable for their own employees regardless of whether
death, disease, loss and/or damage were due to the negligence of the indemnifed party.
In the Caledonia North Sea Ltd v London Bridge Engineering Ltd200 case, Lord Bing-
ham referred to the knock-for-knock indemnity provision that covered employees as
a “market practice [which] has developed to take account of the peculiar features of
ofshore operations”.201 The court held that the knock-for-knock indemnity provision,
properly construed, entitled the operator to be indemnifed by the contractors with
regards to the fatalities and injuries even where they were not liable at common law,
or liable for breach of statutory obligation.202 Knock-for-knock indemnity provisions
are explained in further detail in Chapter 4 of this book.
3.1.2 Under FIDIC

Under the FIDIC model contract, the clause requires each party to indemnify the other
Party from any claims arising out of the contractor’s execution of the works.203 This
obligation extends to any third-party claims. The indemnity clause reads as follows:
17.1 The Contractor shall indemnify and hold harmless the Employer, the Employer’s
Personnel, and their respective agents, against and from all claims, damages, losses
and expenses (including legal fees and expenses) in respect of:
(a) bodily injury, sickness, disease or death, of any person whatsoever arising out
of or in the course of or by reason of the design, execution and completion
200 [2002] UKHL 4.

201 Ibid., para. 7.
202 Ibid.
203 Axel Volkmar Jaeger and Götz-Sebastian Hök, FIDIC: A Guide for Practitioners (Springer 2009).
277
of the Works and the remedying of any defects, unless attributable to any
negligence, willful act or breach of the Contract by the Employer, the Employ-
er’s Personnel, or any of their respective agents, and . . .
The Employer shall indemnify and hold harmless the Contractor, the Contractor’s Per-
sonnel, and their respective agents, against and from all claims, damages, losses and
expenses (including legal fees and expenses) in respect of (1) bodily injury, sickness,
disease or death, which is attributable to any negligence, willful act or breach of the
Contract by the Employer, the Employer’s Personnel, or any of their respective agents,
and (2) the matters for which liability may be excluded from insurance cover, as described
in sub-paragraphs (d)(i), (ii) and (iii) of Sub-Clause 18.3 [Insurance Against Injury to
Persons and Damage to Property].
This research interprets the word “employer” as “operator”. The contractor is

required to indemnify the employer, the employer’s personnel and their agents against
and from all claims, damage, etc. in respect of injury, sickness, disease or death of
any person arising from the contractor’s execution of the works, including any
consequences arising – from the contractor’s design (if any). However, the contractor
is not required to provide indemnity in respect of any negligence or wilful act of
the employer, employer’s personnel or their agents’ damage to or loss of property
to the extent that such damage or loss arising out of or as a consequence of the
contractor’s design, the execution of the works and the remedying of defects or is
attributable to the negligence or a wilful act of the contractor, the contractor’s
personnel or his agents.
On the other hand, the employer is required to indemnify the contractor, the
contractor’s personnel and their agents from all claims in respect of injury, sickness,
disease or death attributable to the negligence of the employer, the employer’s per-
sonnel and their agents; the employer is also required to indemnify the contractor’s
liability which may be excluded from insurance coverage.
It is argued that while Clause 17.1 above is not knock-for-knock indemnity, it is an
example of fair indemnity and hold harmless clauses which bind the contractor. This
is because the clauses do not require the contractor to indemnify the employer, i.e.,
operator for employer’s negligence. In fact, the employer indemnifes the contractor
for the injury, sickness, disease or death which is caused by his negligence. Moreover,
the contractor does not bear uninsurable risk since the employer indemnifes the
contractor for any matters which are not covered by insurance.
3.1.3 Under Operator A’s contract

Operator A drafted the indemnity and hold harmless clauses with regard to the per-
sonal injury or death of employees and loss of property of the parties as follows:
Article X – Liabilities and Indemnities
For the purpose of this article, the COMPANY GROUP shall mean OPERATOR A,
the COMPANY, its CO-VENTURES, its and their respective AFFILIATES, its and their
respective directors, ofcers and employees.
X.1 Personnel and the COMPANY GROUP
The COMPANY shall be responsible for and shall protect, defend, indemnify and hold
harmless the CONTRACTOR and the SUB-CONTRACTOR from and against any and
all claims, liabilities, costs, damages, expenses of every kind and nature, with respect to
278
injury or death or damage to or loss of property of any person employed by the company
group, howsoever arising.
X.2 Personnel of CONTRACTOR
The CONTRACTOR shall be responsible for all and shall protect, defend, indemnify
and hold harmless the COMPANY GROUP from and against any and all claims, liabili-
ties, costs, damages and expenses of every kind and nature, with respect to injury or
death or damage to or loss of property of any person employed by the CONTRACTOR
and/or the SUB-CONTRACTOR, howsoever arising.
X.3 COMPANY’s PROPERTY
The CONTRACTOR shall be liable for and shall indemnify the COMPANY GROUP
against any damage to or destruction or loss of property operated and/or owned by the
COMPANY arising during, and/or as a result of the performance of this CONTRACT,
without regard to whether any act or omission of the COMPANY GROUP contributed
to the loss.
X.4 CONTRACTOR’s EQUIPMENT
The CONTRACTOR shall assume the risk of, and shall be solely responsible for and
in this regard shall indemnify, defend and hold the COMPANY harmless against any
claims arising out of all damage to and/or loss destruction of CONTRACTOR’s EQUIP-
MENT lost or damaged, afects the performance of the SCOPE OF WORKS, the
CONTRACTOR shall replace any lost or damaged CONTRACTOR’s EQUIPMENT
at CONTRACTOR’s sole cost in the most expeditious manner possible and at CON-
TRACTOR’s sole expense.
Based the reading of the above Clause X.1 and X.2, it could be argued that such clause
was drafted on a mutual hold harmless basis, whereby the contractor is required to
indemnify and hold harmless the operator in respect of personal injury and death of
the contractor’s employees. Meanwhile, the company is required to indemnify and hold
harmless the contractor in respect of personal injury and death of the company’s
employees.
This research argues that this provision goes beyond the standard practice as set out
by LOGIC and FIDIC. Under the standard practice only personal injury and death
are covered, whereas the above provision also sets a condition that the contractor is
required to indemnify and hold harmless the company for the “loss of property of
any person employed by the contractor” and the company is required to indemnify
and hold harmless the contractor for the “loss of property of any person employed
by the company”. It could be argued that these extra requirements do not undermine
fair practice as regards indemnity and hold harmless clauses and therefore do not
cause problems for contractors.
However, on further reading of Articles X.4 and X.5, one might realise that
there is imbalanced risk allocation and unfair indemnity hold harmless clauses,
where the contractor is required to indemnify and hold harmless the operator for
the destruction or loss of property operated or owned by the operator irrespective
of whether such loss was caused by the act or omission of the operator. Apart
from that, the contractor is made solely responsible for and required to indemnify
and hold harmless the operator for all damage to and/or loss of or destruction to
property of the contractor’s equipment which occurs in the performance of the
work, at the contractor’s sole cost and expense. These clauses could be unfair to
the contractor.
279
3.1.4 Under Operator B’s contract
Operator B drafted the indemnity and hold harmless clauses with regard to the
personal injury or death of employees and loss of property of the parties as
follows:
Article X – Indemnifcation
X.1 CONTRACTOR agrees to indemnify, defend and save harmless COMPANY
GROUP from and against any and all claims, losses and expenses including without
limitation costs, demands, damages, suits, judgments, fnes, penalties, liabilities, rea-
sonable attorneys’ fees and causes of action whatsoever nature or character, whether
known or unknown, and including without limitation claims, losses and expenses for
property damage, bodily injury, illness, disease, death, pollution or loss of services
wages, consortium or society) in any way directly or indirectly, arising out of, or
related to, the performance or subject matter of this CONTRACT or the ingress,
egress, or presence on any premises (whether land, building, vehicle, platform, aircraft,
vessel or otherwise) owned, operated, chartered, leased, used, controlled or hired by
COMPANY GROUP or CONTRACTOR GROUP, and which are asserted by or
arise in favor of CONTRACTOR GROUP and expressly including any claims, losses
or expenses actually or allegedly caused by the sole, active, passive, concurrent or
partial negligence (of whatever nature or character), fault or strict liability of COM-
PANY GROUP or any other person or the unseaworthiness, unairworthiness or
defective condition of vessels, craft ore premises, whether or not preceding or during
the execution of this CONTRACT.
The above indemnity and hold harmless clause is unilateral, where it is one-sided and
appears to be more favorable to the operator. Under unilateral clauses, the contractor
is required to indemnify and hold harmless the operator for all losses and expenses
such as losses and expenses for property damage, bodily injury, illness, disease, death
and pollution of not only his own employees but also the company’s employees irre-
spective of whether such act was caused by the fault or sole negligence of the
operator.
In Article X, the phrase “including without limitation . . . fnes, penalties” could be
interpreted as granting punitive or exemplary damages. It could also be argued that
punitive or “exemplary damages serve to punish roguish defendants to make examples
of such defendants before society, to protect society’s interest in right conduct and
fair play”.204 Therefore, any indemnifcation given for such punitive or exemplary
damages would not be enforced as it is against public policy.205 Overall, these clauses
seem to be extremely unfair to contractors because it allows the operators to dump
liability and shift it to the contractor.206
204 Roy R Anderson Jr, ‘Indemnity Against Punitive Damages: An Examination of Punitive Damages,
Their Purpose, Public Policy, and the Coverage Provisions of the Texas Standard Automobile Liability
Insurance Policy’ (1973) 27 Sw.LJ 593.
205 Cary A Moomjian, ‘Macondo Litigation Update-How Court Decisions Will Impact Oilfeld Con-
tracting and Insurance Practices’, IADC/SPE Drilling Conference and Exhibition on 4th–6th of March 2014
in Fort Worth, Texas, USA (Society of Petroleum Engineers 2014).
206 Wan M Zulhafz, ‘Recent Trends in Allocation of Risk Post-Macondo: The Growing Tension
Between Oil and Gas Standard Forms of Contract, and Contractual Practice’ (2017) 5 Int.Energy Law Rev.
(IELR) 174.
280
3.1.5 Under Operator C’s contract
Operator C drafted the indemnity clause regarding the personal injury or death of
employees and loss of property of the parties as follows:
X.0 LIABILITIES AND INDEMNITIES

X.1 a) CONTRACTOR PERSONNEL
CONTRACTOR shall be responsible for and shall protect, defend, indemnify and hold
harmless COMPANY, its parent company, subsidiaries, AFFILIATES, consultants and
their respective agents, ofcers, and employees from and against any and all claims,
liabilities, costs, damages and expenses of every kind of nature, with respect to injury
or death or damage to or loss of property of any person employed by CONTRAC-
TOR arising during and/or as a result of the performance of this CONTRACT.
b) COMPANY PERSONNEL
COMPANY shall be responsible for and shall protect, defend, indemnify and hold
harmless CONTRACTOR, its parent company, subsidiaries, AFFLIALTES, consultants
and their respective agents, ofcers and employees from and against any and all claims,
liabilities, costs, damages and expenses of every kind and nature with respect to injury
or death of or damage to or loss of property of any person employed by COMPANY
arising during and/or as a result of the performance of this CONTRACT.
X.2 a) CONTRACTOR’s Equipment and Property
Save as otherwise expressly provide in this CONTRACT, CONTRACTOR shall
assume the risk of, and shall be solely responsible for and in this regard shall indem-
nify, defend and hold COMPANY, its parent company, subsidiaries, AFFILIATES,
consultants and their respective agents, ofcers and employees harmless against any
claims arising out of the damage to the loss, or destruction of, all CONTRACTOR
and its subcontractors’ equipment and property.
b) COMPANY’s Equipment and Property
CONTRACTOR shall be liable for and shall indemnify COMPANY against any
damage to or destruction or loss of property owned by COMPANY, its parent com-
pany, AFFILIATES, Client and other subcontractors and/or Operator B arising during,
and/or as a result of the performance of this CONTRACT, without regard to whether
any act or omission of COMPANY contributed to the loss.
The indemnity clause that has been drafted by Operator C is quite similar to the one
drafted by Operator A. Articles X.1 (a) and (b) were based on mutual hold harmless,
whereby the contractor is required to indemnify and hold harmless the operator in
respect of personal injury and death of the contractor’s employee and on the other
hand, the company is required to indemnify and hold harmless the contractor in
respect of personal injury and death of the company’s employee.
By contrast, it could be argued that there is imbalanced risk allocation and
unfair indemnity and hold harmless clause in Articles X.2 (a) and (b) where the
contractor is required to indemnify and hold harmless the operator and its afli-
ates for not only the damages and loss of the contractor’s property, but also the
operator’s and its afliates’ property regardless any act or omission of operator
contributed to the loss.
3.2 Claims by a third party

Based on the FAIR Principles, each party is best placed to assume legal liability for
losses which he causes to a third party. However, in circumstances where a contrac-
tor is required to perform work on the facility, or in an area of close proximity to
281
any existing facilities, the company should reasonably be expected to indemnify the
contractor group in full for all of the costs arising out of any loss or damage to
such existing facilities.207 Such indemnity would extend to cover the costs of existing
facilities property, third-party personnel, consequential losses and pollution efects
fnancial rewards to the contractor are not increased as a result of undertaking work
in an area which is in close proximity to existing facilities. Signifcant increases in
risk of loss or damage occurring to such existing facilities and the contractor’s own
property and personnel and additional risk in working in a congested area could
therefore be fairly allocated to the company which has the ability to mitigate such
third-party risk by seeking appropriate cross-indemnity agreements with the third-
party owners prior to commencing the project or obtaining sufcient insurance cover
for these signifcant risks as part of the whole project risk management.208
3.2.1 Third-party claims under LOGIC

It is common for oil and gas service contracts to provide for “guilty party pays” recipro-
cal indemnities.209 This means that party A indemnifes party B and party B’s group
against claims in respect of any death of or personal injury to third parties and loss of
or damage to the property of third parties. The indemnity clause covers negligence,
breach of statutory duty or breach of contract of party A or its group. On the other
hand, Party B provides a reciprocal indemnity in favour of party A and party A’s group.
Otherwise, contractors may give an indemnity in respect of third parties’ claims arising
out of the contractor’s operations except to the extent that there is negligence, breach
of statutory duty or breach of contract by the operator or any member of its group.
This places the onus on the contractor to prove such negligence or breach and therefore
flls a gap where it is diffcult to prove which party (if any) caused the death, personal
injury, loss or damage in question. This can be seen in the LOGIC contract form:
22.1 The CONTRACTOR shall be responsible for and shall save, indemnify, defend
and hold harmless the COMPANY GROUP from and against all claims, losses,
damages, costs (including legal costs) expenses and liabilities in respect of:
(c) subject to any other express provisions of the CONTRACT, personal injury
including death or disease or loss of or damage to the property of any third
party to the extent that any such injury, loss or damage is caused by the
negligence or breach of duty (whether statutory or otherwise) of the CON-
TRACTOR GROUP. For the purposes of this Clause 22.1(c) “third party”
shall mean any party which is not a member of the COMPANY GROUP
or CONTRACTOR GROUP.
22.2 The COMPANY shall be responsible for and shall save, indemnify, defend and
hold harmless the CONTRACTOR GROUP from and against all claims, losses,
damages, costs (including legal costs) expenses and liabilities in respect of:
(c) subject to any other express provisions of the CONTRACT, personal injury
including death or disease or loss of or damage to the property of any third
party to the extent that any such injury, loss or damage is caused by the
negligence or breach of duty (whether statutory or otherwise) of the
207 IMCA General Contracting Principles (Rev. 1) (n199).

208 Ibid.
209 Chidi Egbochue, ‘Reviewing Knock for Knock Indemnities Following the Macondo Well Blowout’
(2006) 4 C.L.Int. 7.
282
COMPANY GROUP. For the purposes of this Clause 22.2(c) “third party”
shall mean any party which is not a member of the CONTRACTOR GROUP
or COMPANY GROUP
(d) loss of or damage to such permanent third-party oil and gas production
facilities and pipelines and consequential losses arising therefrom, as speci-
fed in and defned in and in accordance with Appendix 1 to Section I – Form
of Agreement where such loss or damage is arising from, relating to or in
connection with the performance or non-performance of the CONTRACT.
The provisions of this Clause 22.2(d) shall apply notwithstanding the provi-
sions of Clause 22.1(c).
For the purposes of Clauses 22.1(c) and 22.2(c), the reciprocal indemnities between
Company and Contractor relate to the parties included in the Company and Con-
tractor Groups as defned in Clauses 1.2 and 1.9. The Company indemnifes Contrac-
tor Group in respect of loss/damage to permanent third-party oil and gas production
facilities, and consequential losses (as defned) therefrom.
A third party is defned in the LOGIC Contract as “any party who is not a mem-
ber of the Company or Contractor Groups”. Consequently, certain parties, who in
many cases will be present at some parts of the worksite, are third parties for the
purposes of the indemnity clauses. These include in particular other contractors of
the Company. It should be noted, therefore, that the Company’s other contractors
(and the subcontractors of such other contractors) are not included in the “Com-
pany Group” defnition. In this regard, the LOGIC Standard Contracts Committee
upholds the Industry Mutual Hold Harmless (IMHH) as the most appropriate means
of dealing with the allocation of liability for injury to persons, damage to property
and consequential loss between the Company’s contractors and strongly encourages
all contractors to join the IMHH scheme.210
3.2.2 Third-party claims under FIDIC

FIDIC drafts the indemnity clause regarding third-party claims as follows:
17. INDEMNITY
17.1 The Contractor shall indemnify and hold harmless the Employer, the Employer’s
Personnel, and their respective agents, against and from all claims, damages,
losses and expenses (including legal fees and expenses) in respect of:
(a) bodily injury, sickness, disease or death, of any person whatsoever arising
out of or in the course of or by reason of the design, execution and comple-
tion of the Works and the remedying of any defects, unless attributable to
any negligence, wilful act or breach of the Contract by the Employer, the
Employer’s Personnel, or any of their respective agents, and . . .
The Employer shall indemnify and hold harmless the Contractor, the Contractor’s Per-
sonnel, and their respective agents, against and from all claims, damages, losses and
expenses (including legal fees and expenses) in respect of (1) bodily injury, sickness,
disease or death, which is attributable to any negligence, willful act or breach of the
Contract by the Employer, the Employer’s Personnel, or any of their respective agents,
and (2) the matters for which liability may be excluded from insurance cover, as described
in sub-paragraphs (d)(i), (ii) and (iii) of Sub-Clause 18.3 [Insurance Against Injury to
Persons and Damage to Property].
210 Details of title IMHH are available on the LOGIC. See www.logic-oil.com.
283
Under the above Clause 17.1(a), the contractor is required to indemnify and hold
harmless the operator, the operator’s personnel, or its agents from all claims, damages,
losses and expenses in respect of bodily injury, sickness, disease or death of “any
person”. The phrase “any person” includes the employees of the contractor, operator
or any third party. However, the contractor would not be made liable if he is able to
prove that such claims, damages and losses happened because of the negligence, wilful
act or breach of the contract of the operator, the operator’s personnel, or its agents.
3.2.3 Third-party claims under Operator A’s contract

Operator A drafted the indemnity clause with regard to claims by third party as
follows:
X.3 Third Party
The CONTRACTOR shall be responsible for and shall protect, defend, indemnify and
hold harmless the COMPANY GROUP from and against any and all claims, liabilities,
cost, damages and expenses of every kind and nature with respect to injury, illness or
death of, or damage to or loss of property of any third party (including pollution), aris-
ing during and/or as a result of the performance of this CONTRACT without regard
to whether any act or omission of the COMPANY GROUP contributed to such injury,
death or damage to or loss of property.
The above indemnity clause is quite straightforward. It says that the contractor is
required to indemnify and hold harmless the operator, operator’s personnel and its
agent with regard to the claims, liabilities, cost, damage and expenses by a third party
in respect of injury, illness or death of or damage to or loss of property of any third
party (including pollution). This indemnity clause appears to be one-sided. This is
because the contractor would be made liable for the claims and liabilities by the third
party even where such acts were caused by the negligence, wilful act or omission of
the operator, the operator’s personnel, and its agents.
3.2.4 Third-party claims under Operator B’s contract

Operator B drafted the indemnity clause regarding third-party claims as follows:
X.1 CONTRACTOR agrees to indemnify, defend and save harmless COMPANY
GROUP from and against any and all claims, losses and expenses including without
limitation costs, demands, damages, suits, judgments, fnes, penalties, liabilities, reason-
able attorneys’ fees and causes of action whatsoever nature or character, whether known
or unknown, and including without limitation claims, losses and expenses for property
damage, bodily injury, illness, disease, death, pollution or loss of services wages, con-
sortium or society) in any way directly or indirectly, arising out of, or related to, the
performance or subject matter of this CONTRACT or the ingress, egress, or presence
on any premises (whether land, building, vehicle, platform, aircraft, vessel or otherwise)
owned, operated, chartered, leased, used, controlled or hired by COMPANY GROUP
or CONTRACTOR GROUP, and which are asserted by or arise in favor of CON-
TRACTOR GROUP and expressly including any claims, losses or expenses actually or
allegedly caused by the sole, active, passive, concurrent or partial negligence (of whatever
nature or character), fault or strict liability of COMPANY GROUP or any other person
or the unseaworthiness, unairworthiness or defective condition of vessels, craft ore
premises, whether or not preceding or during the execution of this CONTRACT.
The above indemnity clause is expansive and appears to be onerous for contractors.
The reason for this is that it seems that the operator has shifted most of the risks
284
to the contractor. Even though the clause has not mentioned anything about contrac-
tor’s liability in respect of third-party liabilities with regard to property damage,
bodily injury, illness, disease and death, broad phrases such as “any and all claims,
losses and expenses” and “in any way directly or indirectly arising out of, or related
to, the performance or subject matter of this CONTRACT” would imply that third-
party claims, losses and expenses in relation to property damage, bodily injury, illness,
disease, death and pollution are covered under the contractor’s liability.
This research argues that this clause is too wide since it imposes liability to the
contractor in a wide variety of situations. The contractor is to be liable irrespective
of whether such damages was caused by sole, active, passive, concurrent or partial
negligence of whatever nature or character, fault or strict liability of the operator or
the third party itself, or even though due to which could be considered as force majeure
events,211 such as unseaworthiness and unairworthiness. As mentioned previously, this
kind of drafting is an example of an attempt by the operator to dump his liability
by shifting risk to the contractor. This should therefore be regarded as bad drafting.
3.2.5 Third-party claims under Operator C’s contract

Operator C did not specify any indemnity regarding third-party claims.
3.3 Liability regarding pollution

For other areas of risk not caused by the actions of either party, standard form con-
tracts usually share the risk between them.212 For instance, in the case of weather, the
risk is often apportioned according to whether weather conditions are exceptional or
otherwise; it is usually deemed uneconomic to a client for contractors to account for
the risks associated with exceptional weather conditions in the tender process.213
According to the FAIR principles, each party is best placed to indemnify the other
party from the efects of pollution and contamination of whatever kind emanating from
the party’s respective groups’ property and facilities regardless of cause (including negli-
gence). This principle ensures that the parties to the contract have clear, easily understood
and distinct liabilities during the performance of the project. Furthermore, particularly
in relation to the contractor, it ensures that liability levels do not increase as a result of
participating in the project. If the liability levels increase, it would result in additional
insurance or risk management being required during the project term. A key aim is to
avoid duplicate insurance which leads to extra costs and often to gaps in cover.214
3.3.1 Pollution liability under LOGIC

Under the LOGIC contract form, there are reciprocal indemnifcations by both par-
ties in respect of pollution. The contractor would indemnify and hold harmless the
company group against claims in respect of pollution occurring on or emanating
from contractor group’s premises, property or equipment. On the other hand, the
211 Edward G Hinkelman, Glossary of International Trade (5th edn, World Trade Press 2009) 75.
212 Edwards (n134).
213 Ibid.
214 IMCA General Contracting Principles (Rev. 1) (n628).
285
company would indemnify and hold harmless the contractor group against claims
in respect of pollution emanating from the reservoir or the company group’s property.
The respective indemnity clause is written as follows:
22.3 Except as provided by Clause 22.1(a), Clause 22.1(b) and Clause 22.4, the COM-
PANY shall save, indemnify, defend and hold harmless the CONTRACTOR
GROUP from and against any claim of whatsoever nature arising from pollution
emanating from the reservoir or from the property of the COMPANY GROUP
arising from, relating to or in connection with the performance or non-perfor-
mance of the CONTRACT.
22.4 Except as provided by Clause 22.2(a) and Clause 22.2(b), the CONTRACTOR
shall save, indemnify, defend and hold harmless the COMPANY GROUP from
and against any claim of whatsoever nature arising from pollution occurring on
the premises of the CONTRACTOR GROUP or emanating from the property
and equipment of the CONTRACTOR GROUP (including but not limited to
marine vessels) arising from, relating to or in connection with the performance
or non-performance of the CONTRACT.
22.5 (a) Subject to Clause 22.5(b) below, the CONTRACTOR shall be responsible for
the recovery or removal and when appropriate the marking or lighting of any
wreck or debris arising from or relating to the performance of the WORK or the
property, equipment, vessels or any part thereof provided by the CONTRACTOR
GROUP in relation to the CONTRACT, when required by law, or governmental
authority, or where such wreck or debris is interfering with COMPANY opera-
tions or is a hazard to fshing or navigation and shall, except as provided for in
Clause 22.2 and Clause 22.3, save, indemnify, defend and hold harmless the
COMPANY GROUP in respect of all claims, liabilities, costs (including legal
costs), damages or expenses arising out of such wreck or debris, whether or not
the negligence or breach of duty (whether statutory or otherwise) of the COM-
PANY GROUP caused or contributed to such wreck or debris.
22.6 (b) Notwithstanding the provisions of Clause 22.1, where the COMPANY provides
transportation for the property of the CONTRACTOR GROUP to the ofshore
WORKSITE, and the COMPANY elects to, or is required by law or governmental
authority to recover or remove or mark or light any wreck or debris of such property,
the COMPANY shall, except as hereinafter provided, save, defend, indemnify and
hold harmless the CONTRACTOR GROUP from and against any claim of whatever
nature relating to the costs of such recovery, removal, marking or lighting. Provided,
however, that the foregoing indemnity and hold harmless shall not apply to the
extent that the recovery, removal, marking or lighting arises as a result of the neg-
ligence or breach of duty (statutory or otherwise) of the CONTRACTOR GROUP.
3.3.2 Pollution liability under FIDIC

No indemnity clause under FIDIC covers pollution. However, under Clause 4.18 in
the FIDIC model contract, it specifes that the contractor is required to take all
reasonable steps to protect the environment both on and of the site and to limit
damage and nuisance to people and property resulting from pollution, noise and any
pollution otherwise resulting from his operations. According to the FIDIC model
contract, contractor’s liability should only be limited to the values indicated under
the employer’s requirements. The respective clause is written as follows:
4.18 Protection of the Environment
The Contractor shall take all reasonable steps to protect the environment (both on and
of the Site) and to limit damage and nuisance to people and property resulting from
pollution, noise and other results of his operations.
286
The Contractor shall ensure that emissions, surface discharges and efuent from the
Contractor’s activities shall not exceed the values indicated in the Employer’s Require-
ments, and shall not exceed the values prescribed by applicable Laws.
3.3.3 Pollution liability under Operator A’s contract

Operator A drafted the indemnity clause regarding pollution as follows:
X.6.1 Pollution from the COMPANY’s Wells and Facilities
The CONTRACTOR shall indemnify and hold harmless the COMPANY from and
against such liability resulting from or for cost incurred or payments made by the COM-
PANY to control or clean up the pollutant or to prevent the threat of pollution or as
compensation for damage sufered by others from any pollution originated from the wells,
where applicable, owned or operated by the COMPANY arising from the CONTRAC-
TOR’s and/or SUB-CONTRACTOR’s performance of the SCOPE OF WORKS under
this CONTRACT howsoever caused.
X.6.2 Pollution from Equipment and/or Property and/or Vessel
The CONTRACTOR shall protect, indemnify and hold harmless the COMPANY
from and against all liability for pollution emanating from equipment and/or property
and/or vessel owned, leased, chartered or hired by the CONTRACTOR in connection
to the performance of the SCOPE OF WORKS and shall reimburse COMPANY for all
control and/or cleanup costs, and/or claims related to any such pollution. The CON-
TRACTOR undertakes that equipment or waste in any form originating from the SCOPE
OF WORKS shall not be dumped overboard.
A literal reading of the above indemnity clause suggests that it is unilateral and
onerous to the contractor. This is because the contractor is solely liable and required
to indemnify and hold harmless the operator for pollution whether emanating from
the operator’s well and facilities, or “from equipment and/or property and/or vessel
owned, leased, chartered or hired by the contractor in connection to the performance
of the scope of works”.
However, this is the example whereby the clause is unilateral in nature, but it
is in fact fair to the contractor, and reciprocity does not necessarily apply in this
case. In the above clause, the contractor indemnifes the operator for pollution at
the operator’s well and facilities which is “arising from the CONTRACTOR’s and/
or SUB-CONTRACTOR’s performance of the SCOPE OF WORKS”. Thus, there
is no reason for the operator to be liable for the damages, unless the operator has
contributed towards such pollution.
3.3.4 Pollution liability under Operator B’s contract

Operator B drafted the indemnity clause regarding pollution as follows:
X.1 CONTRACTOR agrees to indemnify, defend and save harmless COMPANY GROUP
from and against any and all claims, losses and expenses including without limitation
costs, demands, damages, suits, judgments, fnes, penalties, liabilities, reasonable attorneys’
fees and causes of action whatsoever nature or character, whether known or unknown,
and including without limitation claims, losses and expenses for . . . pollution . . . in any
way directly or indirectly, arising out of, or related to, the performance or subject matter
of this CONTRACT or the ingress, egress, or presence on any premises (whether land,
building, vehicle, platform, aircraft, vessel or otherwise) owned, operated, chartered,
leased, used, controlled or hired by COMPANY GROUP or CONTRACTOR GROUP,
and which are asserted by or arise in favor of CONTRACTOR GROUP and expressly
287
including any claims, losses or expenses actually or allegedly caused by the sole, active,
passive, concurrent or partial negligence (of whatever nature or character), fault or strict
liability of COMPANY GROUP or any other person or the unseaworthiness, unairwor-
thiness or defective condition of vessels, craft ore premises, whether or not preceding or
during the execution of this CONTRACT.
The above indemnity clause appears to be one-sided. It was drafted as widely as

possible to lessen the liability of the operator and to shift greater risk and liability
to the contractor including liability in relation to pollution. Under Clause X.1, the
contractor is required to indemnify and hold harmless the operator against any claims,
losses and expenses for pollution whether it directly or indirectly arose from the
contract irrespective of whether the act was caused by the sole, active, passive, con-
current or partial negligence, fault or strict liability of the operator.
3.3.5 Pollution liability under Operator C’s contract

Operator C did not specify any indemnity regarding pollution.
3.4 Insurance coverage

In Malaysia, there are no specifc types of insurance that are required by law in relation
to the provision of oil and gas services.215 Even though the maintenance of certain insur-
ance policies is often required by the contract, where the contractors will typically be
obligated to carry, among other policies, contractors’ all risks insurance covering general
liability to the public – which are generally dependent on the type of contracts.216 Franklin
notes that the recent trend with regard to Construction All Risk is that some operators
prefer to take out self-insurance rather than follow the traditional method of extending
its coverage to the contractors and subcontractors in order to cut costs for the operator.217
Therefore, the operator is always on the safer side. For example, the Production Sharing
Contracts usually require PETRONAS to be named as co-insured with the company in
insurance policies that have been taken out by the contractor in relation to the relevant
oil and gas services.218 Thus, it could be argued that without any law to monitor this
obligation, the contractors are exposed to the inherent risks, as it may not be covered by
the insurance industry where the risk is regarded as operator’s risk.
FIDIC presupposes that risk exists and should be covered by insurance, to the extent
that is possible to cover such risk.219 It is in this context that Clauses 17 (Indemnity)
and (Insurance) should be read together. However, insurance cover is only available
for insurable risk, which is a risk depending on fortuity, which means that the event
or circumstance has to be sudden and accidental and that comprehensive data for
the purposes of premium calculation exists. Obviously, some of the risk inherent
to a construction contract do not depend on fortuity – it is therefore not insurable.
This type of risk is also referred to as speculative risk. In principle, speculative risks
215 Hewitt (n19) 331.

216 Ibid.
218 Hewitt (n19) 366.
219 Jaeger and Hök, FIDIC: A Guide for Practitioners.
288
are foreseeable. An experienced contractor should be able to foresee potential risks.
The question is whether he should also make allowance for the event that the risk
occurs. Unforeseeability means the risk is not reasonably foreseeable by an experienced
contractor by the date for submission of the tender. The second paragraph of Sub-
clause 18.1 requires a meeting between the Employer and the Contractor in order
to agree the terms of insurance prior to the issue of the letter of acceptance. This is
a critical point, which should not be ignored. However, it is desirable that the terms
and issues of insurance are clearly identifed before the submission of the tender.
Under the LOGIC contract form, the insurance required under the provisions of
the insurance clause will vary depending on the scope of work for each particular
contract. The exact values of insurances required by the company must be specifed
in appendix 1. The clause also requires that subcontractors carry appropriate amounts
of insurance as may be relevant to their work. Although not provided in the standard
wording, the contractor is recommended to consider the need for a reciprocal com-
mitment from the company to also insure.220 In this respect, the main criteria will
of course be determined by the size and fnancial stability of the company in each
case. Hence the insurance provisions under FIDIC and LOGIC could be regarded
as models for ideal risk allocation between the parties.221
4 Conclusion
Based on the empirical study conducted in Malaysia discussed in this chapter, this
research argues that there is unequal bargaining power between operators and con-
tractors in the Malaysian oil and gas industry. This inequality of bargaining power
leads to imbalanced risk allocation and unfair indemnity hold harmless clauses in
oilfeld service contracts.
In Malaysia, the absence of a law to regulate imbalanced risk allocation and unfair
indemnity and hold harmless clauses in oilfeld service contracts should be perceived
as a serious problem because it leads to the problem of inequality of bargaining
power resulting from the dominant position of the operators over the contractors. To
date, there is no statutory restriction on contractual provisions purporting to exclude,
limit or indemnify one or both of the parties in relation to liability and indemnity.
Generally, contract law in Malaysia is governed by the Contract Act 1950. However,
section 77 of the Contract Act 1950 merely discusses the meaning of “contract of
indemnity”. Meanwhile, section 78 of the Contract Act 1950 merely talks about
“right of indemnity holder when sued”. Based on the case analysis in Malaysia, the
courts have not addressed this problem. Furthermore, the judicial position on the
doctrine of unconscionability and inequality of bargaining power remains unclear
in Malaysia since the Court of Appeal has taken inconsistent views in the cases of
220 ‘General Conditions of Contract (Including Guidance Notes) for Construction’ in Standard Contract
for U.K. Ofshore Oil and Gas Industry, vol October (2nd edn, LOGIC 2003) 2.
221 Wan M Zulhafz, ‘A Comparative Analysis on the Enforceability of Knock-for-Knock Indemnities
in Thailand and the United Kingdom’ (2017) 44 J.Malaysian Comp.Law (JMCL) 33.
289
Saad Marwi222 and American International Assurance Co Ltd.223 As a result of this
gap in the law which results in lack of legal protection to the contractor, the operator
might use his dominant position and continuously shift greater risk to the contrac-
tor. This uneven and therefore unfair risk allocation may cause signifcant fnancial
setbacks to the contractor.
This problem deserves attention from the Malaysian government. The problem
must be resolved. If left unresolved, it presents a threat to the commercial develop-
ment of the Malaysian oil and gas industry. Even though insurance provisions are
sometimes provided in the contract, there are no guidelines available to govern the
conduct of the parties. Additionally, the insurance requirements are not mandatory
for the parties. Some operators opt to take out self-insurance rather than to buy a
premium that will extend its coverage to cover the operators and subcontractors, for
example in respect of Construction All Risk. This scenario can cause the contractor
to assume uninsured risks, which could lead to detrimental fnancial exposure in the
occasion of a catastrophic incident.
The situation might get worse for contractors in the event that contractors have
to assume double jeopardy contractual risk, whereby the contractors not only need
to assume operators’ risk and also subcontractors’ risk. To solve this problem, it is
argued that a specifc legal mechanism should be adopted in Malaysia to protect and
limit the liability of the contractors under oilfeld service contracts. As discussed above,
sections 3 and 5 of the Civil Law Act 1956 provide for the application of English
law in Malaysia unless another provision has been or shall be written into law. The
English legal principles are therefore applicable and ought to be applied to cover the
gap in Malaysian law. However, the applicability of the English legal principles will
be subject to the circumstances in the Malaysian oil and gas industry.
222 Saad Marwi v Chan Hwan Hua & Anor [2001] 3 CLJ 98.
223 American International Assurance Co Ltd v Koh Yen Bee (f) [2002] 4 MLJ 301.
290
INDEX
Note: Page numbers in italic indicate a figure and page numbers in bold indicate a table on the
corresponding page.
accident costs 65–68, 85–86; deterrence versus 285–288; perceptions of the contractual
compensation 84–85; harm to environmental formation process 265–268; perception
interests 82–84; illustrating the operation of of risk allocation and indemnity clauses
knock for knock 69–71; industrial setting 69; 268–273; research design 262–263
litigation externalities 80–82; private ordering “caveat” clause (Germany) 196–197
71–73; risk interdependencies 80; social norms censoring: knock-for-knock regime 145–150;
71–73; social welfare 73–79 provisions limiting exposure 134–135
admiralty jurisdiction (United States) 216–217 choice-of-law 106–111; and Brazil 114–116;
Africa: Francophone African legal systems and United States 111–114
239–240; see also France and Francophone civil liability regime (Brazil) 205–209
civil law systems co-insured 148–149
AI endorsements 100–102 commercial contracts versus consumer
anti-indemnity statutes (United States) 113–114 contracts (Germany) 198
Australia: contribution to historical evolution of company groups 49–50
K4K clauses 13 compensation: deterrence versus 84–85
“consequential” loss: liability for 59–62
back-to-back indemnity: oil and gas contracts construction contracts see fabrication and
30–31, 32 construction contracts (Norway)
balanced regime 120–121 consumer contracts versus commercial
BGB (Germany) 200–201 contracts (Germany) 198
BIMCO SUPPLYTIME 2017 5–6 content test (Germany) 195–196
Brazil 204–205, 212–213; and choice-of-law contextualism 39–40
114–116; civil liability regime 205–209; Contract Act § 36 (Norway) 178–181
contribution to historical evolution of K4K contractor 141–143; gross negligence 138–141;
clauses 11–12; knock-for-knock clauses liability for breach 120–125
211–212; limitation of liability clause 209–211 contractual formation process (Malaysia)
breach: contractor’s liability for 120–125; of 265–268
statutory duty 40–46; validity of agreed contractual liability (Brazil) 205–206
limitations for 137–145 contra proferentem 38–39; and the problem
of negligence and breach of statutory duty
carve-outs: gross negligence 92–97 40–46
case law: Germany 202; see also English case law control 171–172
case studies (Malaysia) 261–262; analysis of courts: carve-outs for gross negligence 92–97;
274–289; claims by a third party 281–285; review of general terms and conditions
description of 263–274; findings 265; (Germany) 194–197
insurance coverage 288–289; liability
regarding personal injury or death of damage: to the contract object 156–158;
employees and loss of property of the parties provisions limiting exposure resulting from
276–281; liability regarding pollution 134–137; risks of 125–126
291
index
death of employees (Malaysia) 276–281 family zones 130–131
debate: negotiation of terms demands more than fault, nature of (French law) 248
debating (Germany) 196 fault-based liability (Brazil) 206–207
Deepwater Horizon: and choice-of-law federal court jurisdiction (United States)
111–113; contribution to historical evolution 215–217
of K4K clauses 10–11; see also litigation post federal maritime law (United States) 111–113
Deepwater Horizon FIDIC: personal injury or death of employees
default: provisions limiting exposure resulting and loss of property of the parties 277–278;
from 134–137 pollution liability 286–287; third-party
deterrence: compensation versus 84–85 claims 283–284
development of knock-for-knock clauses see France and Francophone civil law systems:
knock-for-knock clauses, development of Francophone African legal systems 240–241;
distribution of risk: normal presumptions of French Civil Code of 1804 239–240; French
35–36 law 240; indemnification of third-party losses
diversity jurisdiction (United States) 215–216 249; mutual hold harmless clauses 241–249
drafting: contra proferentem and the problem freedom of liability, indemnity and subrogation
of negligence and breach of statutory duty 169–170
40–46; definitional issues 49–50; “full and French Civil Code of 1804 239–240
primary” 48–49; multi-party issues 47–48; “full and primary” 48–49
words delimiting the circumstances in which
the indemnity and hold harmless provision general terms and conditions (Germany): are
will take effect 46–47 knock-for-knock clauses general terms and
conditions? 197; court review of 194–197
efficiency: basis for liability in contractual geographical extent 55
relationships 174–175; insurance coverage Germany: case law 202; contribution to
172–173; liability rules 173–174, 174 historical evolution of K4K clauses 12–13;
employees 50; personal injury or death of use of knock-for-knock clauses in 193–194,
(Malaysia) 276–281 202–203; validity issues 194–202
enforceability 3; United States 224–226 gross negligence: carve-outs for 92–97
English case law and legal context 107, “group concept, the” 167–169
109–111; evolution of K4K clauses in 7–10
entry into force 54–55 harm to environmental interests 82–84
environmental interests: harm to 82–84 health and safety regulation 77–78
exceptions 56; French law 246–247 hold harmless clauses: and indemnity 22–24;
exclusion of liability: French law 246–248; insurance coverage (Malaysia) 288–289;
Germany 199–202 interpreting 37–40; Malaysia 274–289; and
exposure see provisions limiting exposure mutual indemnity 24–25, 24; oil and gas
extension of benefits to groups 55 context 25–34; personal injury or death of
employees and loss of property of the parties
fabrication and construction contracts (Norway) (Malaysia) 276–281; pollution liability
117, 153–156; censoring the knock-for-knock (Malaysia) 285–289; statutory control
regime 145–150; contractor’s liability for of 34–35; third-party claims (Malaysia)
breach under offshore construction contracts 281–285; words delimiting circumstances
120–125;damage to the contract object 46–47; see also mutual hold harmless clauses
156–158; do NTK 15 provisions on exclusion (France and Francophone civil law systems)
and limitation of liability hold good 150–152;
knock-for-knock system 125–134; offshore implementing K4K clauses: issues in 3
project insurances 158–163; the problem indemnification 21–25; French law 249–250;
118–120; risk zones 156; setting aside indemnifiable losses (Brazil) 207–209;
provisions limiting exposure resulting from normal presumptions about the distribution of
default or damage 134–137; validity of agreed risk 35–36; Norway 154–155, 156; position
limitations of liability for breach of contract of third parties 36–37; statutory control of
137–145; waivers and indemnities shall apply indemnity and hold harmless clauses 34–35
regardless of cause 158; waivers of liability, indemnity 22–24, 98–99; anti-indemnity statutes
indemnification and waiver of subrogation 156 (United States) 113–114; apply regardless
fairness test (Germany) 198 of cause 158; back-to-back 30–31, 32; and
292
index
general liability insurance 99–100; industry regulation 175–181; waivers and indemnities
mutual hold harmless agreement (IMHH) shall apply regardless of cause 158; waivers
53–54; insurance coverage (Malaysia) of liability, indemnification and waiver of
288–289; interpreting 37–40; limitations in subrogation 156
insurance policies 103–104; Malaysia 255– intent, exclusion of liability for (Germany)
261, 268–274, 274–289; mutual 24–30, 24; 199–202
oil and gas context 25–34; personal injury or interpretation: contra proferentem and the
death of employees and loss of property of the problem of negligence and breach of statutory
parties (Malaysia) 276–281; pollution liability duty 40–46; definitional issues 49–50; “full
(Malaysia) 285–288; qualified 31–34; simple and primary” 48–49; indemnity and hold
25–26; statutory control of 34–35; third-party harmless clauses 37–40; multi-party issues
claims (Malaysia) 281–285; United States 47–48; words delimiting the circumstances
224–229; words delimiting circumstances in which the indemnity and hold harmless
46–47; see also indemnity (Norway) provision will take effect 46–47
indemnity (Norway) 117, 153–156; censoring invalid terms (Germany) 199–202
the knock-for-knock regime 145–150;
contractor’s liability for breach under offshore jurisdiction see Australia; Brazil; France and
construction contracts 120–125; damage to Francophone civil las systems; Germany;
the contract object 156–158; do NTK 15 Malaysia; Nordic countries; Norway; United
provisions on exclusion and limitation of States
liability hold good 150–152; freedom of
169–170; knock-for-knock system 125–134; knock-for-knock clauses, development of
offshore project insurances 158–163; the 1–3; Australia 13; Brazil 11–12; conclusive
problem 118–120; risk zones 156; setting critique 17–19; Germany 12–13; issues in
aside provisions limiting exposure resulting using, implementing and enforcing 3; modern
from default or damage 134–137; validity of use 5–6; Nordic countries 13–17; offshore
agreed limitations of liability for breach of oil and gas sector 6–7; shipping sector 6–7;
contract 137–145; waivers and indemnities standard contracts and English case law
shall apply regardless of cause 158; waivers 7–10; US jurisprudence 10–11; various uses
of liability, indemnification and waiver of 3–5
subrogation 156 knock-for-knock clauses, effect of choice-of-law
“individually negotiated” (Germany) 196–197 on 106–111; and Brazil 114–116; and United
industrial setting 69 States 111–114
industry mutual hold harmless agreement knock-for-knock clauses, implementing: issues
(IMHH) 52–59 in 3
industry-negotiated model contracts (Norway) knock-for-knock clauses, regulation of 87–88,
188–189 97; carve-outs for gross negligence 92–97;
insurance 4; co-insured 148–149; global literature 88–89; Norway 125–128; when
view of 128; incorporation of indemnity useful 89–92
limitations in insurance policies 103–104; knock-for-knock clauses, using 3–5: BIMCO
insured perils 160–162; Malaysia 288–289; SUPPLYTIME 2017 5–6; insurance 4; issues
“other insurance” clauses 102–103; perfect in 3; offshore contracts 4–5; P&I clubs 4;
pooling equilibrium 78–79; and regulation SUPPLYTIME 2005 3–4; when clauses are
127–128; United States 227–229; see also useful 89–92
insurance (Norway); liability insurance
insurance (Norway) 131–134, 153–156, landward areas 57
164–165; content and structure of the law of obligations (French law) 239; exceptions
knock-for-knock principle 166–171; damage linked to 246–247
to the contract object 156–158; freedom of liability: Brazil 205–209;“consequential” loss
liability, indemnity and subrogation 169–170; 59–62; exclusion of (Germany) 199–202;
“the group concept” 167–169; insurance law French 239, 241–244, 246; Malaysia
166; insurance regulation 170–171; offshore 276–281, 285–288; see also liability
project insurances 158–163; rationale for (Norway); liability insurance; limitation
the knock-for-knock principle 171–175; risk of liability
zones 156; tort law 165–166; type of loss and liability (Norway) 164–165; censoring the
basis for liability 166–167; validity of the knock-for-knock regime 145–150; content
293
index
and structure of the knock-for-knock parties 276–281; liability regarding pollution
principle 166–171; contractor’s liability for 285–288; perception of the contractual
breach under offshore construction contracts formation process 265–268; perception
120–125; freedom of liability, indemnity and of risk allocation and indemnity clauses
subrogation 169–170; “the group concept” 268–274; research design 262–263
167–169; insurance 131–134; insurance maritime law, federal (United States)
law 166; insurance regulation 170–171; the 111–113, 219–223; indemnity in 224–229;
problem 118–120; provisions on 128–131; supplemented with state law 223; tension
rationale for the knock-for-knock principle between OCSLA and 223–224
171–175;and regulation 125–128; setting market practice (Germany) 198
aside provisions limiting exposure resulting multiple parties 50–59, 51; drafting and
from default or damage 134–137; statutory interpretation 47–48
regulation of risk 182–187; tort law 165–166; mutual hold harmless clauses (France and
type of loss and basis for liability 166–167; Francophone civil law systems): and
unlimited no-fault 184–185; validity of the English terminology 241–242; Francophone
regulation 175–181 African legal systems 240–241; French Civil
liability insurance 98–100; incorporation of Code of 1804 239–240; French law 240;
indemnity limitations in insurance policies indemnification of third-party losses 249; legal
103–104; “other insurance” clauses 102–103 nature of 242–246; legal regime governing
licencee: channelling liability to (Norway) 246–250; mutual hold harmless clauses
186–187 241–249; origins and growing prevalence 241
limitation of liability 62–64; Brazil 209–211; mutual indemnity 24–30, 24
Norway 124–125, 137–145, 150–152
literature 88–89 nature of the fault (French law) 248
litigation externalities 80–82 nature of the loss (French law) 248–249
litigation post Deepwater Horizon (United negligence: carve-outs for gross negligence
States) 235–237; applicable law 214–224; 92–97; and contra proferentem 40–46
federal court jurisdiction 215–217; indemnity negotiation: “individually negotiated”
in maritime 224–229; maritime law 219–224; (Germany) 196–197; negotiation of terms
Outer Continental Shelf Lands Act 217–219, demands more than debating (Germany) 196
223–224; state law, application of 229–234; NL 5–1–2 (Norway) 176–178
state waters 219 no-fault liability (Norway) 184–185
LOGIC: personal injury or death of employees non-contractual liability (French law) 246
and loss of property of the parties 276–277; Nordic countries: contribution to historical
pollution liability 285–286; third-party evolution of K4K clauses 13–15;
claims 282–283 see also Norway
Longshore and Harbor Workers’ Compensation Norway see fabrication and construction
Act (United States) 226–227 contracts (Norway); indemnity (Norway);
loss: indemnifiable (Brazil) 207–209; nature of insurance (Norway); liability (Norway)
(French law) 248–249; prevention of Norwegian Petroleum Act 182–187, 192
173–175, 174; property of the parties NTK 15 150–152
(Malaysia) 276–281; third-party (French law)
249–250; type of 166–167 obligations, law of see law of obligations
Louisiana Oilfield Indemnity Act (United (French law)
States) 113–114, 229–233 obligation to notify 123–124
offshore sector 4–5, 164–165; content and
Malaysia 289–290; analysis of indemnity structure of the knock-for-knock principle
and hold harmless clauses 274–289; case 166–171; freedom of liability, indemnity and
studies 263–274; claims by a third party subrogation 169–170; “the group concept”
281–285; empirical study of oilfield service 167–169; historical encounter of the K4K
contracts 261–262; findings 265; general clauses in 6–7; insurance law 166; insurance
overview of legal system 251–252; insurance regulation 170–171; rationale for the
coverage 288–289; law of contract 252–261; knock-for-knock principle 171–175; tort law
liability regarding personal injury or death 165–166; type of loss and basis for liability
of employees and loss of property of the 166–167; validity of the regulation 175–181;
294
index
see also fabrication and construction normal presumptions about the distribution
contracts (Norway) of risk 35–36; Norway 187–192; overall
operation of knock for knock: illustration of limitation of liability 62–64; position of third
69–71 parties 36–37; statutory control of indemnity
operators 57 and hold harmless clauses 34–35; words
order of precedence 55 delimiting the circumstances in which the
“organized financing of loss” 147–148 indemnity and hold harmless provision will
“other insurance” clauses 102–103 take effect 46–47
Outer Continental Shelf Lands Act (United
States) 217–219; tension between maritime sanctions against breach 121–122
law and 223–224 secondary obligations: breach of 123
self-regulation 75–76
P&I clubs 4 shipping sector: historical encounter of the K4K
perfect pooling equilibrium 74–79 clauses in 6–7
perfect separating equilibrium 74 simple indemnity: oil and gas contracts
perils of knock for knock 79–85 25–26
personal injury or death of employees social norms 71–73
(Malaysia) 276–281 social welfare: effect of knock for knock on
personnel 50; see also employees 73–79
Piper Alpha litigation 9–10 standard contracts: evolution of K4K clauses
pollution: Malaysia 285–288; Norway 184–187 in 7–10
private ordering 71–73 state law (United States): Louisiana Oilfield
property 50; loss of the property of the parties Indemnity Act 113–114, 229–233; and
(Malaysia) 276–281 maritime law incomplete 223; Texas Oilfield
provisions limiting exposure 134–137 Anti-Indemnity Act 114, 233–234
state waters (United States) 219
qualified indemnity 31–34 statutory duty see breach of statutory duty
strict liability (Brazil) 206–207
“reasonable” arrangements (Norway) 149–150 subrogation: freedom of 169–170; see also
reasonableness test (Germany) 197–199 waiver of subrogation
recovery, basis for 162–163 SUPPLYTIME 2005 3–4
regulation: insurance 170–171; liability
136–137, 182–187; see also knock-for-knock terms and conditions see general terms and
clauses, regulation of conditions; invalid terms
repeat interactions 76–77 Texas Oilfield Anti-Indemnity Act (United
research design (Malaysia) 262–263 States) 114, 233–234
right to defend 56 third parties 36–37; French law 243–246,
risk: of damage 125–126; global view of 128; 249–250; Germany 201; Malaysia 281–285;
interdependencies 80; need for contractual Norway 131
control of 171–172; risk insurances 155–156; tort defaults 65–68, 85–86; deterrence versus
“risky” terms (Germany) 198–199; risk compensation 84–85; harm to environmental
zones 156; statutory regulation of (Norway) interests 82–84; illustrating the operation of
182–187; see also risk allocation knock for knock 69–71; industrial setting
risk allocation 20–21; contra proferentem 69; litigation externalities 80–82; private
and the problem of negligence and breach ordering 71–73; risk interdependencies 80;
of statutory duty 40–46; definitional social norms 71–73; social welfare
issues 49–50; French law 243, 244; “full 73–79;
and primary” 48–49; indemnification and tort law (Norway) 165–166
related concepts 21–25; indemnity and hold tort liability (Brazil) 205–206
harmless provisions in the oil and gas context transportation by air excluded
25–34; interpreting indemnity and hold 56–57
harmless clauses 37–40; known problems of type of loss 166–167
drafting and interpretation 40–50; liability
for “consequential” loss 59–62; Malaysia UKCS: normal presumptions about the distribu-
268–274; multiple parties 47–48, 50–59, 51; tion of risk 35–36; position of third parties
295
index
36–37; statutory control of indemnity and vessel services, contracts for (Norway)
hold harmless clauses 34–35 164–165; content and structure of the knock-
United States 235–237; applicable law for-knock principle 166–171; freedom of
214–224; and choice-of-law 111–114; liability, indemnity and subrogation 169–170;
contribution to historical evolution of K4K “the group concept” 167–169; insurance law
clauses 10–11; federal court jurisdiction 166; insurance regulation 170–171; rationale
215–217; indemnity in maritime 224–229; for the knock-for-knock principle 171–175;
maritime law 219–224; Outer Continental tort law 165–166; type of loss and basis for
Shelf Lands Act 217–219, 223–224; state liability 166–167; validity of the regulation
law, application of 229–234; state waters 219 175–181
unlimited no-fault liability (Norway) 184–185
using knock-for-knock clauses 3–5: BIMCO waiver of liability (Norway) 156
SUPPLYTIME 2017 5–6; insurance 4; issues waiver of recourse (French law) 246–248;
in 3; offshore contracts 4–5; P&I clubs 4; effects of 247–248
SUPPLYTIME 2005 3–4; when clauses are waiver of rights of subrogation 55–56
useful 89–92 waiver of subrogation 104–105; maintenance of
148–150; Norway 156
validity: agreed limitations of liability for waivers apply regardless of cause
breach of contract 137–145; French law (Norway) 158
246–247, 249; Germany 194–202; Norway “willful misconduct” 143–145
175–181 words delimiting circumstances: indemnity and
variation mechanism 123–124 hold harmless 46–47
296

Knock For Knock Indemnities and The Law Contractual Limitation and

Uploaded by

Copyright:

Available Formats

You might also like

Knock For Knock Indemnities and The Law Contractual Limitation and

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Knock For Knock Indemnities and The Law Contractual Limitation and

Uploaded by

Copyright:

Available Formats

KNOC K-F OR-KNOC K IN D EMN ITIES

Double Insurance and Contribution Comparative Analysis of Interim Measures:

For more information about this series, please visit:

C ON T RAC T UAL LI MI TATI O N

First published by Informa Law from 2023

and by Informa Law from Routledge

British Library Cataloguing-in-Publication Data

ISBN: 978-1-032-07408-5 (hbk)

Typeset in Times New Roman

List of contributors xix

PART I SELECTED TOPICS

CHAPTER 2 AN INTRODUCTION TO RISK ALLOCATION IN

CHAPTER 3 CONTRACTING AROUND TORT DEFAULTS:

CHAPTER 4 ON KNOCK-FOR-KNOCK CLAUSES AND

CHAPTER 5 KNOCK-FOR-KNOCK INDEMNITY PROVISIONS

CHAPTER 6 THE EFFECT OF CHOICE OF LAW ON

CHAPTER 7A LIMITING AND CHANNELING LIABILITY UNDER

CHAPTER 7B THE OBLIGATION UNDER NORWEGIAN

CHAPTER 8 LIABILITY AND INSURANCE CLAUSES IN

CHAPTER 9 STATUTORY LIABILITY REGULATION

CHAPTER 10 APPLYING KNOCK-FOR-KNOCK IN GERMANY 193

CHAPTER 11 KNOCK FOR KNOCK UNDER BRAZILIAN LAW 204

CHAPTER 12 THE IMPLICATIONS OF LITIGATION POST

CHAPTER 13 MUTUAL HOLD HARMLESS CLAUSES IN

List of contributors xix

PART 1 SELECTED TOPICS

CHAPTER 2 AN INTRODUCTION TO RISK ALLOCATION

CHAPTER 3 CONTRACTING AROUND TORT DEFAULTS:

CHAPTER 5 KNOCK-FOR-KNOCK INDEMNITY PROVISIONS

CHAPTER 6 THE EFFECT OF CHOICE OF LAW ON

CHAPTER 7A LIMITING AND CHANNELING LIABILITY

CHAPTER 7B THE OBLIGATION UNDER NORWEGIAN

CHAPTER 8 LIABILITY AND INSURANCE CLAUSES IN

CHAPTER 9 STATUTORY LIABILITY REGULATION

CHAPTER 10 APPLYING KNOCK-FOR-KNOCK IN GERMANY 193

CHAPTER 11 KNOCK FOR KNOCK UNDER BRAZILIAN LAW 204

CHAPTER 12 THE IMPLICATIONS OF LITIGATION POST

CHAPTER 13 MUTUAL HOLD HARMLESS CLAUSES IN

CHAPTER 14 LESSONS FROM THE APPLICATION OF

The knock-for-knock indemnity: concepts, efects and questions

Kristofer Svendsen, Endre Stavang and Greg Gordon

11 E. E. Caledonia v Orbit Valve [1995] 1 All ER 174.

Brazil Civil Code (Napoleonic Code)

The development of knock-for-knock

1.2 Essential meaning of a K4K clause

3.1 SUPPLYTIME 2005

3.2 Insurance and P&I clubs

3.3 Use in ofshore contracts

4.1 The BIMCO SUPPLYTIME 2017

5 A historical encounter of the K4K clauses in the shipping

6.1 The signifcance of the Piper Alpha litigation

36 A. Iyer, “Excluding ‘Consequential Loss’ in Ofshore Contracts”, 23 October 2020, www.iylegal.com/

7 The contribution of other jurisprudence to the historical

7.2 Deepwater Horizon

7.7 Later judicial trends

7.8 The way forward

71 C. Kumarasinghe, “Case Update: Transocean v Providence”, May 2016, HFW, www.hfw.com/Trans-

An introduction to risk allocation in oil and