Study on the influence of Plant Lifetime Extension (PLEX) on nuclear liability/010. For. GreenPeace Nordic. 11 December 2013

Transcription

1 Study on the influence of Plant Lifetime Extension (PLEX) on nuclear liability/010 Prof. Dr. Michael G. Faure LL.M. and Dr. Tom Vanden Borre γ For GreenPeace Nordic 11 December 2013 Maastricht University and Erasmus University Rotterdam. γ Institute for Energy and Environmental Law, Catholic University of Leuven.

2 TABLE OF CONTENT 1. Introduction The international nuclear liability conventions Origins of the international nuclear liability conventions The Preliminary Report The Harvard Report Overview of the international conventions on nuclear liability Scope of the international nuclear liability conventions Scope ratione materiae Territorial scope Principles underlying the international liability conventions of the first generation Strict liability Legal channelling of liability to the nuclear operator The principle of legal channelling Exceptions to the legal channelling Right of recourse for compensation paid to third parties Right of recourse for on-site property damage Limitation of liability Compulsory insurance Exclusive jurisdiction obligation of having one single forum Public funding Changes after Chernobyl Specific provisions with regard to ageing of a nuclear power plant The applicability of the conventions to ageing Legal channelling and ageing Global system? Limited adherence to nuclear liability Conventions EU Member States that are not party to any of the nuclear liability Conventions Consequences

3 3. Comparison of current limits and amounts available Overview of amounts available in EU Member States US nuclear liability system The Price-Anderson Act and its evolutions over the years The 1957 regime : a shift from public to private funding regime and beyond: further increase The NRC property rule : Ratification of the Convention on Supplementary Compensation Conclusions Amounts of damage caused by a nuclear accident Economic consequences Distortions created by the subsidy Artificial competitiveness Prevention of nuclear accidents Reduced victim compensation Insurance of the nuclear risk The nuclear European insurance pools Third party liability insurance in the US The nuclear mutual insurance systems Limits of the pools Alternatives: pooling by operators Cost of insurance Financial security The need for a new compensation model The basic principles of a new compensation model Alternative financial schemes Advantages of risk sharing Risk-pooling for marine oil pollution Toward an EU nuclear risk pooling scheme?

4 5.2.4 An EU Price-Anderson Act: retrospective pooling? Munich Re proposal Compensation for oil spills Proposal for nuclear liability Pooling as a solution to generate higher amounts of cover? Conclusions List of Literature

5 1. Introduction 1. Nuclear power largely developed after the Second World War. It was especially in the 1950s that possibilities of the commercial development of nuclear energy were explored. It is important to have in mind that the development of nuclear power has been facilitated by the creation of various international nuclear liability conventions which were largely created on the request of the suppliers of nuclear material. Indeed, since nuclear technology at that time was to an important extent an American monopoly, American suppliers feared liability if they would deliver material, known-how and services to European operators and an accident would follow. Nuclear liability conventions were hence created that exclusively channelled liability to the operator of the power plant (thus excluding the liability of the supplier) and substantially limited the liability of the licensee of the power plant (both in amount and in time). Many of the principles of these conventions paved the way for a strong development of nuclear power, especially in Europe, but also in other continents. 2. Although there may be technical differences between the types of reactors and installations, many of the nuclear installations that were constructed in the 1960s and 70s 1 now are in their forties or fifties and some of those hence reach the time limit that was originally foreseen for their exploitation. Some countries like Finland, the UK and France have chosen not to continue the operation of these older power plants and have decided to build new nuclear power plants. 3. Other countries opt more for the extension of the lifetime of several nuclear power plants. Several nuclear power reactors have already been granted lifetime extension. This means that after the nominal design lifetime, varying between 25 to 40 years, licenses of these plants have been renewed for, typically, a 10 or 20 years period. Consequently, the lifetime of such a renewed reactor can be up to 60 years. The decision as to whether or not grant lifetime extension lies in the first place with the nuclear safety authorities of the country where the nuclear power plant is located. In the United States, over 70 nuclear power plants have been granted license renewal; also Russia is granting lifetime extensions. design. 1 Worldwide, the most commonly used type of reactor is the Pressurised Water Reactor (PWR); these are indeed US 5

6 4. Generally speaking a lifetime extension application will require investments in the safety of the power plant: replacing worn parts, upgrading of various systems etc. In Belgium, where a political decision has been taken with regard to granting lifetime extension to Tihange 1, the Belgian nuclear safety authority (FANC-AFCN) has indicated that in assessing the lifetime extension, particular attention will be on the ageing and the design. 2 With regard to ageing, the Agency refers to a program for managing older installations in line with international standards. In this respect, special focus will lie on those parts of the reactor that cannot be replaced like the reactor vessel. A monitoring of these parts will be required in order to be able to detect possible failures in an early stage. With regard to design, the Agency refers to a program for modernizing or upgrading of the installation based on a safety assessment of the oldest units. The mere fact that several nuclear power plant operators ask for lifetime extension indicates that there are sufficient economic reasons to choose lifetime extension rather than building a new power plant. Part of this will most likely also relate to difficulties in terms of public acceptance for constructing new nuclear plants. 5. It is precisely this nuclear Plant Lifetime Extension (abbreviated as PLEX) that causes major worries to stakeholders such as Greenpeace. As a result, Greenpeace has launched a request for research into the consequences of this nuclear Plant Lifetime Extension. Obviously, an important part of the research investigates the risks involved with nuclear aging, as well as the economics of nuclear ageing and the extent to which public participation should be made possible when decisions are made on nuclear aging. Those aspects are crucially important and will be addressed in other parts of the project that has been commissioned by Greenpeace. 6. However, one particular question is the relationship between plant lifetime extension and nuclear liability. Nuclear liability will be the particular focus of this study. In this study, we hence do not examine the question whether the extension as such increases the probability or the potential damage in case of a nuclear accident. Those aspects are dealt with in other parts of this project. However, one of the questions that obviously arise, also with regard to nuclear plant lifetime extension, is whether the nuclear operators will be liable for the costs of an accident in case of such a lifetime extension. To an important extent, so we will argue in this report, this question cannot be answered by merely focusing on plant lifetime extension, 2 FANC en BEL V publiceren technische nota verlenging levensduur kerncentrale, website FANC, 6

7 but has to be addressed by critically examining the entire legal liability framework for nuclear accidents. 7. Indeed, the international regime dealing with nuclear liability has to an important extent been seriously criticized in legal and economic literature, precisely for the already above mentioned characteristics, i.e. the exclusive channelling of liability to the licensee of the nuclear power plant (thus excluding liability of all others who could equally influence the accident risk) and for providing very low limits on the liability of the licensee of the power plant. We will hence argue that these features already today constitute an important problem as far as providing effective compensation to victims is concerned, but potentially also as far as providing incentives for the prevention of nuclear accidents is concerned. These problems are hence not typical for PLEX; however, to the extent that PLEX leads to a potential increase in the probability of a nuclear accident or of the potential damage, the mentioned negative features of the international nuclear liability regime would only be exacerbated. Given that the other studies in this project do indeed provide reasons to argue that nuclear ageing could lead to increased risks, PLEX would hence be all the more an important reason to be very cautious in deciding which liability rules should apply in case of a nuclear accident in a power plant which has been awarded a lifetime extension. Which rules and principles should govern the liability of the nuclear operator and of those who have participated in the extension (e.g. suppliers of certain parts or companies having made safety calculations etc.)? 8. Hence, our study will to an important extent focus on the current international legal framework and its (in)adequacy in providing adequate compensation and prevention, arguing that those problems might potentially only become worse in case of PLEX and thus call for a reform of the regime. We will therefore not only critically review the current regime based on the existing literature, but we will equally look at proposals to reform nuclear liability and critically evaluate those. 9. It is in that respect interesting to mention that the criticism on the nuclear liability regimes is definitely not new. Faure and Van den Bergh already argued in 1990 that the international legal framework for nuclear accidents in fact only provided protection for the nuclear industry and could hence only be explained as the result of effective interest group 7

8 politics. 3 In 1997, Trebilcock and Winter provided an economic analysis of nuclear accident law and were equally critical on the channelling and limitation of liability. 4 Also American studies had already pointed at the fact that the nuclear liability regime (even the one in the US which is in our view has some interesting features and provides for higher compensation to victims than the international regime) in fact constitutes an implicit subsidy to the nuclear industry. This point was especially made by Dubin and Rothwell in Faure and Skogh also criticized (in 1992) the current international legal framework and argued in favour of an international pooling regime which would be able to generate substantially higher amounts. 6 Also Vanden Borre provided in his dissertation (of 2001) a critical analysis of the international nuclear liability regime, calling for an urgent need for reform It is in fact striking that although those critical voices with respect to the international nuclear liability framework were already formulated since the early 1990 s, these voices did not lead to a fundamental rethinking of this framework. As we will argue below, a revision of the international conventions did indeed take place after the Chernobyl accident, but first these have, in fact, not entered into force yet. Moreover, notwithstanding suggestions for a fundamental reform, 8 the amounts were increased but are still largely insufficient to cover the costs of an average, let alone a serious, nuclear accident. 11. This is all the more striking since in related domains (in the sense that it also concerns catastrophic accidents) such as e.g. marine oil pollution, evolutions equally took place, but these led to the generation of substantially higher amounts for damage caused by vessel induced marine oil pollution. It is, just to provide an illustrative example, striking that today for a country like France, only a total amount of compensation would be available in a case of a nuclear accident of 381 million euro of which, moreover, only 91 million would be paid by the operator and the remainder through a subsidy provided by the installation state (in that example France) and all contracting states. 9 Today the compensation available in case of marine oil pollution (the type of Prestige or Erika disasters) amounts almost to one billion euro, which is moreover entirely financed by industry (the shipping industry and the oil 3 Faure and Van den Bergh (1990). 4 Trebilcock and Winter (1997). 5 Dubin and Rothwell (1990). 6 Faure and Skogh (1992). 7 Vanden Borre (2001). 8 See inter alia Faure (1995). 9 See Faure and Fiore (2008a),

9 industry), without state intervention. 10 The amending protocols of the Paris and Brussels conventions of 2004, as well as the Protocol to the Vienna Convention and the Convention on Supplementary Compensation do continue to provide for state intervention in case of a nuclear accident. Even the highest amount under these Conventions, 1.5 billion euro, 11 is clearly insufficient to deal with the damage resulting from a large nuclear incident. 12. The comparison with marine oil pollution is interesting since in the latter domain, there has been a strong involvement of the European Commission after the Prestige and Erika incidents. It was in fact under the threat of the creation of a separate European compensation fund that the parties at the International Maritime Organisation (IMO) agreed to increase the amounts available via the International Oil Pollution Compensation Fund. They feared indeed that the EU may create a separate fund to avoid that the parties within the IMO agreed that the amounts would be increased to the current levels. 12 In the nuclear domain, until recently, European activism has been largely absent as far as the liability regime is concerned. Europe to a large extent just required that the Member States joined one of the international conventions. Over the past three years several EU Member States already have increased the liability of the nuclear operator to the levels required by the revised Paris and Brussels Supplementary Convention ( 700 million or 1.2 billion) without these revised conventions being in force. At the same time the European Commission seems to become more active with regard to nuclear liability. Even before the Fukushima accident of March 2011, the European Commission had created a working group in order to analyse possible revisions in the nuclear liability regime and the role that Europe could play in that respect. Fukushima of course put nuclear liability high on the European agenda. Interesting in that respect is that now voices are also heard especially from international reinsurers that, provided the nuclear industry is willing to participate, substantially larger amounts could be generated for compensating victims of nuclear accidents than the amounts currently provided for in the nuclear liability conventions. 13. This shows that now, at least at a European level, the nuclear liability framework is in full evolution. It is therefore the right moment to critically analyse the nuclear liability regime 10 See Wang (2007) and see below We will discuss this in detail in section Wang (2007). 9

10 and point at potential shortcomings and reform possibilities. Nuclear ageing and PLEX make such an analysis only more urgent. 14. In this study, we will analyse the possible impact of nuclear plant lifetime extension (PLEX) on nuclear liability. We will address the following questions: Does the current legal framework on nuclear liability address nuclear ageing? Would it be a good idea to have a specific provision addressing nuclear ageing? What is the liability of suppliers of upgrades of such a lifetime extension? 15. At the end of this introduction, we should stress that we merely focus on nuclear liability in land-based nuclear installations. The prevention of nuclear accidents is, however, only to some extent dependent upon the scope of liability provisions. Safety regulation and standardization by industry may play a far more important role. Even though there are important evolutions in those domains as well, 13 analysing the regulatory framework aiming at prevention is beyond the scope of this study and will therefore not further be addressed. Also nuclear liability issues related to the transport of nuclear substances is left outside the scope of this study. 16. In the remaining of this study, we will focus on the following issues: first, we sketch the nuclear liability provisions, in general as they follow from the international legal framework, addressing specifically the potential consequences of ageing of nuclear installations (2). Then, we go in more detail on the amounts currently available for compensation and the economic consequences of the limitation of liability (3). An important problem also constitutes the fact that the current nuclear insurance regime also has important shortcomings. It is important to address those since the limited insurability of the nuclear risk has often been proposed as a justification for the liability limits (4). We then argue that it is possible to develop a new compensation model which would create higher amounts of compensation for victims and thus a better risk exposure (providing incentives for prevention) to operators and other stakeholders (5). We conclude by arguing that the new model we propose could also better deal with specific questions related to ageing and to PLEX (6). 13 See e.g. the proposal for a council directive of 17 October 2013 amending Directive 2009/71/Euratom establishing a community framework for the nuclear safety of nuclear installations, COM(2013) 715 final. 10

11 2. The international nuclear liability conventions 17. In this section, we will explain the international nuclear liability regime. In order to understand the philosophy of that regime, it is important to recall its origins (2.1). Second, we will give an overview of the existing international nuclear liability conventions (2.2). After that, the scope and principles of these conventions will be discussed (2.3 and 2.4). We will highlight the changes after Chernobyl in 2.5. This analysis will allow us to assess any specific provisions applicable to ageing nuclear power plants, c.q. whether the conventions provide for specific provisions for claims following damage caused by ageing nuclear power plants (2.6). Finally we will answer the question whether the international conventions do provide for a global regime (2.7). 2.1 Origins of the international nuclear liability conventions 18. In the 1950s, many countries decided to use nuclear electricity to become part of their energy mix. At that time, only two countries were able to supply the necessary know-how and equipment: the United States and the former Soviet-Union. Western European countries wanting to have access to the necessary technology and know how, had therefore no choice than to rely on American knowledge and technology. For the U.S., the Western European countries clearly were an interesting and promising market for the expansion of their nuclear industry. 19. However, from the outset, the American nuclear industry feared liability for damage in Europe resulting from nuclear accidents in installations where they have supplied know-how, technology and equipment. In order to address this fear the first bilateral agreements between the U.S. and Europe contained a so-called hold-harmless clause; under such a clause, the (European) nuclear operator held the (American) supplier harmless for all claims resulting from his activities As business expanded, the American companies were unsure whether this holdharmless clause provided sufficient and adequate protection. The US Atomic Industrial Forum therefore conducted two studies on the possible liability claims of victims of a nuclear accident against U.S. suppliers. The first study was the Preliminary Report on 14 See, inter alia, R. Gautron (1967) at, 59, M. Isenbergh, (1967), at 279, A.W. Murphy (1962), at 166 and J.P. Piérard (1963), at

12 Financial Protection against Atomic Hazards (hereinafter called the Preliminary Report ) 15. The second study was entitled International Problems of Financial Protection against Nuclear Risk (hereinafter called the Harvard Report ) The Preliminary Report 21. According to the authors of the Preliminary Report, the major problem in assessing the potential liability in case of a nuclear accident was the lack of knowledge concerning the nuclear risk. A liability claim after a nuclear accident would be many times the amount available on the insurance market. 17 The insurance industry could not offer sufficient cover given the insufficient capacity on the insurance market and given the lack of data on the probability of a nuclear accident; therefore it was difficult to calculate the insurance premium. Thus, the possibilities for a nuclear operator to spread his risk were limited to the means available on the insurance market; consequently, for that part of the damage that was not covered by insurance, he could be held liable with his entire assets The Report stipulated that given the undisputed fact that there is a vital national interest in the development of atomic power 19 it was logical that also the authorities (government) had an important role to play. The government indeed needed not only to make sure that a potential liability claim did not hinder the development of the nuclear industry, but also needed to protect the safety and the security of the population. Therefore, it was concluded that a government program had to be created in order to meet two basic requirements: first, the protection of the industry against unknown liability claims and second, the need to protect the public against the damage they would suffer. 23. This protection of the industry was nothing else than a limitation of the liability. The authors of the Report added that a limitation of liability would only be acceptable if there was an additional compensation available, on top of the liability amount. 20 Any damage in excess of that amount should be covered by the government. The aim of this government intervention 15 Prepared by the experts of Columbia University and published in March Prepared by Harvard Law School and published in Preliminary Report, Preliminary Report, Preliminary Report, Preliminary Report,

13 was to prevent private companies from not investing in nuclear activities for fear of an enormous liability claim The Preliminary Report concluded that the interests of both the industry and the public could be met by limiting the amount of liability of the nuclear operator to the amount available on the insurance market and by providing for public funds for damage not covered for by the operator (or his insurer). 25. As we will discuss, the U.S. nuclear liability law (called the Price-Anderson Act) adopted in 1957 indeed limited nuclear liability and provided for government invervention on top of this limited amount. 22 The U.S. legislator thus clearly followed the conclusions of the authors of the Preliminary Report. The Price-Anderson Act, however, only covers liability for nuclear accidents in the U.S. It does not and cannot cover nuclear liability issues in Europe. This transboundary issue was analysed in the Harvard Report The Harvard Report 26. According to the Harvard report, the American industry could play an important role in the development of nuclear energy in the whole world; on the other hand, the American companies that would support the nuclear energy industry were exposed to considerable risks. 23 It was feared that victims of nuclear accidents in Europe would sue American companies in the U.S. The Report indicated that these companies could indeed be held liable for defects of the material/products delivered to the operator (product liability). The Harvard Report identified three reasons why these companies should not be held liable. 27. First, in case of a nuclear incident, the suppliers feared being prosecuted instead of or jointly with the nuclear operator, even if their role was limited to calculations or supervision of specific parts of the reactor, because the victim of a nuclear incident could be compelled to sue as many companies as possible. It would most certainly be lucrative for a West European victim to sue American suppliers in the US. This could, according to the Harvard Report, 21 R. Lowenstein, Indemnification against liability for nuclear accidents in licensed atomic energy activities, in J.L. Weinstein (ed.), Progress in Nuclear Energy Law and Administration, Pergamonn Press, vol. 3, 1962, Cfr. infra, section Harvard report, 5. 13

14 result in substantial legal problems, 24 and subsequently lead to lengthy trials of which a large part would be futile. 25 Secondly, according to the Harvard report, it would be unfair to hold the suppliers liable since, after the delivery of goods and services, they generally lose control thereover. 26 Moreover, the authors of the Harvard Report considered the nuclear operator to be in a better position than the different suppliers to control the risk. Thirdly, the nuclear operators were more capable of obtaining insurance. By concentrating liability on the operator, pyramidal insurance costs could be avoided. 27 As a result, not every supplier would have to buy a separate insurance coverage, limiting insurance costs. 28. The authors of the Harvard Report concluded that the best solution was to make all liability claims against atomic suppliers impossible by legislative intervention. To abolish every legal claim against the supplier would be the most simple and most effective solution ; the new social and economic circumstances inter alia resulting from the peaceful use of nuclear energy necessitated such a legislative intervention; these circumstances were more important than legal dogmatic objections. 28 The Report therefore proposed the introdcution of twelve measures which eventually were largely adopted by the nuclear liability conventions (channelling of liability, strict liability, limitation of the liability amount, limitation of the liability in time, the mandatory insurance of liability and the exclusive competence of the court of the country where the incident occurred, etc.). 29 Moreover, the draft text of what later became the 1960 Paris Convention on Third Party Liability in the Field of Nuclear Energy was annexed to the Harvard Report. Both the Harvard and the Preliminary Report have thus strongly influenced the compensation system enacted by the nuclear liability conventions. 2.2 Overview of the international conventions on nuclear liability 29. In the beginning of the development of private nuclear industry, two international regimes were created that regulate the civil liability for damage caused by nuclear accidents, i.e. that create compensation mechanisms for nuclear damage. 30 The first treaty regime has been established under the auspices of the OECD Nuclear Energy Agency (NEA) and consists 24 In particular, problems of an international private law nature: competent court, applicable law and enforceability of foreign judgments in the US. At the same time, there existed the risk that a victim would try to avoid the law of the installation State by forum-shopping. 25 Harvard report, Ibid., Ibid., 57; Norbert Pelzer (1994) at 9. Exactly the same reasoning is found in the Exposé des Motifs, Motif Harvard report, Cfr. infra, section For a useful overview, see, inter alia, Liability and Compensation for Nuclear Damage: An International Overview (OECD: 1994). 14

15 of the Convention on Third Party Liability in the Field of Nuclear Energy of 29 July 1960 (hereinafter referred to as the Paris Convention ) and the Brussels Supplementary Convention to the Paris Convention on Third Party Liability in the Field of Nuclear Energy of 31 January 1963 (hereinafter referred to as the Brussels Supplementary Convention ). The Paris Convention introduces the five major principles of international nuclear liability conventions. The Brussels Supplementary Convention provides for additional compensation of damage in case the coverage of the operator under the Paris Convention is inadequate or insufficient. 30. The second nuclear liability treaty regime has been developed under the aegis of the International Atomic Energy Agency (IAEA): the Vienna Convention on Civil Liability for Nuclear Damage of 21 May 1963 (hereinafter referred to as the Vienna Convention ). The Vienna Convention introduces, as does the Paris Convention, the five major principles of international nuclear liability law. 31. The NEA treaty regime is regionally confined (i.e. Western Europe, Slovenia and Turkey), while the IAEA treaty regime is worldwide in scope. 32. There have been no significant changes to the different nuclear liability treaties until after the Chernobyl accident. 31 The accident triggered a revision process for both the NEA and the IAEA regime, resulting in the adoption, on three different dates, of several new international conventions 32. On 21 September 1988 the Joint Protocol Relating to the Application of the Vienna Convention and the Paris Convention was agreed upon. On 12 September 1997, two new legal instruments were opened for signature in the IAEA regime: the Protocol to Amend the 1963 Vienna Convention on Civil Liability for Nuclear Damage (hereinafter called: the Protocol to the Vienna Convention ) and the Convention on Supplementary Compensation for Nuclear Damage (hereinafter called: the Convention on Supplementary Compensation, of CSC). Whereas the Protocol to the Vienna Convention quite logically modifies the Vienna Convention, the CSC is an independent stand-alone 31 Both the Paris and Brussels Supplementary Conventions have been supplemented by a few additional protocols. The Protocol to the Paris Convention of 16 November 1982 adjusts some of the definitions and imposes liability on the operator for damage to the means of transport; the Protocol to the Paris Convention of 16 November 1982 changes the unit of account into SDR and increases the liability amounts of the three tiers from the initial 120 million up to 300 million SDRs. Although the basic text is always that of the Paris Convention, when referred to the latter it will include the additional protocols as well. In this respect, see, inter alia, M. Lagorce (1993), at After the Chernobyl accident several conventions concerning nuclear safety issues have been adopted as well. In this paper, we only discuss the international conventions dealing with nuclear liability and compensation. 15

16 convention, since any country can join the Convention without having to be a Party to the Paris or Vienna Convention. 33. Finally, the revision process of the NEA-regime resulted in the opening for signature of two new instruments on 12 February 2004: the Protocol to amend the Convention on Third Party Liability in the Field of Nuclear Energy of 29 July 1960 (hereinafter referred to as the Protocol to the Paris Convention ) and the Protocol to amend the Convention of 31 January 1963 supplementary to the Convention of 29 July 1960 on Third Party Liability in the Field of Nuclear Energy (hereinafter called the Protocol to the Brussels Supplementary Convention ). 34. In order to be able to make the distinction between the different treaties, it is useful to distinguish the nuclear international liability regime of the first generation on the one hand (being the Paris Convention, the Brussels Supplementary Convention and the Vienna Convention) and the nuclear international liability regime of the second generation on the other hand (all the other Conventions mentioned above). Whereas the nuclear liability conventions of the first generation consisted of three conventions, the nuclear liability conventions of the second generation added five more treaties. Thus, as a result of the revision process, the international nuclear compensation system consists of no less than eight international conventions. The table hereunder gives an overview of the different conventions. Table 1: Overview of the international nuclear liability conventions NEA-regime IAEA-regime Overview of the international nuclear liability conventions First generation Second generation Paris Convention on Third Party Protocol to amend the Convention on Third Liability in the Field of Nuclear Party Liability in the Field of Nuclear Energy of 29 July 1960 Energy of 29 July 1960 of 12 February 2004 Brussels Supplementary Convention Protocol to amend the Convention of 31 to the Paris Convention on Third Party January 1963 supplementary to the Liability in the Field of Nuclear Convention of 29 July 1960 on Third Party Energy of 31 January 1963 Liability in the Field of Nuclear Energy of Vienna Convention on Civil Liability for Nuclear Damage of 21 May February 2004 Protocol to Amend the 1963 Vienna Convention on Civil Liability for Nuclear Damage of 12 September Joint Protocol Relating to the Application of the Vienna Convention and the Paris Convention of 21 September Convention on Supplementary Compensation for Nuclear Damage of 12 September

17 35. Hereunder we will now first briefly explain the scope (2.3) and the principles underlying both the Paris and Vienna Conventions as they were laid down in the nuclear liability conventions of the first generation (2.4). After that, we will explain how this regime has been changed by the conventions of the second generation, thus equally indicating how the compensation regime is organized today (2.4). 2.3 Scope of the international nuclear liability conventions 36. Before analysing the major principles of the nuclear liability conventions, it is useful to determine their scope: to what types of accidents do the conventions apply and until where and when do they apply Scope ratione materiae 37. According to Article 3.1 of the Paris Convention, the operator of a nuclear installation shall be liable, in accordance with the Convention, for: damage to or loss of life of any person; and damage to or loss of any property other than the nuclear installation itself and any other nuclear installation, including a nuclear installation under construction, on the site where that installation is located; and any property on that same site which is used or to be used in connection with any such installation, upon proof that such damage or loss (hereinafter referred to as "damage") was caused by a nuclear incident in such installation or involving nuclear substances coming from such installation. We find a similar definition in Article II.1 of the Vienna Convention: The operator of a nuclear installation shall be liable for nuclear damage upon proof that such damage has been caused by a nuclear incident in his nuclear installation ( ). Two elements are crucial in determining the scope of the conventions: nuclear installations and nuclear incident. 38. A nuclear incident is being defined as: means any occurrence or succession of occurrences having the same origin which causes damage, provided that such occurrence or succession of occurrences, or any of the damage caused, arises out of or results either from the radioactive properties, or a combination of radioactive properties with toxic, explosive, or other hazardous properties of nuclear fuel or radioactive products or waste or with any of them, or from ionizing radiations emitted by any source of radiation inside a nuclear installation. (Article 1.1 of the Paris Convention). According to Article I.l of the Vienna Convention a nuclear incident means any occurrence or series of occurrences having the same origin which causes nuclear damage. The Paris Convention thus links the occurrence or 17

18 the damage to radioactive properties of radioactive material used in a nuclear installation. The Vienna Convention has the same link, via the definition of nuclear damage This definition implies that there is a nuclear incident, according to the Vienna and Paris Convention, if a damage was caused whereby either the damage itself or the occurrence arises out of or results from the radioactive properties, or a combination of radioactive properties with other properties of substances inside a nuclear installation. In principle, one could say that as soon as either the damage or the occurrence have a radioactive nature or cause, the definition of nuclear incident of the Conventions will apply. Therefore, if both the occurrence and the damage are due to radioactivity, compensation may be claimed under the Conventions. However if the occurrence and the damage are conventional, this will fall outside their scope. Compensation may, however, be claimed under the Conventions either where an occurrence due to radioactivity causes conventional damage or injury or where an occurrence of conventional origin causes radiation damage or injury Note that the text of the Conventions does not require as such there to be a sudden occurrence; therefore one can assume that the Conventions apply to accidental and gradual pollution. This is confirmed by the Exposé de Motifs of the Paris Convention: an uncontrolled release of radiation extending over a certain period of time is considered to be a nuclear incident if its origin lies in one single phenomenon even though there has been an interruption in the emission of radioactivity The Paris Convention defines a nuclear installation as follows: reactors other than those comprised in any means of transport; factories for the manufacture or processing of nuclear substances; factories for the separation of isotopes of nuclear fuel; factories for the reprocessing of irradiated nuclear fuel; facilities for the storage of nuclear substances other than storage incidental to the carriage of such substances; and such other installations in which there are nuclear fuel or radioactive products or waste as the Steering Committee for 33 According to Article I.k of the Vienna Convention nuclear damage means (i) loss of life, any personal injury or any loss of, or damage to, property which arises out of or results from the radioactive properties or a combination of radioactive properties with toxic, explosive or other hazardous properties of nuclear fuel or radioactive products or waste in, or of nuclear material coming from, originating in, or sent to, a nuclear installation; (ii) any other loss or damage so arising or resulting if and to the extent that the law of the competent court so provides; and (iii) if the law of the Installation State so provides, loss of life, any personal injury or any loss of, or damage to, property which arises out of or results from other ionizing radiation emitted by any other source of radiation inside a nuclear installation. 34 Exposé des Motifs of the Paris Convention, nr Exposé des Motifs of the Paris Convention, nr

19 Nuclear Energy of the Organisation (hereinafter referred to as the "Steering Committee") shall from time to time determine; any Contracting Party may determine that two or more nuclear installations of one operator which are located on the same site shall, together with any other premises on that site where radioactive material is held, be treated as a single nuclear installation. (Article 1.2 of the Paris Convention). According to Article I.j of the Vienna Convention a nuclear installation means (i) any nuclear reactor other than one with which a means of sea or air transport is equipped for use as a source of power, whether for propulsion thereof or for any other purpose; (ii) any factory using nuclear fuel for the production of nuclear material, or any factory for the processing of nuclear material, including any factory for the re-processing of irradiated nuclear fuel; and (iii) any facility where nuclear material is stored, other than storage incidental to the carriage of such material; provided that the Installation State may determine that several nuclear installations of one operator which are located at the same site shall be considered as a single nuclear installation. 42. The definition of nuclear installation in the Paris Convention is a negative definition: if a reactor does not fall within the list of Article 1.2, it is to be considered as a nuclear installation in the sense of the Conventions. Basically, the result is very similar under the Vienna Convention: the Convention applies only to nuclear incidents occurring at or in connection with certain nuclear installations. As such, the definition of nuclear installation is quite broad. Especially risks associated with these kinds of activities are, in the spirit of the Conventions, considered to represent an important risk, and thus will require the application of the special nuclear liability regime. 43. Therefore, the Conventions will not apply to all accidents with nuclear material and releasing radioactivity. E.g. if radioactivity release takes place in a hospital in the course of a medical treatment or diagnosis, this will not be considered as an incident in a nuclear installation as defined by the Conventions. In that case common tort law will apply. 44. It seems that the criterion of the drafters of the Conventions has been the criticality of the installation and, associated to that, the risk of release of important quantities of ionising radiation and thus capable of touching the population. The Exposé des Motifs of the Paris Conventions refers to mining, milling and the physical concentration of uranium ores and factories for the manufacture or processing of natural or depleted uranium, facilities for the storage of natural or depleted uranium: they do not involve high levels of radioactivity and 19

20 therefore fall outsides the exceptional régime of the Convention. 36 For the same reason, the Steering Committee of the NEA has indicated that sub-critical assemblies should not be included in the definition of reactor within the meaning of Article 1 (a) (ii) of the Paris Convention. 37 However, installations for the disposal of nuclear substances shall, for the preclosure phase, be considered as a nuclear installation. 38 Moreover, the Steering Committee has indicated that the provisions of the Paris Convention should be interpreted as covering nuclear installations in the process of decommissioning The Exposé des Motifs of the Paris Convention indeed states at several occasions that the Convention provides for an exceptional régime and its scope is limited to risks of an exceptional character for which common law rules and practice are not suitable. Whenever risks, even those associated with nuclear activities, can properly be dealt with through existing legal processes, they are left outside the scope of the Convention. 40 We read the same in the Official Records of the Vienna Convention: the special rules of the Convention are necessarily only applicable with respect to damage which, as compared to that resulting from conventional activities, is of an extraordinary nature and cannot be covered by conventional insurance arrangements Territorial scope 46. According to Article 2 of the Paris Convention, the Convention does not apply to nuclear incidents occurring in the territory of non-contracting States or to damage suffered in such territory, unless otherwise provided by the legislation of the Contracting Party in whose territory the nuclear installation of the liable operator is situated. 47. The Vienna Convention does not have any specific provision with regard to its territorial scope. The draft Vienna Convention, debated at the Diplomatic Conference contained however a similar provision as the one in the Paris Convention. This provision was rejected stating that it was contrary to international law that the benefits of the Conventions should be applied to non-contracting States. 42 Despite the absence of any specific provision in 36 Exposé des Motifs of the Paris Convention, nr Interpretation of the Steering Committee of 8 June 1967 (NE/M(67)1]. 38 Decision of the Steering Committee of 11 April 1984 (NE/M(84)1]. 39 Interpretation of the Steering Committee of 28 April 1987 (NE/M(87)1]. 40 Exposé des Motifs of the Paris Convention, nr Official Records Vienna Convention, p Official Records Vienna Convention, p

21 this regards, it seems to be accepted that the Vienna Convention, like the Paris Convention, only applies to nuclear accidents or damage suffered in the territory of a Contracting Party. 48. Note that the Protocol of the Vienna Convention states that the Convention shall apply to nuclear damage wherever suffered. However, the legislation of the Installation State may exclude from the application of this Convention damage suffered (i) in the territory of a non- Contracting State; or (ii) in any maritime zones established by a non-contracting State in accordance with the international law of the sea. Such an exclusion may apply only in respect of a non-contracting State which at the time of the incident (i) has a nuclear installation in its territory or in any maritime zones established by it in accordance with the international law of the sea; and (ii) does not afford equivalent reciprocal benefits. 2.4 Principles underlying the international liability conventions of the first generation 49. The compensation regime introduced in the Paris and Vienna conventions is based upon a variety of principles: strict liability, legal channelling of the liability to the nuclear operator, limitation of liability, compulsory insurance and exclusive jurisdiction Strict liability 50. According to Art. 3 of the Paris Convention, the operator of a nuclear installation shall be liable for: damage to or loss of life of any person; and damage to or loss of any property other than the nuclear installation itself and any other nuclear installation, including a nuclear installation under construction, on the site where that installation is located; and any property on that same site which is used or to be used in connection with any such installation. This liability is established upon proof that such damage or loss was caused by a nuclear accident in such installation or involving nuclear substances coming from such installation. Thus, a victim wanting to introduce a claim against a nuclear operator does not have to prove a fault committed by the operator. This type of liability is generally known as strict liability. 51. However, the Paris Convention leaves in place liability of any individual for damage caused by a nuclear incident for which the operator, by virtue of Article 3(a)(ii)(1) and (2) or Article 9, is not liable under the Paris Convention and which results from an act or omission of that individual done with intent to cause damage. 21

22 52. Article 3(a)(ii)(1) and (2) of the Paris Convention refers to damage to the nuclear installation(s) itself and on-site property, whereas Article 9 refers to situations of force majeure which is, except if further confined by national law, limited to nuclear incidents directly due to an act of armed conflict, hostilities, civil war, insurrection or a grave natural disaster of an exceptional character. 43 Therefore, in such cases, the operator is not liable, nor will its liability insurance coverage be triggered for compensating damage as a result of these events. In this respect, is has to be noted that the nuclear liability regime in the Paris Convention provides for a quite stringent liability of the operator, given the fact that the operator will still be liable in cases where, under general tort law, classic exonerations would apply. 53. According to the Exposé des Motifs of the Paris Convention: the absolute liability of the operator is not subject to the classic exonerations such as force majeure, Acts of God or intervening acts of third persons, whether or not such acts were reasonably foreseeable and avoidable. ( ) The only exonerations lie in the case of damage caused by a nuclear incident directly due to certain disturbances of an international character such as acts of armed conflict and hostilities, of a political nature such as civil war and insurrection, or grave natural disasters of an exceptional character, which are catastrophic and completely unforeseeable, on the grounds that all such matters are the responsibility of the nation as a whole.. The nuclear operator is therefore liable for damage caused by acts of terrorism The Exposé des Motifs of the Paris Convention qualifies this type of liability as absolute liability. This type of liability was motivated by the fact that already in the fifties, there was a long-established tradition of legislative action or judicial interpretation that a presumption of liability for hazards created arises when a person engages in a dangerous activity. It was further noted that, because of the special dangers involved in the activities within the scope of the Convention and the difficulty of establishing negligence in view of the complex techniques of atomic energy, this presumption has been adopted for nuclear liability We find a similar provision in Art. IV.3(a) of the Vienna Convention: No liability under this Convention shall attach to an operator for nuclear damage caused by a nuclear incident directly due to an act of armed conflict, hostilities, civil war or insurrection. 44 Nathalie Horbach, Omer F. Brown, II, and Tom Vanden Borre, (2002), at Exposé des Motifs, n