Greater London Authority Liability Paper
1. Introduction
This note addresses the legal liabilities that might arise in the event of two possible scenarios for incidents involving radioactive waste or spent fuel transported by rail through London:
This analysis commences with a general description of the statutory framework for liability for nuclear accidents. It then considers the extent to which the current framework applies to the scenarios concerned and then lastly considers the shortcomings, i.e the extent to which the current framework either has no application or is deficient. It includes an appendix (Annex 2) with costs that have been attributed to a number of actual and theoretical accidents involving nuclear materials.
2. The Legal Framework
2.1 General
The legal position is governed by statute, the Nuclear Installations Act 1965 as amended. This limits and regulates the liability of the operator of a nuclear site for events on and, to a degree, off that site in the event of a relevant incident. The State assumes financial responsibility for the consequences of accidents over a monetary limit currently set at £140 million .
2.2 The International background
To understand the Nuclear Installations Act 1965 (the "NIA") it is necessary to appreciate a little about the international context. The relevant legal provisions for the UK are contained in sections 7 to 21 of the NIA. This legislation is intended to meet the requirements of the Paris Convention on Third Party Liability in the Field of Nuclear Energy of 1960 ("the Paris Convention") as supplemented by the Brussels Supplementary Convention of 1963 (the "Brussels Convention").
2.2.1 The Paris and Brussels Conventions
These Conventions were concluded under the auspices of the Paris-based Organisation for Economic Co-operation and Development (OECD). The United Kingdom is a party to both Conventions. The purpose of the Paris Convention of 1960 was to harmonise national legislation with regard to third party liability and insurance against atomic risks and to establish national regimes for liability and compensation in the event of a nuclear incident. The Convention establishes the liability of the operators of nuclear plant for personal injury and damage to property as a result of a nuclear incident. That liability will arise if the damage is caused by the accident, so that no proof of fault or negligence on the part of the operator is required . This corresponds to what is usually described as absolute liability in the English law of tort.
The Paris Convention limits the liability of the operator to 15million Special Drawing Rights ("SDRs", a unit of account used by the International Monetary Fund) in respect of a nuclear incident. Contracting States might establish greater or lesser amounts but in no event less than 5 million SDRs. The Brussels Supplementary Convention increased the total compensation available to 300 million SDRs per incident, equivalent to some £260 million. The Paris Convention requires that operators shall take out insurance cover or other financial security up to the limit of liability specified .
2.2.2 The Vienna Convention
The Vienna Convention on Civil Liability for Nuclear Damage of 1963 was negotiated under the auspices of the Vienna-based International Atomic Energy Agency (IAEA), an Agency of the United Nations. It is of potentially worldwide application rather than being confined to OECD countries. The purpose of the Convention, like that of the Paris Convention, is to provide for minimum protection under national law against damage resulting from certain peaceful uses of nuclear power. The Vienna Convention follows the same principles as the Paris Convention in that it provides for absolute liability on the part of the nuclear operator and permits the State within which the installation exists to limit the extent of the liability but not to a sum less than 5 million US dollars in respect of any one incident. The Vienna Convention also states that the operator shall be required to maintain insurance or other financial security covering his liability for nuclear damage in the amount or type which the State in which the installation exists shall specify.
None of the parties to the Paris Convention are parties to the Vienna Convention although the United Kingdom is a signatory and did express the intention to ratify the Convention but never has.
2.2.3 Revisions after the Chernobyl Treaty.
The Chernobyl accident exposed major systemic deficiencies in the application and content of the Vienna Convention and the Paris Convention.
(Almost all of these persist and are addressed in section 4.) The IAEA subsequently initiated a review of the Vienna Convention.
One deficiency was the very poor geographic coverage. Another was the absence of any reciprocity between the two Conventions. There were also major problems regarding (a) the limited definition of what damage qualified for compensation and (b) the totally inadequate amount of compensation. The Protocol to Amend the Vienna Convention on Civil Liability for Nuclear Damage of 1997 seeks to address both these points. The definition of nuclear damage has been made extendable beyond personal injury and damage to property to cover
But these additional heads of compensation apply only "to the extent determined by the law of the competent court" of the State party. So no common agreement has been reached on this and there is no binding requirement whatever to cover such losses and the Protocol is not yet in force in any event.
As for compensation levels, these are to be increased to 300 million SDRs and to be supported by an international inter-State pooling mechanism created by the 1997 Convention on Supplementary Compensation for Nuclear Damage, when this comes into force. This is designed to provide assistance to the State with primary liability from other State parties. This is a similar mechanism to that contained in the Brussels Convention.
The Protocol will not come into force until it has five contracting parties. Recently it had only two. The Convention on Supplementary Compensation also requires five contracting parties with significant nuclear capacity before it comes into force: so far it has only two.
By contrast with this IAEA activity, the OECD’s Nuclear Energy Agency and its constituent Member States including the UK have not amended the Paris Convention but are still, some 14 years after the Chernobyl accident, discussing matters.
2.2.4 General Principles
The Paris and Vienna Conventions thus have four principal common features.
operator.
As explained below each of these features has become a part of the United Kingdom law.
2.3 UK law
2.3.1 Liability of site operator
Section 7 of the NIA imposes a basic obligation on every holder of a nuclear site licence. It requires the licensee to ensure that no occurrence on the site involving nuclear matter causes (a) injury to any person or (b) damage to any property except that of the licensee. The liability is absolute: a person who has suffered damage need not establish any fault or negligence on the part of the licensee. The liability also extends to any occurrence involving nuclear matter in the course of carriage on behalf of the licensee and to nuclear matter that has been on the site as described in part 3.
2.3.2 Compensation
The Act provides that compensation shall be payable to anyone who has suffered injury or damage caused by a breach of the duty imposed on persons with a nuclear site licence and on certain others. The compensation is made payable if the injury or damage was incurred in the UK.
The amount of compensation which any operator may be liable to pay in respect of any one occurrence is limited to £140 million. The Secretary of State for Energy may prescribe an increased limit . The operator is required to maintain insurance cover (or some other means of cover) in respect of his potential third party liability up to £140 million. If claims are brought which in aggregate exceed the £140 million limit of liability for any one occurrence, then the Government must make available funds to ensure that claims are satisfied up to a total of 300 million SDRs . If the total claims exceed 300 million SDR the Act provides that they shall be satisfied by the Secretary of State for Energy to such extent and out of funds provided by such means as Parliament may determine . In other words the matter is left for the discretion of Parliament. The existence of and the amount of any further claims against the Secretary of State under this second stage may be determined by the High Court .
A limitation period of 30 years is imposed on claims for compensation, ie the claim cannot be entertained if made more than 30 years from the date of the occurrence which gave rise to the claim . There is an overall limitation period of 30 years for claims under the statutory provisions. Claims made more than ten years after the occurrence but within the overall 30 year limitation period are made to the Secretary of State for Energy and are satisfied out of the 300 million SDR sum or out of such further funds as Parliament may determine. It is because of these prescribed periods that the compulsory insurance has to provide cover up to £140 million not only in respect of the current cover period but also in respect of any cover period within the previous ten years .
The liability under the Nuclear Installations Act 1965 replaces (subject to minor exceptions) any liability which would have existed at common law as a result of a nuclear accident . So claims against the operator in tort at common law are extinguished in favour of the statutory liability.
In summary if anyone can show that injury or damage has been caused within the UK by an incident in the United Kingdom, that person has a claim which can be asserted within the next 30 years after the incident
available, and
(3) on such other sums as Parliament may determine.
All other liability in respect of the incident is excluded.
3. Application of UK liability regime to the scenarios.
3.1 a terrorist attack on a train leading to the dispersion of radioactive material.
The "occurrences" for which the site operator is to be liable are of three kinds:
Plainly the event suggested does not into the first category but might fall into the second. Does it fall into the third? This category refers to
- an occurrence
- on the relevant site or
- in the course of carriage.
An "occurrence" for the purpose of section 7(2) is not defined. In principle an attack would be "an occurrence".
Nuclear matter is defined as
"(subject to any exceptions which may be prescribed)
(a) any fissile material in the form of uranium metal, alloy or
chemical compound (including natural uranium), or of plutonium metal, alloy or chemical compound, and any other fissile material which may be prescribed; and
(b) any radioactive material produced in, or made radioactive by
exposure to the radiation incidental to, the process of producing or utilising any such fissile material as aforesaid;"
Waste or spent fuel would fall within that definition.
Excepted matter is defined as
"nuclear matter consisting only of one or more of the following, that is to say—
(a) isotopes prepared for use for industrial, commercial, agricultural, medical, scientific or educational purposes;
(b) natural uranium;
(c) any uranium of which isotope 235 forms not more than 0.72
per cent;
(d) nuclear matter of such other description, if any, in such
circumstances as may be prescribed (or, for the purposes of
the application of this Act to a relevant foreign operator, as
may be excluded from the operation of the relevant international agreement by the relevant foreign law); "
Waste or spent fuel would not fall within that definition and do not fall within the prescribed exceptions.
Assistance in determining whether the notion of an occurrence extends to a terrorist act can be obtained by posing the following questions.
(a) What is the plain meaning of the words?
Conventions, do they cover this kind of incident?
(c) What do other legal experts have to say on this?
The plain meaning
An occurrence involving a terrorist act certainly qualifies as an occurrence in the sense of an event, incident or happening as defined by the OED. When the 1965 Act was introduced as a Bill it was understood that this was to provide "that occurrences covered are no longer restricted to those which take place on the licensed site or in the course of carriage within the United Kingdom;".
Consistency with the Conventions
When the Minister introduced the Bill containing this extension, he said
"The Bill deals with the establishment of an international system governing civil liability for injury or damage arising from the use of nuclear energy. It is necessary so that we can amend our law, to enable us to ratify three international Conventions in the field of nuclear energy."
He explained that the three Conventions which the UK had signed were the Paris Convention, the Vienna Convention and the Brussels Convention supplementary to the Paris Convention .
Although the UK ratified the Paris and Brussels Conventions but not the Vienna Convention, it is clear that the Nuclear Installations Act was intended to be consistent with both Paris and Vienna conventions. This intention to achieve consistency with both Conventions was confirmed when the Nuclear Installations Act 1969 was passed to correct inconsistencies with the Treaties: "… our ratification of the Conventions ... means that our domestic law must be consistent with their provisions.". So what do the Treaties say?
(i)Article 3 of the Paris Convention provides that
"(a) The operator of a nuclear installations shall be liable, in accordance with this Convention…upon proof that … (the damage) was caused by a nuclear incident in such installation or involving nuclear substances coming from such installation …" (otherwise than in circumstances of authorised carriage or other).
Article II of the Vienna Convention provides for the operator to be liable for
"a nuclear incident
This seems to confirm that it is sufficient if the material came from the installation even if this was not authorised by the operator, as would occur in the case of theft or hijacking of material followed by its use in an explosive device. These Convention articles appear to be implemented through sections 7(2)(c) and 10(1)(b) NIA.
(ii)The Paris Convention exempts the operator from liability in the event of "damage caused by a nuclear incident directly due to an act of armed conflict, hostilities, civil war or insurrection..".
The Vienna Convention also provides that "No liability … 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."
These Articles make it clear that a nuclear incident can be such in principle even where the act is due to an act intended to cause damage (whilst exempting the operator from liability for particular instances).
These Convention requirements are drawn more narrowly in section 13(4)(a) NIA which excludes liability for an occurrence "attributable to hostile action in the course of any armed conflict, including any armed conflict within the UK"
(iii) Article 8(b) of the Paris Convention currently deals with time limits for claims " (i)n the event of damage caused by a nuclear incident involving nuclear fuel or radioactive products or waste, which at the time of the incident have been stolen…".
Article VI 2 of the Vienna Convention contains a similar provision.
This also establishes that an incident can involve nuclear fuel or radioactive products that have been stolen.
The requirement can also be found similarly expressed in section 15(2) NIA.
(iv) The Paris Convention refers to "damage caused by an incident result(ing) from an act or omission done with the intention of causing damage.."
The Vienna Convention contains a similar reference.
This also establishes that an incident may consist of an act of intentional damage.
The NIA contains a parallel provision in section 13(6).
Other legal experts
According to Michael Barnes QC, the Inspector at the Public Inquiry into the last proposal to construct a nuclear reactor in the UK,
"Injury or damage caused by hostile enemy action is excluded from the ambit of liability, but accidents caused by terrorism or sabotage could, it seems, give rise to liability." .
According to Professor Lopuski,
"The wording of the war exoneration as determined by the Vienna Convention and the Paris Convention does not cover explicitly terrorism. While generally speaking terrorism should be distinguished from "civil war or insurrection", in some cases, it may be difficult to draw a distinction between terrorism and civil war because terrorism may be one of the ways of waging the civil war."
On the basis of the above it seems reasonable to conclude that an incident involving an explosion utilising stolen radioactive material from the Sellafield plutonium store or from MOX fuel rods produced at Sellafield would be the liability of the Sellafield site operator, the wholly Government-owned BNFL, provided the incident fell short of being part of an insurrection or civil war.
The issue of a liability here is more straight-forward: an incident during authorised transport is clearly provided for within the legislation. Some complexity however does attend the allocation of liability between UK operator and foreign operator.
In principle the division of responsibility for incidents involving nuclear non-excepted matter during carriage within UK territory where the carriage takes place away from a nuclear site is as follows:
(i) The UK site operator is liable where the carriage is undertaken either
(ii)A Paris-convention foreign operator is liable where the carriage is either
(iii) A non-Paris-convention foreign operator is liable where the carriage is on that foreign operator’s behalf (section 11)
On this basis, plutonium being carried on behalf of BNFL would be that company’s liability. Plutonium being carried on behalf of a German customer of BNFL would be the German company’s responsibility and plutonium being carried on behalf of a Japanese company, would be the Japanese company’s responsibility. Where BNFL is returning plutonium to a customer in Germany or Japan, liability would seem to turn on the question of whether that carriage is on behalf of BNFL or on behalf of the overseas customer. Arguments might turn on the nature of the return contracts.
4. The deficiences of the current liability regime.
The Chernobyl accident revealed the following deficiencies in the Paris/Vienna liability regime:
All these points are very relevant to the present scenarios and to deficiencies in the UK liability regime. It is worth emphasizing that the UK does not need to wait for international Conventions or Protocols to be agreed before it can introduce improvements to modernise its own nuclear liability regime but it has chosen to take no legislative action since Chernobyl apart from altering the financial liability ceilings.
1. Insufficient compensation
Section 2.3.2 above explained the limited liability of the operator of £140 million and the taxpayers residual responsibility through the Government for up to some £260 million (300 SDRs). It is not appropriate that tax-payers’ should shoulder the responsibility for risks attendant upon a commercial operation above £140 million. In any event the ceiling of £260 million is far too low. For example the cost of the 1979 Three Mile Island accident was £3800 million, and the cost of the Chernobyl accident has been put at between £1950 million and £200,000 million. Modernising the liability regime would involve abolishing the ceiling on liability. Annex 2 to this Note considers some of the figures for compensation suggested for real and hypothetical accidents.
2. An overly restrictive definition of "nuclear damage"
The following losses were not covered by the Paris/Vienna Conventions or the Nuclear Installations Act 1965 at the time of Chernobyl and remain uncovered today:
Modernising the liability regime would involve explicit provision for all these heads of damages including damage to the environment and natural resources.
In Merlin v. BNFL, a Cumbrian homeowner sought compensation from BNFL for plutonium contamination of his property . The court found that the mere presence of radionuclides in his house did not constitute physical damage. Despite the loss in value of the house, compensation was not recoverable. By contrast in Blue Circle Cement v AWE intermingling of radioactive substances in soil amounts did constitute physical damage so that compensation for the damage and the loss in sale value of premises was recoverable. The cases are in many ways contradictory and illustrate the uncertainty complained of.
3. Time limits for submitting a claim
The Paris/Vienna regimes provide for ten year limitations. In the UK this applies to claims against the operator but as the operator is only insured for the first ten years, for a further 20 years claims may be brought directly against the Government.
Modernising the liability regime would involve abolishing any time limit for bringing claims given the very long periods that can run before impacts manifest themselves.
4. Difficulties in the proof of causation and damage
(i)ill health and death
Proof of causality is notoriously difficult to establish. The extent of physical harm may not become apparent for decades and when cancers do appear they may be indistinguishable from cancers with other causes.
In Reay v BNFL the Plaintiffs were unable to prove that the acute lymphatic leukaemia and the non-Hodgkins lymphoma which had affected two children - the first fatally - had been caused by BNFL’s operations. An excess number of cancers in the Sellafield area and epidemiological research associating leukaemia with paternal preconception irradiation at Sellafield were not sufficient to establish causation.
Without presumptions that eg presence at a particular incident will be presumed to be responsible for any subsequent cancer unless shown otherwise, claims are very difficult to establish. At Sellafield a BNFL occupational scheme provides for employees to obtain compensation for radiation injuries without proof of cause. Merely being exposed to a particular radiological hazard ought to be the basis for modernising statutory liability as well.
(ii)damage to property
The level of contamination sufficient to constitute "damage" is not adequately defined. This is illustrated by the cases of Merlin and Blue Circle mentioned above.
5.Exonerations
It is not satisfactory that there should be uncertainty about whether an occurrence falls within or without the concept of "insurrection" as in the present hypothetical case. The same act may or may not establish liability depending on the existence of hostile activity against the background of which the particular act takes place.
6.Lack of priorities in the distribution of compensation
There is no indication in the Conventions or the Nuclear Installations Act as to how funds are to be distributed in the event of insufficient funds as between e.g early and late claimants, those severely injured and those with property damaged.
7. The difficulty and expense of private law suits.
The length, procedural complexity and expense involved
in a private law suit are significant obstacles as the case of Reay showed.
This was one of the longest running personal injury cases on record which
had major impacts on legal aid policy in view of its expense and which
has contributed to the recent curtailment of funds for all civil legal
aid cases. Obtaining and retaining access to appropriate lawyers, scientific
experts, technologists over perhaps a period of fifteen years for a complex
law suit presents formidable problems.
The courts are not well equipped to deal with a massive number of claims arising out of a disaster as events in the 80s and 90s showed. Here the major complicating factor is causation. The physical, if not the pychological. effects of a ship sinking or a fire or a football disaster are very much more readily demonstrable than radiation injuries which may arise at any time over the lifetimes of the victims. The scope for disagreement over claims is huge.
5.Conclusion
Both scenarios appear to be alluded to in the UK liability legislation, but with some uncertainty attaching to the terrorist act. However the applicability in principle of the legislation is of very little comfort in view of its multiple and manifest deficiencies. The implications for a community affected by a nuclear accident contaminating a wide area are very worrying. Businesses appear to have no remedy for the purely economic impacts and decontamination costs appear irrecoverable from the nuclear operator in the absence of "physical" damage to property or person. The cost of evacuation, the damage to a city’s economic prosperity, the loss of use of buildings pending decontamination where this was possible, their loss in value, the economic disruption, in fact all the major consequences of such an accident do not appear to be catered for. It follows that a location affected by such an incident would be massively disadvantaged. A radical overhaul of the legislation is well overdue, as is the full assumption of the risks to communities created by BNFL of which the two scenarios are examples.
Annex 1
Scenarios for incidents involving civil plutonium
Two possible scenarios for incidents involving civil plutonium separated from spent reactor fuel elements at Sellafield are (1) the illegal acquisition of plutonium or mixed-oxide (MOX) fuel and the construction of a primitive nuclear explosive and (2) an accident involving an aircraft carrying plutonium in its cargo.
Nuclear terrorism
Having obtained plutonium oxide from the plutonium store at Sellafield or a MOX fuel assembly by diversion or theft, a terrorist group would have little difficulty in making a crude atomic bomb. If plutonium oxide (PuO2) itself is stolen it could be used directly to produce a nuclear explosive or converted into plutonium (Pu) metal, which could then be used to make a nuclear explosive.
MOX contains a mixture of uranium oxide and PuO2. The necessary steps of separating PuO2 from uranium oxide (UO2) in MOX are straightforward chemistry. And so is the conversion of PuO2 into plutonium metal.
The construction of a primitive nuclear explosive from the Pu metal or PuO2 is not technologically demanding. These operations do not require materials from specialist suppliers. The information required to carry them out is freely available in the open literature.
The operations would require some sophistication. Terrorist organizations are, however, certainly capable of sophisticated planning and the application of scientific principles. The construction of the conventional explosive device that destroyed the PanAm jumbo jet over Lockerbie on 21 December 1988, required considerable planning and scientific skills. And so did the construction of the nerve gas weapon used in the Tokyo underground by the AUM group on 20 March 1995. It is a sobering fact that the fabrication of a primitive nuclear explosive using reactor-grade plutonium, obtained from MOX, would require no greater skill than that for the production and use of the two terrorist weapons mentioned above.
The size of the nuclear explosion from a crude nuclear device made by a terrorist group device is impossible to predict. But even if it were only equivalent to the explosion of a few tens of tonnes of TNT it would completely devastate the center of a large city. Such a device would, however, have a strong chance of exploding with an explosive power of at least a hundred tonnes of TNT. Even one thousand tonnes or more equivalent is possible, but unlikely.
Effects of the explosion of a primitive nuclear explosion
with an explosive yield equivalent to that of 100 tonnes of TNT.
The largest conventional bombs used in warfare so far had explosive powers equivalent to about ten tonnes of TNT. The largest terrorist explosion so far has been equivalent to about two tonnes of TNT. A nuclear explosion equivalent to that of 100 tonnes of TNT in an urban area would be a catastrophic event, with which the emergency services would be unable to cope effectively. Exploded on or near the ground, such a nuclear explosive would produce a crater, in dry soil or dry soft rock, about 30 metres across. For small nuclear explosions, with explosive powers less than a few kilotons, the lethal action of radiation covers a larger area than that affected by blast and heat. The area of lethal damage from the blast produced by a 100-tonne nuclear explosion would be roughly 0.4 square kilometres; the lethal area for heat would be about 0.1 square kilometres; and that for prompt radiation would be roughly 1.2 square kilometres. Persons in the open within 600 metres of such an explosion would very probably be killed by the direct effects of radiation, blast, or heat (25). Many other deaths would occur, particularly from indirect blast effects from the collapse of buildings, from being thrown into objects or from falling debris. Heat and blast will cause fires, from broken gas pipes, petrol in cars, and so on. The area and extent of damage from fires may well exceed those from the direct effects of heat.
A nuclear explosion at or near ground level will produce a relatively large amount of early radioactive fall-out. Heat from fires will cause the radioactive particles to rise into the air; they will then be blown downwind, eventually falling to the ground under gravity at rates and distances depending on the velocity of the wind and the weather conditions. The area significantly contaminated with radioactive fall-out will be uninhabitable until decontaminated. The area concerned may be many square kilometres and it is likely to take a long time to decontaminate it to a level sufficiently free of radioactivity to be acceptable to the public.
An explosion of this size, involving many hundreds of deaths and injuries, would paralyse the emergency services. They would find it difficult even to deal effectively with the dead. Many, if not most, of the seriously injured would die from lack of medical care. In the UK, for example, there are only a few hundred burn beds in the whole National Health Service. There would be considerable delays in releasing injured people trapped in buildings, for example.
And, even for those not trapped, it would take a significant time to get ambulances through to them and then to transport them to hospital. Therefore, a high proportion of the seriously injured would not get medical attention in time to save them. Experience shows that, when large explosions occur in an urban area, panic sets in which also affects the trained emergency personnel. This panic would be considerably exacerbated by the radioactive fall-out accompanying a nuclear explosion.
Effects of a 1000-tonne nuclear explosion
The British Cabinet Office has calculated the effects of a primitive nuclear explosive detonated at ground level in a typical city. The explosion was equivalent to that produced by 1,000 tonnes of TNT, a possible but unlikely explosive yield from a crude nuclear weapons. Within one minute, people outdoors on near windows inside houses would be killed by thermal radiation (heat) up to a distance of 200 metres from the point of detonation. Within one minute, blast would kill people up to a distance of 800 metres, and initial nuclear radiation would kill people up to a distance of 1 kilometre. People within two kilometres would be injured by blast and those within one kilometre would be injured by heat. Communications equipment would be damaged by the nuclear electromagnetic pulse up to a distance of about two kilometres and electronic equipment would be damaged or disrupted up to a distance of about ten kilometres, with severe consequences for fire services, police headquarters, and hospitals. The electromagnetic pulse would affect motor vehicles out to about ten kilometres.
Assuming a 24 kilometre per hour wind, ionising radiation levels from radioactive fallout within an area of about 15 square kilometres would be high enough to cause radiation sickness in the short term to those exposed in the open, and in some cases to those in buildings. This area would extend some ten kilometres downwind nd would have a maximum width of about two kilometres.
Furthermore, radiation levels in an area of about 400 square kilometres would be such that certain counter-measures would have to be taken to protect people from the long-term effects of exposure to radiation - for example, fatal cancers. This area would extend some 80 kilometres downwind.
The most serious source of radioactive contamination from any crude nuclear explosive device is likely to arise from the dispersal of plutonium. If one kilogram of plutonium is uniformly distributed it will contaminate about 600 square kilometres to a level of one micro-curie per square metre, the maximum permissible level allowed for plutonium by international regulations. This means that a very large area will have to be evacuated and decontaminated, an expensive procedure which could take years.
(2) An accident involving an aircraft carrying plutonium
If an aircraft carrying plutonium crashes and catches fire, the plutonium (Pu) is likely to be scattered far and wide. This would have serious onsequences for people in the area concerned. Pu is a very toxic material. This toxicity arises mainly from its radiological effects; its chemical toxicity is similar to that of the heavy metals and is considerably less than its radiological toxicity. The radiological hazard of Pu arises mainly from the ionising radiation delivered to various internal organs of the body when Pu is ingested or inhaled into the body. Plutonium delivers a negligible external radiation dose to the skin because it emits mainly alpha particles that do not generally have sufficient energy to penetrate the skin.
Generally speaking, for the intake of a given amount of Pu, that which is inhaled is much more hazardous than that which is ingested. Pu is more easily absorbed into the blood stream through the lungs than through the gastrointestinal (GI) tract. Inhaled Pu will irradiate the lung; ingested Pu will irradiate the walls of the GI tract. Ingested and inhaled Pu may migrate via the blood stream to concentrate selectively in the liver and bones.
The health effects of Pu may be short-term (acute) or long-term (chronic). Inhalation, for example, may lead to acute pulmonary oedema. Long-term effects include an increased risk of cancer. Inhalation of Pu will expose the lung tissue to irradiation by alpha particles, increasing the risk of lung cancer. Some plutonium may eventually be carried from the lung to other organs (mainly the liver and the skeleton) where the radiation will increase the risk of cancer at these new sites.
The inhalation of 0.047 mg of Pu normally produced in a nuclear-power reactor (reactor-grade Pu), has a very high probability (virtually a certainty) of producing a cancer. A PuO2 particle containing 0.047 mg of Pu would have a diameter of 200 microns (1 micron is 1/1,000,000 metre or 1/10,000 centimetres.) It would not be possible to inhale such particles deep into the lung, where the lung tissue can absorb them. Any particle bigger than about 3 microns is not inhaled and is said to be not respirable. To inhale a total of 0.047 mg, it would be necessary to inhale about 300,000 3-micron PuO2 particles. This is a lot of particles.
But a very large number of respirable particles could be produced if a fire disperses PuO2 and explosion caused by an aircraft accident, about 30 per cent of the PuO2 may be respirable. In a hot fire, up to a few per cent of the Pu metal burnt will be converted into respirable PuO2. If, for example, 2 kilograms of PuO2 is dispersed, about 600 grams might be respirable. Assuming that all the particles are 3 microns in diameter, about 4 trillion (million million) of them will be produced.
An averagely active person breathes about 1.5 cubic metres of air a minute. Assuming that 2 kilograms of PuO2 are blown up in an explosion and that the 600 grams of respirable PuO2 are uniformly dispersed through a cube of air, 500 metres on the side. The concentration of Pu in the air is then 0.0048 mg per cubic metre. A person breathing this air for 1 hour would inhale 1.5 x 0.0048 = 0.0072 mg. A person breathing for 6 hours would inhale about 0.047 mg of Pu, enough to cause a fatal cancer.
The level of land contamination with Pu isotopes which would require decontamination (by, for example, the removal of top soil) depends on the circumstances. The National Radiological Protection Board (NRPB) recommends that land contaminated by more than about 1,000,000 Bq per square metre. Of relatively insoluble radioactive fine particles, like PuO2, will require evacuation until it is decontaminated. If evenly distributed, a kilogram of reactor-grade Pu will contaminate nearly 550 square kilometres to this level.
After the accident, the particles, which have fallen to the ground, are still a potential health hazard. If the particles are disturbed, or blown by the wind, they can become airborne again and will remain a health hazard until the area is decontaminated.
Annex 2
Estimated costs of real and theoretical accidents
A. Actual accidents
1/2. Palomares and Thule
Accidents involving nuclear weapons occurred at Palomares, Spain, in 1967 and Thule, Greenland, in 1968. Each cost around $500 million. More than one bomb was involved on each occasion although plutonium was dispersed not by explosion but by impact with the ground and fire. Also sparsely populated regions were involved . It was reported that at Palomares, around 1500 tons of topsoil and vegetation had to be removed for safe disposal and at Thule thousands of tons of ice and snow.
3.Three Mile Island
This accident caused only very small releases of radioactivity but is estimated to have cost between US$ 2 and 4 Billion .
4.Chernobyl
The official Soviet Economic Forecasting Agency calculated direct costs to the Soviet Union of between £1.95 and £3.1 Billion. Western commentators have estimated the direct cost to be in the order of £6 Billion. The Supreme Soviet originally set aside 26 billion roubles (around £26 billion) to try to cover all the associated costs. The Head of the Soviet Fire Service, making an estimate including long-term costs of treating those suffering from radiation sickness and other illnesses, calculated a figure of £200 Billion. Medever in "The Social and Environmental Impact of The Chernobyl Accident", reported that the total costs of the accident had increased from 2 billion roubles in 1986 to 17 billion roubles in 1991. Attempts to assess the costs up to the year 2000, have suggested figures ranging from 170-215 billion roubles. In the Ukraine the size of the 1993 "clean up" budget for the ongoing effects of the Chernobyl disaster has been greater than the entire defence budget (Wood). The cost to countries outside the former USSR has been between 1.5 to 4 billion dollars. Medevev concludes that the Chernobyl accident was the "most expensive industrial accident in modern history". The very lowest estimate of the total costs of the accident by 1991 was $6 billion US.
The massive operation to try to make settlements affected by the Chernobyl accident reinhabitable, highlighted many of the technical problems which would be encountered as a result of a nuclear accident in a populated part of the UK.
5. Goiana
If an urban environment is involved, then the problems of decontamination could be far more complex and costs correspondingly greater. In 1987 at Goiana, Brazil the dispersal of only 100 grams of Caesium-137 of 1400 curies radioactivity from a medical radiography machine required the removal of 3,500 cubic metres of soil and the demolition of 7 houses.
B. Theoretical Accidents
1.US Estimates
The US Government Rasmussen Report in 1975 estimated that a major accident could cause $14 Billion in property damage and 3300 immediate deaths. This report included a prediction of 4500 cancer deaths in the subsequent 10-40 years and 5100 genetic defects in later generations. This report was criticised from within the Government and by the independent scientific community because long term health effects cited were considered to be 50 times too low.
The Pace University Center for Environmental Legal Studies (US) in an estimate produced for the US Department of Energy and the New York State Energy Research and Development Authority put potential costs (as at 1990) at between US$ 613 billion and US$ 652 billion . However this estimate excluded:
The US government has assessed that around 100 square kilometers
could be contaminated if the conventional explosives of a nuclear warhead detonated, costing around $500 million dollars (£260 million) to clean up.
2. UK estimates
Estimates of costs of a major nuclear accident in the UK, were provided to the Sizewell Inquiry. Earth Resources Research and Friends of the Earth estimated 3000 early cancer deaths and short term attributable costs of £15 Billion to the U.K. alone .
3.German estimates
A report produced by Prognos AG for the German Federal Ministry of Economics estimated the cost for a worst case accident scenario for the Biblis-BPWR (soviet designed) power station including the expected costs of cancer deaths at US$ 6.8 Trillion.