How Not to Reduce Plutonium Stocks
The Danger of MOX-fuelled Nuclear Reactors

Corner House Briefing 17

by Frank Barnaby

first published 30 December 1999


Plutonium is radioactive by-product of nuclear reactors and one of the most toxic substances known. The nuclear industry argues that producing mixed-oxide (MOX) nuclear fuel would reduce plutonium stockpiles. It is unlikely to do so and instead would encourage the risk of nuclear terrorism and the spread of nuclear weapons.



Plutonium is a radioactive by-product of nuclear reactor operation and one of the most toxic substances known. The world would be a safer place if the governments of countries with stocks of it, including Britain, would adopt effective policies for reducing and managing them.

Two recent authoritative reports recommend that the British government take urgent action to reduce its "civil" plutonium stock - currently one quarter of the world's total and set to rise to about two-thirds by the year 2010.

The March 1999 House of Lords report, Management of Nuclear Waste, concludes that British government policy on plutonium "should be the maintenance of the minimum strategic stock, and the declaration of the remainder as waste".1 A report from the Royal Society, Britain's main learned society, meanwhile states that:

"In addition to disposing of some of the plutonium already in the stockpile, steps should be taken to reduce the amount added to it each year, primarily by reducing the amount of reprocessing carried out."2

The government's reply to the House of Lords is expected to be followed by a public consultation before changes in legislation are proposed.3

But, at the same time, the government is considering an application from British Nuclear Fuels Limited (BNFL), the government-owned company which separates plutonium from spent nuclear fuel rods, for a licence to operate a new plant at Sellafield in Cumbria to produce mixed-oxide (MOX) nuclear fuel from its plutonium stockpile.4 The nuclear industry justifies the Sellafield MOX plant as one way of reducing plutonium stocks.

But critics point out that this is not a rational way to manage plutonium. This briefing aims to contribute to an informed debate during the current flurry of British government nuclear policymaking by explaining why.

The Dangers of Plutonium

Plutonium, created in civil and military nuclear-power reactors as uranium fuel degrades (see below Box 1 "Producing Plutonium"), presents a major international problem: what should be done with it? Plutonium is dangerous for two main reasons: it can be used to manufacture nuclear weapons; and it is highly toxic mainly because of its radioactivity.

Inadequate control of plutonium will frustrate efforts to prevent the spread of nuclear weapons, particularly to countries that do not now have them. It will also make it easier for terrorist groups to acquire some of the material and construct nuclear explosives of their own.5 Moreover, plutonium is so toxic that it must be kept out of the environment as far as possible. The inhalation of microgramme quantities has a very high probability of causing fatal lung cancer, while ingestion can cause bone and liver cancer. The most common plutonium isotopes, moreover, stay radioactive for extremely long periods of time. Plutonium-239, for example, has a half-life of about 24,000 years. To all intents and purposes, once it is in the environment, it stays there permanently.

Box 1: Producing Plutonium

Almost all plutonium is produced in nuclear reactors which are usually fuelled with uranium. Uranium has two important isotopes - uranium-235 and uranium-238. Uranium-235 is a fissile isotope. When a nucleus of an atom of a fissile isotope captures a neutron travelling at any speed, fast or slow, it undergoes fission, that is, the nucleus splits into two nuclei, called fission products, of different chemical elements. The energy released is used to generate electricity.

During fission, two or three neutrons are emitted from the original uranium-235 nucleus. If one of these neutrons is captured by the nucleus of an atom of the more stable uranium-238, it will cause fission only if it is travelling at a very high speed. If it is not, as is normally the case, a nucleus of the radioactive isotope neptunium-239 will be produced that will decay into plutonium-239, another fissile isotope. Therefore, as the uranium fuel undergoes fission and is gradually used up in the reactor, an increasing amount of plutonium-239 is inevitably produced.

But plutonium-239 can also capture neutrons to become plutonium-240, which in turn can capture neutrons to become plutonium-241, and so on. Consequently, a mixture of plutonium isotopes is gradually produced in the uranium reactor fuel. This then has to be extracted chemically from the spent fuel.

There are various grades of plutonium, each with different isotopic compositions depending on the way in which the reactor producing it is operated. Plutonium produced in commercial or civil nuclear-power reactors operated for the most economical production of electricity is called reactor-grade plutonium. Plutonium produced in military plutonium production reactors, specifically for use in nuclear weapons, is called weapons-grade plutonium.

Plutonium Stocks

Plutonium was first produced in significant amounts as part of the Manhattan project, set up by the United States during the Second World War to manufacture nuclear weapons. The bomb dropped on Nagasaki was a plutonium bomb. Since 1945, the world has produced a huge amount of plutonium - a total of about 1,500 tonnes. About 250 tonnes of this plutonium were produced for use in nuclear weapons. The other 1,250 tonnes are "civil" plutonium, produced as an inevitable by-product by civilian nuclear reactors while they are generating electricity.

The established nuclear weaons powers are no longer producing weapons-grade plutonium - they now have more than they need. But the amount of "civil" plutonium is increasing significantly. The world's nuclear power reactors (437 are operating in 32 countries) are producing an additional 70 tonnes of plutonium a year.

About 250 tonnes of "civil" plutonium have already been chemically separated from spent nuclear power reactor fuel elements in factories called chemical reprocessing plants. These plants take as their raw materials spent reactor fuel elements containing unused uranium and fission products as well as plutonium. These three substances are chemically separated from each other by dissolving them in nitric acid.

Six commercial scale reprocessing plants are currently operating in four countries: two at Sellafield, Britain; two at La Hague, France; one at Chelya-binsk, Russia; and one at Tokai Mura, Japan. A second Japanese plant, at Rokkasho Mura, is scheduled to start operating soon after the year 2003.

About 80 tonnes of plutonium separated in reprocessing plants are in France, about 50 in the UK, about 50 in Japan, and about 40 tonnes each in Germany and Russia. Smaller amounts (less than 8 tonnes per country) are in Belgium, India, Italy, the Netherlands, Spain, Switzerland, and the USA. If current reprocessing plans go ahead, there will be about 550 tonnes of separated "civil" plutonium by the year 2010. This amount of plutonium could produce a staggering 110,000 nuclear weapons.

Sources: Albright, D., Berkhout, F. and Walker, W., Plutonium and Highly Enriched Uranium 1996; World Inventories, Capabilities and Policies, Oxford University Press for Stockholm International Peace Research Institute, Oxford, 1997: Leventhal, P., "How to Verify a Comprehensive Production and Safeguard All Stocks", presentation at International Network of Engineers and Scientists Against Proliferation seminar on Fissile Materials and Tritium, Geneva, June 1995.

Pressures for a Plutonium Economy

After the explosion at the Chernobyl nuclear-power reactor in the Ukraine in 1985, the nuclear industry has been in decline, at least in the West. The accident in October 1999 at the Tokai Mura fuel-manufacturing plant in Japan has not helped its fortunes. Only Asian countries, mainly China, India, Japan, Pakistan, South Korea and Taiwan, are building significant numbers of new nuclear-power reactors, (see Table 1 below). To keep themselves going, therefore, Western nuclear businesses rely increasingly on providing fuel to existing reactors, decommissioning nuclear facilities and managing radioactive waste.

The industry believes its main hope for long-term survival in the West is commercial "breeder reactors" - nuclear-power plants fuelled primarily with plutonium. In theory, breeders would produce more nuclear fuel than they would consume and would eventually become self-sufficient in fuel. This was one of their main justifications in the late 1960s and 1970s when it was feared that there would eventually be a shortage of uranium to fuel nuclear-power reactors which could be mined at economic prices. But plutonium-containing fuel cannot be "bred" at the reactors themselves. It requires special chemical reprocessing plants to recover plutonium economically from spent reactor fuel elements. The nuclear industry, in other words, is looking forward to a complex plutonium economy complete with commercial reprocessing plants and worldwide plutonium transport between reactors and plants.

Given the abundance of relatively cheap uranium, however, it will be several decades at least, if not a century, before breeder reactors could be economically viable, if they ever will be. Moreover, breeders face other serious problems. They are cooled with a liquid metal, such as sodium, and the technology that would allow large quantities of liquid metal to be used safely and effectively has yet to be developed. The experimental breeder reactors built so far cannot generate electricity at competitive prices. And breeder reactors are, of course, also subject to the pressures forcing the decline of the rest of the nuclear industry.

Until breeders become technically and economically viable, the nuclear industry argues that plutonium oxide (the form in which plutonium exists after reprocessing) should be mixed with uranium oxide (the form of uranium usually used in reactors) and the resulting mixed-oxide (MOX) mixture used to part-fuel ordinary reactors. The industry sees this as an interim survival strategy - a way of maintaining the momentum of nuclear technology and investment until commercial breeder reactors come on line.

Thus MOX is currently part of the fuel used in civil nuclear-power reactors in Belgium, France, Germany and Switzerland, while Italy, Japan, The Netherlands, and Sweden have used it for trial purposes. Germany began using MOX in 1966 and currently loads five light-water nuclear-power reactors with it. Switzerland started using MOX in 1978 and now employs the fuel in three reactors, while France, which began in 1987, uses it in thirteen. Belgium has two reactors which have used MOX since 1995. Thus a total of 23 nuclear-power reactors worldwide are using MOX fuel (see Table 2 below). None are loaded exclusively with MOX; typically, MOX accounts for about 30 per cent of the nuclear fuel in the core of the reactor.

Commercial MOX fabrication plants, meanwhile, are operating or will soon operate in five countries - at Sellafield, Britain; Dessel, Belgium; Marcoule and Cadarache, France; Tokai Mura and Rokkasho Mura, Japan; and Chelyabinsk, Russia. All these countries, except Belgium, also have repro-cessing plants from which to obtain plutonium oxide.

MOX producers, like reprocessors, generally rely on foreign customers to boost their income, the significant potential users being Japan, Switzerland and Germany (although the future of Germany's nuclear programme is unclear since the Green Party became part of the coalition government in September 1998). Available estimates suggest that MOX supply will be about two times greater than MOX demand up to the year 2015.

It should be emphasized that 75 or 80 per cent of the plutonium still contained in spent civilian reactor fuel elements will have to be disposed of without reprocessing the elements. Only about 20 per cent of the plutonium contained in the 180,000 tonnes of spent fuel rods discharged by civilian reactors has been separated in reprocessing plants, and, according to global plans for civil reprocessing, this percentage is unlikely to increase significantly in the foreseeable future.

Table 1: Nuclear Power Reactors in Asia
Country Reactors Total Electricity Reactors under Total   in Operation Generated(MW) Construction Electricity                 Generated
China 3 10,000 6 4,420
India 10 1,695 4 808
Japan 53 43,690 2 1,863
South Korea 15 12,340 3 2,550
Pakistan 1 125 1 300
Taiwan 6 4,900 1 1,300
Table 2: Nuclear Reactors using MOX fuel
Country Reactor
Belgium Doel-3
France Le Blayais-1
Le Blayais-2
St. Laurent B1
St. Laurent B2
Germany Brokdorf
Gundremmingen B
Gundremmingen C
Switzerland Beznau-1

Disposing of the Surplus? Or Creating an Even Bigger One?

Because breeder reactors remain unviable, the reprocessing plants which separate out plutonium from spent nuclear fuel rods are now producing an embarrassing surplus of plutonium. The nuclear industry tries to justify MOX by arguing that its production and use will help dispose of this surplus. This reasoning, however, is incorrect. The MOX route does not decrease stocks of plutonium but rather increases it - unless MOX fuel is used in a very much larger fraction of reactor cores than is considered safe or practicable, even by the nuclear industry.

The following figures explain why the MOX route would increase stocks of plutonium. A typical MOX fuel assembly - about 289 fuel rods arranged geometrically - contains about 435 kilogrammes of uranium and 25 kilogrammes of plutonium. The MOX fuel remains in the reactor for three years. A spent MOX fuel assembly contains about 19 kilogrammes of plutonium. There are 48 MOX assemblies in, for example, a typical pressurized water reactor (PWR) generating 900 megawatts of electricity. The 48 assemblies are divided into three sets of 16 assemblies, each one of which can be assumed to be replaced each year. A 900-megawatt PWR will, therefore, dispose of about 96 kilogrammes of plutonium per year - (25 - 19) x 16 kilogrammes per year, or about one tonne of plutonium every ten years.

But the non-MOX fuel assemblies in the remaining 70 per cent of the reactor core will meanwhile produce an even greater amount of plutonium as the uranium oxide fuel is burnt up (see above Box 1 "Producing Plutonium"). Overall, the typical ratio of plutonium "out" to plutonium "in" would be about 1:1.17 for a 900-megawatt PWR using MOX and uranium fuel assemblies and containing 5.2 per cent of plutonium in 30 per cent of the core.

Thus if current plans for reprocessing and for MOX use are enacted, the world's stock of plutonium, now about 250 tonnes, will grow. Plutonium stocks could decrease only if MOX fuel were used in a very much larger proportion of the reactor cores. Yet this is not possible or safe for reasons explored below.

Box 2: From Reactors to Bombs: The Illegal Use of MOX

Having obtained a MOX fuel assembly by diversion or theft, a terrorist group would have little difficulty in making a crude atomic bomb. The necessary steps of chemically separating the plutonium oxide from uranium oxide, converting the oxide into plutonium metal, and assembling the metal or plutonium oxide together with conventional explosive to produce a nuclear explosion are not technologically demanding and do not require materials from specialist suppliers. The information required to carry out these operations is freely available in the open literature.

Some recent official statements imply that plutonium produced in nuclear-power reactors - and therefore that which could be obtained from MOX - cannot be used in nuclear weapons or nuclear explosive devices. For example, Ryukichi Imai, former Japanese Ambassador for Non-Proliferation, has stated that:

"Reactor-grade plutonium is of a nature quite different from what goes into the making of weapons ... Whatever the details of this plutonium, it is quite unfit to make a bomb."

This statement is incorrect, as Robert Seldon of Lawrence Livermore Laboratory explains:

"All plutonium can be used directly in nuclear explosives. The concept of ... plutonium which is not suitable for explosives is fallacious. A high content of the plutonium 240 isotope (reactor-grade plutonium) is a complication, but not a preventative."

The Director General of the International Atomic Energy Agency, Hans Blix, stresses that his organization:

"considers high burn-up reactor-grade plutonium and in general plutonium of any isotopic composition ... to be capable of use in a nuclear explosive device. There is no debate on the matter in the Agency's Department of Safeguards."

Declassified Admissions

And at a conference in Vienna in June 1997, Matthew Bunn, who chaired the US National Academy of Sciences analysis of options for the disposal of plutonium removed from nuclear weapons, made a crucially important statement based on recently declassified material "of unprecedented detail on thisa#ubject":

"For an unsophisticated proliferator, making a crude bomb with a reliable, assured yield of a kiloton or more - and hence a destructive radius about one-third to one-half that of the Hiroshima bomb - from reactor-grade plutonium would require no more sophistication than making a bomb from weapon-grade plutonium. And major weapon states like the United States and Russia could, if they chose to do so, make bombs with reactor-grade plutonium with yield, weight, and reliability characteristics similar to those made from weapon-grade plutonium. That they have not chosen to do so in the past has to do with convenience and a desire to avoid radiation doses to workers and military personnel, not the difficulty of accomplishing the job. Indeed, one Russian weapon-designer who has focused on this issue in detail criticized the information declassified by the US Department of Energy for failing to point out that in some respects if would actually be easier for an unsophisticated proliferator to make a bomb from reactor-grade plutonium (as no neutron generator would be required)."

Exploding the Myths

That reactor-grade plutonium can be used to fabricate nuclear weapons was proved by the British who exploded such a device in 1956 and by the Americans who exploded at least one such device in the 1960s.

Terrorists are probably most likely to acquire MOX while it is being transported; this is when it is most vulnerable. A single MOX assembly typically contains about 25 kilogrammes of plutonium, enough for at least two nuclear weapons. The use of MOX, therefore, significantly increases the risk of nuclear terrorism in particular and the proliferation of nuclear weapons in general.

Source: Imai, R., Plutonium, No. 3, October 1994; Lovins, A. B., "Nuclear Weapons and Power-Reactor Plutonium", Nature, 28 February 1980, pp.817-823 and typographical corrections, 13 March 1980, p.190; Mark, J. C., Reactor-Grade Plutonium's Explosive Properties, Nuclear Control Institute, Washington D.C., August 1990; Selden, R. W., Reactor Plutonium and Nuclear Explosives, Lawrence Livermore Laboratory, California, 1976; Blix, H., Letter to the Nuclear Control Institute, Washington DC, 1990; Bunn, M, Paper at International Atomic Energy Agency Conference, June 1997.

Safety of Reactors Using MOX

Reactor operators and MOX producers generally claim that burning MOX in light-water reactors which have been designed to use ordinary uranium oxide fuel does not pose any additional safety problems. These claims are usually based on the fact that plutonium is produced continually anyway during the operation of a reactor fuelled conventionally with uranium oxide and that some of this plutonium undergoes fissions, typically accounting for approximately one-third of the total fissions. It is concluded that plutonium fissions in light-water reactors do not constitute a new problem.

Such arguments are faulty. In a typical uranium oxide fuel element, the amount of plutonium accumulated in the fuel element while it is in the reactor will be about one per cent of the weight. In a typical new MOX fuel element, however, plutonium will account for five per cent or more. MOX fuel and partially-used uranium oxide fuel are therefore not comparable.

Moreover, there have been, to say the least, an inadequate number of safety studies for reactors containing MOX fuel, and it is very difficult to obtain those that have been done. The nuclear industry claims commercial confidentiality to keep them secret. The industry also claims that safety analyses have shown that the behaviour of MOX fuel assemblies is very similar to that of uranium oxide assemblies, but these analyses have not been subject to objective independent assessment.

Reactors burning MOX fuel are less safe than those burning uranium oxide fuel for two reasons. Firstly, the fact that MOX fuel pellets are constructed from two actinide6 oxides rather than one makes fabrication and quality control considerably more difficult for MOX pellets compared with uranium oxide fuel pellets. Secondly, differences in the properties of plutonium and uranium in the core of a MOX-burning reactor alter the functioning of the reactor with adverse consequences for safety.

Quality Control Problems

Pellets of MOX fuel are prepared at BNFL's MOX Demonstration Facility (MDF) at Sellafield.7 The pellets are placed end to end in sealed tubes filled with the inert (or chemically inactive) gas argon (often also used in electric light bulbs and TV tubes). The resulting fuel rods are held in geometric array by spacers to form a fuel assembly for a nuclear-power reactor.

Quality control of reactor fuel elements is crucial for safety. MOX fuel pellets must be produced to very demanding standards. Linear dimensions, density, bulk composition and homogeneity should all be assured to within very narrow limits. A lapse in any one of these parameters could result in a reactor accident. Radioactivity could be released into reactor buildings, and accidents even on the scale of Three Mile Island or Chernobyl cannot be ruled out. Defective pellets may also lead to distortions in fuel rods which require a great deal of time and money to rectify.

Recent revelations of lapses in quality assurance procedures, including the falsification by forgery of inspection data at the British MOX Demonstration Facility (MDF) are of considerable concern.8 But these lapses represent only part of the problem of assuring the quality of MOX fuel. The quality control procedures themselves are at fault, not just their implementation. The very nature of the fuel pellets and the way they are made precludes adequate quality assurance procedures from being implemented economically.

If the attritor mill at the MOX plant at MDF is operating correctly, it should produce fine, uniformly mixed particles a micron (a millionth of a metre) in diameter which should flow like a liquid through subsequent processing stages until they are pressed and heated to form the final heat-fused cylindrical pellet. But experience in other industries, such as the pharmaceutical industry, indicates that processes that depend on the flow of powders are far from totally reliable, particularly when the mixing of different constituents is involved.

From the point of view of reactor safety, one of the most important properties of a MOX pellet is its plutonium content - the weight of plutonium in the pellet as a percentage of the total weight. If the oxide powders are inadequately mixed before they are fed into the attritor mill, the plutonium content will vary from pellet to pellet. Too much plutonium in a pellet could produce excessive local heating, possibly damaging the cladding of the fuel rod, with adverse safety consequences. Too little plutonium would mean that customers are not getting what they paid for.

More serious is the problem of inadequate mixing of the powder fed into the attritor mill or inadequate mixing in the mill itself. Variations in water content, composition and initial size of the particles used to make the pellets, or wear and tear of the attritor mill, could cause inadequate mixing and partial or total clogging of the mill. The flow of powder in the subsequent stages of pellet fabrication can be affected by all these factors. One result could be inhomogeneous distribution of plutonium within pellets, leading to plutonium "hot spots" - larger-than-average plutonium oxide particles on the surface of pellets.

It is difficult to predict where these inhomogeneties will turn up. Brief fluctuations in the efficiency of mixing the oxides and powders would not be detected unless virtually all the pellets were inspected; even extended fluctuations would be missed if the samples taken for inspection were not large enough.

Yet such hot spots could damage the cladding of MOX reactor fuel rods, causing uneven local heating.9 This is a risk particularly given that more fission (and hence more heating) occurs at the surface of pellets than at their centres. Large plutonium oxide particles could also accumulate together in one place to produce aggregates. The larger the aggregate, the greater the heat produced by fission of the plutonium and the more likely damage would result to the cladding of the rod.10 In some circumstances, the result could be reactor accidents, which are more likely in fuels with higher plutonium content.

BNFL checks for inhomogeneities in pellets using alpha-autoradiography. A section is cut from each sample pellet and then polished. It is then placed in contact with a photographic film for some days, after which the film is developed and examined so that the size and number of clumps of plutonium oxide particles made visible can be assessed.

This method of testing is labour intensive, destructive to the pellet and time-consuming - which may be why BNFL routinely inspects just one pellet taken in about every 40,000. Of the pellets inspected, about 20 per cent typically do not meet the standards.

Besides the variable composition of the pellets, variable size and weight may also be a threat to safety. A pellet which is too small may rattle about within the fuel rod and give rise to serious wear and tear in the fuel cladding; the swelling of a pellet as a consequence of neutron irradiation may cause a pellet that is too large to damage the cladding as well.

Higher-density pellets, meanwhile, may swell excessively, and pellets with lower density may split. Finally, too many impurities, for example of carbon or fluorine, may lead to fuel-rod corrosion; and too much trapped air or gas expelled from pellets could increase the risk of rupturing the cladding of a rod, and of an accident.

In addition, each fuel rod should be weighed to help ensure that the correct number of correctly-sized pellets have been introduced into it, and to determine when MOX pellets have been replaced with ceramic blanks of a similar size but containing no plutonium, by, for example, a person wanting to steal plutonium. This procedure is of considerable importance in ensuring that all the plutonium entering the process can be accounted for in the completed fuel assemblies. Unless this is done carefully, it will be nigh on impossible to know whether plutonium has been lost or stolen during fuel fabrication and assembly by terrorist or other groups (see above Box 2 "From Reactors to Bombs").

Checks must also be made on the metal (plutonium and uranium) content of the pellets and the pellet's oxide-to-metal ratio. These measurements give further information about the plutonium content.

Box 3: Primitive Device, Devastating Impact

If a terrorist group or government agents from a country lacking nuclear weapons acquired one or more fuel assemblies it would be relatively easy for them to extract the plutonium from MOX pellets and fabricate a nuclear explosive. This would require no greater skill than that required for the construction of the nerve gas weapon used in the Tokyo underground by the AUM group on 20 March 1995.

Fuel pellets in MOX fuel assemblies are normally composed entirely of pure reactor-grade plutonium and depleted uranium, both present as dioxides and fused together by heat to form a ceramic. The material is designed to be soluble in fairly concentrated nitric acid for ease of reprocessing. The plutonium oxide could be chemically extracted from the MOX using a procedure devised by a second-year undergraduate, using one of a number of methods described in standard reference works and scientific journals. Plutonium metal could then be produced from the plutonium oxide.

A nuclear explosive could be fabricated by two or three people with appropriate skills, using either the plutonium oxide or plutonium metal. The nuclear-physics data needed to design a crude nuclear device are readily available in the open literature.

Nuclear Basics

The critical mass of a fissile material, such as plutonium, is the minimum mass necessary to sustain a nuclear-fission chain reaction. If the mass is more than critical (that is, super-critical), the fission chain reaction is sustained for as long as the mass of plutonium remains super-critical.

A nuclear explosion occurs when a mass of plutonium or plutonium oxide that is less than critical is compressed by a symmetrical shock wave to make it super-critical. Such compression reduces the volume of the mass of plutonium, increasing its density. As the critical mass is inversely proportional to the square of the density, increasing the density by compression will reduce the mass required for critical assembly. If the mass of plutonium is held together long enough, enough fission reactions will take place to produce an explosion.

The critical mass of reactor-grade plutonium oxide crystals is about 35 kilogrammes if the mass is spherical, while that of reactor-grade plutonium metal is about 13 kilogrammes. A sphere of plutonium oxide with a critical mass would be about 18 centimeters in diameter; a sphere of plutonium metal with a critical mass would be about 12 centimeters in diameter. Such a sphere would be placed in the center of a mass of a conventional high explosive to compress it to super-criticality.

If the sphere of plutonium oxide or plutonium metal is surrounded by a shell of beryllium or uranium, neutrons which escape from the sphere without producing a fission event are reflected back into the sphere. Such a reflector, therefore, reduces the critical mass - a thick reflector by a factor of two or more.

Because the shell of reflecting material is heavy, it also acts as a tamper. When the high explosives are detonated, the shock wave causes the tamper to collapse inwards. Its inertia helps hold together the plutonium during the explosion to prevent the premature disintegration of the fissioning material and thereby to obtain a larger explosion.

To compress the plutonium, it is likely that a terrorist group would use a plastic explosive such as Semtex, since it is easy to handle. About 400 kilogrammes of plastic explosive, molded into a sphere around the reflector/tamper placed around the sphere of plutonium, and incorporating 50 or so symmetrically-placed detonators to be set off simultaneously, would be sufficient.

Tens of Tonnes of TNT

The size of the nuclear explosion from such a crude device is impossible to predict. But it could easily be on the order of that yielded by a few tens of tonnes of TNT - enough to devastate the centre of a large city - or even a hundred tonnes or more. By comparison, the largest conventional bombs used in warfare to date have had explosive powers equivalent to about ten tonnes of TNT, and the largest used in terrorist attacks, 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. Such an explotion would produce a crater, in dry soil or dry soft rock, about 30 meters across. The area of lethal damage from the blast would be roughly 0.4 square kilometers and from heat about 0.1 square kilometers. The direct effects of radiation would very probably kill persons in the open within 600 meters of such an explosion. Many other deaths would follow, particularly from the collapse of buildings, falling debris, and other impacts. Heat and blast would cause fires from broken gas pipes, petrol in cars, and so on. An explosion of this size, involving many hundreds of deaths and injuries would paralyze the emergency services. In the UK, for example, there are only a few hundred burn-beds in the whole National Health Service.

Even if the device, when detonated, did not produce a significant nuclear explosion, the dispersal of its plutonium through blast and fire would have disastrous effects. Plutonium particles smaller than three microns in diameter could be breathed into, and retained by, human lungs, where they would be very likely to cause lung cancer by irradiating the surrounding tissue with alpha-particles.

Much dispersed plutonium would also remain in surface dusts and soils until removed. If only one kilogramme of plutonium were uniformly distributed, it could contaminate about 600 square kilometers to a level of one micro-curie per square meter, the maximum permissible level allowed by international regulations. Decontamination following even a non-nuclear explosion involving plutonium, therefore, could take many months or years.

The threat of dispersal of many kilogrammes of plutonium makes a crude nuclear explosive device a particularly attractive weapon for a terrorist group, the threat being enhanced by the general population's justifiable fear of radioactivity.

Source: Lovins, A. B., 'Nuclear Weapons and Power-Reactor Plutonium', Nature, London, 1990, pp.283, 817-823 and typographical corrections, pp.284, 190; Rotblat, J., Nuclear Radiation in Warfare, Taylor and Francis, London, 1981.

Inadequate Checks

As far as is known, BNFL applies 15 checks to its MOX pellets.11 This comprehensive set of checks should identify many of the problems outlined above. But the frequency of most of the checks is totally inadequate. In some checks, only one sample is taken per 40,000 or so pellets. In no check is more than one sample taken from 1,500 pellets.

Such a low sampling rate is bound to allow flawed pellets to get through the checking procedures. Of particular concern is the total inadequacy of the checks to detect inhomogenieties in plutonium distribution within pellets. Yet the cost of checking properly for inhomogenieties in the distribution of plutonium in fuel pellets is likely to be prohibitively expensive, meaning that adequate checking of MOX fuel produced in commercial MOX plants is not economically feasible. The reasonable conclusion is that MOX fuel as currently produced is not safe to use.

Box 4: Other Arguments Against MOX

MOX cannot contribute to a reduction in the world's surfeit of plutonium. It is beset by many other drawbacks as well:

  • The use of MOX increases the risk of nuclear-weapon proliferation by countries and by terrorist organisations.
  • Differences between plutonium fission and uranium fission, together with other factors, may make reactors fuelled by MOX less safe.
  • The cost of MOX fuel is typically several times higher than that of normal (low-enriched uranium) fuel.
  • There is no foreseeable shortage of economically-viable uranium because demand and prices have dropped.
  • The need to protect MOX fuel elements kept at nuclear reactors will involve reactor operators in new physical security problems.
  • International safeguards designed to prevent nuclear proliferation are very difficult to enforce at facilities associated with MOX, particularly plutonium bulk-handling plants, like reprocessing and MOX fabrication plants.
  • The use of MOX increases the risk of serious accidents since transportation of sensitive materials must be organized from reactors to reprocessing plants to MOX fabrication plants and back to reactors.
  • Transporting MOX fuel is riskier than moving other nuclear materials because it is more of a target for theft and sabotage.
  • The need to guard MOX fuel closely during transportation can involve the use of large numbers of armed police (as happened at Barrow, Cumbria, Britain during a recent shipment of MOX from Sellafield to Japan), giving rise to justifiable fears of the development of a police state.


Three main ways have been proposed to reduce the world's stock piles of plutonium:

  • Plutonium oxide could be mixed with uranium oxide to produce MOX fuel for nuclear-power reactors;
  • The plutonium could be stored until decisions are made about its final disposal; and
  • The plutonium could be incorporated into glass blocks and disposed of in geological repositories.

The MOX route has many disadvantages. Among the most serious are the following:

  • It reduces plutonium stocks extremely slowly or not at all;
  • It encourages the spread of nuclear weapons and the risk of nuclear terrorism (see above Box 3 "Primitive Device - Devastating Impact");
  • Reactors using MOX fuel are less safe than reactors using normal uranium fuel.

The use of MOX in a nuclear-power reactor cannot, therefore, be said to be a solution to the problem of excess plutonium stocks, at least for the foreseeable future. A more rational solution would be to stop reprocessing of spent nuclear fuel rods to separate out the plutonium in the first place and to store existing stocks of plutonium as safely as possible until they can be permanently disposed of. Many argue, of course, that the storage and disposal of plutonium cannot be made safe; but MOX is not the answer.

If the British government is serious about its commitment to reduce the risk of nuclear-weapon proliferation and nuclear terrorism, it should stop separating plutonium from spent reactor fuel elements by closing its reprocessing plants, stop producing MOX nuclear fuel and dispose permanently of its existing plutonium stocks.

Notes and References

1 House of Lords, Select Committee on Science and Technology, Management of Nuclear Waste, HL Paper 41, March 1999. Minimum "strategic stock" refers to the four tonnes of plutonium needed to fuel a fast breeder reactor generating 1,000 megawatts of electricity.

2 Royal Society Study Group, Management of Separated Plutonium, Royal Society, 1998.

3 The British government's reply to the House of Lords' report is expected, in the spring of the year 2000, in the form of a Green Paper (consultation paper) followed by a public consultation, leading to a White Paper (proposed changes in legislation).

4 When operating at full capacity, BNFL's Thermal Oxide Reprocessing Plant (THORP) at Sellafield can produce about seven tonnes of plutonium a year from spent nuclear fuel rods.

5 For more details, see Barnaby, F., "Nuclear Legacy: Democracy in a Plutonium Economy", Corner House Briefing 2, The Corner House, November 1997.

6 An actinide is any of a series of 15 radioactive chemical elements beginning with actinium of atomic number 89 and ending with lawrencium of atomic number 103, indicating that they have similar properties to the element actinium. Plutonium is atomic number 94 and uranium atomic number 92.

7 This is done by mixing uranium oxide and plutonium oxide powders together and feeding the mixture into an attritor mill which rubs and pulverises them together and then into a spheroidiser which, without melting the mixture, heats and fuses it before pressing it out into pellets

8 Connor, Steve, "Inspectors Sent in at Sellafield Admit to Serious Safety Lapses", The Independent, 14 September 1999.

9 Schmitz, F. and Papin, J., "High Burn-up Effects on Fuel Behaviour under Accident Conditions: The Tests, CABRI REP-Na.," J. Nuc. Materials, Vol. 270, 1999, 55-64.

10 Gouffon, A. and Merle, J. P., "Safety Problems Related to the Use of MOX assemblies in PWRs", Paper for International Working Group on Water Reactor Fuel Performance, International Atomic Energy Agency, Vienna, 1990.

11 The pellets are checked for isotopic composition; plutonium enrichment; metal content; oxide/metal ratio; impurities; gas contents; appearance (visual check); outer diameter; height; dish dimension (the punch used to produce the pellet produces a dish-like indentation at each end of the pellet); chamfer dimensions (the top and bottom surfaces of the pellet are beveled); end squareness; density; alpha-autoradiography (to identify plutonium hot spots); and solubility. According to BNFL, the dimensions of the fuel pellets are measured, some automatically by laser micrometry.


End Note

This briefing paper was written by Dr. Frank Barnaby, former director of the Stockholm International Peace Research Institute. See also Corner House Briefing 2, "Nuclear Legacy: Democracy in a Plutonium Economy", November 1997.