The proliferation of nuclear weapons to countries that do not now have them is a crucial international issue. Currently, there are five established nuclear-weapon powers: the USA, Russia, the UK, France and China. These powers regularly test nuclear weapons, and we have significant information about the numbers, types and technical characteristics of their nuclear arsenals. Israel is known to have nuclear weapons, although the Israeli government has consistently refused to confirm this.
The total number of nuclear weapons in the world's arsenals is about 50,000. The ex-Soviet arsenal has about 27,000; the USA has about 20,000; France has about 600; the UK has about 400; China has about 250; and Israel has about 150 (Norris et al. 1991). A number of other countries are on the threshold of becoming nuclear-weapon powers.
India tested a nuclear device in 1974, demonstrating its capability to produce a nuclear force, but the Indian government claims that it has no stockpile of nuclear weapons. Pakistan is generally believed to be on the threshold of producing nuclear weapons and may have already done so. Argentina and Brazil are believed capable of producing nuclear weapons, but both countries have apparently moved away from a political decision to do so.
Other Third World countries who could, if they take the political decision to do so, produce nuclear weapons in a relatively short time include Taiwan, South Korea and possibly North Korea. Iraq was suspected of developing nuclear weapons before the Gulf War, and Iran is also suspected of having ambitions to produce a nuclear force.
Any country with a peaceful nuclear-power programme has the plutonium and the expertise to produce nuclear weapons. The peaceful atom and the military atom are intimately linked - 'Siamese twins', in
49
the words of one Nobel Prizewinning nuclear physicist. The spread of nuclear-power reactors is, therefore, of fundamental importance to any consideration of the spread of nuclear weapons.
Between President Eisenhower's Atoms for Peace programme in the mid-1950s and the early 1980s the nuclear industry expanded rapidly. The oil-price increase in 1973 was a particular boost for it. A number of important countries like France and Japan became intent on reducing their dependence on oil imports by installing nuclear-power reactors. Another boost was the concern over the contribution to the greenhouse effect, and hence to global warming, from the atmospheric pollution produced by the burning of fossil fuels in power stations. Nuclear electricity was promoted as being environmentally friendly.
During the 1980s, however, the nuclear-power industry suffered a series of setbacks. The first was the realization that nuclear electricity was relatively very expensive. And in 1986 came the Chernobyl nuclear accident. Reactor safety became a second major challenge to the nuclear-power industry. The difficulty of finding a politically and publicly acceptable solution to the problem of the disposal of high-level radioactive waste and concern about the health effects of low-level radiation added to doubts about nuclear power.
In 1970, the world's nuclear-power reactors were generating a total of about 20 giga-watts of electricity (1 giga-watt of electricity, or GWe, is 1,000 million watts of electricity). Five years later, this total had about quadrupled, to 75 GWe in 1975. It took another fifteen years for the total to quadruple again. Today, the total world's nuclear generating capacity is 326 GWe, generated by 423 power reactors in twenty-five countries - the USA, France, the Commonwealth of Independent States, Japan, Germany, Canada, the UK, Sweden, South Korea, and Spain (IAEA 1991). These top ten account for 91 per cent of the total. The other countries operating nuclear-power reactors are Argentina, Belgium, Brazil, Bulgaria, Czechoslovakia, Finland, Hungary, India, Mexico, the Netherlands, Pakistan, South Africa, Switzerland, Taiwan and Yugoslavia. Four other countries - China, Cuba, Iran and Romania - which now have no nuclear power are constructing power reactors.
In the year 2000 the world's nuclear generating capacity will be about 400 GWe. There are at present eighty-three nuclear-power reactors under construction. When completed, these will add 66 GWe to the world's nuclear capacity.
As the uranium fuel elements in a power reactor are burned to produce heat from the fission process (which is used to produce steam from boiling water to run turbines to generate electricity) fission products
50
and plutonium are produced. Spent reactor fuel elements are so radioactive, because of the fission products in them, that they are self-protecting. But if the plutonium is removed from the fuel elements and separated from the fission products, in a chemical reprocessing plant, it is in a form that can be relatively easily handled. As the amount of plutonium produced worldwide in civilian nuclear-power reactors and separated from spent reactor fuel elements in commercial reprocessing plants increases it will become easier for governments (and sub-national groups) to obtain it illegally. Whether or not to reprocess is, therefore, a crucial political decision.
The world's civilian nuclear-power reactors have so far produced about 700 tonnes of plutonium, and are producing about 75 tonnes a year. By the year 2000, the world's civilian reactors will have produced a total of about 1,700 tonnes of plutonium and will be producing some 100 tonnes a year (Albright and Feiveson 1991).
About 90 tonnes of civilian plutonium have so far been chemically separated from spent reactor fuel elements in reprocessing plants outside the Commonwealth of Independent States. By 2000, according to current plans, some 300 tonnes will have been separated. For comparison, the amount of military plutonium in the world's nuclear arsenal is about 220 tons. By 2000, large commercial reprocessing plants will be operating in the UK, France, Russia and Japan.
Plutonium is produced as an inevitable by-product in all nuclear reactors. But the isotopic composition of the plutonium produced in reactors operated for different purposes varies. The plutonium produced specifically for military purposes is rich in the isotope plutonium-239. 'Weapons-grade' plutonium typically contains more than 93 per cent of the isotope plutonium-239 and is the material which produces the most efficient nuclear weapons. Plutonium produced in nuclear-power reactors operated to produce electricity in the most economical way, known as reactor-grade, typically contains only about 60 per cent plutonium-239. About 25 per cent is plutonium-240 (in weapons-grade plutonium the amount is typically about 7 per cent) and about 10 per cent is plutonium-241.
Can this reactor-grade plutonium be used to produce nuclear explosions? This is an important question because, if it can, countries operating nuclear-power reactors for peaceful purposes, particularly electricity production, have access to plutonium that could be used to produce nuclear weapons. And, as the quantity of reactor-grade plutonium in the world increases, it becomes easier for a country to acquire it illegally - on a plutonium black market, for example - and produce nuclear weapons. That reactor-grade plutonium can be used to produce a nuclear weapon has been shown in the USA, where at least two such devices have been built and tested.
51
The critical mass of typical reactor-grade plutonium in the form of a bare metal sphere surrounded by a natural uranium reflector, about 10 centimetres thick, is about 7 kilograms. Reactor-grade plutonium is usually stored, after reprocessing, in the form of plutonium oxide and is, therefore, most likely to be available in this form. The oxide can, however, be easily converted to the metal form.
The removal of plutonium from reactor fuel elements is a relatively straightforward chemical process. If operated commercially, for a profit, a plutonium-reprocessing plant is a complex and costly chemical establishment and, because the capital cost is relatively independent of the size of the plant, economic reprocessing can only be achieved if a large-scale plant is used to serve many reactors. A typical modern commercial reprocessing plant will reprocess about 1,200 tons of spent reactor fuel a year, containing about 12 tons of plutonium, and service about forty modern nuclear-power reactors.
To obtain plutonium for military purposes, where money is no object, reprocessing can easily be done on a small scale. In fact, a country could obtain plutonium clandestinely from a nuclear reactor acquired especially for the purpose. The components for a small reactor, capable of producing enough plutonium to make a few nuclear weapons a year, can be easily, and secretly, obtained on the open market for roughly $30 million (about the same cost as a modern fighter aircraft).
The reactor and a small reprocessing facility to remove the plutonium from the reactor fuel elements could be clandestinely constructed and operated. These units, and room to design and construct nuclear weapons from the plutonium, could be effectively disguised and hidden in a building or underground. Israel used a clandestine reactor and reprocessing facility to produce plutonium for its nuclear force.
Some countries choose enriched uranium rather than plutonium as the fissile material for nuclear weapons. Three methods are available for enriching uranium: the gas diffusion, the gas centrifuge and the jet nozzle techniques. The technique mainly used so far is the gas diffusion method. But, given the materials now available, particularly carbon fibre, countries wishing to enrich uranium would probably opt for the gas centrifuge method. Pakistan is using this route to produce fissile material for nuclear weapons, and Iraq was in the process of doing so before the Gulf War.
A gas centrifuge for uranium enrichment consists of a vacuum tank containing a long rotating drum with a nozzle at one end and an orifice at the other. Uranium hexafluoride gas is pumped in via the nozzle and, as the gas moves up inside the rotating drum, molecules
52
of the uranium hexafluoride gas will tend to be flung outwards by the centrifugal force. Molecules of uranium hexafluoride gas in which the uranium is uranium-238 are slightly heavier than molecules of uranium-235 hexafluoride. There will, therefore, be a difference in the centrifugal force acting on the molecules of different masses when the gas is rotated at very high speed. Molecules of the lighter uranium-235 isotope will diffuse towards the centre. The inner portion thus becomes enriched in uranium-235, and this is collected at the exit orifice.
A plant containing many gas centrifuges in a cascade is needed to enrich a useful quantity of uranium. The slightly enriched flow of uranium gas from the first centrifuge is fed into the nozzle of the next centrifuge in the cascade and so on. The uranium is circulated around the cascade until the desired degree of enrichment is obtained. A gas-centrifuge plant for a nuclear-weapon programme could, like a military plutonium facility, be constructed clandestinely.
Because the basic design of nuclear weapons using nuclear fission is now well known and the nuclear data needed are in the open literature, and because the facilities to produce plutonium and enriched uranium for nuclear weapons can be built and operated secretly and simply, we do not know for sure which countries have nuclear weapons bombs and which do not. A nuclear-weapon programme could be kept secret until a nuclear test was performed.
A nuclear test removes all ambiguity about a country's nuclear-weapon programme. This is, of course, why any country wishing to hide nuclear-weapon developments would want to avoid testing.
An important question is whether or not the military and political leaders would be prepared to accept nuclear weapons unless and until the designs have been tested with full-scale nuclear tests. In particular, the military may require to know the precise explosive power of any weapons under its control and may demand tests to check that estimated yields can, in practice, be achieved within relatively narrow limits.
So far as today's nuclear-weapon designers are concerned, they will probably be so confident that they could design and construct ordinary nuclear weapons using implosion (i.e., a Nagasaki-type design) that they would not need a full-scale test - at least, when weapons-grade fissile material is used. Nor would testing be requested if enriched uranium is used in a gun-type design. Testing is, however, likely to be called for if reactor-grade plutonium were used.
The designers would probably be sure that the weapons would produce explosive yields within their predicted range. Provided that
53
the fissile material used was weapons-grade uranium or plutonium, the designers could predict the explosive power rather precisely, within a narrow range. Also, the scientists and engineers who built the weapons would be confident that they would explode according to the design.
Whether or not the military will take the word of the nuclear scientists and engineers that the nuclear weapons they design and build will work reliably according to their predictions will depend on the attitude of the military to science and technology. If a significant fraction of the senior military officers are technically minded, it is likely that the military will accept the word of the scientists.
But nuclear weapons that include an element of nuclear fusion are a different matter. The designers of boosted nuclear weapons and of full-scale thermonuclear weapons will want to test them. Even today, the design of a thermonuclear weapon is a very complex matter.
The test of a thermonuclear weapon need not involve testing the entire assembly at full explosive power. It would normally be enough to test the fission trigger plus a small section of the fusion component to test that the fusion process was set off. The yield of the test may be relatively low. If the scaled-down device produces some fusion, it can be assumed that the full-scale weapon will work effectively.
To hide a nuclear explosion with an explosive power greater than, say, 10 kt is a difficult task, even if it is set off deep below ground. Such an explosion can normally be detected by seismic monitoring equipment operated outside the country in which the nuclear explosion takes place. The Indian explosion in May 1974 of a 12 kt nuclear device was, for example, easily detected. And the Indian explosion, set off at a depth of about 100 metres underground in the Rajasthan desert, produced a crater in the ground of 150 metres in diameter which was easily observed by reconnaissance satellites.
The key components of a nuclear weapon could be tested without a full-scale nuclear test. Pakistan is said to have done so. The design of a fission weapon could be tested by, for example, using only a small sphere of fissile material at the centre of the conventional chemical explosive lens system. The sphere could be sufficiently small that, when the core is exploded, the nuclear fission yield produces an explosion of a power about the same as that of the conventional explosive (i.e., 200 kilograms or less). This amount of fission would be sufficient to be detected by radiation detectors placed around the test assembly.
The detection of a burst of radiation, particularly neutrons, would show that an effective implosion has taken place. But the explosion would be insufficiently powerful to be detected by seismic monitors outside the country concerned.
54