Keeping the Lid on Nuclear Arms

New Scientist
August 17, 1991, p. 26-30

In an historic accord last November, Argentina and Brazil agreed not to produce nuclear weapons and pledged to open their secret nuclear sites to mutual inspections. The accord. the first by two developing countries that are not signatories to the 1970 nuclear Non-Proliferation Treaty, sets a powerful precedent for other areas where the prospect of the spread of nuclear weapons is amplifying local tensions. The agreement suggests that non-NPT countries can move towards international safeguards without sacrificing national pride or autonomy. Pakistan hailed the agreement as a potential model for future negotiations with India, and Israel has shown interest in Latin American arms-control techniques as a model for the Middle East.

But however good the intentions, formidable obstacles have still to be overcome before the agreement is implemented. In particular, it is proving difficult to establish an independent inspection system, one of the most important steps in convincing the world of the two countries’ commitment after decades of clandestine operations. Neither Argentina nor Brazil has so far agreed to accept unconditionally verification on the scale required by the International Atomic Energy Agency, which is charged by the United Nations with ensuring that nuclear material for civilian purposes is not diverted to military uses. Next month the two countries will sit down with the IAEA for the fifth round of a series of negotiations to try to agree on a foolproof verification method. IAEA officials are optimistic that both countries will eventually accept thorough inspections of their nuclear facilities.

The governments of the two countries face considerable opposition from their armed forces, which want to keep the military options open. Last November’s accord, signed at the border city of Foz do Iguacu, followed a spectacular revelation the previous month by Brazil’s secretary of science and technology, Jose Goldemberg. For 15 years, he disclosed, the Brazilian military establishment had mounted a covert effort to make nuclear weapons. Called the “parallel” programme, it secretly paralleled the civilian nuclear power programme under which Brazil was openly importing technology from West Germany. Brazil’s previous governments had denied the parallel programme’s existence, even after the country returned to civilian rule in 1985 and began improving relations with its neighbour. Argentina, too, had secretly built a uranium enrichment plant suitable for a nuclear weapons programme.

Argentina and Brazil have mounted civil nu-clear power programmes built on imports, but technical snags and lack of money have plagued them both. Argentina’s two power reactors, which have a total capacity of less than 1000 megawatts, have not run reliably, and Brazil’s single 600-megawatt reactor breaks down so often it has been dubbed “the firefly”. More reactors are planned and being built, but budget cuts are forcing the two countries to reconsider their options. The Foz do Iguacu agreement appears to be part of a reappraisal of commitments to nuclear energy.

Two projects from Brazil’s parallel programme have attracted particular attention: the Aramar uranium enrichment plant at Ipero, operated by the navy, and the army’s plans to build a nuclear reactor for producing plutonium. For military use, 5 kilograms of plutonium is as effective as 15 to 20 kilo-grams of uranium, which means that plutonium bombs can be lighter or more powerful than uranium ones. According to Goldemberg, the army has scaled down the reactor from its original rating of 20 megawatts to 2 megawatts. reducing its capacity for plutonium production from one bomb’s worth, or 5 kilograms a year. to I kilogram, which is hardly enough for serious nuclear weapons manufacturing. But Brazilian critics, including Social Democrats in the legislature. say that the army should not he doing such work at all. None of the exotic uses suggested for the plutonium, including nuclear-powered satellites and buoys, and fast breeder reactors, appears feasible or relevant to Brazil’s energy or security needs.

Criticism of the Aramar plant focuses on plans to give it the capability to produce enough uranium to fuel 20 Hiroshima-sized bombs a year by 1996. Aramar will have 958 centrifuges by the end of this year, according to its director Rear Admiral Othon Luiz Pinheiro da Silva. enough to fuel about one bomb every three years. But the navy wants to spend more than $300 million to boost the plant’s output by a factor of 50 within the next five years. A budget proposal strongly supports the navy’s goal.

Aramar’s centrifuges spin uranium hexafluoride gas at high speed to separate the heavier uranium-238 isotope. which is stable, from the lighter uranium-235 isotope, which is unstable and produces the neutrons that sustain the chain reaction in a nuclear fission bomb. This enrichment increases the concentration of fissile uranium-235 from the natural level of 0.7 per cent to as much as 93 per cent, the strength generally used for nuclear weapons.

Navy officials say that Aramar’s purpose is to give Brazil the capability to produce its own supply of partially enriched uranium (less than five per cent uranium-235) to fuel power reactors. They add that the plant will also produce the fuel for Brazil’s planned nuclear-propelled submarines, which would require uranium enriched to anywhere between 7 and 93 per cent. Though the navy could buy reactor fuel more cheaply on the world market, it can plausibly claim that Brazil should aim for self-sufficiency in strategic materials.

It is less easy to excuse the navy’s desire for the submarines themselves. Since a British nuclear submarine sank the Argentine cruiser General Belgrano during the South Atlantic war of 1982, both Argentina and Brazil have viewed nuclear-powered submarines as crucial for maritime defence. But the main advantage of nuclear submarines over diesel-powered vessels is their ability to travel at sea for long periods, a feature more suited to strategic nuclear deterrence. In The Lessons of Modern War, published last year by Westview Press, American military analysts Anthony Cordesman and Abraham Wagner observe that Argentina’s single diesel submarine effectively harassed the superior British navy during the conflict.

On the Argentine side, the most threatening project is a gaseous-diffusion enrichment plant at Pilcaniyeu, near the border with Chile. The diffusion process passes uranium hexafluoride gas through porous membranes that are more readily penetrated by uranium-235 than by uranium-238. Diffusion is technically simpler than enrichment using centrifuges but requires more energy. Operating at its projected capacity, the Pilcaniyeu plant would be capable of fuelling about four nuclear weapons per year. Budget cuts seemed to have delayed its completion, however.

Argentina’s other sensitive project, a pilot-scale plutonium extraction plant at Ezeiza in Buenos Aires, was put on indefinite hold in March 1990, also for lack of funds. Its planned capacity would allow Argentina to extract 15 kilograms of plutonium per year—enough for two to three bombs—from the spent fuel produced by the German-supplied Atucha-I power reactor north-west of Buenos Aires, and the Canadian-supplied Embalse power reactor at Rio Tercero in the centre of the country. Germany or Canada would have to approve the extraction, however, because both countries reserved this right in their sales contracts.

The November accord thus came at a critical moment. Simply by spending more money on existing programmes, both Brazil and Argentina could soon have begun producing material suitable for building a nuclear weapons arsenal.

The two countries have also been willing to sell dangerous technology to other developing nations. From 1984 to 1990, Argentina sponsored the Condor II medium-range missile project jointly with Egypt and Iraq. using a network of companies established in Europe to provide the necessary equipment. Pressure from the US forced Argentina to withdraw from the project early last year. Similarly, international pressure forced a team of Brazilian engineers to leave Iraq late last year. during the Gulf crisis. The Brazilians had been teaching Iraqi rocket scientists how to extend the range of their Scud missiles and how to build a rocket launcher to put satellites in space.

Brazil also seems to have been involved in the development of Iraq’s nuclear weapons programme. Last year, two former Brazilian government ministers told the national newspaper Jornal do Brasil that Brazil sold natural uranium to Iraq in the early 1980s. helped Iraq prospect for uranium de-posits. analysed samples of Iraqi uranium, supplied nuclear material and equipment for laboratory tests. and designed an Buenos An’ Aires underground plant to produce the uranium dioxide used to make uranium hexatluoride for enrichment.

It is even possible that Brazil taught Iraqi engineers how to run centrifuges. Several Iraqi technical teams visited Brazil during the 1980s, and at least one had access to the Aramar enrichment plant. In 1990, the contract manager of the Brazilian firm that designed Iraq’s uranium dioxide plant told Jornal do Brasil: “It was always in our minds that the real objective of the Iraqi nuclear programme was to construct an atomic bomb.” Although neither Brazil nor Argentina is a party to the NPT, and so has not flouted an international commitment, Fernando Collor de Mello, president of Brazil, apologised late last year for the technology transfer that he described as “potentially significant”. Iraq is a signatory of the NPT, though for political and commerical reasons its suspected development of a nuclear weapons programme went unchallenged until the invasion of Kuwait.

Developing friendship

The November accord between Argentina and Brazil follows a decade of improving relations between the two countries—after years of competition for political and economic influence in South America. An agreement in 1979 on development and border issues in the Plata basin opened the way to a series of nuclear agreements in the 1980s. First came technical cooperation, then reciprocal ceremonial visits to sensitive plants and, finally, the beginnings of a mutual inspection system. According to John Redick, a specialist in Latin American nuclear politics at the University of Virginia: “The Foz, do Iguacu agreement only makes sense in the context of broader economic cooperation. Without the economic ties, the nuclear agreement doesn’t mean a lot.”

Argentina and Brazil have pledged to set up a common accounting and control system for nuclear material: exchange lists of all their plants and materials: submit their accounting system to the IAEA: and permit reciprocal inspections. The first three steps were carried out by mid-December, says Goldemberg. The two countries also promised to allow the IAEA to inspect their nuclear sites. In addition, they agreed to conform to the terms of the Treaty of Tlatelolco, which bans nuclear weapons from Latin America and has been signed and ratified by every other country in the region except Cuba and Chile.

According to the IAEA, it is still unclear exactly what kind of inspection the two countries will propose. or what degree of intrusion they will accept. The most open and comprehensive form of inspection system, enshrined in the NPT and the Treaty of Tlatelolco, encompasses what the IAEA calls “full-scope safeguards”, which entail monitoring all nuclear facilities and materials, whether imported or domestic. But not even all parties to the NPT accept such a degree of intrusion because the treaty operates a two-tier system that distinguishes between countries that have nuclear weapons and those that do not. The full-scope safe-guards are applied only to countries that do not already have nuclear weapons, a distinction that Argentina and Brazil have rejected as discriminatory.

In December last year, a Brazilian congressional committee investigating the implications of the ac-cord voted against allowing the IAEA into Brazil’s ground nuclear research sites. The committee, with the support of the military establishment, wants the monitoring done by an internal commission responsible only to the Brazilian congress. In response, Goldemberg has eschewed the term “full-scope safeguards” and called instead for “very comprehensive” inspections. In Argentina, some officials are reluctant to grant access to all parts of the Pilcaniyeu enrichment plant. They say they want to protect technical information.

If the IAEA is allowed only to review the results of a bilateral accounting system, without being able to make its own independent inspection, there can be no confidence that all nuclear material will be accounted for. For a start, the two countries have no plans to use seals, cameras, or other electronic detection methods to prevent tampering between visits, as the IAEA does (“Arms and the ban”, New Scientist, 29 July 1989). Bilateral inspections would also be vulnerable to simple changes of heart: either country could stop cooperating at any time, either for military reasons or because of friction between them. An agreement with the IAEA, however, would be an international obligation. Breaching the agreement would expose the non-cooperating country to censure by the UN.

Since May, however, the IAEA has been discussing a “comprehensive” safeguards agreement with the two countries and says it expects full-scope standards to be accepted eventually. But before any kind of inspection can begin, it will be necessary to account for the nuclear material each country has already produced, starting with a beginning balance. The last point is crucial, because both countries have been running secret plants with unknown outputs. To calculate what Pilcaniyeu and Aramar have already made, inspectors will need records such as the history of each plant’s production levels, the raw materials used, energy consumption and waste produced.

Clandestine imports will also have to be disclosed. For in-stance, in the early 1980s, China secretly sold Argentina about 70 tonnes of heavy water without imposing any restrictions on how the material could be used. Heavy water (deuterium oxide) is used to sustain a chain reaction in reactors that use natural rather than enriched uranium fuel. China also sold Brazil about 200 kilograms of uranium enriched to a concentration of 3, 7 and 20 per cent uranium-235, again without imposing any restrictions on its use.

Most of these inspection questions would be solved if both countries conformed to the Treaty of Tlatelolco, which they have signed but not implemented in full. But even the treaty has some loopholes, such as allowing “peaceful nuclear explosions”. Most parties to the treaty regard such explosions as indistinguishable from nuclear weapon tests and therefore barred, but until late last year Argentina and Brazil were not prepared to go along with this view.

Another loophole in the treaty is that, in defining nuclear weapons, it states that an instrument “used for the transport or propulsion of [a nuclear weapon] is not included . . . if it is separable from the device”. This could be interpreted as exempting submarine reactor fuel from inspection. The NPT has a similar loophole: it allows enriched uranium to be withdrawn from inspection for “non-proscribed military purposes”, which has been interpreted to include submarine fuel. Goldemberg has promised that submarine fuel will be included in inspections, but the Brazilian navy may resist this to maintain politically sanctioned access to enriched uranium.

Pressure from the countries that supply Argentina and Brazil with their nuclear equipment and materials helped to forge the November accord. Germany, a key nuclear supplier to both countries, announced earlier in 1990 that it would stop exporting to countries that did not accept full-scope safeguards. Canada, another important supplier, has had such a policy for years. And the United States was holding up the sale of a much desired IBM supercomputer to Brazil because of reports in the international press that Brazil was helping Iraq develop ballistic missiles.

Some US government officials say privately that they regard both Collor and Carlos Menem, president of Argentina, as sincere in their plans to renounce nuclear weapons. The questions are whether the two presidents can carry out this pledge, and whether the Foz do Iguacu agreement will out-last the governments that signed it. Redick says that the Brazilian air force includes “a maverick group that planned to go the Indian route and test a bomb in a supposedly peaceful nuclear explosion”.

A prudent course for supplier countries would be to reward Argentina and Brazil with non-nuclear benefits in the short term, but to withhold the more sensitive exports until the two countries put all nuclear facilities under civilian control, reveal their entire nuclear inventories, set up a thorough inspection system, and pledge publicly to stop making dangerous exports. Canada has set the tone: despite its desire to sell Argentina a fourth power reactor, the Canadian government has said that its reaction to the Foz do Iguacu agreement will depend on “the nature and scope of agreements to be negotiated with the IAEA and any changes proposed for the Tlatelolco Treaty”.

Unfortunately, the US has not followed Canada’s lead. President George Bush approved the export of the IBM supercomputer in December, overriding concern in the Arms Control and Disarmament Agency, the Department of Energy and Congress that the machine would boost Brazil’s ability to design nuclear missiles. To complicate matters, Brazil then refused to guarantee that it would not use the computer for military projects or re-export the machine to another country.

In arms control as elsewhere, actions speak louder than words. Full “nuclear transparency” and a thorough, reliable inspection system are musts for international confidence in the November accord. Implemented properly, the Foz de Iguacu agreement could reach around the world.

Gary Milhollin is professor of law at the University of Wisconsin and director of the Wisconsin Project on Nuclear Arms Control, Washington DC. Jennifer Weeksa former researcher with the Wisconsin Project, is now military policy analyst with the Arms Control and Foreign Policy Caucus of the US Congress.