Testimony of Gary Milhollin
Director, Wisconsin Project on Nuclear Arms Control
Before the House Committee on Foreign Affairs,
Subcommittee on International Security and Scientific Affairs
And on International Economic Policy and Trade
June 12, 1985
I am grateful for the opportunity to be here today, and to testify before these two Subcommittees on the question of plutonium use in the world.
Plutonium is the preferred material for making atomic bombs. It is in most of the warheads in the U.S. stockpile, and made the world’s first atomic explosion in 1945. Twelve pounds of it destroyed the city of Nagasaki. To make a bomb with plutonium, one places a sphere of it in the center of a sphere of high explosives. Once the high explosive is ready, it doesn’t take long to insert the plutonium — a few days will do. So very little time exists between the moment a competent weapons maker — or someone working with him — steals or diverts plutonium and the moment such a person has an atomic bomb. This fact — the quick convertibility of plutonium into atomic bombs — makes plutonium unique as a world commodity.
But is plutonium a world commodity? Should it be one? That is the broad issue before this hearing. The Ford and Carter administrations tried to discourage the use and availability of plutonium. However, some of the closest U.S. allies and trading partners wanted — and still want — to create plutonium in large quantities for commercial use. They wish to extract plutonium from spent nuclear reactor fuel, and then use it or sell it to others as new reactor fuel. If they do, it will mean that weapons-useable plutonium will be created in large quantities, shipped in large quantities, and stored in large quantities, and that all of this will happen in several foreign countries which do not now have nuclear weapons. Tons of plutonium will be involved. It will also mean that this type of commerce will become a precedent among the advanced countries, which the developing countries will want to follow.
Since the Ford and Carter days events have moved along, France made one shipment of plutonium to Japan in 1984. It now wants to begin shipments to Switzerland, and is considering the general sale of plutonium to fuel the present generation of power reactors. For the United States, these developments pose two questions. The first is simply one of security. The national security of the United States, and the individual security of each person who lives here, is placed at risk by the diversion of plutonium anywhere in the world. The second question is whether the United States should oppose this spread of plutonium, or simply go along with it. The United States still controls a large portion of the spent reactor fuel in the world — from which this plutonium is sought to be made — and has important influence over the very allies and trading partners which wish to market plutonium.If the United States wanted to, it could stop this spread of plutonium from happening.
I have been asked to comment today upon three specific questions. First, whether the United States would receive timely warning if U.S.-origin plutonium were stolen or diverted somewhere in the world. A corollary of that question is whether the Reagan administration’s interpretation of timely warning is consistent with the Nuclear Non-Proliferation Act of 1978. Second, I have been asked to comment upon the proposed transfer of U.S.-origin plutonium from France to Switzerland. The Department of State approves of this transfer; the Department of Defense opposes it. Third, I have been asked to comment upon the practice — which is now United States policy — of approving the use of plutonium in certain countries, which do not appear to be a proliferation risk, and of opposing its use in other countries, which do appear to be a proliferation risk. I will take these questions up in order.
The concept of “timely warning” is central to any use of U.S.-origin plutonium. When the United States exports fresh nuclear reactor fuel — in the form of low-enriched uranium, which is not useable in weapons — the recipient must promise not to retransfer the fuel beyond its jurisdiction or reprocess it without United States consent. Reprocessing consists of chopping up the spent fuel, dissolving it in acid, and extracting (separating) its plutonium in pure form. Plutonium in this form can be put directly into a weapon. Before reprocessing, a country having spent reactor fuel is still several months away from being able to make a nuclear weapon. If it has not built a facility in which to do the reprocessing, it is years away. After reprocessing, it is days or weeks away. The difference is crucial, and is the reason for requiring U.S. consent. When a recipient requests U.S. consent to either reprocess or retransfer, the request must be judged under the criteria in Section 131 of the Atomic Energy Act (Section 303 of the NNPA). To approve the request, the Secretaries of State and Energy must find that granting it “will not result in a significant increase of the risk of proliferation ….” When making this finding, the “foremost consideration” is whether the retransfer or reprocessing “will take place under conditions that will ensure timely warning to the United States of any diversion„..” To be timely, the warning must be “well in advance of the time at which the non-nuclear-weapon state could transform the diverted material into a nuclear explosive device….”
These words describe the basic goal of timely warning. It is simply this: that plutonium must be held and processed in such a way that, if a diversion of it occurs, the United States will be warned of the diversion well before the plutonium can be made into a weapon. The warning must be “well in advance” of the moment of possible fabrication. Why “well in advance”? Because it has always been assumed that the United States would, upon learning of a diversion, intervene diplomatically or by force to prevent the diverter from achieving a weapon. If the diverter achieves one or several weapons before the United States can react, the reaction becomes far more difficult. This “response time” has always been thought essential to national security.
With respect to plutonium, there can be timely warning in some cases, but not in others. When a foreign country holds spent fuel, the plutonium in it cannot be used in a weapon without reprocessing, which takes time. Months at least. That is time enough for a response, if a diversion is detected quickly. Spent fuel rods are visible in storage, can be counted easily, and are not simple to move. If spent fuel is diverted, one can reasonably hope for warning in time to react. Separated plutonium is quite different. First, it cannot be measured accurately when handled in bulk. The limit of accuracy in measuring the plutonium in a reprocessing plant is 1%. At Tarapur, where India hopes to reprocess two tons of U.S. plutonium, this measurement error means that 20 kilograms could be diverted without detection. Twenty kilograms can make up to 5 fission bombs with reflectors, depending upon the purity of the plutonium. The second problem with separated plutonium is its short conversion time. That time is from seven to ten days if the plutonium is finished metal, and from one to three weeks if the plutonium is an oxide. The International Atomic Energy Agency, which is responsible for safeguarding plutonium, will not give timely warning of its diversion. The IAEA cannot, and does not promise to give timely warning. In the words of its legal counsel, “it is a misconception.. .that it is the prescribed task of the agency to deliver timely warning. That is not the task under NPT safeguards.”
If the IAEA does not deliver timely warning, what does it deliver? It delivers “timely detection.” The IAEA promises only to detect a diversion within the conversion time of the material being safeguarded. That means, for separated plutonium, that the IAEA will inspect it every seven to ten days if it is in metallic form and every two to three weeks if it is in oxide form. The IAEA would not notify the United States, or anyone else, of a diversion within that time. Notice would come only after a report of material unaccounted for had made its way up through IAEA channels. At the IAEA, a report of material unaccounted for requires a period of time for evaluation, a period to report the diversion to the Director General, and a period for him to report it to the Board of Governors, who would then meet and decide: a) whether to ask the country concerned to remedy the discrepancy, b) whether to report the discrepancy to all members of the Agency, or c) whether to report it to the United Nations. It is obvious that this will take longer than one to three weeks.
So, for separated plutonium, IAEA safeguards will not provide timely warning. How then, did the retransfer of separated U.S. plutonium to Japan satisfy the Atomic Energy Act? The Japanese transfer was the first major shipment of separated plutonium to a non-weapons state since passage of the NNPA. The shipment was approved because the Act allows factors other than timely warning be considered in deciding whether the retransfer causes “a significant increase of the risk of proliferation…”
The other factors are listed in the Senate Report on the NNPA. They are whether the recipient is committed to non-proliferation, whether the recipient has a security agreement with the United States, whether it has a stable government, whether it is militarily secure, and whether it has a need for nuclear energy. The Senate Report also states that if it is not clear that timely warning will be provided, “a strong combination of other factors is necessary to compensate for this weakness in safeguards.”
The Japanese case has raised a dispute over what timely warning means. The Departments of State and Energy took the position that timely warning existed in that case. They said that timely warning could be based upon Japan’s governmental system, its non-proliferation policies, and its security relationship to the United States, rather than simply upon the scientific facts determining the conversion time of separated plutonium. The implication is that Japan’s stable government, its avowed policy against nuclear weapons, and its close relation to the United States Would all have to change in order for Japan to divert plutonium. These changes would be noticed by the United States, so that there would then be timely warning that a diversion might occur. This position was rejected by the Nuclear Regulatory Commission. The NRC pointed out that under State and Energy’s interpretation, “it would be necessary to consider the same non-technical factors both in connection with the timely warning analysis and in connection with the overall increase in the risk of proliferation’ finding.” The NRC said that Congress intended for timely warning to be “essentially a technical matter involving such factors as safeguards measures applied to the material and the technical ease of incorporating the material into a nuclear explosive device.”
It is pretty clear that the NRC is right. When the NNPA was adopted, all the Congressional discussions of timely warning — in committee reports, in markups, and floor debates — assumed that timely warning was purely a matter of scientific fact. That is, that the only factors in timely warning were whether a diversion would be detected, and whether it would be reported in time to intervene. A thorough study of the legislative history on this point has recently been done by Leonard Weiss, who is Minority Staff Director of the Senate Subcommittee on Energy, Nuclear Proliferation, and Governmental Processes. Mr. Weiss played a key role in framing the timely warning language. His study shows, beyond any doubt, that NRC’s view is the one Congress intended. The NRC view is also consistent with logic and common sense. Why would Congress provide that political factors could be considered in addition to timely warning, if Congress intended for those very same political factors to determine whether timely warning existed? ‘Congress could not have intended for the political factors to be considered twice. What Congress did intend was for the executive branch to make a separate finding on timely warning,
distinct from the political factors, to weigh the political factors along with timely warning to determine the risk of proliferation, and to give timely warning ” foremost consideration.” It is manifestly impossible to give it “foremost consideration” unless it is different from the other factors to which it is being compared. By not making a separate finding on timely warning, the executive branch has reduced the importance of timely warning, and made it impossible to give it foremost consideration. There is also another, less obvious reason why the executive branch cannot be right. If one follows the scenario which the executive branch poses, Japan could not, because of the stability and openness of government, divert plutonium without first changing its policy on non-proliferation, or its security alliance with the United States. Such changes, the argument goes, would be overt and the United States would learn of them. But suppose Japan did change one, or all of these factors. What would the United States do? Would the United States deem itself to be timely warned, and intervene at that point to get its plutonium back? The answer is that it would not, because it could not. There is absolutely no link between any of these political factors and the U.S. right to retrieve its plutonium. The only way the United States can legally demand the return of its plutonium from Japan is if Japan violates safeguards, or otherwise breaches the U.S.-Japan agreement for cooperation. In other words, if Japan diverts the plutonium. The U.S.-Japan nuclear trade agreement does not require Japan to have a particular form of government, be enthusiastic about non-proliferation, or maintain a security relation with the United States. Japan can change all of these without affecting in any way its right to keep the plutonium. It only gives up that right when it breaches safeguards by diverting. Because the right to intervene is triggered only by diversion, the previously-occurring political factors are irrelevant. The only important factors — those which affect intervention time and thus timely warning — are the rapidity of detection (safeguards adequacy) and the time it will take the diverter to make a weapon. Congress understood this very well when it enacted the NNPA. It is only the Departments of State and Energy which do not seem to understand it.
The remedy for this is to amend Section 131. Congress should do this: first, require the NRC to make the finding on timely warning, and to base its finding simply upon the factors Congress originally intended: that is, the adequacy of safeguards and the conversion time of the material in question. This is the definition now used by the NRC and the one Congress intended when it passed the NNPA. If the NRC is not able to find that timely warning would be provided, then the request would have to be denied under current law unless the absence of timely warning were outweighed by the political factors. When the case reaches this point, the question changes from science to politics. A non-weapons state will receive weapons-ready material by U.S. export, there will not be timely warning of a diversion, and the approval depends upon political factors. It could be that the approval should be granted. But, in a democracy, political questions such as this are important enough for public involvement and public debate. They should not simply be left to the experts.
Congress should require that in any retransfer or reprocessing case in which NRC finds that timely warning will not be given, Congressional approval is necessary before the retransfer or reprocessing is allowed to occur. Congressional approval will insure that the political factors are properly weighed, and that timely warning is given its true importance. It will take some time to do this, but that time will be more than justified by the importance of timely warning to national security. Once material is exported without provision for timely warning, it effectively goes beyond U.S. reach. It is worth taking some time to insure that such a risk is worthwhile.
This same analysis applies to the Swiss retransfer. About two hundred kilograms of U.S.-origin plutonium are involved. Because the plutonium is separated, the IAEA will not be able to give timely warning of its diversion. The approval depends upon political factors. What are they? Switzerland has a stable government. It has signed the Non-Proliferation Treaty. But it is not an ally of the United States and has no security relationship with the United States. It may or may not be militarily secure (is anyone militarily secure?); it may or may not need nuclear energy (the United States no longer needs it enough to order new reactors). One could argue over these factors for a long time.
The real importance of the Swiss case lies elsewhere. It resides in two facts. First, that Switzerland has one of the worst records on nuclear explorts of any country in the world. And second, that by exporting separated plutonium to Switzerland, the United States will be abandoning the line — which it has drawn for four years — beyond which U.S. separated plutonium was not supposed to go.
The record on Swiss exports is dismal. One need only look at the facilities in the world which now pose the greatest threat of proliferation. First, there is Pakistan’s unsafeguarded enrichment plant at Kahuta. Over United States objections, Switzerland supplied a giant, specially built gasification and solidification unit for that plant, together with high-vacuum values. Switzerland could have been under no illusions about that export. Second, there is the unsafeguarded enrichment plant in Argentina, built in secret for five years at Pilcaniyeu. There is strong evidence that Swiss equipment was used in that plant, and therefore that the Swiss knew the plant was being constructed. There could have been no illusions about that export either. In addition, the Swiss sold Argentina a heavy water plant in 1980 without requiring full-scope safeguards, again over U.S. protests. Third, there is the unsafeguarded enrichment plant in South Africa, at Valindaba. The Swiss sold compressors for that plant. Also, in 1984, the United States was forced to hold up action on the current retransfer request because of reports that Switzerland planned to sell a heavy water plant to South Africa. There would have been no peaceful use for such a plant in South Africa’s program. This is Switzerland’s export record, or at least the part of it which is public. Because of it, there is Swiss equipment in most of the dangerous facilities now operating in the world. Also because of it, Switzerland does not have “good non-proliferation credentials.” I wish to emphasize that these examples are the ones which are public. There are others which are not public. Before any U.S.-origin plutonium goes to Switzerland, these subcommittees should ask for a presentation of the classified part of these exports. The Subcommittees are likely to find things which are even more revealing than the ones I have mentioned.
The second problem with the Swiss case is one of precedent. When the Reagan administration entered office, the U.S. policy on plutonium changed. The administration said it would not “…inhibit reprocessing.. .in nations with advanced nuclear power programs where it does not constitute a proliferation risk…” The policy is frankly discriminatory and assumes that discrimination among countries makes sense when countries are truly different in their relations to the United States, and in their commitment to nuclear non-proliferation. There is nothing wrong with the policy in theory; the question is whether it can work in practice. Before the Swiss case, the administration had drawn a line around Euratom and Japan for the purpose of deciding who got separated plutonium from the United States. In the Swiss case, we go over that line. Switzerland has signed the Non-Proliferation Treaty. But it is not a United States ally, as Japan and the Euratom countries are, and has no security relationship with the United States. Nor does it have “good non-proliferation credentials.” South Korea has signed the Non-Proliferation Treaty, is an ally of the United States, has a large nuclear program, and has been important enough for United States blood to be shed in its defense. Taiwan has signed the Non-Proliferation Treaty, has a larger nuclear program than Switzerland’s, and has long had a security relationship with the United States. Yet, the United States vigorously suppressed the efforts of both Korea and Taiwan to acquire separated plutonium. Moreover, it is still not the policy of the United States to allow these countries to get separated plutonium. Finally, there is Mexico, our close neighbor and good friend to the south. Mexico has great nuclear ambitions and a much better record on non-proliferation than Switzerland. Is the United States going to furnish Mexico with separated plutonium when Mexico reaches the point of requesting it? If not, why not? Are Mexico, South Korea, and Taiwan less trustworthy, or less valuable friends of the United States, than Switzerland? The problem with the Swiss case is that for four years we have had a clear line marking off the countries which could get separated plutonium from the United States and now that line is gone. So, where is the line? A policy of discrimination requires the ability to draw a line. A line which moves is not a line, and a line which no one can define is not a line either.
The solution to the Swiss case is suggested by the issue it poses. That issue is whether a policy of discrimination can really work in practice. A U.S. decision that Switzerland is more trustworthy, and should receive more favorable U.S. treatment, than South Korea, Taiwan and Mexico is basically insulting to South Korea, Taiwan and Mexico. It does not seem possible to formulate any convincing principle according to which Switzerland is on the “good” side of a line and these other countries are on the “bad” side.
The remedy is for Congress to put a hold on the retransfer until there is a clear, defensible, and operational line beyond Euratom and Japan. Ad hoc treatment of each country is not a policy. A Congressional hold would have a great many benefits. First, it would allow a review of Switzerland’s export record, both public and classified, to see what Switzerland’s “non-proliferation credentials” really are. Second, it would allow Congress to obtain a statement of where the administration now draws the line on separated plutonium. Such a line is essential if U.S. policy is to be taken seriously by other countries. Third, it would allow Congress to explore what Switzerland has promised in exchange for the favorable treatment,. Is there, for example, a promise by Switzerland to require full-scope safeguards on its exports? That is a promise worth striving for. It is also something which, if not promised, says a lot about a country’s commitment to non-proliferation.
My last point — and the third question I was asked to address — is whether one can really distinguish countries which are a proliferation risk from those which are not. U.S. policy now hinges upon the faith that this can be done. One looks at the stability and openness of a country’s government, whether it adheres to the Non-Proliferation Treaty, whether its nuclear program is highly developed enough to make its request reasonable, and so forth. One assumes that one can judge a country’s intentions by its outward behavior. The example of Sweden proves that is false.
This past April, a history of Sweden’s nuclear weapons program was published in a Swedish technical journal, My Teknik. The history is based upon numerous interviews with persons in charge of the program, and upon previously classified government documents. The Swedish government has not denied the history’s essential findings. The history reveals a number of interesting things. First, that Sweden’s civilian and military programs worked together from the beginning, and that their relation was kept a secret. Sweden insisted, for example, upon mining uranium, even though imported uranium was far cheaper, in order to avoid the peaceful use restrictions on imports. Also, Sweden acquired the land for and tried to build a reprocessing plant despite the fact that its civilian program could not justify it. When, in the mid-1960’s, Sweden became interested in switching to the light-water reactors marketed by the United States, Sweden began secretly to study the use of reactor-grade plutonium (the grade of plutonium which light-water reactors produce) in nuclear weapons. Sweden actually conducted weapons experiments using this plutonium. In the period 1968-1970, U.S. inspectors, looking for U.S.-origin plutonium, discovered that it had been transferred to the weapons research program. In 1968, 20 kilograms of weapons-grade plutonium in spent fuel from the Agesta reactor was sent to Belgium for reprocessing and on to the Federal Republic of Germany. This plutonium was sent on an unrestricted basis, but may have been subject to U.S. controls. Sweden manufactured numerous bomb parts, such as neutron initiators, electrical detonators, and explosive lenses. In 1972, two years after depositing its ratification of the Non-Proliferation Treaty, Sweden conducted ten secret nuclear explosive tests with plutonium in an underground laboratory.
What conclusions can one draw from this? First, that there is a great temptation to use a nuclear power program to boost a nuclear weapons program. Sweden was unable to resist the temptation. Second, that such a use can be kept secret, and can continue in the teeth of the most solemn assurances to the rest of the world. The NPT forbids a non-nuclear-weapons state to “manufacture…nuclear weapons or other nuclear explosive devices…” Sweden’s actions from 1970 to 1972 violate the spirit of this language, and probably the letter as well. According to the Ny Teknik, the 1972 tests were done with metallic plutonium “setups” weighing 5-10 grams. The plutonium and high explosives were placed in a special room and ignited by remote control. Instruments recorded what happened when the plutonium was compressed. Were these “nuclear explosive devices?” They were surely devices using nuclear material, which exploded. Their purpose was military, not civilian. It is unclear whether the plutonium itself exploded, but whether it exploded or not, Sweden was building “nuclear explosive devices” in a way which could not be squared with what one expects from a non-weapons member of the NPT. Sweden also made numerous bomb parts over the course of its secret program. Did it stop making them in 1970, or throw them all away in 1970? If it was still making detonators, explosive lenses, or neutron initiators, after 1970, it violated the letter of the Treaty. Weapons parts are weapons under Article II of the NPT. Finally, under Article III of the Treaty, Sweden promised to maintain IAEA safeguards on all plutonium within its jurisdiction. After 1970, did Sweden inform the IAEA of the whereabouts of the plutonium used in the tests?
The NRC has now asked the executive branch whether Sweden used U.S.-origin plutonium in the tests. The United States exported heavy water to Sweden from the 1950’s through the 1960’s, and may have exported the uranium core for the Agesta reactor. Also, the U.S. has supplied Sweden with separated plutonium for many years. In 1972, the year of the ten tests, the U.S. apparently sold Sweden eight kilograms of weapons grade plutonium.
The example of Sweden shows that a country simply cannot be judged from its outward behavior. From 1970 to 1972, after Sweden had ratified the NPT, its “non-proliferation credentials” were excellent. An open, stable government, an advancing nuclear program, a reputation as a neutral. Above all, a treaty commitment not to make “nuclear explosive devices.” These appearances were unreliable in the early 1970’s. Are the appearances in other countries more reliable in the early 1980’s? They are not more reliable and the reason is this: the promise not to make nuclear weapons goes to the heart of a country’s national security. Such a promise will be kep only so long as a country’s national security
is not jeopardized by it. When it is Jeopardized, the promise will be abandoned and probably abandoned in secret. That is simply how countries act on matters of national security. Sweden did not abandon its secret program until its national security no longer required it. It therefore seems wrong, or at least very risky, to rely upon these outward appearances, as the U.S. policy now does, rather than a practical assessment of what a country can in fact do with what it is being given.
Finally, there is one additional point on plutonium policy which I would like to add; it concerns India. There have been recent statements to the effect that India may have a “bomb in the basement.” Mr. Gandhi has said that India can make nuclear weapons on short notice if it wishes. I think it is important to realize that India does not now have any fissile material out of which it would be legal to make a weapon. India’s separated plutonium is safeguarded except for the plutonium from the CIRUS reactor, which is subject to a guarantee of peaceful use. That guarantee was given to Canada for its supply of the CIRUS reactor, and to the United States for its supply of the CIRUS heavy water. India has no high-enriched uranium, and has not yet begun to reprocess the unrestricted plutonium from the MAPP-1 reactor. India’s test in 1974 was billed as “peaceful, but a weapons deployment could not be. The point here is that if India were to declare next week, next month or in the next six months that it were deploying a weapon, that would mean it had violated its peaceful use guarantee to Canada and the United States. It is quite remarkable that no one in the executive branch, or apparently in Congress, finds this fact disturbing. An effective plutonium policy requires enforcement of peaceful use guarantees. It also requires monitoring of U.S. materials. The CIRUS plutonium is not safeguarded, so no IAEA inspectors are keeping track of it. But that does not mean that the United States should not itself demand an accounting adequate to show that the pledge of peaceful use is being kept. The need for this accounting will not go away. CIRUS plutonium has been used in the core of India’s experimental breeder reactor, which is designed to breed “supergraden plutonium in its blanket. That supergrade plutonium will have been made from U.S. material, and should be restricted to peaceful use. This apparent lack of interest in the CIRUS plutonium, and the apparent failure to appreciate its importance, shows that U.S. plutonium policy lacks thoroughness and perseverance. I recommend that before we launch an Indian astronaut into space, we make sure India is not using our plutonium in atomic bombs.