Security for Weapons-Usable Nuclear Materials: Expanding International Cooperation, Strengthening International Standards

Limited access to fissile materials— the essential ingredients of nuclear weapons— is the principal technical barrier to nuclear proliferation in the world today. With access to a sufficient quantity of separated plutonium or highly enriched uranium (HEU), many nations and even some terrorist groups would be capable of building a nuclear bomb.1 The amounts required are small: the U.S. Department of Energy (DOE) has officially declassified the fact that four kilograms of plutonium is potentially enough for a nuclear weapon; the amount of weapon-grade HEU required is roughly three times as large. Were such material to fall into the wrong hands, a proliferator''s bomb program could potentially be shortened from a decade to months or weeks. The international community could be faced with a terrifying new threat with virtually no warning— and virtually no time to dissuade the proliferator from building a bomb.

Hence, ensuring that all weapons-usable materials are secure and accounted for is an absolutely fundamental nonproliferation responsibility of all states handling such materials.2 Given that proliferating states have been willing to spend billions of dollars on their efforts to produce fissile material— and given that a single bomb could threaten tens of thousands of lives— the level of effort devoted to securing and accounting for stocks of even a few kilograms of fissile material should be even higher than that devoted to protecting stores of millions of dollars worth of cash, gold, or diamonds. Indeed, as argued later, a strong case can be made that the essential ingredients of nuclear weapons should be protected roughly as rigorously as nuclear weapons themselves are. As the DOE regulations on physical protection put it, "use of weapons of mass destruction by a terrorist(s) could have consequences so grave as to demand the highest reasonably attainable standard of security."3

While security for nuclear materials has traditionally been seen as solely a national responsibility, the potential effects of a theft of plutonium or HEU would threaten the entire international community, not just the state where the theft occurred— even more than in the case of a nuclear accident. Thus, the international community has an overwhelming interest in seeing that all such material is secure and accounted for. Ensuring that nations have effective security and accounting programs for such materials should occupy as central a place on the international agenda as ensuring that they have effective export control systems and that non-nuclear-weapon states open their nuclear facilities to full-scope safeguards.4

Today, a broad range of factors, from documented seizures of kilogram-quantities of weapons-usable fissile material to the newly demonstrated capability and will of terrorists to use weapons of mass destruction, suggest that there is an unprecedented urgency to ensuring effective security for all weapons-usable material worldwide. The threat of nuclear theft appears to be higher than ever before, and the systems designed to meet that threat are under unprecedented stress. The need for increased international cooperation and strengthened international standards for security and accounting for these dangerous stockpiles has never been greater. Fortunately, the opportunities for such expanded cooperation and revised standards have probably also never been greater. The remainder of this article will be devoted to describing these needs and opportunities, making the case for a substantial expansion of international cooperation in physical protection, and for new efforts to put in place more stringent international standards and to increase the role of the international community in ensuring they are met.

New Threats, New Stresses on the System:
The Need for Improved Security and Accounting for Nuclear Material

Inevitably, approaches to security and accounting for nuclear materials must adapt to changing conditions. Today, the threat of nuclear theft appears greater than ever before. There is far stronger evidence than ever before that there are actors on the international scene who would be extremely interested in acquiring stolen stocks of fissile material:

• The Aum Shinrikyo incident in Japan graphically demonstrated that terrorists with both the capability and the will to use weapons of mass destruction— and which governments fail to deal with until after they have struck— are not merely the stuff of nightmares5;
  
• The Oklahoma City and World Trade Center incidents in the United States, among others, demonstrated that Aum Shinrikyo was not alone in seeking to cause mass destruction, rather than simply to gain attention through terror, as was typically the case with terrorists in the past;
  
• Iraq, Iran, and North Korea have all been reported by U.S. and European intelligence services to be seeking illegally acquired fissile material from abroad— and Iraq''s willingness to spend billions of dollars on its fissile material production program demonstrated just how much some states are willing to pay for the wherewithal to produce a nuclear arsenal;
  
• While past proliferant nuclear weapons programs have been based on building an indigenous fissile material production capability, Iraq''s "crash" program to build a nuclear bomb after the invasion of Kuwait, based on the safeguarded HEU research reactor fuel in its possession at the time, provides a clear demonstration of a state''s willingness to base its weapons effort on a small amount of diverted or stolen material, in order to acquire one or two bombs while a larger production capability is being built— and the U.N. Special Commission has repeatedly expressed concern that even its extremely intrusive ongoing monitoring system might not be enough to detect a bomb program based on stolen fissile material.

At the same time that the threat appears to have increased, the systems to manage nuclear material have been placed under unprecedented stress:

• The collapse of the Soviet Union, and the subsequent political, social, and economic transformations in the post-Soviet states, have created enormous stresses, with inadequate funding to maintain security systems or even pay nuclear custodians, fundamental changes in management and oversight of nuclear activities, and weapons-usable material left in states that have never before required a national system for dealing with it— even in areas with active armed conflict (such as the facility at Sukhumi).
  
• The post-Cold War dismantlement of tens of thousands of nuclear weapons in the United States and Russia is freeing hundreds of tons of excess fissile material, which must be secured, accounted for, opened to monitoring, and ultimately used or disposed of;
  
• The opening of the first modern, large-scale reprocessing plants in France and Britain, combined with the continued operation of smaller facilities, has led to a drastic expansion in processing, storage, transport, and use of separated weapons-usable plutonium in the civilian cycle. There are now more than 160 metric tons of civilian separated plutonium in storage around the world, and since reprocessing continues to outpace fabrication of this material into uranium-plutonium mixed oxide (MOX) fuel, this figure is expected to continue to increase. Tens of tons of this material are processed each year, and tons of it are shipped across international borders.

Most of the physical protection systems in place today were designed before these new threats and stresses arose. It is perhaps not surprising, therefore, that a variety of reports, analyses, and accounts of visits to facilities suggest that physical protection in many countries may not be sufficient to deal effectively with these new conditions. There are many facilities in the world today where the chance that a determined attack by a small but well-armed and well-trained terrorist group would succeed in stealing weapons-usable material remains unacceptably high. The possibility of conspiracies of knowledgeable insiders working together to covertly steal material is even more difficult to address. Judging the vulnerability of individual facilities to such threats is quite difficult; U.S. experience has repeatedly shown that systems that are excellent on paper can sometimes perform poorly in realistic tests, particularly when the "adversary" in the test has detailed knowledge of the strengths and weaknesses of the system.

Different countries have very different physical protection and safeguards cultures, and these affect both how physical protection is implemented and the effectiveness of the resulting systems. In Russia, for example, there remains a heavy emphasis on armed force, though a new emphasis on modern technology is slowly taking root. The U.S. approach places heavy emphasis on both well-armed protection forces and modern technology. In Japan, by contrast, where possession of firearms by private citizens has been forbidden for centuries, nuclear facilities do not have armed guards— even if hundreds of kilograms of weapons-usable separated plutonium are present. Instead, reliance is placed on detection and barrier technologies to provide warning and then delay any attempted theft until nearby police forces could arrive. A variety of other countries, such as Canada, also have no armed guards at their nuclear facilities. Experts from some countries argue that they require less stringent protection measures than other countries do because they face a smaller terrorist threat. Approaches to ensuring that the personnel assigned to manage and guard nuclear material are reliable -- a critical element of effective physical protection -- very widely from one country to the next.

In recent years, however, terrorist and criminal threats have become increasingly global. Aum Shinrikyo, for example, operated on at least four continents. Terrorists have been arrested in New York who were planning terrorist acts on the other side of the world. Organized crime connections span continents and cross oceans. Thus, it is essential to ensure that variations in safeguards cultures do not result in some physical protection systems being substantially easier to overcome than others; the threat is composed of intelligent and mobile adversaries who may well be able to identify and strike the weakest link. This again highlights the need for stringent international standards of physical protection.

Today, the risk of theft of weapons-usable nuclear material is particularly acute in the former Soviet Union. As of 1994 (when current cooperative programs to address these issues first got underway on a substantial scale), essentially no former Soviet nuclear facilities had effective detection equipment (known as "portal monitors") at the gates to sound an alarm if a worker were carrying out plutonium or HEU. Fences at many facilities had holes or were overgrown with vegetation. The principal devices in use to indicate whether materials had been tampered with were easily-faked wax seals (and most workers with access to the material had the appropriate stamp needed to create a new seal). No accurate, measured inventories of the material on hand at most sites existed. And no accurate national accounting systems or effective regulatory frameworks were in place. While work is underway to correct these deficiencies, it will take years to complete. For these reasons, the U.S. Director of Central Intelligence has testified that weapons-usable nuclear materials "are more accessible now than at any other time in history— due primarily to the dissolution of the former Soviet Union and the region''s worsening economic conditions," and that none of the facilities handling plutonium or HEU in the former Soviet states has "adequate safeguards or security measures" in place.6

Vulnerabilities are not limited to the former Soviet Union, however. A few examples from the 1990s suggest the difficulties that may exist elsewhere7:

• One U.S. team visiting a non-Soviet facility handling several hundred kilograms of separated plutonium was greeted by a single armed guard; when they left, after visiting the plutonium-handling areas, the guard was on break, and nowhere to be seen.
  
• At another non-Soviet facility with several kilograms of fresh HEU, security was sufficiently poor that the facility''s managers themselves urgently asked for international help, possibly including removing the material for safekeeping.
  
• At yet another non-Soviet facility, lightly irradiated HEU research reactor fuel, which had been cooling long enough that it was no longer self-protecting, was in a storage pond secured only with an ancient chain-link fence and a single watchman.

These accounts are all anecdotal; in none of these incidents was a full vulnerability analysis done to confirm or deny the apparent vulnerability. But at a minimum, these incidents and countless others like them suggest that there needs to be a better way for the international community to know whether problems really do exist, and where additional resources for physical protection might best be focused.

This description of the problem should not be interpreted as an American complaint about the rest of the world; the fact is that U.S. physical protection systems have had serious weaknesses over the years as well. Early on, little account was taken of the risk of subnational theft, and it was perfectly legal, for example, to ship kilogram-quantities of plutonium by commercial freight, or to store separated plutonium in facilities without 24-hour guard forces.8 As recently as 1986, inspections revealed that there were no portal monitors to prevent nuclear material from being carried out at some of the exits to Pantex, the U.S. nuclear weapons assembly and disassembly facility, perhaps the most sensitive facility in the entire U.S. complex. (Monitors were installed within days of the inspection, along with other new measures.) In that time period, in a test at the Savannah River Site, the guard force failed to prevent a mock terrorist force from gaining access to the facility and making off with mock plutonium— even though the guard force had received unauthorized warning as to exactly when and where the terrorists would attack— and the guards were still shooting at each other 45 minutes after the terrorists had left.9 Incidents such as these, combined with withering Congressional investigations, and the successful attack on the U.S. Marine barracks in Lebanon, provoked a large-scale effort to improve security (and later, material control and accounting) throughout the U.S. complex. (An early and fast-paced part of this effort was dubbed "Operation Cerberus," after the mythical guardian of the gates of hell.) Total spending on safeguards and security within DOE doubled.

Today, U.S. MPC&A programs are probably among the most stringent and effective in the world. Nevertheless, improvement is a never-ending process, and important issues continue to arise that require correction.

To take a just few essentially random examples reported in recent years:

• A 1995 inspection at Pantex found that security problems identified there in 1988— shortly after the period when it had not even had portal monitors— had still not been fixed; 10
  
• In early 1997, the security chief at Rocky Flats quit "in disgust," complaining that he could not guarantee security at the site, and the Secretary of Energy acknowledged that "significant" security problems existed at the site (most of which DOE officials believe have since been addressed);11
  
• In 1995, DOE''s Office of Security Evaluations issued a report on accounting for nuclear material which found that no accurate inventories exist for thousands of kilograms of scrap plutonium and HEU in the U.S. complex, and that even though physical protection was expected to be reliable, "an accurate inventory is necessary for continued assurance against theft or diversion." (Here, too, a major effort is underway to address the issues identified.)12 In short, the need to increase physical protection for nuclear material in the face of the new threats and new stresses is a global problem, not limited to any one country or region.
  
• In late 1997, a series of internal government reviews and harsh press reports pointed to serious weaknesses in DOE physical protection programs, including excessive reductions in the size of armed guard forces, aging alarm systems that no longer functioned properly, and the like. As a result, Congress mandated the creation of a new "Department of Energy Security Management Board," including senior officials of the Energy and Defense Departments, the CIA, and the FBI.13

As a result of the inadequacy of some physical protection systems— particularly some of those in the former Soviet Union— the 1990s have seen a disturbing level of genuine incidents of theft and smuggling of weapons-usable nuclear material. While the vast majority of such reports are scams, or involve material with no relevance to nuclear weapons, this should not obscure the importance and urgency of the documented seizures of stolen weapons-usable material that have occurred. The following represent the confirmed cases involving hundreds of grams or kilograms of weapons-usable material:

• 1.5 kilograms of weapon-grade HEU from the "Luch" production association in Podolsk, Russia, in 1992;
  
• 1.8 kilograms of 36% enriched HEU from the Andreeva Guba naval base near Russia''s Norwegian border in July 1993;
  
• 4.5 kilograms of material enriched to over 19% U-235 from the Sevmorput naval shipyard near Murmansk in November 1993;
  
• Over 360 grams of plutonium seized in Munich on a plane from Moscow as a result of a German "sting" operation in August 1994; and
  
• 2.73 kilograms of essentially weapon-grade (87.7% U-235) HEU seized in Prague in December 1994.14

In short, theft of kilogram-quantities of directly weapons-usable nuclear material is an ongoing reality in the 1990s. While there is no evidence that enough material for a bomb has yet fallen into the hands of states such as Iran, Iraq, Libya, or North Korea, it is impossible to know what has not been detected. It is essential to ensure, as quickly as practicable, that all weapons-usable nuclear material worldwide is secure and accounted for— to prevent such a catastrophe from occurring— through expanded international cooperation and strengthened international standards.

Opportunities I: Expanding International Cooperation
The most important immediate step is to expand real, on-the-ground efforts to improve security and accounting for nuclear materials. International cooperation can be a critically important tool, spreading both MPC&A resources and expertise.

International cooperation to improve MPC&A measures has expanded to unprecedented levels in recent years. The largest single cooperative effort is the U.S. cooperation with the states of the former Soviet Union, a program which is expected to receive $137 million in U.S. funding during fiscal 1998; the U.S. goal in this effort is to work cooperatively with experts from the former Soviet states to ensure that by the end of 2002, modern security and accounting systems are in place for all separated plutonium and HEU throughout the former Soviet Union. Work is already underway at more than 40 sites, involving both work under formal government-to-government agreements and more informal lab-to-lab cooperation, and many hundreds of U.S. and former Soviet experts are actively involved. Several European nations and Japan are also engaged in limited MPC&A cooperation with the former Soviet states, as are the International Atomic Energy Agency (IAEA)--which has played a useful coordinating role for international cooperative efforts in the former Soviet states outside of Russia— and EURATOM.

There are many other examples of international cooperation in this area all over the world. The United States, for example, has long sponsored international courses in physical protection, as have some other countries, and the United States has worked cooperatively with many countries to which it exports nuclear technology and materials to ensure that their material is effectively secured and accounted for. Following on the model of measures taken to improve nuclear safety, where international peer review teams organized by the IAEA and the World Association of Nuclear Operators (WANO) have played an extremely helpful role in reviewing individual facilities and making suggestions for improvement, the IAEA has recently begun organizing similar international peer reviews of physical protection, at the invitation of individual member states.15 The MPC&A community— represented to a large degree by the Institute for Nuclear Materials Management— is increasingly working together on an international basis.

There are still substantial opportunities, however, for further expanding international cooperation:

• Accelerating cooperation in the former Soviet Union. In testimony to the Russian Duma in November 1996, senior officials of the Russian Ministries of Atomic Energy and Defense pointed out that only a small fraction of the amount needed for physical protection of nuclear materials and nuclear weapons was available from the Russian federal budget, and estimated that the total cost to upgrade physical protection in Russia to the levels they believed necessary would be several hundred million dollars per year— far beyond what the United States and Russia are currently spending on the task.16 The United States and Russia should certainly increase the funding they provide for their cooperative MPC&A modernization programs. But other countries should pitch in more substantially as well. So far, the contributions from other potential donor countries do not come close to matching the U.S. contribution in this area. Major nuclear states such as Germany, Britain, France, Japan, and others should be encouraged to make larger contributions to this effort, which is critical to the security of each of these states. Many of the facilities that require modernized MPC&A systems are completely civilian facilities— such as HEU-fueled research reactors— where non-nuclear-weapon-states such as Germany and Japan could participate as readily as weapon states could. The overall MPC&A cooperative program should be accelerated as much as practicable, with the goal of having modern security and accounting systems in place at all the sites where separated plutonium and HEU are located as soon as that can be achieved, and of building a new safeguards culture based on modern safeguards technology. Moreover, there is a substantial broader agenda of cooperation on other measures to reduce the threat posed by nuclear theft, from training and equipping forces to deal with nuclear smuggling, to monitoring and disposition of excess fissile material, to helping Russia''s closed nuclear cities diversify to civilian tasks, that requires substantially more funding than it is currently receiving, and in which many countries in addition to just the United States and Russia could potentially take part.17
  
• Expanded IAEA-organized peer reviews. The IAEA initiative to organize physical protection peer reviews is so far a small, nascent effort. Every effort should be made to expand this program over time, making physical protection peer reviews as regular a part of international nuclear activity as nuclear safety peer reviews are becoming; ultimately, occasional physical protection peer reviews should become a normal part of the operations of major nuclear facilities. In many cases, countries will have security concerns regarding allowing foreign experts to conduct such peer reviews, but with time and experience, it should be possible to work out means to allow meaningful peer reviews to be conducted even at relatively sensitive facilities without compromising information that must be protected. In particular, the major weapon states, including the United States, should voluntarily invite the IAEA to organize international peer reviews of physical protection arrangements at selected facilities on their soil. As part of this effort, the IAEA should make a much more energetic effort to discuss physical protection issues with member states, to identify states that would be interested in taking part in MPC&A cooperation, and to identify sites where modernization of MPC&A systems may be required.
  
• New MPC&A cooperation with China. The People''s Republic of China has extensive military nuclear activities and plans to rapidly expand its small civilian nuclear power program. In recent years, China has strengthened its MPC&A standards and regulations. Chinese experts have expressed interest in cooperating with U.S. experts in modernizing MPC&A systems. In the Chinese case, there is the opportunity for such a cooperative program to install modern safeguards and security systems while the nuclear system remains under firm central control. Moreover, as the Chinese nuclear complex is comparatively modest in size, the cost of a complete program to ensure that all weapons-usable nuclear material in China was protected by the best available security and accounting systems would be comparatively small. Lab-to-lab MPC&A cooperation between U.S. and Chinese facilities would be an especially promising approach. There may also be a role for other regional states with physical protection expertise— such as Korea and Japan— to expand their cooperation with China as well, though the fact that most of China''s weapons-usable material resides within its military programs will inevitably limit the possible role for non-nuclear-weapon states. Now that the U.S. administration has decided to implement the Agreement for Cooperation permitting civilian nuclear technology transfers to China, the door is even more widely open for the United States and China to move forward quickly in developing wide-ranging MPC&A cooperation.
  
• Expanding MPC&A cooperation in other countries. There are, of course, many other countries where cooperation to improve and modernize MPC&A systems could and should be expanded. A wide variety of countries around the world have separated plutonium or HEU on their soil. Outside of the most developed countries, these materials are usually present only in small quantities, often simply a few kilograms of HEU fuel for a research reactor or critical assembly; few people outside the scientific community using that facility may even be aware of the material''s existence, let alone its potential implications. There is therefore a fertile field for expanding U.S. and other international cooperative efforts, seeking to ensure that every country that possesses even a kilogram of this material takes appropriate precautions to protect and account for it.

Opportunities II: Strengthening International Standards

Today, there is no international mechanism in place to ensure that all countries using weapons-usable materials provide an effective and consistent level of security and accounting for them. Such a mechanism is urgently needed, and growing international support is creating new opportunities to take the first steps toward creating it.

A 1994 report of the Committee on International Security and Arms Control (CISAC) of the U.S. National Academy of Sciences recommended that the United States pursue new international arrangements to improve safeguards and physical security over all forms of plutonium and HEU worldwide. In particular, because gaining access to fissile material is by far the most difficult technical obstacle to producing nuclear weapons, the CISAC report recommended that, to the extent possible, weapons-usable materials, whether military or civilian, should be guarded and accounted for as though they were nuclear weapons— a goal the report called the "stored weapons standard" --and that inter-national standards should be updated to meet this goal. CISAC also recommended that an international organization be given "authority to inspect sites to monitor whether the standards are met."18

Meeting the stored weapon standard would mean that all areas with weapons-usable materials, military or civilian, would be within highly secure vaults or work areas, with multiple layers of protection to prevent any insider or outsider theft, continuous monitoring, and substantial armed guard forces. As in the U.S. and Russian nuclear weapon security systems, no individual would be permitted to be alone with weapons-usable nuclear material without another person present, and individuals with access to such material would be carefully screened for reliability (including examination of their financial status) --and rescreened at periodic intervals. The protection systems would be designed with the goal of providing reliable protection against insider theft by individuals in any position— even in collusion with outside forces— and against covert or forcible outsider theft, even by teams of well-armed and well-trained attackers. None of these objectives would be impossible to achieve for weapons-usable nuclear material worldwide, and indeed, substantial quantities of weapons-usable nuclear material are already protected to such standards. The CISAC report acknowledged, however, that the bulk processing of material "will unavoidably make accounting more difficult than in the case of nuclear weapons, and it may also be institutionally difficult to preserve the strict security arrangements associated with nuclear weapons themselves." But the report argued that "precisely because of the difficulty of the task, it is important to preserve the goal."

Current international standards fall far short of these objectives. Although an attempt to set international standards for nuclear materials security was made in the 1980 Convention on the Physical Protection of Nuclear Material, that convention was drafted at a time when today''s threats --from nuclear smuggling to the use of weapons of mass destruction by terrorist groups— did not yet exist.19 U.S. approaches have changed radically since then, resulting in more than a doubling of annual spending on safeguards and security; other countries'' approaches have changed significantly as well. The Convention is quite vague in its requirements, applies primarily to international transport of materials, does not cover military materials at all, and has no provisions for verification or enforcement. Moreover, many countries, including some who possess significant quantities of weapons-usable material, are not parties to the Convention. Similarly, although the IAEA has published more detailed guidelines for physical protection of nuclear materials, these are purely advisory— and even they do not set any standards for how well a physical protection system should perform (that is, what threats it ought to be designed to defeat).20 Neither the IAEA nor any other organization monitors or compiles accurate, up-to-date information on physical security procedures worldwide; thus, there is no means for the international community to know where remedial action may be necessary, or where the next marginal dollar for international physical protection cooperation could best be spent. Moreover, no comparable convention setting standards for material control and accounting systems exists.

A number of countries, including the United States, have attempted to impose some international standards— and procedures for checking physical protection arrangements— through bilateral agreements. Under U.S. law, the executive branch must periodically certify that countries using U.S.-origin nuclear materials are providing them with adequate physical protection, and to meet this requirement, the United States has long undertaken programs in which selected facilities handling U.S.-origin materials are occasionally visited by U.S. experts to examine their physical protection arrangements. Often, this becomes part of a joint cooperative effort to modernize and improve these arrangements. Following the U.S. lead, the Nuclear Suppliers Group has adopted guidelines for physical protection of material originating within its member states.21 But these efforts are no substitute for a broad international approach to ensuring effective physical protection for weapons-usable material worldwide.

A major international effort to improve security and accounting for weapons-usable nuclear materials worldwide would be costly, probably adding tens of millions of dollars a year to the costs currently paid for such activities. Physical protection would remain, however, a tiny fraction of the overall costs of nuclear activities worldwide; meeting the "stored weapon standard" would not require measures that would add in any significant way to the overall costs of nuclear-generated electricity (particularly as most such electricity is today generated without the use of directly weapons-usable material). In any case, the cost of mitigating proliferation risks should be considered an essential part of the cost of operating a facility that uses weapons-usable materials— an externality that should be internalized, just as the costs of pollution prevention and mitigation should be paid by the polluters.

There appears to be growing international support for the idea that more stringent international standards, and a greater international role in seeing that they are met, are needed. The U.S. DOE has largely adopted CISAC''s "stored weapon standard" for its excess plutonium disposition program, including, for example, a decision to transport MOX fuel with the same vehicles and security measures used to transport nuclear weapons. This will provide a useful demonstration that the stored weapon standard, or something close to it, can be achieved in practice without a substantial increase in overall nuclear operations costs.

In 1995, a Special Panel of the American Nuclear Society, including senior representatives not only from the United States but from Russia, Japan, France, Germany, and Britain as well, issued a unanimous report, Protection and Management of Plutonium, which also called for increased international attention to physical protection, specifically endorsing the concept that separated plutonium should be protected as rigorously as nuclear weapons are, and the idea of giving the IAEA the authority to inspect physical protection arrangements.22

Most recently, in mid-1997, a broad international committee established by the IAEA to consider new steps in international cooperation in preparation for the IAEA''s June 1997 international symposium on "Nuclear Fuel Cycle and Reactor Strategy: Adjusting to New Realities," called for new agreements to provide "strengthened assurance that national physical protection standards and performance meet high standards and that remedial action to correct any deficiency will be taken."23 This committee recommended making it possible, "though an international convention or other means," for the IAEA to assess physical protection measures at individual sites and "offer advice and assistance to correct deficiencies." Such assessments would be carried out either at the request of individual states, or "in a systematic manner," with the IAEA given mandatory authority under the provisions of a new agreement. The group pointed out that while there are great sensitivities associated with physical protection measures, "the IAEA has established an outstanding record in protecting sensitive State information and this problem would appear to be soluble."

The time has come to begin the hard work of translating this nascent but growing consensus into more stringent international standards and additional authority for the IAEA to take a larger role in physical protection. It is critical, however, that this effort not be allowed to distract from the even more urgent task of correcting known physical protection deficiencies: improvements on the ground are more important, for the moment, than creating stricter standards on paper. But over the long term, stricter standards will have a fundamental role to play— and it should in principle be possible for governments to muster the energy to pursue both.

Creating a regime of greatly strengthened international standards for physical protection will inevitably be difficult, and require the expenditure of considerable diplomatic capital. Many countries still consider their own physical protection arrangements a matter of exclusive national sovereignty, not a subject for international discussion. Cultural differences in approaches to physical protection will complicate discussions of specific strengthened approaches or standards. Nuclear industries using weapons-usable materials will object to the potential for increased costs, and these industries typically have strong influence on their governments with respect to issues perceived to be technical nuclear issues, such as this one. While a regime such as that recommended in the 1994 CISAC report should be the ultimate goal, it is likely that this goal can only be approached step-by-step.

Collecting Information
The first step in creating stringent global standards is to collect information as to where countries are today in approaching such standards. As noted above, no such repository of information currently exists, leaving the international community with no way of knowing where the most significant risks may lie. The IAEA should be given the authority to begin keeping a database of current information on physical protection, provided voluntarily by member states; a questionnaire could ask about countries'' specific policies with respect to the individual items recommended in the IAEA''s guidelines on physical protection, INFCIRC 225. A request by the IAEA Board of Governors would be sufficient to initiate such a voluntary data-collection effort. At least general information from this database should be made available to member states and to the public, perhaps in an annual publication; more detailed information— particularly information that suggested vulnerabilities at individual sites— may have to be treated as safeguards confidential, at the request of the providing state. At a later stage, this database could be supplemented with data gathered during IAEA visits or inspections, some of which would certainly be safeguards confidential (see below).
Undertaking Binding Pledges
Binding pledges by individual states to meet certain levels of physical protection— and to allow some managed form of access by IAEA-organized peer review teams— could be a useful step toward building momentum toward revised agreements, just as individual states'' nuclear testing moratoria played a key role in leading to the successful conclusion of the Comprehensive Test Ban. Individual states could pledge to implement all of the recommendations of INFCIRC 225 for all of their weapons-usable material (both military and civilian), and to allow occasional access by international peer review teams, who could review the physical protection arrangements at an individual facility chosen by the IAEA from a list of all facilities handling weapons-usable material provided by the state making the pledge. These initial pledges would be expected to include a statement that the pledging state reserved the right to choose the level of access to be provided, balancing the need for effective peer review with the need to protect certain information in the interests of nonproliferation and national security. Since the purpose of these voluntary visits would be to help ensure that material at the site was not stolen for military purposes by unauthorized parties, it should be possible for IAEA staff to conduct such visits, even at sites conducting military activities, within the limits of the IAEA''s statute and its limitation to verification of peaceful use; if this was considered to be a problem, the IAEA could simply serve as the organizer for international peer review teams, which would not be considered to be IAEA inspectors. If states did not feel comfortable with opening certain facilities to international teams organized by the IAEA, they might work out suitable peer review arrangements within a group of allied or regional states with whom they felt more comfortable, at least as a first step. Such pledges could readily be made binding, by means of a brief document that each pledging state could sign, representing an agreement between itself and the IAEA. This approach would allow states that saw an urgent need for stricter standards to take the lead by first applying these standards to themselves. If a core of key countries could be convinced to make such pledges early on, such an approach might rapidly build momentum toward greatly strengthened international standards.
Clarifying the "Stored Weapon Standard"
If, in the long term, the objective is to convince those who use weapons-usable materials to protect them to the "stored weapon standard," it would be helpful to clarify what exactly that would entail, and to work out specific measures that would be applicable to bulk material rather than only to item-accountable nuclear weapons. An interested state or group of states could organize a meeting at which the weapon states would describe the security procedures for stored nuclear weapons— at a level of generality appropriate for a forum that would include non-nuclear weapon-states— and experts could discuss the measures that would be needed to ensure a comparable level of security for weapons-usable nuclear material. This would not be a negotiating session, but a discussion of the specifics of what the "stored weapon standard" would mean if accepted in the future— providing a template from which negotiators could later draw, if and as desired.
Strengthening INFCIRC 225: Design Basis Threats and Armed Guards
While there have been some revisions of the INFCIRC 225 recommendations in recent years, there has been no attempt to carry out a "bottom up" review of these recommendations in light of changed world circumstances. There are a wide range of ways in which the specific recommendations in INFIRC 225 might be strengthened, which should be considered. One possibility would be to incorporate some elements of a "performance-based" approach, specifying the level of performance that should be achieved, rather than relying solely on the current prescriptive approach, specifying individual measures to be taken. A performance-based approach would include at least a minimal "design basis threat" that everyone agreed physical protection systems should be designed to defeat. With the increasingly global nature of terrorist groups and organized crime, it is difficult to argue that there is any country in the world where theft attempts by well-placed insiders or attacks by small groups of well-trained and well-armed terrorists are not serious possibilities; hence, all physical protection systems for weapons-usable material should be defined to be capable of defeating such threats. Regular and realistic performance testing is critical to understanding the real performance of a physical protection system, and recommendations for such testing should be included in INFCIRC 225.

Another important agenda item, given the stakes, is to include provisions for armed guard forces for all weapons-usable material; it is simply never likely to be possible for technology alone, backed up by police forces some distance away, to provide the same level of protection that can be achieved by combining technology, on-site armed guard forces, and remote police (and military) backup. Work should begin now to design culturally acceptable approaches to providing armed guard forces for weapons-usable material facilities in countries that do not yet have them. Britain, for example, which has prohibitions on private firearms similar in some respects to those of Japan, has resolved this problem by establishing an armed special-purpose national police force for nuclear facilities; Japan and other countries could consider similar approaches.
Providing Authority for IAEA Inspection
As noted earlier, the IAEA is already organizing voluntary "peer reviews" of physical protection arrangements in a few countries. Over time, however, efforts should be made to increase the IAEA''s authority to inspect to ensure that effective physical protection measures are being implemented. As a first step, the IAEA should be given the authority to collect information related to physical protection observed by its safeguards inspectors in the normal pursuit of their business. This could potentially be done through a new direction approved by the Board of Governors. Over time, however, this authority should be further increased, allowing the IAEA to request inspections at designated facilities specifically designed to examine physical protection arrangements, and requiring states to permit the inspections, with some form of managed access balancing the benefits of such inspections with the protection of proliferation-sensitive information. This might be accomplished through the adoption by individual states of a new protocol— as in the case of the recently approved protocol giving the IAEA the authority to implement those 93+2 safeguards measures not already within its existing authority24— or as part of a new agreement. As noted above, since the purpose of these inspections would be to ensure that appropriate measures are in place to prevent removal of material for unauthorized military or terrorist purposes, and no direct access to materials in military use or in military forms would necessarily be required, it should be possible to carry out such inspections at both civilian and military facilities within the bounds of the IAEA''s existing statute. Of all these measures, creating a binding inspections regime may be the most difficult to achieve, provoking the greatest political resistance from some states; while this measure is important, and steps should be taken in this direction, resistance to binding inspections should not be allowed to stymie progress toward the other elements of a strengthened international physical protection regime.
Modifying the Physical Protection Convention?
A more ambitious and far-reaching goal would be to negotiate amendments to the Convention on the Physical Protection of Nuclear Material. Suggestions have been made ranging from simply extending the coverage of the convention''s current provisions to cover domestic material (rather than almost exclusively material in international transport, as is currently the case), to extending them to cover military material as well, to radically revising the entire document to include new standards— up to and including the "stored weapon standard." Ultimately, such a modified international convention— or a new one, as described below— will be needed. But the negotiation of such amendments is likely to be a difficult and painful process, and is not likely to be successful until other measures that can be implemented in the nearer term have built up a stronger base of international support for change.
A new nuclear terrorism convention?
Yet another approach would be to start from scratch and negotiate a new convention, designed to include a range of measures to reduce the threat of nuclear smuggling or nuclear terrorism— which could include new standards for physical protection of weapons-usable material, the most fundamental and cost-effective measure to prevent these dangers. The Russian government, for example, has recently proposed a draft of a new convention on nuclear terrorism, which includes at least a requirement that states adopt "all necessary" physical protection measures.25 It is by no means clear whether the framework provided by the existing Convention on the Physical Protection of Nuclear Material would allow the negotiation of more stringent international standards to be accomplished more quickly by modifying that Convention, or whether the obstacles posed by the accumulated inertia of past ways of doing business under that Convention mean that it would be simpler to start from scratch and negotiate a new agreement. Both options should be considered. A new agreement would have the advantage that a variety of measures beyond physical protection— relating to nuclear smuggling cooperation and other related issues— could more readily be included.
Conclusions
Ensuring that all weapons-usable nuclear material worldwide is secure and accounted for is a fundamental nonproliferation priority. The threats of nuclear theft and the stresses on the systems designed to prevent it have never been greater. Considerable opportunities are available to expand international cooperation in improving security and accounting for weapons-usable materials, and to strengthen international standards. In particular, steps should be taken to:

• Accelerate current cooperative programs to ensure that modern MPC&A systems are installed for all weapons-usable nuclear material in states of the former Soviet Union, and to foster the development of a strengthened "safeguards" culture in these states;
  
• Establish similar cooperation to install modern MPC&A systems for all nuclear material in China;
  
• Drastically expand the IAEA program of peer reviews of physical protection, ultimately making such international peer reviews a regular part of the operations of major nuclear facilities worldwide;
  
• Expand other international MPC&A cooperation, with the goal of providing modern physical protection, accounting, and control for all weapons-usable nuclear material worldwide;
  
• Direct the IAEA to begin compiling data on states'' physical protection systems, beginning with voluntarily-supplied data concerning states'' approaches to the individual elements of INFCIRC 225, and continuing to include data relevant to physical protection acquired during IAEA safeguards inspections;
  
• Initiate a series of binding pledges by individual states to meet stringent standards of physical protection, and to allow peer review of their physical protection arrangements, with appropriate managed access to protect national security;
  
• Convene a conference of experts to identify specific measures that would be included in a program designed to meet the "stored weapons standard";
  
• Strengthen INFCIRC 225, giving consideration in particular to including performance-based approaches and performance testing, and strengthened recommendations for armed guards for all weapons-usable nuclear material;
  
• Begin taking steps toward creating a binding requirement to accept IAEA inspections or IAEA-organized reviews of physical protection arrangements, with managed access to protect national security;
  
• Negotiate modifications to the Physical Protection Convention, or a new agreement, to greatly strengthen international standards for physical protection of nuclear material.

None of these steps will be easy. Taking advantage of the opportunities that now exist will require sustained, energetic, and high-level leadership from several countries, to overcome diplomatic obstacles and ensure that the necessary resources are provided. Few tasks, however, could be more deserving of the effort than ensuring that the essential ingredients of nuclear weapons do not fall into the wrong hands.

NOTES

1 For a useful unclassified discussion of the possibility that terrorists could build nuclear explosives, see J. Carson Mark et al., "Can Terrorists Build Nuclear Weapons?" in Leventhal, Paul, and Yonah Alexander, eds., Preventing Nuclear Terrorism. Lexington MA: Lexington Books, 1987. That article, drafted by U.S. nuclear weapons designers, represented a negotiated statement agreed to by both advocates and skeptics of the thesis that terrorists could readily build nuclear explosives; unfortunately, it concludes that the answer to its title question is yes. Subsequent work in the U.S. design laboratories has tended to confirm this conclusion. In addition, it must be remembered that subnational thieves may provide stolen fissile material to states, where it can then become the basis for national nuclear weapons programs— probably the most plausible and dangerous threat posed by potential theft of nuclear material. Indeed, all of the seizures of stolen weapons-usable material to date involved thieves planning to sell the material to others, not thieves belonging to a subnational group planning to produce a bomb itself.

2 In this article, I use the term "weapons-usable" fissile material to refer to all separated plutonium, of whatever grade, and to all HEU (that is, uranium enriched to more than 20 percent U-235). While much of this material is not "weapon grade," and would not be preferred by nuclear weapons designers, it can be used to make nuclear explosives. In particular, reactor-grade plutonium can be used to make a bomb with an assured, reliable yield in the kiloton range, and a probable yield significantly higher than that, using technologies no more sophisticated than those used in first-generation nuclear devices such as the Nagasaki bomb; sophisticated weapon states could also use reactor-grade material such as that coming from light-water reactors today to make weapons with reliability, yield, weight, and other characteristics comparable to those made from weapon-grade plutonium; and states with intermediate design sophistication could build weapons from reactor-grade plutonium with intermediate capabilities. The United States has recently declassified the most detailed information yet about what could be done with reactor-grade plutonium; see U.S. Department of Energy, Office of Arms Control and Nonproliferation, Final Nonproliferation and Arms Control Assessment of Weapons-Usable Fissile Material Storage and Excess Plutonium Disposition Alternatives. Washington DC: DOE, January 1997, pp. 37-39.

While this article focuses on directly weapons-usable material, which requires the highest level of physical protection, the need for adequate security and accounting of material in less attractive categories should not be ignored. Producing weapon-grade HEU from low-enriched uranium (LEU), for example, requires dramatically less enrichment work than producing it from natural uranium; if a proliferator were able to acquire a secret stockpile of LEU, the plant needed to complete the remaining enrichment would be far smaller, less costly, and easier to hide. Similarly, it would be far easier to keep secret a program to separate plutonium from a stock of illegally acquired spent fuel than it would to keep secret the combination of such a separation program and a nuclear reactor for producing the plutonium in the first place.

3 U.S. Department of Energy, "Protection and Control of Safeguards and Security Interests," Order 5632.1C. Washington DC: DOE, July 15, 1994.

4 In this article, following International Atomic Energy Agency (IAEA) practice, I use the term "diversion" to refer to actions by the legitimate owners of the material to incorporate it into nuclear weapons in violation of pledges not to do so, and the term "theft" to refer to acquisition of the material by parties other than its legitimate owners. Similarly, as in IAEA practice, I use the term "safeguards" to refer to international safeguards designed to detect diversion of nuclear material by states (or, in more recent programs to detect clandestine nuclear activities at undeclared sites); in this article, the terms "security" and "physical protection" are used interchangeably, referring to states'' primarily domestic security systems, designed to prevent theft of nuclear materials by non-state actors (possibly working in conjunction with foreign states). Increasingly, some modern national systems seek to combine material protection, control, and accounting (MPC&A) into an integrated system of "domestic safeguards" to ensure that all material is secure and accounted for.

5 As is well known, Aum Shinrikyo was also exploring nuclear material, particularly uranium, in a preliminary way.

6 Deutch, John, "The Threat of Nuclear Diversion." In Global Proliferation of Weapons of Mass Destruction, hearings before the Permanent Subcommittee on Investigations, U.S. Senate Committee on Government Affairs. S. Hrg. 104-422, Vol. 2, 104th Congress, 2nd Session. Washington DC: Government Printing Office, 1996. Since nuclear weapons themselves are easily counted, are typically stored in heavily guarded vaults, and are managed by a military organization (the 12th Main Directorate of the Ministry of Defense) that continues to have substantial cohesion, Deutch indicated that the CIA''s judgment was that nuclear weapons themselves remained secure, but that "the threat from within the Russian military and the deteriorating economy mean that this judgment could change rapidly."

7 Based on interviews with U.S. government and IAEA officials. I have not named the facilities, in order to focus the debate on the general point concerning the need for stringent standards worldwide, rather than on the specific circumstances at individual facilities.

8 For a journalistic critique of these early practices that was influential in changing them, see John McPhee, The Curve of Binding Energy. New York, NY: Farrar, Strauss, and Giroux, 1974; see also Willrich, Mason, and Theodore B. Taylor, Nuclear Theft: Risks and Safeguards. Cambridge, MA: Ballinger, 1974.

9 See Nuclear Weapons Facilities: Adequacy of Safeguards and Security at Department of Energy Nuclear Weapons Production Facilities, hearings before the U.S. House Committee on Energy and Commerce, H. Hrg. 99-143, 99th Congress, 2nd Session. Washington DC: Government Printing Office, 1986.

10 This was mentioned in the DOE Office of Evaluations'' summary of the Pantex site, in its December 1996 update (available from the author). That update is no longer available at DOE''s web site, having been replaced by an October 1997 update, which can be found at http://tis.eh.doe.gov/web/eh2/profiles/ptxspfn.html.

11 See Jim Carrier, "Flats Security Lax, Ex-Officials Warn," The Denver Post, May 20, 1997; Mark Eddy, "Flats Theft Risk Discounted— But DOE Confirms Security Problems," The Denver Post, June 10, 1997; and Jim Kirksey, "Flats Security Rating Higher— Facility Ranked ''Satisfactory''," The Denver Post, September 12, 1997. The U.S. DOE rates facilities'' physical protection systems as being "unsatisfactory," "marginal," or "satisfactory." Rocky Flats reportedly had a history of "marginal" ratings, with a mid-1996 review showing "endemic deficiencies," and a series of tests over the years leading to knowledgeable individuals succeeding in removing test "plutonium" from the facility without authorization or detection. In January 1997, matters got even worse and the facility was rated "unsatisfactory;" the security director at the site quit in April of 1997. The site received its first "satisfactory" rating in several years in September 1997. However, a DOE report in mid-1997 reportedly pointed out that there was "reason for skepticism" about the improvements, as Rocky Flats "has a long history of improving its protective systems only to allow them to degrade once again when priorities shift or external pressures abate." Perhaps it was not surprising, then, that a November 1997 DOE review described security at Rocky Flats again as "less-than-adequate," indicated that "DOE protection standards were not beeing met with respect to the protection of special nuclear materials against theft," and pointed to "on again, off again" management attention to physical protection there. See U.S. Department of Energy, Office of Oversight, Interim Report on the Status of Safeguards and Security in the Department of Energy, November 1997. This phenomenon of shifting levels of attention and priority, of course, is not limited to Rocky Flats, or to the United States.

12 Increasing Fissile Inventory Assurance Within the U.S. Department of Energy, Deputy Assistant Secretary for Security Evaluations. Washington DC: DOE, January 1995. (Available at http://www.tis.eh.doe.gov/web/eh2/reviews/fiss_rep.html.)

13 See, for example, Peter Eisler. "Reduced Budgets Erode Security at Nuke Plants." USA Today, October 22, 1997; Peter Eisler. "Unit to Probe Nuke Safety is Approved." USA Today, November 7, 1998; and U.S. Department of Energy, Office of Oversight, Interim Report on the Status of Safeguards and Security in the Department of Energy, November 1997.

14 Potter, William C. "Before the Deluge? Assessing the Threat of Nuclear Leakage From the Post-Soviet States." Arms Control Today, October 1995, pp. 9-16.

15 This program is known as the International Physical Protection Assistance Service. The IAEA''s Board of Governors approved the initiation of such a program in March 1995, and the IAEA announced the service was available in march 1996. So far, the countries that have requested such peer reviews have primarily been countries in Eastern Europe with relatively modest quantities of weapons-usable material on their soil, if any.

16 For an edited English translation of the transcript, see Yaderny Kontrol Digest, No. 5, Fall 1997 (Center for Policy Studies in Russia, Moscow.)

17 For a review of the current status of this broad agenda, see Matthew Bunn and John Holdren, "Managing Military Uranium and Plutonium in the United States and the former Soviet Union," Annual Review of Energy and Environment, Vol. 22, 1997.

18 U.S. National Academy of Sciences, Committee on International Security and Arms Control. Management and Disposition of Excess Weapons Plutonium. Washington DC: National Academy Press, 1994, pp. 31, 136-137. I was the study director for this report.

19 International Atomic Energy Agency. "The Convention on Physical Protection of Nuclear Material," INFIRC 274. Vienna, Austria: IAEA, 1980.

20 International Atomic Energy Agency. "The Physical Protection of Nuclear Material," INFIRC 225, Rev. 3. Vienna, Austria: IAEA, 1993.

21 International Atomic Energy Agency. "Communications Received From Certain Member States Regarding Guidelines for the Export of Nuclear Material, Equipment, or Technology," INFCIRC 254. Vienna, Austria: IAEA, 1978.

22 American Nuclear Society, Protection and Management of Plutonium, Special Panel Report (Glenn Seaborg, honorary chairman). La Grange Park, IL: ANS, August 1995.

23 "Key Issue Paper 6: International Cooperation," (M.B. Kratzer and I. Kouleshov, co-chairmen), in International Atomic Energy Agency, International Symposium on Nuclear Fuel Cycle and Reactor Strategy: Adjusting to New Realities. Vienna, Austria: IAEA, 1997.

24 "93+2" is the name given to the package of safeguards measures developed in recent years to provide inspectors with greater access to information and locations, in an effort to improve the efficiency and effectiveness of the safeguards system, and particularly to improve the system''s ability to detect undeclared nuclear activities. Development of the new measures was formally initiated in 1993 and was expected to take two years, hence the name; in the end, it took four years before a protocol embodying those aspects of the new measures that were not already within inspectors'' authority was approved. See International Atomic Energy Agency. "Model Protocol Additional to the Agreement(s) Between State(s) and the International Atomic Energy Agency for the Application of Safeguards," INFCIRC 540. Vienna, Austria: IAEA, 1997.

25 Delegation of the Russian Federation, "Convention on the Suppression of Acts of Nuclear Terrorism," UN General Assembly Document A/AC.252/L.3, January 28, 1997.