Annex A

The Canberra Commission on the Elimination of Nuclear Weapons


The elimination of nuclear weapons will not be possible without the development of adequate verification. A political judgement will be needed on whether the levels of assurance possible from the verification regime are sufficient. All existing arms control and disarmament agreements have required political judgements of this nature because no verification system provides absolute certainty. This situation has not prevented the international community acting in the area of nuclear and other weapons of mass destruction first with the Treaty on the Non-Proliferation of Nuclear Weapons and the International Atomic Energy Agency safeguards system, then the Chemical Weapons Convention and the Comprehensive Test Ban Treaty. Nor has it prevented negotiation and implementation of bilateral nuclear arms control agreements including the Intermediate-range Nuclear Forces Treaty which eliminated an entire class of nuclear weapons.

The nature of nuclear weapons, the secrecy that has surrounded their development and uncertainties about total amounts of nuclear material produced for weapons will make it very difficult, or in the view of some impossible, to be confident that states which have operated large scale military nuclear programs have made full declarations of their holdings of nuclear weapons and fissile material.

This potential uncertainty should not deter reductions to small residual arsenals. At that point the verification system can be re-evaluated and the benefits and risks of further reductions compared. Development and implementation of the verification arrangements needed for each step toward elimination will provide immediate benefit through reducing the dangers posed by nuclear weapons and the threat of nuclear proliferation including nuclear terrorism. And a world of small residual arsenals would still be a safer place than the present world although the dangers of nuclear proliferation and a renewed arms race would remain.

Because no verification system can be perfect it is inevitable that some risk will have to be accepted if the wider benefits of a nuclear weapon free world are to be realised. The international community will need to determine the level of risk acceptable. This decision will be influenced by a range of factors, particularly the global circumstances applying when the elimination stage is reached. That the verification system for a nuclear weapon free world will involve a small probability that attempted breakout might go undetected does not alter the fact a nuclear weapon free world would be, fundamentally, a safer place, as Part One of this report makes clear. Furthermore, in an era in which the accuracy, penetrating power, and destructive force of conventional weapons are increasing rapidly, and economic interdependence is growing, the development of an illegal nuclear force would, in all probability, be self-defeating. It is nevertheless essential that there be a wide and politically acceptable level of confidence in the verification system. For this to be achieved the results of verification activities will need to be transparent to the international community both at the level of states and at the public level.

It should be recognised that a verification regime is composed of both its material and technical features, which should be of the highest order attainable, and the common political and legal commitments which support it. This creates the climate of confidence essential to any successful verification regime. Further, an inclusive approach to verification can increase levels of assurance. In the case of verification for a nuclear weapon free world, technical verification can be supplemented by measures such as transparency in nuclear activity, relevant national intelligence information passed to verification bodies, an enhanced role for individuals in verification and application of effective export controls.

A number of factors will assist development of adequate verification arrangements for a nuclear weapon free world. First, the nuclear weapon scientific/industrial complex is a tightly regulated governmental enterprise, so extensive records of nuclear weapons and weapons fissile material production should be available. Second, nearly 30 years of experience has been accumulated in verifying compliance with the NPT, and IAEA safeguards offer a proven and evolving system for delivering a high degree of assurance that safeguarded nuclear material remains in peaceful use. And, third, there is the experience of the SALT, START, INF, CFE and CWC agreements which individually and collectively demonstrate the powerful influence that political will can exert over what is desirable and possible in terms of verification.

The nuclear disarmament process will be progressive with new verification arrangements required at various stages. Because of the importance of adequate verification it is likely that progress with verification will dictate the timetable for the last stages of disarmament. Verification is likely to involve bilateral US/Russian measures, the nuclear weapon states and the IAEA at various stages of the dismantlement and elimination of nuclear weapons. The undeclared nuclear weapon states and threshold states will have to be involved in nuclear disarmament. Verification measures appropriate to these states' nuclear status at that time will have to be applied. Bilateral or regional involvement could be employed as a means of providing additional assurance and confidence building above and beyond international inspections.

This annex concentrates on measures which may make up a verification regime to provide assurance that states are complying with nuclear disarmament obligations. In addition, it is of crucial importance that there be very high physical security against diversion or theft of nuclear weapons, fissile material (whether of military or civil origin) and nuclear weapon non-nuclear components and materials. A breakdown in physical security could result in nuclear weapons, nuclear material or components coming into the possession of would-be proliferator states or sub-state groups, including terrorists which would jeopardise the disarmament process. Nuclear disarmament will at various stages of the process involve monitored storage of weapons and weapons components including fissile material. It is imperative that the highest standards of physical security be applied to such items and material. Consideration of how this can best be achieved should form part of the nuclear disarmament process.

Developmental work on verification arrangements should begin soon to ensure that movement toward a nuclear weapon free world is not delayed by lack of adequate verification.

The political commitment to eliminate nuclear weapons must be matched by a willingness to make available the resources needed for nuclear disarmament, including for effective verification. The amounts involved are likely to be considerable, especially for the dismantlement of weapons and disposition of their fissile material content, but very much less than developing, maintaining and upgrading nuclear arsenals.

This annex does not seek to be a definitive plan for the verification arrangements for a nuclear weapon free world. Its purpose is to identify some of the issues which will need to be addressed and to offer some comments on these issues. Questions of the mechanisms for applying the verification arrangements are mostly left open as it will be for the countries concerned and the international community as a whole to define these as the process unfolds.

Verification Tasks

The disarmament process will be progressive with new verification arrangements required at various stages. Few facilities in the nuclear weapon states are safeguarded at present and a number of other states operate unsafeguarded fissile material production facilities. The first stage of extending safeguards in these states is likely to be verification of facilities and material covered by a convention to end fissile material production for weapons. Systems will be needed to verify that nuclear warheads are dismantled and destroyed and that their fissile material content cannot be reintroduced to weapons use. To ensure that a nuclear force of strategic significance cannot be reconstituted quickly, a staged process for verified destruction of the nuclear weapons infrastructure is likely to be considered necessary. An intrusive inspection regime and new techniques will be needed to ensure a high probability that significant undeclared nuclear activity would be detected. Development of verification arrangements for each step toward a nuclear weapon free world will, in addition, be of immediate benefit to the existing non-proliferation regime.

Verifying the 'completeness' of declared stocks of warheads and fissile material will be a crucial and difficult operation. The IAEA has expertise in verifying declarations of previously unsafeguarded nuclear programs including its work in Iraq, the DPRK and South Africa after that country renounced nuclear weapons. The extent to which this is transferable to the very large military programs of the nuclear weapon states is to be established.

Another problem for a verification regime lies in the physical characteristics of current nuclear weapons and the fissile materials that are used in the core of the weapon. Many weapons are small, readily transported and readily concealed. The fissile material cores are smaller and thus even more easily concealed. While radiation emitted from these cores can be detected at close range, it is not clear that they would always be detected if in properly shielded storage facilities, even through environmental sampling. However, nuclear weapons in storage deteriorate with time and the ongoing maintenance needed for a secret cache of weapons would carry a risk of exposure or detection.

If a nuclear weapon free world is to be credible and stable, it clearly will have to place prohibitions on much more than just weapons. Irreversibility of nuclear disarmament will also require verified elimination or conversion to exclusively civil use of the facilities used to develop and construct nuclear weapons and dedicated nuclear delivery vehicles. In the transitional period some of the facilities used to develop and construct weapons are likely to be needed to dismantle them, so the nuclear weapon states will need to keep a part of their plant operational until the very last items in the residual stockpiles are disassembled.

Confidence to move to the final elimination phase would be enhanced if by that time all delivery vehicles built primarily for nuclear weapons are eliminated, leaving only the residual arsenals of bombs or warheads in monitored storage. It is therefore important that verified elimination of such delivery vehicles occurs in tandem with elimination of nuclear warheads. Means could be devised to make the removal of any weapons from monitored storage and their installation on improvised delivery vehicles as difficult and time-consuming as possible.

Other components which play an important role in nuclear weapons such as tritium should also be subject to a verification regime. Non-nuclear components of a weapon may also need to be taken into account. These are a collection of diverse materials: plastics, metals, chemical high-explosives and also extremely sophisticated electronics and various other items all organised inside the weapon to produce the optimum explosive output from the fissile material. These non-nuclear parts are in some cases made in or near the final assembly facility, but others come from far away, from specialised workshops or enterprises most of whose output may be civilian.

Measures must be taken to preclude leakage of sensitive information during the dismantlement process. Practical options for doing this include requiring states which own nuclear weapons to dismantle them within a containment boundary with monitored inputs and outputs. It may also be possible for international inspectors to estimate, with sufficient accuracy, the fissile material content of the stored fissile material 'pits' from dismantled nuclear warheads without revealing sensitive information. This will depend on a judgement of what constitutes sufficient accuracy and what would be reasonable assumptions about the measures that might be used to defeat such verification.

In the transition to a nuclear weapon free world it will be important to find the right balance between bilateral (US/Russia), plurilateral (nuclear weapon states) and appropriate international inspection of nuclear material made excess to military requirements. Bilateral and plurilateral inspections may be less transparent in the assurance they offer to the non-nuclear weapon states than international inspections. But bilateral or plurilateral inspections may be considered preferable for the verification of material in sensitive forms. The transparency issue could be addressed by the nuclear weapon states perhaps as part of increased accountability at NPT meetings. In areas of regional nuclear tension, bilateral or regional involvement in inspections on nuclear facilities and in monitoring the dismantlement of any nuclear weapons could be employed as a means of providing additional assurance and confidence building above and beyond international inspections.

Components of a Verification Regime

For any verification system the basic requirement is to establish what is to be prohibited or controlled. The verification regime for a nuclear weapon free world would need to bring under safeguards fissile material currently contained in weapons and military stockpiles, and to provide the most credible assurance that all such material has been accounted for; to provide a very high level of assurance that no weapons or stocks of fissile material have been concealed during the disarmament process; to ensure that all nuclear weapons facilities have been dismantled or converted to peaceful use; and to verify destruction of strategic delivery vehicles developed primarily for nuclear purposes.

Current, prospective and future treaties could provide the legal authority for application of the verification regime.

The IAEA has wide experience in application of safeguards to provide high assurance that nuclear material remains in peaceful non-explosive use. Subject to strengthening of its safeguards system the IAEA would seem the logical body to verify non-proliferation undertakings in a nuclear weapon free world. The development of concepts for CTBT verification is well advanced. Bilateral US/Russian agreements such as START and INF are a model for a verification regime for elimination of nuclear delivery vehicles. The other main elements of the verification regime, especially verification of the elimination of nuclear warheads, are less well developed and should be afforded greater priority to ensure that progress toward elimination of nuclear weapons is not held up by delays in developing and proving the verification system necessary.

Non-Proliferation Undertakings

Verification of non-proliferation undertakings en route to a nuclear weapon free world and after this is achieved will require a highly developed capacity to detect undeclared nuclear activities at both declared and undeclared sites.

Iraq demonstrated that a state with sufficient determination and resources may be able to establish a self-contained clandestine military nuclear program. This prompted a reappraisal of IAEA safeguards as it was clear there was a need to improve the safeguards system's capacity to detect undeclared nuclear activity. As a result the IAEA and its member states have worked to strengthen the effectiveness and improve the efficiency of the safeguards system. Since 1993 this effort has focused on a comprehensive program known as '93+2'. The 93+2 program is aimed at enhancing the legal and technical capability of the IAEA safeguards system with respect to its ability to detect undeclared nuclear activities. The 1995 NPT Review and Extension Conference also gave strong political support to strengthening IAEA safeguards including explicit support for the 93+2 objectives. And at the Moscow Nuclear Safety and Security Summit in April 1996 the participating countries (US, Russia, UK, France, Germany, Japan, Canada and Italy) agreed as well to work vigorously to strengthen IAEA safeguards.

The elements of the 93+2 program of most obvious application to verification of a nuclear weapon free world are increased IAEA access to information, expanded access for IAEA inspectors and use of environmental sampling. Regarding the first of the 93+2 elements, the need for maximum transparency about a state's nuclear program is fundamental. In a nuclear weapon free world transparency will be essential, especially for those states which formerly had a nuclear weapons capacity. As nuclear disarmament proceeds, doubts are bound to arise about some states' commitment to a nuclear weapon free world including whether full declarations of fissile material production have been made, whether nuclear weapons have been hidden or are being developed clandestinely and so on. Information provided by a state about its nuclear program, such as its plans for future nuclear fuel cycle activities or its fuel cycle research and development activities, together with other safeguards information such as fissile material production records, can contribute to determining whether such doubts have foundation. The state's declarations would be systematically evaluated in the light of all the other information available to the IAEA about a state's nuclear activities, and any questions or inconsistencies would be followed up.

To provide the levels of non-proliferation assurance needed in a nuclear weapon free world IAEA inspectors will need to have access to any location in a state, at very short notice or no notice and with no right of refusal. The expanded managed access arrangements being negotiated as part of the 93+2 program are a starting point in the development of access arrangements which will be needed in a nuclear weapon free world. The application of a program such as 93+2 would be central not only to effective non-proliferation arrangements but also to the ultimate development of effective verification arrangements for a nuclear weapon free world. Acceptance of the more demanding access rights will be facilitated by the universality of the nuclear weapon free world verification regime.

New technologies proposed as part of the 93+2 program have the potential to contribute significantly in this area. Of particular promise is use of environmental sampling which through air, water and soil sampling can detect characteristic radionuclide and chemical emissions from a broad array of nuclear and other industrial activities. Environmental sampling is thereby able to provide important information about the presence or absence of specific nuclear activities. Such information will be vital for verification of a nuclear weapon free world. To maximise the contribution of environmental sampling the IAEA must have the right of access to any location.

As the world moves toward a nuclear weapon free world the differences in application of safeguards in the nuclear weapon states and non-nuclear weapon states will have to diminish with the end point being universal application of the same safeguards in all countries. Verifying all nuclear weapon state facilities, including former nuclear test sites, will cause the costs of the safeguards system to rise sharply because most of the nuclear weapon states have extensive civil nuclear power programs of which only a few facilities are currently safeguarded. Improvements in safeguards procedures which have been demonstrated in the earlier phases of nuclear disarmament may allow development of alternative and more cost-effective safeguards approaches. Such approaches may moderate the increase in resources needed, for example improvements in the IAEA's capacity to detect undeclared nuclear activity may allow reduction or elimination of routine inspections at reactors.

Sharing of information between the IAEA, the chemical weapons verification regime, the prospective Comprehensive Test Ban Treaty Organisation and the biological weapons verification regime (when developed) should be explored as a means of strengthening the weapons of mass destruction non-proliferation regime. For similar reasons the weapons of mass destruction verification regimes need a flow of information on international trade in relevant sensitive items. States and individuals should also do all they can to maximise the information base of the international bodies about possible clandestine nuclear activity, including the provision of information obtained from national export licensing systems and other national technical means. Care would be needed in sharing proliferation relevant information not to breach the conditions of confidentiality under which states supply information on their own activities to verification agencies.

Verifying a Production Cut-Off

The process of developing verification for elimination of nuclear weapons will be aided by progressive controls on nuclear activity in the nuclear weapon states, the undeclared weapon states and threshold states. A cut-off convention would be the first step toward extending the safeguards applied in non-nuclear weapon states to these states, including establishing a legal basis for IAEA inspections to verify compliance with the convention. Because it would effectively cap the amount of nuclear weapon raw material, a cut-off agreement is essential to ensure the irreversibility of nuclear reductions. As with the CTBT, a cut-off convention would be open to universal adherence by all states so that these agreements can draw the states presumed to have a nuclear weapons capacity into the nuclear disarmament process.

The extent of verification required by a cut-off convention will be determined largely by its scope, which is not yet resolved. The main options are a wide scope agreement which would apply to all nuclear facilities involved in fissile material production, processing or use as well as existing stocks of fissile material and future production, or an agreement concentrating on the sensitive fissile material production facilities, i.e enrichment and reprocessing plants, and the product from these plants.

Cut-off verification will require at least application of IAEA safeguards at all enrichment plants capable of producing highly enriched uranium, all plutonium separation (reprocessing) plants, all highly enriched uranium and mixed oxide fuel fabrication plants and research reactors and critical assemblies using large quantities of highly enriched uranium or plutonium in the states joining the treaty. There might also be safeguarding of reactors and other nuclear facilities, but the elements noted above are generally accepted as the most effective and efficient ways of ensuring any fissile material produced is under safeguards that would bar its use in weapons.

The unilateral nuclear weapon states' action to end production of fissile material for weapons suggests that a cut-off agreement limited to production can be achieved within a reasonable timeframe. This approach would also moderate the increase in the IAEA's resources needed to enable it, as appropriate, to verify a cut-off convention as a verification regime concentrated on production facilities and their products would suffice. In contrast a wide scope agreement covering all facilities and all fissile material would require application of fullscope type safeguards similar to those currently applied in the non-nuclear weapon states.

Acceptance of a commitment to cease production of fissile material should not imply that existing stockpiles are to exist in perpetuity. Arrangements should be found to have stocks verified and safeguarded as early as possible in the disarmament process.

This limited verification, confined to safeguarding of enrichment and reprocessing facilities, highly enriched uranium and separated plutonium, is technically adequate, assuming that there are no clandestine, undeclared enrichment or reprocessing plants. Such an assumption will be supported by the increased capabilities of the IAEA safeguards regime for the detection of clandestine facilities.

Measures to build confidence that all activity has been declared should be developed concurrently with negotiation of a cut-off convention and might include declarations by all states of all their nuclear activities, military as well as civil, possibly with ongoing reporting on all activities, monitoring for environmental signatures indicating possible undeclared activities, application of remote surveillance techniques and access arrangements to enable the IAEA to investigate possible undeclared enrichment or reprocessing activity.

Safeguarding of enrichment and reprocessing facilities is complex, and considerable time will be needed to develop IAEA safeguards. To prepare the ground for verification of a cut-off convention the NWS should begin cooperative work with the IAEA on developing safeguards approaches for their facilities to be covered under cut-off.

Verifying Nuclear Warheads Dismantlement and Elimination

Existing nuclear arms reduction treaties provide for destruction of missiles and other delivery systems but do not address elimination and destruction of nuclear warheads. This situation is reflected in the state of nuclear arms control verification. Methods for monitoring the destruction of strategic and shorter range missiles and strategic bombers are well established and have been used to verify destruction of heavy missile launchers (silos and submarine launch tubes) and heavy bombers under the SALT and START treaties, and intermediate-range missiles under the INF treaty. In the case of nuclear warheads, methods for verifying their dismantlement have been worked out on a general level but no comprehensive verification regime is in place.

The United States and Russia have taken some preliminary steps to ensure warheads are dismantled and the process made irreversible but these two states are yet to agree on specific technologies and procedures which could be employed. Higher priority should be given to bilateral and multilateral development of the techniques needed to verify nuclear warhead elimination. Bilateral procedures should in due course be shared with the other three nuclear weapon states, perhaps with agreed modifications, as they prepare to join the disarmament process.

The first step toward a verification system for the elimination of nuclear weapons will be for the nuclear weapon states to declare their holdings of nuclear warheads and weapons grade material. In the first instance this could involve a US/Russian exchange as part of preparations for further bilateral reductions.

A possible model is the nuclear stockpile data exchange under discussion between the United States and Russia. A data exchange of this type could initially provide information on numbers of nuclear stockpile weapons added, retired, dismantled and remaining in service, broken down by categories. Information on total masses of military plutonium and highly enriched uranium again broken down by categories should also be provided. Subsequent to a US/Russian exchange of stockpile data, whether public or not, the other nuclear weapon states could make similar declarations as preparation for joining the nuclear disarmament process. As reductions proceed the initial data exchange should be expanded to provide a comprehensive picture of a state's military nuclear activity. The undeclared nuclear weapon states and threshold states will also have to end their nuclear ambiguity and to provide data on their programs to establish a basis for their involvement in nuclear disarmament.

Confidence building would be served by openness about weapons stockpiles. It is essential that states move promptly toward full disclosure of production and stocks of nuclear warheads and unsafeguarded fissile material.

One problem that must be addressed is the poor quality of accounting procedures applied during the early years of fissile material production. For example the United States recently admitted to a measurement error problem resulting in an inventory difference or material unaccounted for of 2.8 tonnes of weapons grade plutonium. A difference of this magnitude in the civil plutonium cycle would be cause for great concern. In the military cycle measurement uncertainties could be used to disguise retention of stocks of nuclear weapons material.

The United States and Russia are already cooperating on measures to improve accountancy and control of weapons material. All states producing unsafeguarded fissile material must ensure they are in a position to establish the most credible baseline data possible for their fissile material production. Techniques such as study of enrichment plant records and tails assays should be employed to reduce to the minimum any uncertainties about past production of fissile material.

The more information that can be exchanged regarding the specific locations, amounts, and forms of materials, the greater the potential synergistic benefit in terms of developing a full picture of fissile material production. This process should be applied to each phase of the life cycle of military fissile materials: production and separation of the materials; fabrication of fissile material weapons components; assembly, deployment, retirement, and disassembly of nuclear weapons; and storage and eventual disposition of fissile materials. These measures would be mutually reinforcing, building confidence that the information exchanged was accurate and that the goals of the regime were being met.

A sufficiently inclusive approach would make it difficult to falsify the broad range of information exchanged in a consistent way. Nevertheless, because of the large amounts of fissile material involved, a small measurement uncertainty would represent sufficient material for many nuclear weapons. Resulting doubts that some nuclear material and/or nuclear bombs may have been hidden may delay final elimination of nuclear weapons but should not prevent movement toward this objective.

When information on warhead numbers and types has been established a next step would be to seal warhead containers and indelibly tag them using suitable verification techniques such as bar codes, tamper indicating seals, metal surface 'fingerprints', measurement of mass, dimensions and chemical composition of warheads and active and passive radiation detectors. With some of these techniques there is a danger that warhead design information could be revealed. Approaches in their application are available, however, which should preserve the security of design information, for example through lowering the resolution of radiation detectors.

Inventoried warheads awaiting dismantlement should be inspected periodically to ensure that warhead disposition corresponds with information in the stockpile data exchange and to identify weapons entering a dismantlement facility. Use of tagging techniques should ensure that fake warheads cannot be substituted for weapons awaiting dismantlement and the real warheads diverted.

It would be inadvisable to attempt to apply IAEA inspections at the dismantlement process unless verification techniques are available which protect sensitive information. Alternatives which would allow monitoring of dismantlement without revealing design information are available, such as application to the dismantlement facility of the containment principle whereby a boundary would be established around the dismantlement facility. Actual dismantlement would be carried out by citizens of the state owning the weapons. All portals with access through this boundary would be monitored visually and using techniques outlined above to ensure there was no passage of unauthorised items into or out of the facility. The main inputs would be the tagged warheads. The main outputs would be accurately measured quantities of highly enriched uranium and plutonium in forms which do not reveal design information and which can be made subject to IAEA safeguards. Non-nuclear components would be destroyed within the containment boundary by the state owning the weapons.

Another option for protecting sensitive information could be to ensure that inspectors monitoring the dismantlement process come from countries with a similar level of weapons program to the weapons being dismantled.

The two main fissile materials, highly enriched uranium and plutonium, are at the heart of every weapon. In any phased elimination arrangement, both should be safeguarded downstream from the point where weapons are dismantled to their eventual disposal. In the interests of speed, monitoring of storage could initially be conducted by the nuclear weapon states but the IAEA should be brought into the process rapidly. In the case of plutonium stored as 'pits' the US National Academy of Sciences (NAS) report on management and disposition of excess weapons plutonium concluded that adequate safeguards could be provided without compromising sensitive weapons design information by declassifying the mass of plutonium in the pits, and allowing IAEA monitors to assay the sealed containers holding the pits without observing the components' dimensions. While this procedure needs to be scrutinised in the light of reasonable assumptions about what a state might do to attempt to defeat such verifications, as a concept it is worthy of further investigation.

Although intermediate storage is an inevitable step preceding all longer-term disposition options, such storage should be minimised. Maintaining vast stocks of excess material in a readily weapons usable form over the long term would send negative political signals for non-proliferation and for the elimination of nuclear weapons. It should also be noted that the security against the risks of diversion and theft is entirely dependent on the durability of the political arrangements under which storage is conducted. One of the key criteria by which disposition options should be judged is the speed with which they can be accomplished, and thus how rapidly they curtail these risks of storage.

Disposition of Warhead Uranium and Plutonium

Verification arrangements will be needed for monitoring the long term disposition of fissile material removed from warheads. The NAS report referred to above recommends that the United States and Russia pursue long term disposition options that:

  • Minimise the time during which this material is stored in forms readily usable for nuclear weapons
  • Preserve material safeguards and security during the disposition process, seeking to maintain the same high standards of security and accounting applied to stored nuclear weapons (which the NAS report termed the 'stored weapons standard')
  • Result in a form from which the uranium would be as difficult to recover for weapons use as ordinary commercial low enriched uranium, and the plutonium would be as difficult to recover for weapons use as the larger and growing quantity of plutonium in commercial spent fuel (which the NAS report termed the 'spent fuel standard')
  • Meet high standards of protection for public and worker health and the environment.

In the case of highly enriched uranium, achieving these goals is technically straightforward. Highly enriched uranium can be blended with other forms of uranium to produce proliferation resistant low enriched uranium for commercial fuel. The United States has agreed to purchase 500 tonnes of excess Russian highly enriched uranium, blended to low enriched uranium, over 20 years. The United States is planning to undertake a similar blending process for most of its own stockpile of excess highly enriched uranium.

Speeding up the rate of blending down of highly enriched uranium would have the advantage of reducing the time during which this material remained in weapons usable form. Even if the commercial market cannot absorb the material more rapidly, or sufficient facilities for blending the material more rapidly to a commercial quality product cannot be made available, it would be highly desirable to blend the material rapidly to an intermediate level below 20 percent enrichment, or even below 10 percent so that it was no longer usable in weapons.

Plutonium raises more difficult issues. Because, at least in principle, all mixtures of plutonium isotopes could be used to make a nuclear explosive device, plutonium cannot be blended to a highly proliferation resistant form in the same way that highly enriched uranium can. The NAS study identified two leading candidate approaches for reducing the accessibility of weapons plutonium to a level corresponding to the 'spent fuel standard'. They are:

  • The current reactor/spent fuel option, which would use light-water reactors or Canadian deuterium-uranium reactors of currently operating types or evolutionary adaptations of them, employing mixed-oxide fuel in a once-through mode, to embed the weapons plutonium in spent fuel similar to the larger quantity of such fuel that will exist in any case from ordinary nuclear electricity generation
  • The vitrification with wastes option, which would immobilise the weapons plutonium together with intensely radioactive fission products in heavy glass logs of the type planned for use in the immobilisation of military high level radioactive wastes.

All options should be evaluated carefully to determine which offers the best solution for long term disposition of former weapons plutonium, including any new possibilities that emerge as nuclear disarmament proceeds.

The security risks of plutonium in spent fuel are not zero, and this is so whether the plutonium is of military or civilian origin. So while it is very worthwhile to provide for weapons plutonium, as rapidly as possible, the same chemical and radiological barriers to diversion and theft for weapons use as exist for reactor grade plutonium in spent fuel, it is also important that safeguards and protections applied to all spent fuel are adequate in relation to the residual security risks posed by such material.

Work should be accelerated on development of techniques for IAEA safeguarding of former weapons use fissile material. The nuclear weapon states should work closely with the IAEA to develop methods which provide the high level of assurance needed without compromising sensitive information.

Civil Fissile Material

In principle, plutonium of any isotopic composition (apart from plutonium containing 80 percent or more of the isotope Pu-238) can be used in nuclear explosive devices, and for IAEA safeguards purposes all plutonium (other than Pu-238) is regarded as a 'direct use material' that can be used in the manufacture of nuclear explosives. Because of the short time needed to convert direct use material into components for a nuclear explosive device it has been suggested that verification of a nuclear weapon free world would be simplified if plutonium recycle did not occur. Use of plutonium in civil power programs is not proscribed by the NPT, however, and a number of countries have formed the view that they have no alternative to plutonium use in their civil fuel cycle if they are to meet their electricity supply needs. Such states have invested large sums in civil plutonium use.

Plutonium is produced as a natural consequence of the irradiation of U-238. The production of plutonium in a conventional reactor is therefore unavoidable. In practice, plutonium for nuclear weapons purposes is produced in dedicated reactors where burn-up levels, hence Pu-240 and Pu-238 content, can be minimised and there is no doubt that plutonium at a suitably low burn-up level is extremely attractive for nuclear weapons purposes, and that 'reactor grade' plutonium is less so. History shows that reactor grade plutonium has not been a material of choice for weapons use.

Nevertheless, plutonium use in the civil fuel cycle raises a number of issues including the requirement that strict controls be applied through application of safeguards, physical protection and rigorous national accountancy and control. Because of the sensitivity of this material, any stockpiling of plutonium by a non-nuclear weapon state beyond legitimate energy needs would be of security and proliferation concern and could result in doubts about the viability of a nuclear weapon free world.

It is essential that the control regime for civil plutonium use continue to deliver high levels of confidence that such material remains in exclusively peaceful use. States using civil plutonium also have a duty to ensure that by doing so they are not creating regional or wider tensions. This obligation is especially cogent regarding assurances that they are not stockpiling fissile material in excess of normal civil operational requirements for nuclear energy requirements. One means of doing this would be for such states to increase transparency regarding their management and use of fissile material by publishing details of their projected fissile material needs and fissile material holdings. Once a comprehensive, voluntary arrangement is operating steps could be taken to develop a treaty requiring all states to declare and account for their stocks of fissile material.

A correct balance must be struck by the international community between the interests of states using weapons grade or direct use material for civil purposes and the wider general interest in ensuring that use of such material does not result in proliferation pressures or frustrate achievement of a nuclear weapon free world.

One possibility may be to draw a distinction between plutonium of different isotopic grades and to use this distinction both for safeguards purposes and for a proscription on the separation of plutonium of an isotopic composition which makes it attractive for weapons use. If combined with a prohibition on production of uranium at or near weapons grade and the cut-off convention (which would apply only to fissile material produced for explosive use), this would stop production of all nuclear material at or near weapons grade. This would constitute an important confidence building measure in support of the nuclear non-proliferation regime and the elimination of nuclear weapons. Weapons grade nuclear materials have very limited use in civil nuclear activities and therefore a prohibition on their production should not cause practical difficulties with any ongoing legitimate civil (or military) requirement for such materials being met from existing stocks.

Were a state to be producing significant quantities of separated plutonium at or near weapons grade, the application of safeguards measures, though technically sound, would not provide the requisite degree of assurance about the future intent of the state concerned. The best way of building confidence is to avoid production of material of this kind. Where reprocessing of low burn-up material is proposed, arrangements could be put in place to ensure such material is reprocessed in stream with high burn-up material, such as normal spent fuel, so that the resultant product will have a sufficiently high proportion of the higher plutonium isotopes.

The clearest example of potential large scale incidence of low burn-up plutonium is the blanket material from fast breeder reactors. Plutonium in fast breeder blankets is the equivalent of very low burn-up, its isotopic composition being similar to weapons grade (or even 'super grade', i.e around 3 percent Pu-240). Since production of blanket material is the major reason for operating fast breeders (i.e to obtain plutonium for recycle), obviously it is not practicable to proscribe the production of such plutonium in irradiated blanket material. It is possible however to avoid the production of low burn-up plutonium as a separated product, by ensuring that irradiated material containing any such plutonium will only be reprocessed in stream with high burn-up material (e.g. fast breeder core fuel, or light water reactor fuel).

It is an unfortunate consequence of the current practice of not differentiating between plutonium grades for safeguards purposes that special attention is not directed to plutonium having the isotopic characteristics of greatest proliferation concern. Where irradiated fuel containing low burn-up plutonium is stored in spent fuel ponds, there is a strong case for subjecting it to particular safeguards attention to provide extra assurance of non-diversion.

A possible risk of drawing a distinction between the various grades of plutonium is that it could result in pressure to consider whether controls on reactor grade plutonium should be reduced. A further consideration is that enhanced controls on low burn-up plutonium would probably increase the costs of safeguarding plutonium from weapons dismantlement. In circumstances where safeguards resources are under great pressure, it would be necessary to determine whether using such resources to increase controls on low burn-up plutonium would be the most cost-effective option in terms of benefit to the non-proliferation regime. Therefore there would be merit in investigating various categories of plutonium in terms of applicable safeguards measures and resulting verification costs.

As to a prohibition on production of uranium at or near weapons grade, apart from minor quantities for laboratory use the only civil requirement for highly enriched uranium (at or above 20 percent U-235) is in certain research reactors and critical assemblies. In recent years there has been a concerted program of converting research reactors from highly enriched uranium to low enriched uranium fuel, and very few still operate on highly enriched uranium. Fewer still operate on highly enriched uranium fuel of weapons grade, the recent decision by Germany to proceed with a new reactor using such fuel being a controversial example. To the extent that use of highly enriched uranium cannot be avoided in advanced scientific research, obtaining this material from the very extensive stocks held by the nuclear weapon states will help run down those stocks and obviate any further production. Highly enriched uranium is also used by some of the nuclear weapon states in marine propulsion reactors for both surface ships and submarines. States using highly enriched uranium for this purpose have adequate stocks and do not require further production.

A prohibition on production of all nuclear material at or near weapons grade may prove a practical step of considerable value in support of the eventual elimination of nuclear weapons and could be included in the proposed cut-off convention or a complementary international agreement.

As nuclear disarmament gathers pace the amount of fissile material to be brought under IAEA safeguards will increase dramatically. This material will be made up of plutonium and highly enriched uranium components from dismantled weapons and fissile material inventories not stored in weapon component form. As a guide, the United States currently has about 84 tonnes of weapons grade plutonium and about 500 tonnes of weapons use highly enriched uranium. Russian stocks are at least equal and could be higher. It is essential that former weapons fissile material be afforded the highest standards of accounting and control and physical protection to ensure that it does not contribute to concerns about cheating or leakage to other actors.

As agreed at the Moscow Nuclear Safety and Security Summit, fissile material removed from weapons should be made subject to IAEA safeguards as soon as practicable. This will have to be done in a way that ensures that sensitive information relating to weapons design is protected. The options for doing this are either to convert the material to forms which do not reveal weapons information when accounted for by traditional IAEA safeguards measurement techniques, or to develop new techniques to account for the material in component form without revealing sensitive information. Both of these methods for protecting weapons design information appear technically feasible but require further development.


Tritium, a radioactive isotope of hydrogen, is an essential ingredient of most modern nuclear weapons, both for initiation of the fission reaction and for enhancing or boosting that reaction. It is subject to rapid radioactive decay; its half-life is 12.3 years; so there is no doubt that an appropriate control regime could play a major part in the elimination of nuclear weapons. While pure fission nuclear weapons can be made without tritium, there would be profound design consequences, e.g. they would have to be physically larger for the same yield, hence less easily deliverable.

Tritium has a number of non-nuclear uses, and Canada, the major civil producer, has established a regime of peaceful use assurances and bilateral accounting for tritium supply. This might form the basis for an international tritium control regime although it is expected that verification arrangements would also be required.

The nuclear weapon states are unlikely to accept inclusion of tritium in the proposed cut-off convention because of the changes to force structures this would require and consequent effect on deterrence. Nonetheless, such controls will be an important part of the disarmament process and associated verification arrangements and it would be surprising if the nuclear weapon states did not come to recognise that it is in their own interests for an appropriate regime to be established in due course. A practical step would be for the nuclear weapon states and other states to commence a detailed study of how such a regime might operate, and what would be acceptable to the nuclear weapon states and to the international community at large.


It is essential that the international community recognise that laying the foundation for a nuclear weapon free world will require additional resources. At the US/Russian bilateral level this will include funding verification measures for bilateral monitoring of the early stages of disarmament such as warhead dismantlement and initial monitoring of fissile material removed from warheads. This process would probably be extended to the other nuclear weapon states when they join the disarmament process with accompanying resource requirements. Resources will also be needed for the multilateral safeguards system in particular to strengthen the IAEA's capacity to detect undeclared nuclear activity and to apply safeguards at nuclear weapon state fissile material production facilities under a cut-off convention.

The IAEA's safeguards budget is approximately US $75 million per year and provides a considerable security benefit for a modest outlay. IAEA safeguards are under great pressure because of the need to apply safeguards at an increasing number of facilities. The demands nuclear disarmament will make of the Agency will add to this pressure. While the 93+2 program is intended to improve the efficiency as well as effectiveness of IAEA safeguards, it is inconceivable that existing levels of funds could be stretched to include the coming demands on Agency safeguards.

The political commitment to eliminate nuclear weapons must be matched by a willingness to make available the resources needed for nuclear disarmament including for effective verification. The amounts involved are likely to be considerable, especially for dismantlement of weapons and disposition of their fissile material content, but very much less than developing, maintaining and upgrading nuclear arsenals. The costs of verification also need to be weighed against the substantial contribution to global, regional and national security effective verification of a nuclear weapon free world would make.

Infrastructure Dismantlement

Part of the penultimate stage before final elimination of nuclear weapons should be the sequential destruction of nuclear weapons facilities. The object of this infrastructure dismantlement would be to discourage any breakout by making it a drawn-out, highly visible, large-scale, costly process.

The infrastructure dismantlement phase would begin with disclosure by the nuclear weapon states and any remaining undeclared weapon states and threshold states of their infrastructure for the production and assembly of the various elements of weapons. They would also need to agree to international monitoring to verify that weapons production has halted and that the capacity to resume production has been essentially eliminated. This would require agreements on infrastructure declarations, monitoring and dismantlement processes. Reductions in stockpiles of warheads would continue.

At this stage the nuclear weapon states and any remaining states presumed to have a nuclear weapons capacity should compile annotated charts tracing each critical element of their weapons back out into the economy through the fabricator up to whatever level technical specialists may designate. At this stage of nuclear weapons elimination there can be no valid reason not to make full disclosure of this supply net and then to eliminate those critical elements whose retention could shorten the time and cost of resuming weapon production.

There may be a need to allow the retention of some facilities on a care and maintenance basis for a period of time to provide reassurance to the nuclear weapon states until their confidence in the process has increased sufficiently to allow them to complete the task. Any cheating at this stage would require clandestine infrastructure which would need to be supplied with fissile material and the verification arrangements for a cut-off convention should be able to detect any clandestine activities at this stage in the process of disarmament.

More should not be expected of infrastructure dismantlement than it can deliver. In the period immediately following elimination any former nuclear weapon state could rapidly reconstitute a few bombs using plutonium recovered from spent fuel, assuming it was prepared to abrogate the relevant treaties and the international community did not act to stop it.

By the time the elimination phase of nuclear disarmament is reached the nuclear weapon states and the undeclared weapon states and threshold states will have halted production of weapon material, accepted safeguards on their facilities for enrichment and for plutonium separation and on material flows from those facilities and dismantled the infrastructure for the production of weapons. Ensuring the maximum degree of transparency during these processes is of central importance.

The Elimination Phase

As the nuclear powers go into their final countdown, there may be resistance to rapid elimination should the nuclear weapon states want a pause of some years to assure themselves (and others) that this elimination of bomb-building capabilities was both genuine and stable. A penultimate step to elimination might be the reduction of nuclear forces to very small residual levels; possibly but not necessarily equal --&shyp; that would be retained until it is clear that a viable support regime for a nuclear weapon free world is in place. The small residual weapon stocks would be reassuring to the nuclear weapon states though not to many others, since it would mean that the world of mutual deterrence had not yet vanished. These residual forces would then be eliminated simultaneously.

Such a stalemate would be less likely if the strengthened safeguards system currently being developed by the IAEA is instituted quickly and further developed over the course of the nuclear disarmament process. In addition, successful operation of verification during the steps toward disarmament on such sensitive tasks as eliminating weapon assembly facilities and reducing weapon stockpiles would enhance confidence that the very last stage was indeed going to be executed in strict compliance with treaty

Verification in a Nuclear Weapon Free World

The most plausible breakout scenario would be for one of the nuclear weapon states, or one of the states thought to have a nuclear capacity, to conceal a few weapons and/or fissile material from the disarmament process. Even with a highly intrusive verification regime, detection of a well shielded weapons/fissile material cache would be difficult with today's technology. But the elimination of nuclear weapons is likely to take some decades so prospects for technical detection of breakout should not be measured against today's technology. It is reasonable to expect substantial increases in the capacity of the technical verification system will flow from the experience gained in verifying the move to a nuclear weapon free world.

Apart from international verification activities of the IAEA and related bodies, the cooperation of all states would be essential as an additional layer of deterrence to any government that might consider concealing fissile material. Any state which through national technical means becomes aware of potential violations of verification regimes should bring this to the attention the appropriate verification authority.

Societal verification, or citizen's reporting, may prove to be an additional means of supporting the verification system for a nuclear weapon free world. Considerable doubts have been raised about societal verification's potential to contribute to verification of a nuclear weapon free world. In today's world or one close to it this scepticism appears to have foundation. Where the sceptics may be wrong is in extrapolating today's world indefinitely into the future. Change is inevitably coming, including in international interdependence, in developments in global communications and within societies. The right of individuals to bring violations of international obligations to the public notice is becoming increasingly recognised. A number of these changes may make societal verification an increasingly meaningful adjunct to more traditional verification methods.

Societal verification, like national intelligence activities, will operate uncertainly and unpredictably. Other governments cannot be confident that a whistle blower will quickly discover any particular violation and act on it. But a government contemplating cheating could never be confident that no whistle blower or agent would ever learn of its misdeed and act on that knowledge. Thus societal verification may prove a useful additional deterrent to any government that might consider an action such as concealing fissile material.

The possible role of societal verification would be enhanced if personal responsibility became an established norm in the area of weapons of mass destruction, i.e if it were accepted that production and use of weapons of mass destruction constituted a personal crime under international law by the individuals involved as well as by the state. In these circumstances there would be a strong incentive for individuals not to participate in or support state weapons of mass destruction programs and an incentive for whistle blowing particularly by persons who might otherwise be seen as being implicated in an illegal activity.

A second source of breakout concern is that states may seek to establish a clandestine nuclear fuel cycle and weapons program. The 93+2 program for strengthening IAEA safeguards is a sound foundation for development of technical arrangements to provide a high degree of probability that undeclared nuclear activity would be detected. As with the concealment scenario, to increase the probability of detection, information from technical verification should be supplemented by a range of other sources including intelligence information, export control regimes and societal verification.

US National Academy of Sciences (NAS): Committee on International Security and Arms Control, NAS, Management and Disposition of Excess Weapons Plutonium, (National Academy Press, January 1994); Reactor Options Panel, Committee on International Security and Arms Control, NAS, Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options, (National Academy Press, July 1995) referred to in Management of Surplus Nuclear Explosive Materials, background paper prepared for the Canberra Commission by Professor John P. Holdren which was drawn on in preparation of this annex.

Last Updated: 1 October 2014