1. INTRODUCTION

Currently, reprocessing and use of separated plutonium in civil nuclear programs are relatively limited, and most spent fuel remains in storage pending decisions on long term disposition. Some governments and commentators continue to advocate the non-proliferation advantages of the "once through" fuel cycle, which eschews reprocessing. However, there is increasing interest, by most states with major nuclear programs, in plutonium recycle based on fast neutron reactors. Fast reactors offer advantages in uranium utilisation and radioactive waste management – the challenge is to establish technical and institutional conditions to avoid the potential proliferation and terrorism risks of plutonium recycle. This is possible, and is receiving increasing attention internationally. This paper discusses the issues involved.

Plutonium in spent fuel represents an important energy source for the future, through use of fast neutron reactors. Fast reactors can optimise utilisation of uranium resources, and can provide major advantages in radioactive waste management through transmutation of long-lived elements to shorter lived materials. This can substantially reduce the time most high level radioactive waste must be isolated from the environment (from say 100,000 years to around 300-500 years) – bringing management of high level radioactive waste within the capabilities of all states with nuclear power programs.

Using current technologies, recycle of plutonium as further reactor fuel requires reprocessing in order to separate plutonium from spent fuel. The currently used chemical (aqueous) reprocessing technology has the disadvantage, from a non-proliferation perspective, of giving a state the capability to separate plutonium for military use, should it decide to abrogate peaceful use commitments. And from a terrorism perspective, increasing quantities of separated plutonium in commercial use – with increasing numbers of facilities and associated transportation of plutonium – could lead to an increasing risk of terrorists acquiring plutonium.

Proliferation and terrorism risks are compounded with the "fast breeder" reactor design, in which high-fissile (weapons grade) plutonium is produced in a breeder "blanket" surrounding the reactor core. If the combination of fast breeder reactors and plutonium separation were to become widespread, this would present a major challenge to the non-proliferation regime, as well as to nuclear material security.

Today, although there are some 380 reactor years of operating experience with fast reactors, they remain at the research, development and demonstration stage, and plutonium recycle is almost entirely confined to use of MOX fuel in light water reactors (LWRs). The extent of plutonium recycle is relatively limited – globally, MOX fuel provides around 2% of annual nuclear fuel requirements – reflecting that recycle through LWRs offers only limited benefits in terms of efficiency and waste management. Currently large-scale reprocessing plants are operated in five states – France, UK, Russia, Japan and India – and some 40 LWRs, in Europe and Japan, have been licensed for use of MOX fuel.

Now there are two major developments in this area. The first is a revival of interest in fast reactors, as reflected by the Generation IV International Forum (GIF) which is coordinating the R&D programs of 12 states. The second is the growing recognition of the importance of proliferation resistance as an essential criterion for future fuel cycle development. With fast reactors, proliferation resistance requires moving away from the "breeder blanket" design, as well as developing new recycle technologies that avoid production of separated plutonium. Proliferation resistance is being promoted by the three international programs working in this area: GIF, INPRO (the IAEA's International Project on Innovative Nuclear Reactors and Fuel Cycles), and GNEP (the Global Nuclear Energy Partnership).

The importance of proliferation resistance is not limited to horizontal proliferation (efforts to prevent the spread of nuclear weapons to further states). It is also important in the context of revitalised interest in progressing nuclear disarmament. If capabilities are maintained in civil programs to readily produce weapons grade materials, this could undermine the high levels of confidence needed for disarmament to proceed. Measures to halt the spread of such facilities and materials, as well as to ensure effective control of existing facilities and materials, would be a constructive contribution to disarmament efforts.

While the direction of technology development is towards establishing ways of plutonium recycle that minimise both proliferation and terrorist risk, currently there is no binding commitment for states to limit their recycle plans in this way. At least one state, India, is planning a fuel cycle program which has the objective of producing and reprocessing weapons grade plutonium for civil use. A number of other states are considering new reprocessing plants based on current technology. The issue now is how to ensure that the nuclear fuel cycle develops in directions consistent with non-proliferation objectives, and how to reinforce the principle of proliferation resistance through policy decisions and institutional arrangements.

2. PROS AND CONS OF PLUTONIUM RECYCLE

Open fuel cycle Today most nuclear power programs are based on thermal reactors (predominantly light water reactors), operated without plutonium recycle. With the "open" or "once through" fuel cycle, spent fuel assemblies are intended for eventual disposal as nuclear waste. While some states are committed to the once through cycle, most have not taken a firm decision on spent fuel disposition, and spent fuel is being stored indefinitely, keeping open the option of reprocessing/recycle if economic and other circumstances favour this in the future.

Proponents of the once through cycle cite the proliferation risks presented by the separation of plutonium through reprocessing. They argue that plutonium is least accessible for diversion (by states) or seizure/theft (by terrorists) if it remains locked in spent fuel.

The once through cycle is inefficient in the use of uranium resources, and if the anticipated global expansion in use of nuclear energy eventuates, the once through cycle will become more costly as easily recoverable uranium becomes scarcer and prices rise. With thermal reactors the principal source of energy is fissioning of the fissile uranium isotope, uranium-235, which constitutes only 0.71% (i.e. around 1/140th) of natural uranium. On this basis presently identified global uranium resources are sufficient to sustain only 50 years of nuclear power programs at their current scale. Further exploration will result in further uranium resources – uranium is widespread, and if necessary can even be recovered from sea water (albeit at substantial cost) – but uranium is expected to become increasingly expensive.

The once through cycle also has the disadvantage of generating substantial volumes of high level radioactive waste. All the spent fuel has to be disposed of as waste – and the once through cycle does not allow treatment of spent fuel to reduce the period it must be isolated from the environment.

The once through cycle is not free of proliferation concern. Spent fuel repositories resulting from the once through cycle could present a potential proliferation risk to future generations as "plutonium mines". Over some decades radiation levels will reduce, making spent fuel more accessible. Over a period of centuries the higher plutonium isotopes will decay, so that the plutonium in repositories will gradually become more suitable for weapons use.

Closed fuel cycle The "closed" fuel cycle involves recovery of plutonium from spent fuel for re-use as reactor fuel. The basis of plutonium recycle is to convert the predominant, "fertile", uranium isotope, uranium-238 (which comprises over 99% of natural uranium) to plutonium, and to generate energy through fissioning plutonium.

Through conversion of uranium-238 to plutonium, in principle fast reactors can extend energy production from a given quantity of uranium by a factor of 60 or more. In practical terms, the uranium requirements for a fast reactor can be replenished with depleted uranium from enrichment tails. At a consumption of around two tonnes/yr of depleted uranium for a 1,000 MWe fast reactor, the current global stocks of depleted uranium – around 1.5 million tonnes – represent some 750,000 reactor years of operation.

Recycle substantially reduces the quantity of high level radioactive waste. Materials constituting high level waste comprise only 3-4% of spent fuel, compared with the once through cycle, where the whole of the spent fuel has to be disposed of as waste. More importantly, recycle through fast reactors allows for transmutation of longer-lived radioactive materials formed in spent fuel to materials having much shorter half-lives. As mentioned earlier, the period over which most high level waste must be isolated from the environment could be reduced to as little as 300-500 years.

From a non-proliferation perspective, recycle has the advantage of consuming plutonium and in the process degrading its isotopic quality. The disadvantage is that, with current technology, recycle requires separation of plutonium from spent fuel by reprocessing. An important aspect of international efforts to develop a proliferation resistant fuel cycle is the development of reprocessing technology that avoids production of separated plutonium.

3. PLUTONIUM ISOTOPICS

The initial plutonium formed in irradiation is the fissile isotope plutonium-239. Higher irradiation levels (usually equating to longer periods in the reactor) result in additional neutron capture, producing higher plutonium isotopes, e.g. Pu-240, Pu-241 and Pu-242. Increased irradiation also produces quantities of a lower plutonium isotope, Pu-238.

Pu-239 is the plutonium isotope of primary interest for nuclear weapons. Pu-238 and the plutonium isotopes higher than Pu-239 have properties which present technical difficulties for weapons use (high spontaneous fission rate, radiation and heat levels). Weapons grade plutonium is defined as comprising no more than 7% of the isotope Pu-240, i.e. around 93% Pu-239.

Plutonium intended for nuclear weapons use (weapons grade) is produced in reactors designed so that fuel can be readily removed after only a short irradiation time. The object is to produce "low burn-up" fuel, fuel that can be removed after only a few weeks, before too high a proportion of higher plutonium isotopes is produced. By contrast, reactor grade plutonium - plutonium produced through the normal operation of a power reactor – is defined as having 19% or more Pu-240. In fact, reactor grade plutonium typically has around 25% Pu-240, which equates to less than 60% Pu-239. Such plutonium results from fuel being used in a reactor for 3-4 years.

The isotopic quality of plutonium will be degraded through further recycling. Plutonium that has been used as MOX fuel in one LWR operating cycle (3-4 years) will have a Pu-240 content of 30% or more, and a Pu-238 content of more than 2%.

Table 1 – Plutonium isotopic compositions

	Pu-239 %	Pu-240 %
Weapons grade	≥ 93	≤ 7
Fuel grade	< 93 - >77	> 7 - < 19
Reactor grade	≤ 77	≥ 19
Typical LWR fuel	56	24

Non-proliferation and safeguards practice takes a conservative approach to plutonium isotopics. For non-proliferation and safeguards purposes, plutonium of any isotopic composition is regarded as presenting a potential proliferation (and terrorist) risk. This reflects the fact that all isotopes of plutonium are fissionable by fast neutrons, and theoretically a nuclear explosive device – albeit of uncertain yield – could be made using even high burn-up plutonium.

Accordingly, the safeguards system does not formally distinguish between plutonium of different grades, apart from an exemption from safeguards for plutonium containing 80% or more of the isotope Pu-238. To date this has not been a practical issue because almost all plutonium in civil programs is reactor grade. But with the possibility of weapons grade plutonium entering civil programs (if more fast breeder reactors are built, and also with disposition of plutonium released from weapons programs), it is necessary to consider whether the policy of regarding all plutonium as being alike has an unintended consequence – that insufficient attention is given to the dangers of plutonium of higher fissile composition.

History shows that all the states with plutonium-based nuclear weapons have specifically produced weapons grade plutonium for this purpose. Discussion of the possible use of reactor grade plutonium for nuclear explosives reinforces the need for effective non-proliferation and security standards – the diversion or theft of a significant quantity of any separated plutonium would be a major concern. However, this should not distract from the fact that reactor grade plutonium is sub-optimal for weapons use, and has never been so used. There is no doubt that the plutonium quality of choice for both proliferators and terrorists would be weapons grade if it were available – hence it is essential to minimise the availability of this material.

4. "PLUTHERMAL" – RECYCLE WITH THERMAL REACTORS

When the nuclear power industry first underwent expansion in the 1960s, it was envisaged that the fuel cycle would progress from thermal to fast neutron reactors. Despite their advantages, however, the transition to fast reactors did not eventuate, due to a number of factors including: the slowing of economic growth and electricity demand following the Oil Shocks of the 1970s; loss of public and industry confidence in nuclear energy following the Three Mile Island accident; high costs and technical difficulties; and also proliferation concerns regarding the spread of reprocessing.

States that had established civil reprocessing programs turned to plutonium recycle through light water reactors (sometimes termed the "pluthermal" cycle). However, this has not become widespread – currently around 40 LWRs, in Europe and Japan, are licensed for use of MOX fuel. Recycle through LWRs offers only limited benefits in terms of efficiency – typical reactor grade plutonium comprises less than 70% of fissile isotopes. Further, the output of a number of LWRs is required to produce sufficient MOX for loading one reactor. Overall, the potential saving in annual uranium fuel requirements will be no more than around 15%.

While uranium prices remain relatively low, reprocessing is not commercially attractive. Further, the waste management benefits are limited – although reprocessing reduces the volume of high level waste compared with spent fuel, the waste materials remain the same as in spent fuel and need to be isolated from the environment for a similar length of time. Transmutation to reduce the life-span of high level waste requires fast reactors. So the combination of reprocessing and LWRs can still be seen as an interim phase, pending the introduction of fast reactors.

Aqueous reprocessing The problem with plutonium recycle using LWRs is that the plutonium must be separated from fission products through reprocessing – the established technology being the aqueous ("Purex") process. This is a dual-use technology – it originated in military programs where separated plutonium was required for nuclear weapons, and there is nothing inherent about the technology that prevents its use for this purpose. The spread of reprocessing capability presents a potential proliferation risk, and (and putting aside questions of plutonium isotopics) the product of reprocessing – separated plutonium – presents a potential terrorism risk.

MOX fuel Plutonium is not used in civil programs in pure form, but is blended with uranium to make MOX. With some reprocessing plants, plutonium and uranium are co-precipitated, i.e. there is no pure plutonium product. MOX itself is not weapons-usable, but it would not be a major challenge for a state or even non-state actors to chemically process MOX to separate the plutonium, so MOX presents some proliferation and terrorist risk. Although the plutonium in today's MOX fuel is derived from LWR fuel and so is high burn-up, it would still be of considerable concern if such plutonium came into the wrong hands.

To date there have been no significant incidents of theft of MOX. This reflects the strict security measures applied by governments and operators to the processing, transport and handling of MOX fuel. However, if the use of MOX were to become more widespread, a corresponding increase in the risk of an incident might be expected – all the more so as terrorists show increasing interest in acquiring fissile material.

Recycle without reprocessing An interesting example of plutonium recycle without separation is the DUPIC process being developed through collaboration between the ROK, Canada, and the US. DUPIC involves direct re-fabrication of spent PWR (pressurised water reactor) fuel into CANDU reactor fuel, thereby reducing natural uranium requirements and the overall quantity of spent fuel.

The basis of DUPIC is that the fissile content of spent PWR fuel (residual U-235 and produced plutonium) is well suited for use in heavy-water moderated CANDU reactors. No separation of plutonium is involved: dry thermal-mechanical processes are used to reduce spent PWR fuel to a fine powder, which is subject to high temperature to drive off volatile fission products (around 40% of total fission products), pressed into pellets, and fabricated into CANDU fuel bundles.

Since there is no plutonium separation, DUPIC is inherently proliferation resistant. However, its potential application is limited to situations where both PWRs and CANDUs are available. DUPIC does not offer the resource efficiency and waste management advantages of recycle using fast reactors.

5. RECYCLE WITH FAST REACTORS

There is a revival of interest in fast neutron reactors, as demonstrated by the number of national R&D programs being coordinated by the Generation IV International Forum (GIF). Four of the six Generation IV reactor concepts are fast reactors.

The longstanding fast reactor concept is the fast breeder reactor (FBR), in which the reactor core containing the fuel is surrounded by a uranium "blanket" in which neutrons are captured by uranium-238 to produce further plutonium. The plutonium produced in the blanket (as well as in the core) is recovered by reprocessing, and made into fresh fuel. The reactor produces more plutonium than it consumes – the surplus is to be used for fuelling further reactors (enabling a gradual expansion in the number of such reactors).

A major issue from the non-proliferation perspective is that, because of the relatively low neutron activation levels in FBR blankets, plutonium produced in the blanket has a very high proportion of the isotope plutonium-239, well within the weapons grade range. This combination of producing weapons grade plutonium and separating the plutonium by reprocessing presents obvious proliferation concerns. Further, use of separated weapons grade plutonium on a commercial scale would present a significant terrorism risk.

New fast reactor concepts The non-proliferation and security problems associated with the established FBR model are now well recognised internationally. The states engaged in the Generation IV International Forum and the IAEA's International Project on Innovative Nuclear Reactors and Fuel Cycles (INPRO) have specified proliferation resistance as an essential design criterion for future fast reactors.

The conventional FBR is being replaced by new fast neutron reactor designs, in which there is no "blanket" – all plutonium is produced in the fuel core where burn-up levels are high, making it even less attractive for explosive use than the plutonium produced in today's light water reactors. In addition, advanced spent fuel treatments – such as electro-metallurgical processing – are under development which will enable plutonium recycle without separation – plutonium will not be produced as a purified material. Instead, plutonium would remain in a mix with a number of the fission products, resulting in a highly radioactive product, suitable as fuel for fast reactors but which cannot be used for nuclear weapons without the further step of separation, using aqueous reprocessing.

Plutonium will at all times be in a mix having high radiation levels, similar to spent fuel, self-protecting against diversion and theft. A would-be diverter or terrorist could only handle this material with substantial shielding – as with spent fuel – and as noted above aqueous reprocessing would still be required for plutonium separation – again, as is the case for spent fuel.

Thus, these fast reactor designs will incorporate a number of important proliferation resistant features:

• use of a "hot" fuel (plutonium mixed with fission products as well as minor actinides);
• high burn-up, absence of a breeding blanket – all plutonium production occurs in the core, at all times plutonium from the reactor has an isotopic composition unfavourable for weapons;
• in addition, spent fuel treatment avoids any separation of pure plutonium.

It is possible to incorporate further proliferation resistant features in fast reactors. For example, because most fast reactor designs operate under no or low pressure and the core is readily accessible, a potential misuse scenario is undeclared plutonium production through irradiation of uranium targets. The Russian BREST lead-cooled fast reactor design (one of the Generation IV concepts) specifically addresses this scenario. In addition to the proliferation resistant features outlined above, the BREST reactor is designed with an equilibrium core – so there are no excess neutrons available for illicit target irradiation. If a significant amount of target material is introduced into the core, there will be insufficient neutrons to maintain the chain reaction, bringing the reactor to a stop.

Possible engineered proliferation barriers should be studied further. Clearly it would be a major non-proliferation benefit if it were possible to design fast reactors where the core could not be modified, e.g. where it was not possible to reconfigure the reactor as a conventional FBR (i.e. with a blanket), or to increase reactivity margins to allow for target irradiation. One approach may be to design the cooling circuits of fast reactors so as to limit the possibility of reconfiguring the core.

Electro-metallurgical processing Because fast neutrons can be used to fission or transmute a range of elements, plutonium can be recycled without, as with current reprocessing, being fully separated. Electro-metallurgical processing is an advanced spent fuel treatment, in which spent fuel will be melted in a molten salt mix, and most of the uranium separated through electrolysis (electro-refining). Plutonium, a proportion of the uranium, actinides and most fission products would remain mixed together and fabricated as "fresh" fuel. This mix cannot be used in nuclear weapons, and the high radioactivity levels ensure it is self-protecting against theft. Electro-processing is not just a theoretical concept, it has already been demonstrated at a practical scale in the US.

Subject to the need to conduct further research, it appears that direct plutonium separation is not practicable with electro-processing. It is not clear whether it might be possible to arrive at a relatively pure plutonium product through repeated passes through the process, but this would be time-consuming and readily detectable by safeguards. It would be easier and faster for a proliferator to divert the electro-processing product and separate the plutonium using aqueous reprocessing. Further, if electro-processing is used in conjunction with fast reactors as discussed here, the plutonium isotopics would be very unattractive for nuclear weapons use, providing no incentive to attempt plutonium separation with this material.

Thus electro-processing used by itself appears to have high proliferation resistance. However, in analysing proliferation risk it is necessary to consider the potential fissile material acquisition path as a whole. As noted above, a proliferator could use electro-processing in conjunction with aqueous reprocessing – either a clandestine facility or following break-out. In this scenario, the product of electro-processing – the plutonium/uranium/actinide/fission product mix – represents a very substantial quantitative reduction compared with the spent fuel from which it was processed. If a state were planning illicit separation of plutonium (by aqueous reprocessing), electro-processing could reduce the volume of material to be reprocessed illicitly, compared with spent fuel, by a factor of 10 to 25 – a very considerable advantage.

Accordingly, it cannot be excluded that electro-processing could be used as part of a proliferation strategy. However, it can be assumed that if a state decides to proceed with such a strategy, part of the strategy would also be to produce more attractive (i.e. lower burn-up) fuel. As noted above, it is to be hoped proliferation resistant features can be incorporated in fast reactors to exclude the possibility of producing weapons grade plutonium (e.g. using a breeder blanket or uranium targets).

This discussion highlights that proliferation resistant is a relative term, and does not mean proliferation proof. Technical measures for proliferation resistance need to be complemented by institutional measures, including limiting the spread of capabilities at the "back end" of the fuel cycle.

6. SOME POLICY CONSIDERATIONS

Minimising proliferation and terrorism risks Proliferation and terrorism risks from reprocessing and use of separated plutonium can be reduced by using the once through fuel cycle. This will not eliminate risks, however, as substantial quantities of separated plutonium already exist. Further, in view of the level of interest in the potential advantages of the closed fuel cycle for resource optimisation and radioactive waste management – most states with major nuclear programs have R&D programs in this area – it is unrealistic to seek to proscribe plutonium recycle. The challenge is to achieve the benefits of recycle while channeling technical and institutional developments in directions that avoid adding to, and if possible reduce, proliferation and terrorism risks.

The current technologies for plutonium recycle, involving aqueous reprocessing and use of separated plutonium, developed at a time when no practical alternative technologies were available. Today the need is recognised for technical and institutional measures to provide proliferation resistance – and also to minimise risks from terrorism. These factors point to a move away from plutonium separation towards new technologies, such as electro-metallurgical processing.

Until the practicability of new, proliferation resistant, fuel cycle technologies is established, risk minimisation principles suggest that reprocessing and use of separated plutonium should be limited to the extent possible. For example, there is no pressing need to build new aqueous reprocessing plants, and plans for such plants could be deferred pending further technological development. For existing programs for use of MOX fuel, it is essential for governments and operators to ensure the most rigorous security standards for all aspects of producing, transporting and using MOX. Proposals to expand MOX usage to additional facilities need to be very carefully evaluated.

Avoiding weapons grade plutonium in civil programs The plutonium produced and used in civil programs today in MOX fuel is reactor grade. This is unattractive for explosive use. This situation could change in the future, if numbers of fast breeder reactors enter service. Another potential source of weapons grade plutonium for civil programs is plutonium released from military programs as a result of disarmament measures.

An early example of action to avoid separation of weapons grade plutonium was the 1996 decision by JNC (Japan Nuclear Cycle Development Institute) to modify the Recycle Equipment Test Facility (RETF) in Tokai-mura. RETF was designed for reprocessing of FBR blanket assemblies. JNC decided that blanket material will be blended with LWR spent fuel in-process, diluting the fissile content of the plutonium before it is separated.

Paralleling international efforts to eliminate high enriched uranium from civil programs , steps should be considered now, while there is time, to keep weapons grade plutonium out of civil programs. In fact this issue has been recognised and action is being taken at the technical level. The problems associated with the established fast breeder reactor model are now well understood internationally, and a number of states are engaged in developing alternative, proliferation resistant, technical approaches, under the frameworks of GIF and INPRO.

Disposition of ex-military plutonium This is a complex subject. Clearly it is to be welcomed that progress in disarmament is resulting in the release of plutonium from military programs. This plutonium represents a major energy resource, and will be needed for power generation in the future. In addition to the energy produced, use in reactors will have the important benefit of degrading the isotopic quality of this plutonium – through irradiation, low burn-up, weapons grade plutonium will become high burn-up, reactor grade.

However, the question is how to minimise proliferation and terrorist risks from weapons grade plutonium. Such plutonium must be held under the strictest possible conditions of security and verification, and rendered unsuitable for nuclear weapons as quickly as possible. Where possible, weapons grade plutonium should be degraded before entering civil programs, to make explosive use more difficult. A simple approach is blending with reactor grade plutonium at an appropriate ratio. Another approach might be blending with Pu-238 - the practicability of denaturing weapons grade plutonium with Pu-238 or other materials warrants further study.

Where it is not practical to degrade weapons grade plutonium before civil use , very careful attention needs be given to ensuring its security. MOX fuel made with weapons grade plutonium should not be considered equivalent to normal MOX – weapons grade plutonium should be protected at a similar standard as nuclear weapons, so the processing, transport and use of such plutonium must remain under rigorous government control.

Institutional issues The potential problems of fast breeder reactors and production and use of weapons grade plutonium in civil programs are being addressed through technical development. However, not all states with fast reactor programs are committed to proceeding this way. Policy decisions and institutional arrangements are needed to reinforce the objectives of proliferation resistance and security against terrorism.

A possible avenue for this is the proposed Fissile Material Cut-off Treaty (FMCT), for which it is hoped negotiations will commence early in 2010. As currently envisaged, the FMCT will not prohibit production of weapons grade material, provided this is under verification to ensure it is not used for weapons purposes. However, the question of future production of weapons grade material should be seriously considered. Since such material is of particular proliferation and terrorism concern, there is a sound case for not producing any more.

If this is a step too far for the FMCT negotiations, states should explore other mechanisms for securing a commitment against production of weapons grade plutonium in civil programs. The objective would be to secure commitments along the following lines:

• not to intentionally produce weapons grade plutonium in civil programs, and not to reprocess this material as a separate product;
• to phase out or modify reactors that are specifically designed to produce high fissile (weapons grade) plutonium, e.g. fast breeder reactors having breeding "blankets";
• where possible, to degrade weapons grade plutonium released from military programs before use in civil programs. Where this is not possible, processing, transport and use of such plutonium must be under the tightest possible security arrangements, reflecting the risks inherent in this material.

New forms of international cooperation The use of fast reactors and plutonium recycle presents both challenges and opportunities. Addressing the challenges requires the development of new, proliferation resistant technologies. These technologies, however, will not be proliferation proof - appropriate institutional arrangements will also be needed to minimise proliferation risk. This raises a number of issues, especially how the benefits of the new technologies can be made widely available consistently with non-proliferation objectives.

A major benefit of using fast reactors will be with spent fuel management, especially the opportunity to substantially reduce the volume and life-span of high level radioactive waste. This will make disposal of high level waste easier and less expensive. However, it would not be consistent with non-proliferation objectives for capabilities at the back-end of the fuel cycle, even with more proliferation resistant technologies, to become widespread. And for most states, fast reactors and recycle technologies will be too complex and too expensive.

This is where the opportunity arises to develop new forms of international cooperation that will further non-proliferation objectives. The spent fuel management benefits of fast reactors should not be limited to the states operating these reactors, but should be made available to all states with nuclear power programs. This will address an evolving proliferation risk - the growing accumulation of spent fuel around the world. Spent fuel accumulations could provide a pretext for establishing reprocessing capabilities. Internationally, there is increasing recognition of the need to develop institutional arrangements for removal of spent fuel particularly from states of possible proliferation concern.

Arrangements are needed by which states that operate fast reactors will commit to provide spent fuel management services, on a non-discriminatory and equitable basis, to those states with power programs based on thermal reactors. This approach is an important element of GNEP – whatever the future of GNEP, it is an approach that warrants further development.

Another aspect of international cooperation is for those states involved in plutonium recycle to move away from wholly national programs. Safeguards can provide assurance about the peaceful use of sensitive facilities and materials in the near term, but cannot provide assurance about the future. For this, confidence-building measures are needed – one such measure would be to operate programs at the back-end of the fuel cycle on a bilateral, regional or multination basis.

To date it has been assumed that such measures are not required in the case of the nuclear-weapon states, since by definition proliferation considerations do not apply to them. This situation will change, however, with the establishment of the FMCT and with further progress with nuclear disarmament, so in the future confidence-building and transparency measures will be just as important for the nuclear-weapon states.

7. CONCLUSIONS

The increasing use of plutonium recycle, and the prospective introduction of fast neutron reactors, present particular non-proliferation challenges – it is essential for these developments to be pursued in ways which enhance non-proliferation objectives and avoid adding to proliferation and terrorism risks. This requires moving away from current reprocessing technologies, which produce separated plutonium, to new technologies where plutonium can be made into reactor fuel without separation from highly radioactive materials.

The need to proceed in this direction is recognised through the three international programs working in this area: GIF (Generation IV International Forum), INPRO (International Project on Innovative Nuclear Reactors and Fuel Cycles), and GNEP (Global Nuclear Energy Partnership). Currently however there is no binding international commitment for states to limit their recycle plans consistent with the objectives of these programs – and some states with significant nuclear programs are outside these international programs.

A particular area of concern is the potential for use of weapons grade plutonium in civil programs, through further development of fast breeder reactors and through disposition of plutonium released from military programs. The problems of FBRs are now recognised and are being addressed through the international programs mentioned above. Essentially, adoption of proliferation resistance as an essential criterion for future technology development has excluded the FBR – except that at least one state (India) still proposes to proceed with this reactor type. As for use of ex-military plutonium, if it is not possible to degrade this before it enters civil programs it is essential to maintain the most rigorous security measures.

Wider use of plutonium recycle should prompt new forms of international cooperation, both to limit the design and the availability of the technology involved, and at the same time to make the spent fuel management benefits of recycle generally available. All states with nuclear power programs should be able to benefit from the spent fuel/radioactive waste management advantages of fast reactors – but the broader non-proliferation interest requires limiting the number of states operating recycle programs. States that proceed with recycling should accept the responsibility of providing spent fuel management services for other states on a non-discriminatory and equitable basis. In this way recycle will provide a global benefit – eliminating spent fuel accumulations and providing spent fuel management solutions.

As a confidence-building measure, it is desirable for recycling programs to proceed on a multination rather than wholly national basis – even in the case of nuclear-weapon states.

Until the practicability of new, proliferation resistant, fuel cycle technologies is established, risk minimisation principles suggest that reprocessing and use of separated plutonium should be limited to the extent possible. For example, plans for new aqueous reprocessing plants could be deferred pending further technological development. Proposals to expand usage of plutonium fuels to additional facilities should be very carefully evaluated.

It is important for those states engaged in the development of new, proliferation resistant recycle technologies to progress this work expeditiously – the sooner the viability of these technologies can be demonstrated, the sooner the current technologies and practices can be phased out.

Footnotes