The International Atomic Energy Agency (IAEA), in cooperation with the European Commission, the Nuclear Energy Agency of the OECD (NEA) and the Uranium Institute, organised an international symposium in June 1997 with the aim of facing the "new realities" in the nuclear fuel cycle and to come to conclusions on how they should be addressed.
These
new realities are:
- the unexpectedly slow growth of nuclear energy;
- the escalation of back end costs;
- the delay in the introduction of fast reactors;
- the end of the Cold War.
The
consequences of these changes are:
- a surplus of uranium;
- a continued debate over the choice of fuel cycle;
- a surplus of separated plutonium;
- the demilitarisation of weapons plutonium and high enriched uranium (HEU).
In
order to prepare for the symposium, a steering group and six
working groups were formed, with representatives from 12 states
and 4 international organisations. The symposium covered all
aspects of the fuel cycle in six sessions corresponding to the
six working groups:
- global energy outlook;
- present status and immediate prospects of plutonium management;
- future fuel cycle and reactor strategies;
- safety, health and environmental implications of the different fuel cycles;
- non-proliferation and safeguards aspects;
- international cooperation.
Each of the working groups wrote a key issues paper summarising the international common understanding on the issues covered. These papers were published prior to the symposium (Ref 1) and a final, edited version will be published as soon as possible. Furthermore, all orally presented papers will be published as an IAEA Technical Document (Ref 2) and the extended synopses of all oral and poster papers have also been published (Ref 3).
Our report to the Uranium Institute Symposium will concentrate on the two sessions dealing with the issues of plutonium management and the future fuel cycle and reactor strategies. It is, however, necessary to briefly touch on the outlook for nuclear energy for the next 50 years in order to establish a basis for the two major topics of this paper.
Global Energy Outlook
The key issues paper on global energy outlook and a paper from Los Alamos (Ref 4) discussed world scenarios from different viewpoints. The result is shown in Figure 1. Present nuclear capacity is 345 GWe, producing 2200 TWh in 1995, which corresponds to about 17% of the world electricity production. The key issues paper considered three nuclear energy scenarios with the following results in 2050: high variant (IAEA/HV) 1805 GWe, medium variant (IAEA/MV) 1132 GWe, and low variant (IAEA/LV) 333 GWe. The IAEA scenarios were built based on studies by the International Institute for Applied Systems Analysis (IIASA) and the World Energy Council (WEC).
The Los Alamos paper established an entirely new scenario based on a global energy, economics and environmental (E3) model, in which future demands for nuclear power are projected in price competition with other energy sources under a wide range of different drivers. Up to 2050 the Los Alamos curve (labelled "Total" in Figure 1) is not that different from the medium variant of the key issues paper. We are, however, not dealing with predictions but with scenarios. For our purpose - the possible requirements for the fuel cycle - it is enough to know that the truth is likely to lie between the high and the low variants. We will come back to the consequences of these scenarios later when we discuss the fuel cycle options for the next 50 years.
Plutonium Management
This session in the symposium dealt with the present status and immediate prospects of plutonium management up to 2015 - giving special consideration to the problems of plutonium management. If we look back some 20 years to the time of the IAEA's International Nuclear Fuel Cycle Evaluation (INFCE) project, not very much seems to have changed as far as policy is concerned.
INFCE was triggered by the conclusion reached by the United States that a once-through fuel cycle with no separation of plutonium had the least proliferation risk and the lowest cost. Several other countries decided at the time of INFCE that the best approach to the fuel cycle was to pursue programmes to reprocess spent fuel to remove plutonium, and to recycle it to recover its energy value, while relying on safeguards to prevent proliferation. Most of these countries have not changed their positions since.
In the meantime a large and viable recycling industry has been established in France and the UK, and is being established in Japan and Russia. This involves reprocessing plants, facilities for the manufacture of MOX fuel elements, and reactors licensed to use MOX fuel. Some countries, like Germany, have forgone establishing their own recycling facilities but use MOX fuel produced in other countries. Finally, it has become evident in these past decades that fast reactors will be delayed and their commercial introduction is not expected before 2030.
Both the key issues paper and a paper by the NEA (Ref 5) dealt with the experience accumulated in MOX fuel fabrication, both for thermal reactors and fast reactors. In summary, 630 t of MOX fuel have been fabricated and irradiated in reactors; currently 21 thermal reactors in five countries are loaded with MOX fuel and this number is expected to rise to between 36 and 48 by 2000 (Table 1).
The existing MOX fuel fabrication capacity is more than 200 t/y and will quickly rise to about 400 t/y. Therefore, recycling is now an established technology. However, it should be pointed out that in all three scenarios of the global energy outlook, 4000-5000 t of MOX fuel fabrication capacity will be needed by 2010 if all spent fuel is to be recycled.
The controversy over whether recycling creates higher proliferation risks and is more costly than the once-through cycle remains. Several studies have been performed. The NEA paper came to the conclusion that the cost differences between the reprocessing option and the direct disposal option are small (direct disposal being about 10% cheaper in power generation costs) and, considering the uncertainties, this small difference in the fuel cycle costs was considered to be insignificant. A recent study by the US Academy of Sciences, however, concluded that there was no reason to change the current once-through approach because recycling would bring no benefits and would increase the proliferation risk.
The
delay in recycling in either thermal or fast reactors has led
to a build-up of separated plutonium in inventories, which in
the view of many should be reduced - another of the new realities.
The key issues paper gave detailed figures on total generated
and separated plutonium and discussed the question whether and
how this plutonium inventory can be reduced. It may be worthwhile
to quote some of the figures for the end of 1995:
- About 970 t of Pu have been generated in 180 000 t of spent fuel.
- About 185 t of Pu have been separated.
- Of this amount, about 50 t have been recycled.
This means that about 135 t of separated plutonium were in store at the end of 1995. By the end of 1999 the estimated worldwide inventory of separated Pu will reach about 175 t. Through the introduction of MOX fuel this inventory will decrease after 1999 and level out at about 135-140 t (Figure 2). This model calculation assumes that each country is recycling only the plutonium that was produced in its own reactors.
Since there are countries that separate plutonium from their own reactors without recycling (ie the UK, with no plans to recycle, and Russia, with the intention to recycle) the inventory will eventually rise again. This rise in inventories could be averted only by either using the plutonium in these countries or by transferring the plutonium to countries that would use it. In such a case the inventory could be reduced to less than 50 t and would remain at that level, which would be required as the working stocks for the MOX and fast reactor fuel manufacturing plants. However, even if all political obstacles to such a plutonium transfer could be overcome in the time frame up to 2015, the economics of recycling would still affect the viability of this approach.
A very interesting subject discussed was the disposition of weapons plutonium. This was not only part of the key issues paper but was also the subject of two papers presented orally and several poster session papers. The fact that the two major nuclear weapons states ; as a result of the end of the Cold War - are willing to reduce their nuclear arsenals is an important new reality. This, of course, applies not only to plutonium but to HEU as well. The programme to blend down HEU of Russian origin to LEU for use in reactors has aptly been called the "Megatons to Megawatts Programme". Other very positive aspects of the disposition of weapons material are that it constitutes a truly international endeavour and that the material will be subject to international safeguards and verification at an early stage.
The programme for military plutonium disposition described in a paper from the Harvard Center (Ref 6) was the subject of summit meetings between Russia and the USA, and the programme was also discussed by the P-8 Group in Paris in October 1996. However, it will have to overcome many hurdles before it can be implemented. The US has declared 50 t of plutonium as excess to military needs; Russia has declared that it will also transfer fissile material out of its military programme but has not yet declared the quantities to be transferred.
It is widely agreed that the disposition programme should be accomplished as quickly as possible, but so far the necessary funds have not been identified and committed. It was fascinating, however, to hear about the tremendous amount of work that has been done under this joint programme and about the organisational structure that has been built up by the two countries to supervise the programme.
The IAEA symposium had in its title "new realities". The demilitarisation of weapons plutonium is one of the most important new realities, and it is heartening to learn that not only the two major nuclear weapon states are involved, but also other countries. Examples are the Canadian study into using CANDU reactors as an economic alternative for burning MOX fuel containing weapons Pu, the AIDA/MOX Russian/French programme with German participation, and the joint Russian/French/German project for a MOX fuel pilot plant for 1.3 t of weapons plutonium per year. The difficult task of ex-weapons plutonium disposition can only be accomplished as a truly international undertaking. Even so, the disposition of 50 t of Pu will only be completed 20-40 years from when the programme begins.
The US has adopted a "dual track approach" for disposition of former military plutonium, meaning that both using the plutonium in MOX fuel and disposing of it as vitrified waste should be pursued. However, the use of excess plutonium as fuel in reactors has been criticised in the USA as inconsistent with US policy of not supporting reprocessing and recycling of plutonium. The Harvard paper states: "The principal obstacle to implementing plutonium disposition in the United States is politics; the principal obstacle in Russia is money."
In discussing the problem of plutonium disposition it is necessary to keep the magnitude of the problem in mind. The USA has declared 50 t plutonium as surplus; if it is assumed that Russia will dispose of a similar quantity, the total will be about 100 t. In comparison, about 50 t of Pu have already been recycled and by 1999 civil stockpiles will reach about 175 t. Adding to this the ex-military material does not change the order of magnitude of the problem. If the civil material can be handled, then the ex-military material can be handled as well. What is needed is the political will to address and to solve the problem, the necessary funds and effective international cooperation.
If the need arises to dispose of large quantities of plutonium (of military or civil origin), and the capacity of thermal recycling is limited, as it may be in Russia, fast reactors could be used as plutonium burners.
Future Fuel Cycle and Reactor Strategies
This session of the symposium dealt with the time period from 2015 to 2050. What kind of developments in reactors and fuel cycle technologies are we going to see during this period? Will we be faced with entirely new technologies or will we rather have to rely on the continuous development of the existing concepts? How can we deal with the different scenarios for nuclear capacity discussed in the global energy outlook?
In dealing with these questions, the key issues paper first introduced a set of factors influencing the future reactor and fuel cycle concepts, then it discussed the different reactor systems, and ended by defining trends in the nuclear fuel cycle. The orally presented papers deepened or supplemented the analysis in the key issues paper.
It goes without saying that looking at the time period from 2015 to 2050 is more speculative than discussing the immediate future. Nevertheless, there was general agreement that the next 50 years will be dominated by thermal nuclear reactors and that these reactors will continue to play a significant role beyond 2050. So, by 2050 we will very likely see a mixture comprising a large population of thermal reactors with a small population of fast reactors. The number of fast reactors was forecast to grow steadily after 2050.
However, as a paper by an international group (Ref 7) stated, "it may become essential to branch out from familiar designs in order to maintain competitiveness with other energy supply options". In general, we will have a large number of new thermal reactors characterised by the word "advanced", eg ABWR and APWR, the EPR and the advanced CANDU. For the major part of the reactor population we can expect an evolution and not a revolution.
More revolutionary reactor concepts continue to be proposed, including the Radkowski Thorium Fuel (RTF) reactor, the modular high temperature gas-cooled reactor (HTGR), and combinations of nuclear power plants and gas turbines. These concepts will require significant further development before possible commercialisation. However, due to the inertia of nuclear technology it will be difficult for new concepts to succeed.
Let
us concentrate on the direction the evolution may take. The key
issues paper defines seven factors by which future reactor and
fuel cycle concepts must be judged. These are their ability to:
- properly utilise natural resources and national capabilities;
- maximise the economic benefits;
- effectively demonstrate the safety of fuel cycle facilities, and gain government and public approval for the enterprise;
- satisfy national and international policies and goals;
- contribute to sustainable energy supply.
What does this mean for the LWR fuel cycle? An international paper (Ref 8) introduced the concept of the "holistic" fuel cycle, ie considering the complete fuel cycle as a whole rather than as a number of individual stages. This concept integrates the various options for fuel manufacture, reactors, reprocessing, waste management and decommissioning to optimise the whole fuel cycle.
What fuel cycle technologies can be expected in these next 50 years? If we go back to the global energy outlook, it can be shown that uranium resources are sufficient to cover the requirements for the low and the medium variants. In the case of the medium variant, however, there may not be sufficient uranium resources to cover the years after 2050 for reactors existing in 2050 if one assumes that these reactors will have a total lifetime of up to 40 or even up to 60 years. Therefore, ways and means to make better use of resources, and their potential influence, are of importance.
Fuel
utilisation in thermal reactors can be improved in three ways:
- lower the tails assay in the depleted stream of enrichment plants;
- go to higher burnup fuel;
- recycle plutonium.
By reduction of the enrichment tails assay from 0.3% to 0.15% almost 25% of the uranium can be saved, compared to a saving of about 17% by recycling all plutonium in LWRs. Whether all these possibilities will be used will of course primarily depend on economic and political factors.
The gas diffusion enrichment plants will all be closed down in the time period we are considering, and will be replaced by centrifuge and/or laser plants. Centrifuge plants are built in small increments which enables the operators to bring in new, improved designs continuously. The centrifuge technology is therefore a moving target. However, if high uranium prices lead to very low optimum tails assays, or recycled uranium becomes important as a fuel, laser enrichment methods, which can separate isotopes very selectively, will have advantages. This is of course not to say that AVLIS, SILVA or MLIS may not be introduced in any case because of other economic advantages.
For newer reactors the operational lifetime may be 40-60 years; some older designs may, however, be retired earlier. The long term evolution of fuel production and manufacturing technologies will largely be dictated by the mix of reactors that will exist. Figure 3 gives an expiration schedule for existing nuclear power plants. The number of plants will start decreasing quickly after 2010, and by about 2035 all presently existing reactors will have been shut down. The key question for the future of our industry is whether these reactors will be replaced by nuclear plants or by fossil plants.
Figure 4 shows the world nuclear energy capacity from 1970-2050 taken from the medium variant of the global energy outlook. It also shows the amount of additional nuclear capacity required for the medium variant. Note that this figure assumes a 50-year reactor life, whereas a 40-year life was assumed in Figure 3.
Let us go back to the fuel technology. A burnup level of 55-60 GWd/t is viewed as the limit for technical and economic reasons for fuel which is to be recycled; only fuel for the once-through cycle may be driven to burnups in the order of 80-100 GWd/t. However, new cladding materials will be required. Advanced fuel concepts may be introduced gradually between 2015 and 2050. If other thermal reactor concepts like the HTGR are introduced, different kinds of fuel elements (which may have already been developed in the past), such as fuel using coated particles, will have to be produced. This fuel could achieve very high burnups, well beyond 100 GWd/t, and would also be very useful for burning plutonium.
In the case of recycling plutonium in thermal reactors there are limits to the number of recycles possible. Multiple recycling produces degraded plutonium, which limits the number of recycles in thermal reactors to two to three. Such degraded plutonium can, however, be used as a fuel in fast reactors. If such reactors or other effective plutonium burners do not materialise, spent fuel, although in reduced quantities, will still end up in final repositories.
The paper by Dr Ion et al (Ref 8) postulated an evolution in which advanced or new reprocessing technologies will be commercialised in the 2015-2050 period. Existing aqueous reprocessing technology, the 50-year old PUREX process, will remain as the base technology but will be constantly improved to achieve lower activity levels, less waste and cost reductions. Several programmes are being undertaken which are not only innovative and technically impressive but also confront us with many new acronyms, including SPIN, PURETEX, ACTINEX, DIAMEX and SESAME (Ref 8).
New flow sheets based on non-aqueous reprocessing methods could be introduced after 2015, the paper suggests. These processes tend to be shorter and simpler and also offer safety advantages. The methods considered are molten salt, fluoride volatility and, although it does not involve reprocessing, DUPIC (direct use of PWR fuel in CANDUs). The general trend, be it with conventional or with new technologies, is toward reduced costs and improved safety.
The costs of producing MOX fuel are also expected to become lower in the near future. Present MOX fuel fabrication costs are 4-5 times higher than for conventional uranium fuel; increased MOX fuel utilisation should lead to manufacturing costs of less than three times that of uranium fuel. Thus both reprocessing and MOX fuel fabrication will become cheaper.
Industrial fabrication of fast reactor MOX fuel is likely to be based on the proven MOX technology for LWRs. New developments may be closely linked to advanced PUREX and pyrochemical reprocessing methods, which will aim for co-extraction of plutonium and uranium with low decontamination factors. This will require new fabrication technologies, which may lead to the integration of fuel fabrication, reprocessing and waste treatment in one plant.
Several papers at the symposium dealt with partitioning and transmutation (P+T). The aim of P+T is to separate the actinides and long lived fission products and then reduce their toxicity by transmutation. This should reduce the long term hazard of reprocessing wastes. Again, we have different views in different countries: those that follow the recycle route (eg France, Japan and Russia) and expect to introduce fast reactors in the future believe P+T to be of importance; other countries, like the USA, have studied the technique and come to the conclusion that it would not produce major benefits (Ref 9).
Of great importance is of course the radioactive waste management and disposal problem. This concerns not only waste from past and present nuclear power plants, but also the huge amounts of "historic wastes" that are in store from weapons programmes. The general opinion was that disposal, be it direct disposal of spent fuel or disposal of vitrified waste coming from reprocessing plants, is primarily a political and not a technical problem. Several speakers also did not consider it as very urgent, because spent fuel and glass blocks can be stored for extended periods. However, there seems to be general agreement that for reasons of public acceptance at least one repository has to be brought into operation as soon as possible.
Fast reactors are expected to be introduced gradually between 2030 and 2050. The key issues paper gave a figure forecasting that by 2040 only between 10 and 15 GWe of fast reactor capacity will have become operational (Figure 5). Some participants consider this scenario as too optimistic. A paper authored by representatives of the three countries that still have a major fast reactor programme (Russia, France and Japan) (Ref 10) dealt in detail with the research and development targets and the outlook for fast reactors. This paper also comes to the conclusion that fast reactors will not be needed before 2030-2050, and that the major necessary improvements are economics and safety.
Some emphasis by oral and poster papers was been put on systems based on thorium fuel (Ref 11). Thorium represents a vast resource for nuclear energy. Although thorium fuels have been used in trials in several countries, it was the general opinion that widespread use in the next 50 years is going to happen only in a few countries, such as India. The reason for this is that there is a complete infrastructure for uranium fuel in existence and it is unlikely that a complete new infrastructure for thorium will be built up on a worldwide basis. If thorium is used it is probable that the fuel would contain thorium boosted by either uranium or plutonium, in a once through mode.
Summary
This paper dealt with only two of the sessions at the IAEA's symposium in detail (Present Status and Immediate Prospects of Plutonium Management and Future Fuel Cycle and Reactor Strategies) and mentioned briefly another session (Global Energy Outlook). As mentioned above there were three more sessions, and in addition a panel discussion. All in all the symposium covered the entire fuel cycle, comprehensively summarising the present situation and giving an outlook on the developments expected up to 2050.
In
concluding, some of the points that characterised the discussions
in the working groups should be mentioned:
- Being far from having achieved public acceptance, the nuclear community is faced with a new difficulty: in some countries nuclear energy has lost its competitive edge. Regaining competitiveness must be the main target for the near future.
- There is a strongly held view that in the long term the world cannot be supplied with clean energy without nuclear power. In the interim, however, we may witness the diminishing of the reactor population. Only in about 20 years will we know for sure how serious the threat of global warming is. The nuclear industry cannot be switched "on" and "off". Therefore, we will have to retain the facilities, resources, training and competence. This has to be done within an international network.
- The steering group discussed a postulated scenario in which a rapid take off of nuclear energy would be called for some time in the next century from a low starting point - after an extended period of stagnation. The conclusion was drawn that for such a scenario the technical and institutional arrangements would have to be already in place. Therefore it would be appropriate to prepare those arrangements in advance in order to avoid a crisis once the need for rapid deployment of nuclear power plants arose.