The fact that U-238 and Pu isotopes with uneven mass numbers are valuable fuels, in addition to U-235, resulted in early nuclear fuel strategies involving use of both thermal and fast reactors and the reprocessing of spent fuel. However, the low cost of natural uranium and the higher costs of fabricating MOX fuels has made the use of MOX as a fuel less attractive.
The Internationale Ländkommission Kerntechnik (ILK) carried out a study comparing the use of direct disposal and single, complete reprocessing of spent uranium fuel elements on the radiological impact of the nuclear fuel cycle. The study also looked at the impact of reprocessing on nuclear fuel cycle costs and the consequences for interim storage and final disposal.
The ILK study determined that both direct disposal and single reprocessing with subsequent final disposal of the vitrified high-level waste from reprocessing and of spent MOX fuel elements are feasible and can be safely implemented. In particular, the study showed that reprocessing does not lead to significantly higher radiation exposures of either people occupationally exposed or members of the public than the once-through use of fuel.
Direct final disposal and reprocessing have specific advantages and disadvantages. The arguments in favour of reprocessing are:
• Reprocessing all UOX fuel elements once and using plutonium in MOX will save about 20% of the natural uranium used, compared to direct disposal, for the same amount of energy, which can make a substantial contribution towards resource conservation.
• The amount of plutonium arising for final disposal after reprocessing of all UOX fuel elements is reduced by 30%, with the isotope vector of the plutonium contained in the spent MOX, improving the resistance of the material to unauthorised use.
• The uranium derived from reprocessing, which may well become an even more valuable nuclear fuel in the future, does not need to be placed in a final repository.
• The quantity of high-level waste (HLW) arising from reprocessing spent UOX fuel is reduced by a factor of five compared to direct disposal, and further reductions are possible. However, low- and medium-level waste arises during reprocessing which must also be disposed of in a geological repository. Consequently, the balance of waste volumes does not show any advantages for the reprocessing option.
The main arguments against reprocessing include:
• Reprocessing currently engenders fuel cycle costs 10-18% higher than those associated with direct disposal. However, fuel cycle costs only constitute one-third of the total energy production costs, and that rising world market prices of natural uranium will further reduce the cost advantage of direct disposal.
• Larger quantities of minor actinides are produced when MOX is used. However, these minor actinides are not very mobile in the chemical environment of a geological repository.
• Handling separated reactor grade plutonium in the fuel cycle entails additional safeguards. In contrast to the surveillance of fissile material for a geological repository, these measures are required for a comparatively short time – just until other technologies can replace the use of nuclear fission for power generation.
The radiation exposures stemming from final disposal are mainly determined by a few long-lived fission products such as Se-79, Zr-93, or I-129. Comparable quantities of these long-lived fission products requiring transfer to a final repository are produced in both alternatives. Accordingly, no significant differences between direct disposal and reprocessing will arise in terms of the radiation exposures resulting from final disposal.
ILK takes the view that there are no convincing technical and economic arguments for or against reprocessing with the current state of technology.
Reprocessing spent fuel for recycling in LWRs does not represent a suitable means of significantly reducing the radiotoxic inventory of waste because minor actinides are produced by neutron capture in U-238 and in the Pu isotopes. This applies equally to single and multiple recycling in LWRs.
The radiotoxic inventory in the waste stream and the potential radiation exposures from final disposal can only be diminished, and the threat of recriticality in final repositories can only be excluded completely by reprocessing, with advanced separation and transmutation technologies. This will require a combined use of thermal and fast reactors and possibly also subcritical accelerator-driven systems.
ILK states that these discernible future developments do not justify delaying the final disposal of radioactive waste in Germany. Even partitioning and transmutation will not achieve complete destruction of all radioactive waste for technical or economic reasons. Instead, they could very greatly reduce waste quantities, shorten the necessary confinement periods, and improve chemical adaptation of the technical barriers of the repository to the different species of radionuclides. In this regard, final repositories will be indispensable in the future, even though the strict geological requirements placed on such a repository could then be lowered.
Radiological impact of reprocessing
The Nuclear Energy Agency (NEA) of the OECD published a comparative study of the radiation impacts of various fuel cycles, comparing direct disposal and single reprocessing to MOX of all spent uranium fuel elements.
The study assumed a 1000MWe PWR and a burn-up of 40GWd/tHM. For the non-reprocessing option, the entire spent fuel ends up in the repository for high-level, heat-generating waste, while the reprocessing option results in vitrified waste from reprocessing and spent MOX fuel to be stored in a final repository.
The uranium derived from reprocessing was assumed not to be processed into fuel, and the tailings of uranium mining and milling to uranium ore concentrate are taken to be stabilised over long periods, so that the emanation of Rn-222 in particular can be permanently kept at a low level. The reprocessing technique chosen by Germany is the PUREX process (Plutonium Uranium Recovery by Extraction).
The recommendations by the International Commission on Radiological Protection (ICRP) constitute the basis for the radiological assessment of occupational radiation exposure and the radiation exposure of the public. The radiation doses received are described as annual individual doses for occupationally exposed persons or representatives of the most highly exposed groups of the population and as energy-specific collective doses. The collective doses are calculated by summation of the effective radiation doses arising over 500 years.
Individual doses must be viewed against the dose limits recommended by the ICRP of 20mSv/year over a five-year average for occupationally exposed persons, or 1mSv/year for the most highly exposed group of the public. These reference levels can be compared to the annual mean individual dose to the world population of 2.4mSv/year due to natural background radiation.
Limiting the accumulation period to 500 years is based on the consideration that radiation exposures are dominated by the release of the long-lived volatile nuclides C-14 (half-life 5730 years) as a consequence of energy production and reprocessing, and of Rn-222 with the precursor nuclide Th-230 (half-life 77,000 years) following uranium mining. Extending the observation period would not significantly change the radiological relations among the various fuel options while introducing further uncertainties into the generic assessments. There are grounds for assuming that this period of 500 years will see no significant releases from the geological repositories. Releases of long-lived fission products from the final repository only begin to play a role after thousands of years. Within the framework of the study, these releases are not significant in either magnitude or time.
The following detailed radiation exposures for the different fuel options with and without reprocessing are estimated, based on generic analyses of the study or statistical surveys of effective radiation doses in specific parts of the fuel cycle:
• The average measured effective individual doses lie below the ICRP reference levels for all stages of the fuel cycle. Workers in nuclear power plants receive the highest exposures. These exposures do not depend on the use or non-use of MOX.
• For energy-specific collective doses, the highest contributions of 1.0-2.7manSv/GWe result from energy generation in nuclear power plants. Differences between fuel strategies with and without MOX do not exist.
• Collective doses from reprocessing and vitrification of high-level waste are comparatively low at 0.014manSv/GWe.
• Fuel fabrication in a cycle with single reprocessing gives rise to higher collective doses of 0.094manSv/GWe than direct disposal only (0.007manSv/GWe). This difference is not fully compensated by dose reductions due to the 20% reduction in uranium quantities to be mined, milled and enriched in the reprocessing cycle.
• The exposures induced by transport of spent fuel and waste are negligible in every respect.
• According to the study, the specific collective dose of occupationally exposed people in total is 1.04-2.93manSv/GWe without reprocessing and 1.14-2.99manSv/GWe with single reprocessing.
It is safe to say, for both options, radiation exposure is low compared to the ICRP reference levels and to natural background radiation. The differences between the two options are not significant.
Interim storage and final disposal
The radiological consequences of single reprocessing and using MOX on wet or dry interim storage of spent fuel have been considered in the previously described dose levels. The effective doses resulting from interim storage of spent fuel are very low, and the differences between UOX and MOX are not significant.
The higher levels of heat generated by spent MOX fuel compared to uranium fuel do not pose any problems in interim storage. The inventory of fission products in MOX and UOX fuels is almost identical for the same burn-up.
The aim of the management of spent fuel and of medium- and high-level waste from reprocessing is final disposal in deep geological formations. However, no geological repository is currently in operation for these types of waste. There is, currently, no worldwide consensus on the necessary duration of safe confinement of spent fuel and HLW or on whether the fuel should be emplaced in the repository so as to be retrievable or not.
It is evident that a repository must satisfy the following criteria:
• Radiation exposures must be kept low enough to remain low compared to the natural background radiation.
• The repository must be designed and operated in a way that prevents unauthorised access to the fissile material.
• The spent fuel must be emplaced in a way that excludes the possibility of criticality in the repository.
Reprocessing 1t of UOX fuel with a burn-up of 33GWd produces 955kg of uranium, 10kg of all isotopes of plutonium, and about 2m3 of vitrified or cemented waste. The uranium can be used in fabricating reprocessed uranium fuel or can be placed into a final repository. The residual quantities of U-235 can also be recycled in the enrichment step in fabricating UOX fuel from natural uranium. The decay of U-232 and U-234 in uranium separated from reprocessing results in the radiotoxic decay products Th-228 and Tl-208 when that uranium is emplaced in a final repository. The final disposal of uranium from reprocessing would thus not only represent a waste of valuable fuel, but also add to the radiotoxicity and the fissile material inventory of the repository. This also applies, in principle, to the direct disposal of UOX fuel.
Separated plutonium is processed into MOX. A typical MOX fuel element for a PWR contains about 35kg plutonium. Seven UOX fuel elements must be reprocessed to produce this quantity. A spent MOX fuel element still contains some 25kg of plutonium. There is no benefit in radiotoxicity between single reprocessing and using MOX in LWRs, as the radiotoxic inventory of spent MOX can be 6-8 times higher than that of UOX. Nuclide balances very much depend on the fuel management strategy.
The fission products and minor actinides reach the repository in a vitrified or cemented or compacted form. It is estimated that the amount of material containing the radiotoxic inventory can be reduced by a factor of five by recycling uranium fuel as MOX with current reprocessing technologies. The entire waste volume is 1.54 times greater for reprocessing than for direct disposal.
Overall, it can be said that the peculiarities of MOX fuel as a waste form are not as well known as for uranium fuel. However, there are no indications that the final disposal of MOX fuel elements might lead to problems in the final repository. The content of transuranic elements, which is greater by a factor of 3-4 and which dominate the radiotoxic inventory for thousands of years does not pose a special difficulty for the final disposal of MOX. Transuranic elements are relatively immobile in deep geological formations compared to long-lived fission products.
Radiological studies show very low radiation exposures. Studies of a typical scenario for the Gorleben salt dome demonstrate that individual dose rates do not exceed 0.7mSv/year for an equivalent emplacement quantity of 73,000tHM. Radiation exposures from the iodine inventories emplaced dominate up to a disposal period of 3000 years. Up until 100,000 years, the main doses are contributed by Se-79. Subsequently, contributions from the Np decay chain dominate. After 200,000 years, the Np decay chain, on the one hand, and the combined effects of Tc-99 and Cs-135, on the other hand, make comparable contributions totalling 0.2mSv/year. After 106 years, exposures are determined solely by the decay chains of U and Np. By that time, however, a very low level of 0.01mSv/year has been reached.
Multiple recycling of LWR fuel (reprocessing of spent MOX) is possible. The spent MOX fuel is dissolved in nitric acid with spent uranium fuel in order to adhere to the process constraints with regard to dose rate limits and plutonium concentration. However, recycling is possible only for a limited number of times, as plutonium isotopes accumulate in the MOX which cannot be burnt in the thermal spectrum of a LWR or are highly radioactive. In addition, more plutonium is bred continuously as a result of neutron capture in the U-238 in the MOX. This makes the destruction of plutonium in LWRs very inefficient from a radiotoxic point of view.
That the amount of fissile material reaching a repository can be decreased by reprocessing and by using MOX is positive as far as criticality safety is concerned, and also makes unauthorised access to such inventories less attractive. The change in the plutonium isotope vector also impedes any misuse of this fissile material.
There are two scenarios which could, theoretically, lead to the possibility of criticality developing in the repository. In the first scenario, fuel that has not yet reached full burn-up is emplaced in the repository, and water enters the final disposal containers. Such a scenario can be prevented by selecting configurations that cannot go critical even in the case of water ingress, or by lining the containers with neutron-absorbing materials. In the second scenario, water entry into the repository is again assumed. As a possible consequence, the waste disposal containers would be destroyed and Pu-239 – or its decay product, U-235 – would be transported mainly as colloids in the aqueous phase. Another assumption is that the colloids accumulate in waterlogged cracks, fissures and pores of the repository host rock formation, thereby potentially leading to a critical configuration.
The risk of recriticality in the repository is estimated to be very low for all cases.
Cost comparison
There have been several studies comparing the economic viability of direct disposal and single reprocessing.
In the study by the NEA during 1994, the costs for the entire fuel cycle, including costs of natural uranium, conversion, enrichment and fuel element fabrication, were considered. According to the study, a 1390MWe PWR with a mean burn-up of the fuel elements of 42.5GWd/tHM gives a 10% cost advantage for direct disposal over reprocessing.
The 1995 study by the Institute for Energy Economics of the University of Cologne shows that direct disposal is the more economical option under various fuel management strategies for a 1300MWe LWR. This applies when only considering the spent fuel management costs and also when referring to the entire fuel cycle. According to that study, the cost advantage of direct disposal is around 18% when the entire fuel cycle is considered, and remains constant even if uncertainties are taken into account.
These comparisons of economic performance are associated with major uncertainties, as many detailed cost items to be considered, such as the prices of natural uranium and the costs of final disposal, depend on developments which cannot be accurately estimated. Furthermore, the relative cost advantage of direct disposal loses some of its significance as a consequence of fuel cycle costs constituting only one-third of the entire generating costs. At present, however, the stable low uranium fuel costs, the high burn-up attainable, as well as the higher fabrication costs of MOX fuel elements make reprocessing of spent uranium fuel and the use of MOX fairly unattractive to utilities.