In Australia all uranium mining and milling operations are undertaken under the Code of Practice and Safety Guide for Radiation Protection and Radioactive Waste Management in Mining and Mineral Processing. This was drawn up by the national government in line with recommendations of the International Commission on Radiological Protection (ICRP), but it is administered by state health and mines departments. The Code, which was updated in 1995 and again in 2005, sets strict health standards for radiation and radon gas exposure, for both workers and members of the public.

In Canada the Canadian Nuclear Safety Commission is responsible for regulating uranium mining as well as other aspects of the nuclear fuel cycle. In Saskatchewan, provincial regulations also apply concurrently, and set strict health standards for both miners and local people.

Uranium itself is only slightly radioactive. However, radon, a radioactive inert gas, which is a decay product of uranium, is released to the atmosphere in very small quantities when the ore is mined and crushed. Radon occurs naturally in most rocks — minute traces of it are present in the air which we all breathe and it is a significant contributor to the natural radiation dose that we all receive. Because it is airborne, special care must be taken to ensure that mine worker exposure, especially in poorly ventilated mines, is limited.

Open cut mines are naturally well ventilated. The Olympic Dam and Canadian (as well as other) underground mines are ventilated with powerful fans. Radon levels are kept at a very low and certainly safe level in uranium mines. (Radon even in non-uranium mines also may need control by ventilation.)

Gamma radiation may also be a hazard to those working close to high-grade ores such as in Canada. It comes principally from uranium decay products in the ore, so exposure to this is regulated as required. In particular, dust is suppressed, since this represents the main potential exposure to alpha radiation as well as a gamma radiation hazard.

At the concentrations associated with uranium (and some mineral sands) mining, radioactivity is a potential health hazard. Precautions taken during the mining and milling of uranium ores to protect the health of the workers include:

• Good forced ventilation systems in underground mines to ensure that exposure to radon gas and its radioactive daughter products is as low as possible and does not exceed established safety levels.

• Efficient dust control, because the dust may contain radioactive constituents and emit radon gas.

• Limiting the radiation exposure of workers in mine, mill and tailings areas so that it is as low as possible, and in any event does not exceed the allowable dose limits set by the authorities. In Canada this means that mining in very high-grade ore is undertaken solely by remote control techniques and by fully containing the high-grade ore where practicable.

• The use of radiation detection equipment in all mines and plants, often including personal dose badges.

• Imposition of strict personal hygiene standards for workers handling uranium oxide concentrate.

At any mine, designated employees (those likely to be exposed to radiation or radioactive materials) are monitored for alpha radiation contamination and personal dosimeters are worn to measure exposure to gamma radiation. Routine monitoring of air, dust and surface contamination is undertaken.

Canadian mine and mill facilities are designed to handle safely ore grades of up to 26% U.

If uranium oxide is ingested it has a chemical toxicity similar to that of lead oxide. Similar hygiene precautions to those in a lead smelter are therefore taken when handling it in the drying and packing areas of the mill.

The usual radiation protection procedures are applied at an ISL mine, despite the fact that most of the orebody’s radioactivity remains well underground, and there is hence minimal increase in radon release and no ore dust.

An appreciable advantage of thorium-based fuel is the potential to reach very high burn-ups so that the number of fuel assemblies needed to achieve a given energy output is reduced. This would produce savings in fuel manufacture and reduce the amount of waste that was produced although burn-up cannot be increased indefinitely, of course.

The interim storage of thorium spent fuel shows characteristics a little less constraining than those of uranium-based spent fuel because of the relative chemical inertness of thorium. In consequence, maximum acceptable temperatures for dry storage of spent UO2 fuel are lower than for thorium fuel because at higher temperatures, UO2 fuel may oxidize to U3O8 with a volume expansion that may rupture the fuel cladding. Matrix oxidation is not an issue with thorium-based fuels. Further, oxidation of minor solid-solution components such as uranium and plutonium can be easily accommodated within the thorium fuel matrix. Consequently, fuel oxidation is unlikely to be a concern during dry storage of thorium-based fuels and the maximum storage temperature may be limited by other factors such as cladding degradation.11 Similar points concerning the reaction of thorium fuel with water may be made with respect to wet interim storage.

Direct disposal of thorium-based fuels is attractive from the standpoint of long­term behaviour in a geological repository, because thorium oxide is chemically stable and almost insoluble in ground water. The most important chemical difference between thorium and uranium oxides is that thorium is present in its maximum oxidation state whereas uranium is not. Under oxidizing conditions, uranium can be converted into the water-soluble uranyl cation UO2 2+ and its various derivatives. Not only does this produce a mobile radionuclide, it also degrades the fuel, releasing the actinides and fission products, which are contained within it. Conversely, radionuclide release from thorium oxide fuel is expected to be limited by the low solubility of ThO2 and its low cation diffusion coefficient.

No credible aqueous or geochemical process has yet been identified that would greatly accelerate ThO2 fuel-matrix dissolution under disposal conditions.11

Disposal of fission product waste after reprocessing of thorium-based fuel would require treatment similar to that of waste from reprocessed UO2 fuels. Although thorium-based fuel cycles may produce much less plutonium and associated minor actinides than uranium-based fuels, they will instead generate other radionuclides such as Pa-231, Th-229 and U-230, which will have a long­term radiological impact.

Nevertheless, the global radiotoxic inventory (GRI) of waste to be disposed when using a thorium cycle appears to be significantly less than for the standard uranium-plutonium cycle, for the same energy output. This is a real asset for thorium-based fuels which has been confirmed in several studies, such as in a recent one, performed under an EC contract.7 The main findings of these studies are as follows:

• Where only the major actinides are recycled and reused with a thorium matrix (i. e. assuming that all other actinides such as Np, Am, Cm or Pa go to waste), the GRI of the as-disposed thorium cycle waste is reduced by a factor of 10 compared to the uranium-plutonium cycle. As the disposed waste decays, the two GRI values come closer together so that, after 10 000 years, the ‘thorium GRI’ is greater than the ‘uranium GRI’. This is not seen as a major problem, however, because beyond a few tens of thousands of years, both GRI values are relatively low. They are, for example, lower than that of the amount of natural uranium needed to feed a once-through reactor programme of equivalent energy output.

• In the case of recycling of all minor actinides (assuming 0.1% losses to the waste), the ‘thorium GRI’ is less than the ‘uranium GRI’ by a factor of between about 5 to 20 for all times up to 10 000 years. After 20 000 to 30 000 years or so, the ‘thorium GRI’ becomes the greater but, as in the previous case, the absolute values are relatively low being, again, lower than that of the equivalent amount natural uranium used in a once-through cycle.

Being a LWR most of the basic reactivity characteristics of the BWR are similar to those of the PWR (negative void coefficient, etc.). However, the fact that it is a direct cycle boiling system does mean that the overall response characteristics are subtly different. This is best illustrated by considering the response to a step increase in turbine demand.

10.12 Reactor water cleanup system (Source: USNRC).

The initial response to an increase in steam demand is a fall in the steam line pressure and a consequent increase in the steam generation rate. In a PWR this will increase the evaporation in the steam generator leading to a fall in level and increased cooling of the primary coolant. This will reduce the temperature of the coolant entering the core, which will increase the core power, a response which will be reinforced by the control system. The control system will increase the steam generator feed flow and reactor power to match the demand change, but the natural characteristics of the system aid this.

In a BWR the increased evaporation will be in the core, which will increase voidage and thus reduce reactivity and hence core power. Thus the control system must compensate for this power drop as well respond to the increased demand. This can be achieved by increasing the core flow as well as by moving control assemblies. Increasing the flow will raise the boiling boundary increasing moderation and hence power.

Because the coolant boils in the core the power production and reactivity characteristics are not uniform, because the moderator feedback changes with core height. The majority of the power is produced in the lower parts of the core where moderation is more effective. This is the reason why BWR control assemblies enter from the bottom of the core. The control and shutdown assemblies are positioned where small movements have relatively large effects so that they will be immediately effective as they start to insert. A normal gravity driven system with the control assemblies positioned above the core would have poor control characteristics since they would be relatively ineffective until they reached regions of higher water density. Thus the control assemblies are bottom entry to be more effective. These are discussed in the next section.

The Magnox and AGR reactor designs are unusual in their ability not only to detect the presence of a defective fuel element but also to locate it for discharge. In the case of Magnox reactors, the prime reason was the rapid oxidation of uranium metal under defective fuel cladding leading to swelling, loss of heat transfer and possibly fire. In the case of AGRs, the primary concern was release of fission product to the coolant, leading to not only contamination of the circuit but also releases of radioisotopes (especially I-131) to the environment in the event of a depressurisation incident or from routine coolant leakage (typical AGR leak rates are of the order 1-3% of coolant inventory per day).

The detection systems used on both types of reactor were similar. Samples of gas were drawn from the reactor and their activity measured in the burst can (or cartridge) detection equipment (BCD). Measurements of noble gas fission products, which leak from defective fuel pins and elements, are used to detect the presence of failed fuel: the radioactive isotopes of xenon and krypton. Only short­lived isotopes are of use for locating defective fuel, as long-lived isotopes are soon mixed into the general coolant flow with little enhancement in the channel containing the fuel defect.

A bulk activity measurement is not appropriate as the coolant carries standing levels of beta and gamma emitting isotopes such as N-16, O-19 and Ar-41. In order to discriminate the fission gas isotopes from gaseous activation products, use is made of the fact that the majority of the short half-life (tens of seconds) krypton and xenon fission gas isotopes have short half-life radioactive decay products, isotopes of rubidium and caesium respectively.

The coolant sample flow is directed into a ‘precipitation chamber’ containing a wire held at high voltage. The wire attracts charged decay products of the fission gases to its surface. After a ‘soak time’ of a minute or so, the wire is fed into a counting chamber, isolated from the gas sample flow, and the radioactive progeny of the rare gases are detected. This simple system has proved to be robust and sensitive for the detection of failed fuel.

In order to locate the channel containing the failed fuel, quite complex systems of pipework and valves are used in which large sections of the reactor can be sampled, followed by smaller groups of channels and eventually single channels. By scanning sequentially to the zone with the highest signal, the failure channel can be (in principle) rapidly identified and (where on-load refuelling is possible) the fuel can be discharged promptly. The wire is spooled around the BCD precipitator in a continuous loop.

Various supplementary systems have been used, particularly in AGRs, including continuous sampling of coolant and measurement by a high-resolution gamma ray spectrometer (not available when the Magnox and early AGRs were being constructed), and sampling of bulk coolant through a charcoal pack to determine the levels of radio-iodine.

Elastic collisions of fast neutrons with metal atoms in the cladding knock the metal atoms from their lattice sites. The result of this irradiation damage is concentrations of vacancies and interstitials (‘point defects’) in the cladding crystal structure, which are well above those due to thermal effects. The irradiation damage causes hardening (a higher resistance to plastic deformation) and embrittlement (loss of ductility) of the cladding. Additional hardening and embrittlement can be caused by: precipitation, and subsequent growth due to diffusion of helium bubbles; hydrogen pick-up from the coolant (in water-cooled reactors only); diffusion of oxygen from the cladding oxide layer into the cladding metal (in water-cooled reactors only, and only during accident conditions where high clad temperatures are achieved); and selective dissolution of cladding constituents (see 14.2.13). The helium is formed by neutron capture of the cladding alloy constituents and impurities. The hydrogen picked up is primarily created by the chemical reaction of the cladding and the coolant. Diffusion of the point defects can lead to recombination, absorption at sinks (dislocations, grain boundaries and surfaces), formation of two-dimensional dislocation loops, or, in the case of vacancies, formation of three-dimensional clusters known as voids. Nucleated voids can themselves subsequently act as vacancy sinks — the result is void swelling of the cladding.

In reactors with zirconium alloy clad fuel, the combination of the manufacturing process and the hexagonal close-packed structure of zirconium leads to an anisotropic crystal structure (the cladding is said to have ‘texture’), with the basal planes of the unit cells tending to orient in the axial direction. Since dislocation loops composed of interstitials are more favourably formed along the basal planes, while dislocation loops composed of vacancies are more favourably formed normal to this direction, the lattice expansion due to the interstitial loops and the lattice contraction due to the vacancy loops tend to occur in different directions. The net result is an elongation of the cladding such that the cladding volume is maintained constant. In the case of PWR fuel assemblies, this axial growth is also exhibited by the guide tubes, which can lead to deformation of the fuel assembly. (One consequence of this is the operational fault condition known as incomplete rod insertion, whereby bowing of the guide tubes prevents full insertion of control rods.)

Hydrogen pick-up is significant in zirconium alloy clad fuel (~16% for Zircaloy-4 in an LWR (Kaczorowski et al.. 2008)). The hydrogen is primarily generated by the chemical reaction of zirconium and water, i. e. Zr + 2H2O ^ ZrO2 + 2H2 . In the case of fresh cladding and cladding irradiated to moderate burnups, the hydrogen levels in the cladding are low and the hydrogen is generally in the form of a solid solution. However, at higher burnups clad hydrogen contents are greater, and significant quantities of hydrogen can precipitate out as zirconium hydride platelets, which are brittle. Both void swelling and helium generation are negligible in zirconium alloy cladding.

I n reactors with stainless steel clad fuel, the cladding crystal structure is isotropic, so dislocation loops are randomly oriented and there is no axial growth (neither is there radial or circumferential growth). However, in fast reactors the high flux of fast neutrons leads to extensive nucleation of voids, and the clad temperatures are high enough that significant diffusion of vacancies can occur. Thus, void swelling is generally significant (there is a strong dependence on steel type — the austenitic steels tending to swell more than the ferritic or ferritic/martensitic varieties — and the manufacturing process, in particular the amount of cold work). In fact, void swelling is such that it is often the limiting phenomenon with respect to the fast neutron dose that can be accumulated in a fast reactor. Generation of helium via neutron capture is also significant in stainless steel cladding. This is mainly due to thermal neutron capture of nickel and boron, and to fast neutron capture of iron, chromium, nickel, boron and nitrogen. The former dominates in thermal reactors (principally AGRs), while the latter dominates in fast reactors. The result is hardening and embrittlement of the cladding. Hydrogen pick-up is negligible in stainless steel cladding.

P. NETTER, AREVA, France

Abstract: This chapter introduces open and closed fuel cycles. There are discussions about reprocessing targets and constraints. The separation and purification of uranium and plutonium are described. A complete closed cycle for both uranium and plutonium is set out. Finally, the industrial-scale spent fuel reprocessing strategies for selected countries are given.

Key words spent nuclear fuel, uranium reprocessing, plutonium reprocessing, PUREX process.

16.1 Introduction: closed and open cycles

The need to reduce greenhouse gas emissions, the depletion of fossil natural resources and rising energy demand are all factors that support the continued development and deployment of nuclear energy. Additionally, there is the option to go further: by reprocessing the spent fuel, uranium and plutonium can be recovered for recycling. While this releases an otherwise unavailable energy resource, the decision to reprocess is, nevertheless, a political matter in which the policy varies from state to state. Thus we find that some countries that utilize nuclear power reject reprocessing/recycling. Further, the Fukushima accident (14 March 2011) has relaunched the debate over nuclear safety and reuse of reprocessed material (mixed oxide, MOX fuel) that seemed to be fading away with the memory of Chernobyl (26 April 1986).

It is helpful, at the outset, to distinguish between open and closed fuel cycles.

In the open cycle, used fuel is considered as waste, i. e. the owner takes the view that there is no value in the used fuel. Storage is the prevailing practice and it is considered that there is no merit in doing anything other than interim storage followed by disposal. Many countries (e. g. USA, Sweden, Finland) have adopted this approach and used fuel may be stored in pools or in dry storage systems at purpose-built sites. Storage solutions currently on the market enable spent fuel to be managed over a period of several decades or even longer. Permanent disposal of used fuel envisages burial in a deep geological formation where its long-term safety can be assured.

In the closed cycle, used fuel is recycled, Fig. 16.1, i. e. reprocessed to separate the useful materials — uranium and plutonium — from the minor actinides and fission products so that they can be incorporated into new fuel. The unwanted components go into a number of different waste streams. By far the greatest

proportion of the radioactivity (if not the volume), however, goes into high-level wastes (HLW), which are transformed into glass blocks. Countries that have chosen the closed fuel cycle option are France, United Kingdom, Japan, Russia, China and India (Table 16.1). Still in development are reprocessing strategies that aim to make use of the thorium cycle or use advanced techniques for partitioning and transmutation, including the use of dedicated ‘waste burning’ reactors in order to reduce the long-term radiotoxicity of the radionuclides in the waste. With reference to the example of pressurized-water reactors, it can be stated that for every 100 kg of fuel, and depending on burn-up and fuel type, 96-97 kg can be recycled of which 95-96 kg is residual uranium (recyclable as UOX, uranium oxide fuel) and 1 kg is plutonium (recyclable as MOX, mixed oxide U-Pu fuel). The remaining 3^ kg are fission products (high-level waste), which are then incorporated into a glass matrix or with long-lived intermediate waste from spent fuel assemblies (hulls and end-caps). Reprocessing and recycling are energy production: 1g plutonium or 100 g uranium is equivalent to 1 or 2 ton oil. 1000 ton of reprocessed and recycled spent fuel generate about as much energy as 25 million tons of oil. Additional to energy production is a reduction of the volume of the wastes (through compaction or incorporation into a glass).The volume of wastes can be reduced to the range of 0.5 m3/mtU. Overall, reprocessing used fuel to recover uranium and plutonium avoids the wastage of a valuable resource. Recycling can save up to 30% of the natural uranium that would otherwise be required.

Natural uranium

Chemistry

Mining

Recycled uranium

Enriched uranium

Plutonium

Waste final disposal

Recycling MOx fuel fabrication

Fuel fabrication

Reprocessing

16.1 The nuclear fuel cycle (Source: AREVA).

The current global nuclear capability of light water reactors generates about 7000 tons of used fuel per year. Stores of used fuel accumulated across the world amounted to around 172 000 tons U by 2007, of which 32 000 have been recycled.

Throughout the world, the standard method for the separation of uranium and plutonium use aqueous solution processes (wet routes) — liquid-liquid extraction. Alternative routes exist, especially those using pyrochemical processes and dry methods but these are small scale and/or in development.

All commercial reprocessing plants use the well-proven hydrometallurgical PUREX (plutonium uranium extraction) process. This involves dissolving the fuel elements in hot concentrated nitric acid. Chemical separation of uranium and plutonium is then undertaken by solvent extraction steps. The Pu and U can be sent to the input side of the fuel cycle — the uranium to the conversion plant prior to re-enrichment and the plutonium straight to MOX fuel fabrication.

Alternatively, some of the recovered uranium can be left with the plutonium, which is sent to a MOX plant, so that the plutonium is never fully separated out. This is known as the COEX (co-extraction of actinides) process, developed in France as a ‘Generation III’ process, but not yet in use. Japan’s Rokkasho plant uses a modified PUREX process to achieve a similar result by recombining some uranium before denitration, with the main product being 50/50 mixed oxides.

In either case, the remaining liquid after Pu and U has been removed is high-level waste, containing about 3% of the used fuel in the form of fission products and minor actinides (Np, Am, Cm). It is highly radioactive and continues to generate a lot of heat. It is conditioned by calcining and incorporating the dry material into compact, stable, insoluble borosilicate glass, then stored pending disposal.

One of the incentives for recycling spent fuel discharged from nuclear reactor cores is that it meets the dual requirements of a sustainable development policy’.

• Recovery and recycling of reusable materials, uranium and plutonium, so that demand for natural uranium is reduced

• Waste minimization. Reprocessing reduces waste toxicity (both waste volume and radiotoxicity), by conditioning it into stable canisters adapted to its level of activity and half-life for disposal.

In Russia, the Mayak Plant B operated from 1949 to 1960; Plant BB operated from 1957 to 1987. Plant RT-1 (PUREX Process_400 t/year) began operations in 1976. At Krasnoyarsk-26, processing of plutonium production reactor fuel began in 1964 also using the PUREX process. Construction of a new RT-2 plant began in 1972 with an envisaged capacity of 1000-1500 MTHM/yr but plant construction was never completed. The Tomsk-7 plant processed plutonium production reactor fuel, again using the PUREX process, beginning sometime after 1955.

Russian policy3 is to close the fuel cycle as far as possible and utilize recycled uranium, and eventually also to use plutonium in MOX fuel. However, its achievements in doing this have been limited — in 2011 only about 16% of used fuel was reprocessed. All used fuel is stored at reactor sites for at least three years to allow decay of heat and radioactivity. High burn-up fuel requires longer before it is ready to transport. At present the used fuel from RBMK reactors and from VVER-1000 reactors is stored (mostly at reactor sites) and not reprocessed. It is expected that used fuel in storage will build up to about 40 000 tonnes by the time substantial reprocessing gets under way about 2022. The material reprocessed will be burned in fast reactors by 2050, when none should remain.

In late 2007 it was decided that MOX fuel production using recycled materials from both light-water and fast reactors should be based on electrometallurgical (pyrochemical) reprocessing. The goals for closing the fuel cycle are minimizing cost, minimizing waste volume, recycling of minor actinides (for burning), avoidance of separated plutonium and executing all procedures in remote-handled systems.

Used fuel from VVER-440 reactors Kola 1-4 and Rovno 1-2 (in Ukraine), the BN-600 (Beloyarsk) and from naval reactors is sent to the Mayak Chemical Combine’s 400 t/yr RT-1 plant (Chelyabinsk-65) at Ozersk, near Kyshtym, 70 km north-west of Chelyabinsk in the Urals for reprocessing. The original reprocessing plant at the site was hastily built in the mid-1940s, for military plutonium production in association with five producer reactors (the last shut down in 1990). The RT-1 plant started up in 1971 and employs the PUREX process. It is reported to be running at about 100 t/yr capacity, following the loss of foreign contracts, but also that reprocessing does not keep pace with inputs, so some is stored there. About 93% of its feed is from Russian and Ukrainian VVER-440 reactors, about 3% from naval sources or icebreakers and 3% from BN-600. It earlier reprocessed BN-350 used fuel.

Recycled uranium is enriched to 2.6% U-235 by mixing a RepU product from different sources and is used in all fresh RBMK fuel, while separated plutonium is stored. High-level wastes are vitrified and stored. Plans to upgrade the RT-1 plant and enable it to take VVER-1000 fuel, have been approved and were to be completed in 2008. The 2009 federal program has it reaching 500 t/yr from 2012. Used fuel storage capacity there is being increased from 6000 to 9000 tonnes.

VVER-1000 used fuel is sent to the Mining & Chemical Combine (MCC) at Zheleznogorsk (Krasnoyarsk-26) in Siberia for storage. This comes from three Russian, three Ukrainian and one Bulgarian plant. A large pool storage facility

was built by MCC at Zheleznogorsk in 1985 for VVER-1000 used fuel, though its 6000 tonne capacity would have been filled in 2010. The facility was fully refurbished over 2009-10. In December 2009 Rostechnadzor approved expansion to 7200 tonnes and in August 2010 MCC was seeking approval to expand it to 8400 tonnes capacity to allow another six years’ input.

A Pilot Demonstration Centre (PDC) for several reprocessing technologies is under construction by MCC at Zheleznogorsk at a cost of RUR 8.4 billion, to be commissioned by 2015. Its initial capacity will be 100 t/yr, with a later increase to 250 t/yr. The cost of the RepU product is expected to be some EUR 500/kg. (A dual-purpose graphite-moderated reactor principally producing military plutonium, with associated underground reprocessing plant, is also there.)

The partly built larger RT-2 reprocessing plant at Zheleznogorsk was cancelled and was to be dismantled. However, this has been under review and it could form part of the new Global Nuclear Infrastructure Initiative. It is now being redesigned and is expected to operate from around 2025-30.

Since 2004 an 8600 tonne dry storage facility for used fuel (INF DSF-2) has been under construction at Zheleznogorsk and this was completed by the E4 Group at the end of 2011 at a cost of about US$ 500 million for the MCC. It is the largest dry storage facility in the world and will take 8129 tonnes of RBMK fuel, initially from Leningrad and Kursk power plants, followed by Smolensk. RBMK fuel is not presently economic to reprocess so is stored at reactor sites, and when transferred to MCC about 2012 will be stored in sealed shrouds. Further stages of MCC dry storage will take VVER-1000 fuel and increase capacity to 38 000 tonnes by 2016. Used fuel will be stored for up to 50 years, pending reprocessing.

In June 2011, Rosatom announced that it was investing RUR 35 billion in MCC to 2030, including, in particular, MOX fuel fabrication. In February 2012 the figure was put at RUR 80 billion minimum.

Bilibino’s LWGR used fuel is stored on site.

A small MOX fuel fabrication plant has operated at the Mayak plant at Ozersk since 1993. A 60 t/yr commercial MOX fabrication plant is under construction by MCC at Zheleznogorsk (the site of the ADE2 military plutonium production reactor). Another MOX plant for disposing of military plutonium is planned at Seversk (Tomsk-7) in Siberia, to the same design as its US equivalent. (Seversk had the other two dual-purpose but basically military plutonium production reactors, totalling 2500 MWt. One of these — ADE4 — was shut down in April 2008, the other — ADE5 — in June 2008.)

18.1.2 Waste segregation and characterisation

The operational safety case for a disposal facility will not be so very different from that of a waste or spent fuel store. The radionuclides of greatest concern are relatively short-lived gamma emitters (e. g. cobalt-60), which may be detected with a gamma spectroscope or even with a simple hand-held counter. The situation for the long-term (i. e. post-closure) safety of a disposal facility is quite different. Here, the radionuclides with the greatest influence are invariably long lived and often difficult to measure. Typically, they include fission products such as technetium-99, iodine-129 and the actinides. In these cases (with the exception of some actinides) sophisticated laboratory tests are needed to obtain specific activity values. This can be made very much more difficult if the waste has been processed in some way (e. g. mixed with concrete) and, even more so, if different waste streams have been mixed in an uncontrolled way. Consequently, it is essential that waste streams should be kept separate and that characterisation should be performed before any processing is done. Such precautions should allow the radionuclide content of waste streams — and thus waste packages — to be adequately specified so that compliance with waste acceptance criteria can be demonstrated.

Where the radionuclide content of a waste stream is variable, it is usual to assume that, while the total activity may change, the relative proportions of the various radionuclides do not. In this case it may be possible to establish a correlation between difficult-to-measure and easy-to-measure radionuclides. Typically, caesium-137 is used as an indicator for other fission products (e. g. iodine-129) and cobalt-60 is used for neutron activation products such as chlorine-36 and nickel-63.6 There are, however, difficulties with this approach for same radionuclides.7

For spent nuclear fuel, vitrified fission products and decommissioning wastes, the radionuclide content of the material is most often derived by calculation using standard codes and neutron cross sections.

The LCOE calculation makes no allowance for transmission costs on the grounds that these apply more or less equally to all technologies. This is an assumption that may not be entirely valid. Two obvious cases serve to make the point. The first is the case of wind generators that, in Europe, are predominantly located on the west coast. In North America the preferred location is the Great Plains. Both areas are relatively remote from population centres. The second case is nuclear power where, for reasons of safety, remote locations are preferred. In both these cases it may be necessary to install a major transmission line where none existed before. The omission of this cost represents a bias in favour of these technologies

A second factor is that transmission lines are designed and built to carry the maximum power output of the generators, which that the lines serve. Since, on average, renewable generators operate at a fraction of their nominal output, they will not fully use the capacity of the lines and this represents an overinvestment in resources. In this case the effect is likely to be small because we are only concerned with an effect at the margin — i. e. it is only the size of the transmission line, not its existence, that is in question.

Hidden subsidies

LCOE takes no account of hidden subsidies such as limited liability in the event of a nuclear accident. Without this, nuclear utilities would find it impossible to obtain insurance for low-probability, high-consequence accidents and the technology would probably become commercially non-viable. Other hidden subsidies may include government funding for research and development, and regulatory fees.

There are two technologies that have been used for aerodynamic isotope separation, the jet nozzle and the advanced vortex tube. They use similar principles to the gas centrifuge but in these cases the circular wall remains static while the gas travels in a circular path at very high speed.

The jet nozzle process was initially developed in Germany and then transferred to Brazil with continued German participation. A small demonstration plant was built before work ceased in 1994. The process fires UF6 in a hydrogen carrier gas through a slit-shaped nozzle at very high velocity towards a semicircular wall. The fast moving gas and the static wall have much the same effect as a centrifuge, with the heavier 238UF6 tending to stay closer to the wall than the 235UF6. As the gas leaves the wall the stream may be split into slightly enriched and slightly depleted streams, as illustrated in Fig. 7.6.

The principle of the jet nozzle process is simple but the radius of the semicircular wall must be small and the gas velocity must be very high to provide the acceleration forces needed to make it work effectively. The radius of the wall needs to be 0.1 mm or less and the slit width for the nozzle significantly smaller than that. Manufacturing robust, UF6 resistant equipment to the very fine tolerances required and in the quantity needed for an operational scale facility poses a major technological challenge. Furthermore, energy consumption is high and care must be taken to avoid the UF6 reacting with the hydrogen carrier gas. Development work showed that, in essence, the technology was inferior to the gas centrifuge and therefore not suitable for commercial exploitation.

The advanced vortex tube was developed in South Africa and uses similar principles to the jet nozzle, but in this case the UF6 carrier gas mixture is fired at high speed tangentially to a cone-shaped wall, creating a vortex. The circular motion again causes the heavier 238UF6 molecules to concentrate closer to the wall, while the 2 35UF6 becomes enriched further away from the wall. Careful positioning of splitters allows the feed gas to be separated into enriched and depleted components.

As with many enrichment technologies, each stage in the advanced vortex process produces modest levels of enrichment, so that many stages are required to

achieve the enrichment levels needed for practical use. The South African developers were able to combine multiple stages into a single unit known as a helikon, but even then the technology remains inferior to the gas centrifuge. The technology was abandoned in 1990.