Energy Production and Transmutation of Nuclear Waste by Accelerator Driven Systems P K Zhivkov

Size: px
Start display at page:

Download "Energy Production and Transmutation of Nuclear Waste by Accelerator Driven Systems P K Zhivkov"

Transcription

1 This content has been downloaded from IOPscience. Please scroll down to see the full text. Download details: IP Address: This content was downloaded on 30/12/2018 at 21:43 Please note that terms and conditions apply. You may also be interested in: Nuclear Materials Science: The challenges of nuclear waste K Whittle Nuclear Waste: Lords call for a repository Matin Durrani Nuclear Waste: Radioactive railway Sally Croft Energy Production and Transmutation of Nuclear Waste by Accelerator Driven Systems P K Zhivkov Electrical breeder fuels nuclear debate Jean-Paul Schapira LARRMACC Conference, Liverpool, 5 November 1998 S M Clark An accelerator-driven system for the destruction of nuclear waste Jean-Pierre Revol

2 Physics World Discovery Nuclear Waste Management Claire Corkhill and Neil Hyatt Nuclear Waste Management 1 Introduction Globally, nuclear power plants provide almost 11% of the world s electricity. As interest in the generation of electricity by nuclear fission has been renewed by a demand for low-carbon energy sources, this is set to increase. Nuclear power is generated in a process called fission, which involves the splitting of 235 U atoms after impact by a slow neutron (n): U + n A + B + x n MeV Splitting of the 235 U atom releases energy in the form of heat, which is used to create steam that drives electricity-generating turbines. The reaction is self-sustaining because splitting also releases more neutrons, which can go on to split more nuclei of 235 U (called a chain reaction). When a 235 U atom is split, it forms two smaller atoms (labelled A + B for simplicity in the equation above), which are unstable. These smaller atoms include fission products and minor actinides, which are highly radioactive. Like other types of electricity generation, nuclear power produces waste. Managing this waste is similar to the management of toxic chemicals arising from other industrial processes, but with one exception: due to the splitting of 235 U into elements with unstable nuclei, the waste generated is radioactive. This means that the processes used to handle, treat and store the waste must minimise the impacts of radioactivity on human health and the environment. Uniquely, this hazard (the radioactivity) will reduce over time, owing to the radioactive decay of radionuclides (i.e. radioactive isotopes) within the waste to stable isotopes. This process can take of the order of at least one million years, making nuclear waste management one of the most significant, and long term challenges that our society faces. Due to the very large amount of energy generated through nuclear fission from a very small amount of fuel, the quantity of nuclear waste from the last 60 years of civil nuclear power generation is actually relatively small. For example, the average 1 0 doi: / ch1 1 ª IOP Publishing Ltd 2018

3 yearly waste output from nuclear power in the UK, including decommissioning, is ~4000 m 3 yr 1. Coal power stations, by comparison, generate ~ m 3 of ash and sludge per year, in addition to 1.45 million m 3 of carbon dioxide (CO 2 ). Nuclear waste is not only produced from nuclear power, but is also generated from the extensive use of isotopes in medicine (e.g. radioactive tracers ingested to identify tumours) and research. Wastes are also created as by-products of mineral mining. Another source is legacy wastes, which are the remnants of historic military nuclear operations; for example, by-products of plutonium weapon generation at the end of the Second World War and during the Cold War era (particularly in the US, UK and Russia). These wastes were often generated with little regard to their future management, and thus are some of the most challenging and expensive to manage. For example, at the Hanford site in the US, operations are underway to extract sludges and effluents from tank farms where liquid waste from fuel reprocessing was placed in inadequately designed underground storage silos that have begun to leak into the surrounding environment. All types of nuclear waste must be managed properly, now and in future generations, or there will be adverse effects to human health and the environment. The fundamental principles of nuclear waste management are: to ensure the generation of nuclear waste is kept to a minimum; to protect human health and the environment; and to protect future generations (and perhaps, considering the timescales involved, civilisations) while also ensuring they are not burdened with managing nuclear waste generated in our lifetime. 2 Background The nuclear fuel cycle describes the series of industrial processes that produce electricity from uranium fuel typically uranium oxide (UO 2 ) in nuclear power reactors. Figure 1 demonstrates the different types of wastes that are produced throughout the nuclear fuel cycle, from uranium mining, milling, conversion, enrichment and fuel fabrication (front-end processes) to the so-called back-end processes, performed once fuel has undergone fission and is removed from the reactor. At this time, the fuel is known as spent nuclear fuel (SNF). Depending on whether a country operates an open or closed fuel cycle, different types of nuclear waste are produced. In an open fuel cycle, SNF is cooled for several years under water in engineered cooling ponds, then placed within containers and sent for long-term storage (and ultimately disposal in a geological disposal facility). The only additional wastes created in an open fuel cycle are technological wastes (for example, from handling SNF during cooling and placing in containers) and effluents (for example, waste water from SNF cooling ponds). Countries including Sweden, Finland, Canada and the USA operate an open fuel cycle for civil nuclear power. A closed fuel cycle takes advantage of the fact that SNF is composed of 96% UO 2 and 1% plutonium oxide (PuO 2 ), generated by neutron capture of UO 2. Both are fissile materials and, if recycled, can be reused as fuel. To recover this useful material, it must be separated from the remaining 3%, which comprises fission, 2

4 Activation products Fuel corrosion Effluents Nuclear power plant Losses Hex tails Fuel fabrication MOX fabrication Interim storage Activation products Fuel corrosion Effluents Losses F 2, HF, H 2 Effluent Conversion Enrichment U + Pu Reprocessing Immobilisation HLW ILW LLW Volatiles Effluents U Tailings Effluents Milling Conversion Enrichment Immobilisation Disposal Technological waste Effluents Spoil Radon Effluents Mining Spoil Figure 1. Wastes produced through the civil nuclear fuel cycle. Orange arrows indicate processes operating in a closed fuel cycle, while blue arrows indicate those that operate in an open fuel cycle. minor actinide and activation products, representing the most radioactive portion of SNF. Separation is typically achieved via a liquid separations process (such as the Plutonium and Uranium Refining by EXtraction process, PUREX), which involves chopping fuel and separating it from the metal alloy cladding, dissolving the SNF in nitric acid, removing the UO 2 and PuO 2 using an organic solvent and, after drying, immobilising the remaining fission products by vitrification in a borosilicate glass. These reprocessing operations create a large number of different radioactive waste streams. Countries that operate a closed fuel cycle include the UK, France, Japan, Russia and China. Classification of nuclear waste differs according to the regulatory body of each nation state, but generally depends on the radionuclide inventory and half-life. Wastes that comprise concentrations of radionuclides that give rise to radiogenic self-heating are known as high level waste (HLW). These wastes have substantial levels of radioactivity and require shielding, personnel protection, remote handling and consideration of the heat generated. They include SNF and some reprocessing wastes (e.g. vitrified fission, actinide and activation products). Other categories of nuclear waste have lower activity levels than HLW, but may still require special considerations with respect to the level of radioactivity; for example, shielding during processing and handling. Such wastes are known as intermediate level waste (ILW). Collectively, HLW and ILW are referred to as higher activity wastes. Wastes 3

5 1,000,000 Radioactivity (relative scale) 100,000 10,000 1, Total for spent fuel Actinides & daughters Radioactivity of uranium ore Fission & activation products , ,000 Years after processing Figure 2. Graph showing the time required for SNF to radioactively decay to safe levels. that do not generate heat, have a low radioactive inventory and do not require any special considerations in their handling are known as low level waste (LLW). While wastes arising from medical radionuclides can be acceptable for landfill disposal since they are often extremely short-lived, disposal of HLW is more problematic. These wastes contain radionuclides with long half-lives, for example 99 Tc and 239 Pu have half-lives of years and years, respectively. Figure 2 demonstrates the time required for HLW to decay to effectively safe levels. After several hundred thousand years, the radioactivity of HLW will have decayed to the same level of radioactivity as the UO 2 ore from which the fuel was originally mined. After 1 million years, the HLW will still be radioactive, but at a much lower, safer level. Therefore, any disposal option for HLW must be viable for up to 1 million years. Disposal in a deep geological facility several hundreds of metres below the ground is the favoured option for many countries; this is discussed in detail later in this book. The last estimate of the worldwide inventory of nuclear waste, taken in 2007 by the International Atomic Energy Agency, gives the following values for waste generated from nuclear power production: 2.2 million m 3 of ILW and LLW, with a total activity of TBq m 3 vitrified HLW, with a total activity of TBq metric tonnes heavy metal (MTHM) of SNF, with a total activity of TBq. Waste immobilisation The wastes generated at the back-end of the nuclear fuel cycle have great variety in their physical (solid, liquid, gas), chemical (volatile, organic, non-organic, etc) and radiological (heat-generation, half-life) characteristics. To ensure that these diverse 4

6 wastes can be safely handled, stored and disposed of in a way that minimises risks to human health and the environment for the long time periods they will remain radioactive a process of immobilisation into a stable, passively safe wasteform is required. Any given wasteform material should be: a solid material to help with transport and storage; stable under the required temperature range (i.e. will not melt or transform due to radioactive decay heat); stable under the required radiation field (i.e. will not be detrimentally affected by α, β or γ radiation); and durable, to ensure that under conditions of long-term storage and disposal the wasteform is not easily dissolved and the immobilised radionuclides within the wasteform are not released to the environment. A number of materials meet these requirements and are widely used in waste immobilisation, as described below. SNF is dealt with elsewhere in this book. Borosilicate glass glass is an attractive immobilisation matrix, particularly for HLW fission, actinide and activation products separated during reprocessing. Recently, vitrification has been trialled for immobilisation of ILW including decommissioning wastes for example, contaminated masonry and metals from defunct nuclear fuel handling facilities and it will be used to immobilise tank farm wastes from the Hanford site, USA. Vitrification involves melting glass-forming additives (such as boron oxide and silica) with waste so that the final glass product incorporates the radionuclides into the atomic structure. It is possible to incorporate a large number of elements within a given glass matrix and to achieve a relatively high waste loading (~35 wt% for HLW), resulting in a low-volume wasteform. Some elements are difficult to immobilise in glass, such as S, Cl and Mo; the flexibility of borosilicate allows the chemistry of the glass to be fine-tuned to account for this. The addition of boron to silicate glass lowers the melt temperature to a range of C (depending on other additives), which is suitable for a number of different vitrification technologies (such as Joule-heated ceramic melters and cold crucible melters) and also prevents the volatilisation of problematic radionuclides like 137 Cs. Glass materials, being amorphous in nature, show good tolerance to radiation damage and have a high chemical durability; Roman-era glass objects over 2000 years old have been recovered intact from the seafloor, demonstrating the robust nature of this material. Borosilicate glass was first investigated as a waste immobilisation matrix in Canada in the 1950s. In the 1970s and 1980s, France, the US and the UK made the decision to begin vitrifying their HLW in borosilicate glasses, due to their increasingly large inventories of defence and civil nuclear wastes. Borosilicate is not the only glass material to be used in nuclear waste immobilisation; aluminophosphate glasses are also used to immobilise HLW from Russian fuel reprocessing activities. Cement cementation of ILW has been practised for many years in a number of countries. In contrast to vitrification, where radionuclides are chemically immobilised within the atomic structure of the glass, in cementation waste is simply surrounded, or encapsulated, by a wet cement paste that hardens to form a solid block. The advantages of using cement for nuclear waste immobilisation include provision of: a low-cost and simple processing route for a variety of effluents, sludges and dry solids; a final product that is easily handled (with good thermal and physical 5

7 stability); radiation shielding provided by the cement surrounding the waste; alkaline chemistry that lowers the solubility of cationic radionuclides; and sorption of some radionuclides onto the main binder phase, calcium silicate hydrate. The major disadvantage is the large volume increase associated with cementitous encapsulation. This gives rise to a significantly larger volume of waste than prior to immobilisation, increasing the cost of waste management. Furthermore, the encapsulation of metallic wastes has proven problematic due to the ongoing corrosion of metals within the cement matrix; the corrosion products occupy a larger volume than the original metal, and hydrogen is generated during corrosion, resulting in expansion of the waste package in the worst cases. CEM I cements (also known as ordinary Portland cement) are the most common type of cement used. CEM I is often blended with cement additives including blast furnace slag (CEM III) or fly ash (CEM V) to improve properties such as compressive strength or fluidity. Other wasteform materials include geopolymers, bitumen and crystalline ceramics (discussed later in this book). 3 Current directions Geological disposal of high-level nuclear waste With significant inventories of HLW that will be radioactive for timescales outlasting our civilisation, long-term above-ground storage is considered only an interim management measure. A final solution for the long-term disposal of these materials is required to reduce the risk to the environment and future populations. A number of potential solutions have been considered but ruled out for technical reasons. For example, sending the waste into outer space with a target destination of the Sun entails enormous risks (e.g. launch failure) and costs. There is international consensus that the safest option is to remove nuclear waste from the dynamic surface of the Earth, where human intrusion, climate change and tectonic processes may disturb it, and to place it within an underground storage facility several hundreds of metres or more below ground. In a stable rock formation, the environment will remain largely unchanged over the to 1 million years required to allow the waste to safely radioactively decay, isolated from the biosphere. This concept is known as the geological disposal of nuclear waste and is proposed for wastes that are unsuitable for long-term storage in near-surface facilities, including SNF, HLW from reprocessing and other radioactive wastes that generate a significant amount of heat or contain particularly long-lived radionuclides. Multi-barrier concept The general principle of maintaining safety after the facility has been closed relies on the multi-barrier concept. This concept employs engineered and natural barriers that act together to contain the waste and to prevent, or lower, the release of radionuclides to the biosphere, as follows: Several layers of engineered barriers will contain the nuclear waste until most of the radioactivity has decayed; 6

8 The host geology, also known as the natural barrier, will isolate the waste from the biosphere and reduce the likelihood of human intrusion into the facility; The location of the facility, several hundreds of metres below the surface, will ensure long transport pathways to delay any significant migration of radionuclides from the waste to the biosphere until far into the future when much of the radioactivity will have decayed. When considering the great uncertainties associated with the spatial (hundreds or thousands of metres) and temporal (up to a million years) scales of geological disposal, this approach provides confidence that the facility will effectively contain the nuclear waste and retard radionuclide migration to the biosphere. The engineered barrier system, which can be likened to a set of Russian dolls from the inside to the outside, comprises the wasteform, the waste package and the backfill that surrounds them, as shown in figure 3. Arguably, the wasteform is the most important component of the engineered barrier system; it is the dissolution of the wasteform the process by which solid materials dissolve into a solution in groundwater that controls the release of radionuclides to the environment. For example, under ideal conditions of geological disposal, SNF would take significantly more than 1 million years to dissolve completely, by which time most of the radioactivity of the SNF will have decayed. The wasteform is placed within a waste package a metal container composed of a corrosion-resistant metal such as stainless steel or copper which helps to mitigate exposure of the wasteform to groundwater. The final component in the engineered barrier is the backfill, which fills the gap between the waste package and the host geology. The functions of this barrier, typically composed of clay or cement, are to control the movement of groundwater to the waste package for example, by having low porosity and ensuring groundwater movement is only by slow diffusion and to sorb dissolved radionuclides. While the engineered barriers can be considered to be extremely important at mitigating the release of radionuclides from the waste in the relative short-term of Wasteform Matrix dissolves slowly, limits radionuclide release Host geology Disposal at depth in a suitable and stable environment provides isolation Waste Package Prevents access of groundwater until significant radioactive decay has occurred Backfill Slows down water ingress through low porosity Provides sorption capacity for radionuclides Figure 3. The multi-barrier concept. 7

9 geological disposal ( years), the geology and hydrogeology of the host rock controls the long-term containment (> years) of nuclear wastes. It is this part of the multi-barrier system that provides a long travel path for radionuclides to reach the biosphere if they breach the engineered barriers. A host rock of a geological facility for nuclear waste should have minimal fracture systems, low porosity and permeability to reduce the mobility of groundwater and radionuclides, and should have minerals that readily sorb released radionuclides. Several rock types are under consideration for hosting a geological disposal facility for nuclear waste. These fall into three categories: high-strength crystalline rock (e.g. granite), sedimentary rock (e.g. clay) and salt. Each of these rock types has advantages; high-strength rocks tend to have low porosity and are favourable for construction, while clay and salt formations have extremely low permeability (in the case of salt there is no groundwater), are plastic and can form a tight seal around the engineered barrier. The exact design of a geological disposal facility, and the engineered barrier system employed, depends on the type of host rock available in a suitable location (e.g. tectonically stable, away from large population centres) and also on the type and volume of nuclear waste to be emplaced. For example, in one of the proposed designs for a UK facility shown in figure 4, two types of vault and engineered barrier system are envisaged for disposal of cementitious ILW and for HLW, including vitrified HLW and SNF. The former will use a cement backfill and the latter will use clay. In contrast, other countries including Sweden and Finland will dispose only of SNF, necessitating only one type of engineered barrier system (known as the KBS-3 concept, discussed later). Figure 4. UK generic geological disposal concept for nuclear waste in a high-strength crystalline rock, including separate but co-located disposal concepts for intermediate level waste (ILW) and high level waste (HLW)/spent nuclear fuel (SNF). 8

10 International status in 2018 A number of countries have active geological disposal programmes for LLW and ILW. One example is the Waste Isolation Pilot Plant (WIPP) in New Mexico, US, where materials contaminated with by-products of plutonium processing, from historic US military nuclear operations, are being disposed of 660 m below the surface in a salt formation. However, there is currently no operating facility for the disposal of civil HLW. Finland is the most advanced nation with respect to their plans for geological disposal; a licence has been granted to construct an underground facility in a granite host rock at Onkalo and it is envisaged that the first emplacement of SNF will be in In 2011, the Swedish Nuclear Fuel and Waste Management Company (SKB) applied for a licence to build a repository at Forsmark for MTHM of SNF. The review process for this licence is still in progress (as of June 2018). In the US, disposal of SNF and HLW from SNF reprocessing has been proposed in a facility embedded within Yucca Mountain, Nevada, above the water table. This facility has continued to encounter technical and political challenges since 1987 and is yet to be constructed or licensed to accept waste. A number of other countries have either selected sites for disposal but are yet to begin construction (e.g. France, Switzerland), have identified geological disposal as the preferred option for long-term nuclear waste management but are yet to locate a site (e.g. UK, Canada, Japan), or have not yet made a decision on the long-term disposal solution for their waste (e.g. Belgium, Netherlands). Spent nuclear fuel SNF is the most problematic type of nuclear waste; it continuously transforms in terms of its physical and chemical characteristics and it has a susceptibility to corrode in the presence of oxygen, making storage and disposal challenging. After being removed from the reactor, the activity of SNF is six orders of magnitude greater than the original UO 2 fuel, ~10 17 Bq MTHM 1 fuel. A person exposed to this level of radioactivity would absorb a lethal dose in less than a minute if they were standing just a metre away. Until several hundreds of thousands of years have passed (when the radioactivity will be the same as the original UO 2 ore; see figure 2), the SNF is highly hazardous to living organisms and must be safely stored and isolated. Several storage practices are currently utilised. SNF is first stored in cooling ponds for 2 5 years to remove the initial intense heat generated by radioactive decay of fission products. After this time, depending on the type of fuel cycle operated, the SNF is either sent for reprocessing (closed fuel cycle) or stored in dry casks (open fuel cycle). In dry cask storage, SNF bundles are placed within a sealed steel cylinder and surrounded by concrete, providing radiation shielding to the environment around the cask. They may be stored vertically or horizontally, either within a secure warehouse or on a concrete pad at nuclear power stations or other nuclear licensed sites. Convective cooling is still required for dissipation of decay heat. For many countries, there is presently no final disposal solution for the MTHM (metric tonne of heavy metal) worldwide inventory of SNF. This is 9

11 problematic since many new nuclear reactors are under construction (50 reactors in 2018) and commercial SNF reprocessing operations are earmarked for closure (for economic reasons). Hence, the inventory of SNF will continue to grow. There is international consensus that disposal in a geological facility is the most suitable solution. However, as discussed above, only Sweden and Finland have made significant progress towards this goal. In the absence of a final destination, longterm storage of SNF in dry casks is the plan for 100 years or more. However, it is not well understood how the SNF will behave in these conditions over such timescales. It is likely that this option will begin to face significant public and political challenges in the near future as stockpiles of SNF begin to grow. Evolution of spent nuclear fuel with time Within the reactor, UO 2 fuel undergoes a number of transformations driven by nuclear fission that affect the physical and chemical state of the fuel. These processes include: generation of fission, actinide and activation products through radioactive decay; increased temperature; defects caused by radiation; swelling; and restructuring. The distribution of fission products and actinides within SNF after it has been removed from a nuclear reactor is shown in figure 5. The majority of the fission products (e.g. Sr, Zr, Nb and lanthanide elements), and also the higher actinides (e.g. Pu, Np, Cm and Am), are incorporated into the UO 2 structure. Metallic fission products (e.g. Mo, Tc, Ru, Rh and Pd), known as epsilon particles, are found as immiscible precipitates in the grain boundaries. Other fission products precipitate as oxides (e.g. Rb, Cs, Ba and Zr). The final composition of the fuel is typically 96% UO 2,3%fission products and 1% PuO 2, but this can vary depending on the burnup the amount of fission experienced of the fuel. Owing to a steep thermal gradient between the centre (~1700 C) and the outer rim (~400 C) of the fuel pellet, volatile fission products, including Cs and I, tend to migrate towards the gap between the pellet and the cladding. Due to the cooler temperature in the rim, atoms displaced from their lattice positions as a result of radiation damage cannot return through thermal annealing, so this region undergoes a significant restructuring process, forming a structure known as the high burn-up or rim structure. In this region, the grains are significantly smaller than in the centre of the SNF pellet. Fission product gases (e.g. Xe and Kr) form finely dispersed bubbles within the UO 2 matrix, and can coalesce at grain boundaries. The presence of helium bubbles, generated by α-decay can also accumulate in the SNF over time, leading to volume swelling. Swelling may result in a breach in the fuel cladding that surrounds the SNF, which is highly detrimental to the long-term behaviour of SNF in storage and disposal. This may enhance corrosion and release of radionuclides to the surrounding environment. For the first years after removal from the reactor, the dominant source of radioactivity in SNF arises from β, γ-radioactive decay of fission products incorporated within the UO 2 matrix of SNF (figures 2 and 5). After years, α-radiation from the actinide and daughter products is the dominant form of radiation. In the α-decay of SNF, the α-particle dissipates its energy along a μm pathway in 10

12 Figure 5. Schematic of spent nuclear fuel (SNF) microstructure and the distribution of fission products and minor actinides after fission. Reproduced with permission from Bruno J and Ewing R C 2006 Spent nuclear fuel Elements Copyright 2006 Mineralogical Society of America. the UO 2 crystal lattice. Elastic collisions with atoms along this pathway produce several hundreds of displacements per atom. In a collision, the α-particle transfers some of its energy to the collided atom, which then collides with other atoms. This is known as α-recoil. Together, a single α-particle and its α-recoil nuclei can cause ~1200 atomic displacements, which generate a significant effect on the structure of UO 2 in SNF. The displacement of atoms (U and O) in SNF by α-radiation results in the formation of a type of lattice defect known as a Frenkel defect pair, which occurs when a uranium atom leaves its place in the lattice, creating a vacancy, and becomes lodged elsewhere in the lattice as a so-called interstitial. This can occur in the reactor during fission, but here the high temperatures are sufficient to allow defect annhiliation and recovery of the UO 2 lattice (i.e. the interstital returns to the vacancy, promoted by the thermal energy generated through nuclear fission), as shown in figure 6. However, the temperatures experienced by the SNF under storage conditions are much lower and do not allow this process, so that defects tend to accumulate, leading to an increase in the UO 2 lattice parameter and swelling of the SNF. 11

13 Figure 6. The UO 2 lattice (space group Fm3m) with a lattice parameter of a = 5.468(1) Å. The relationship between dose (displacements per atom) and change in the lattice parameter of UO 2 (Δa/a 0 ) as a function of α- damage, α-recoil damage and fission damage. Fission damage has a negligible effect on UO 2 volume swelling due to thermal annealing, while α-damage and α-recoil damage during storage and disposal result in significant lattice parameter increase and volume swelling. Disposal of spent nuclear fuel In a geological disposal facility, dissolution of SNF by groundwater will solubilise radionuclides, making it easier for them to be transported to biosphere, should the eningeered barriers be breached. The dissolution of SNF under geological disposal conditions is summarised in figure 7, and occurs in two main stages. The first is an initial fast dissolution step, where volatile fission products (I, Cs, Cl) and fission gas bubbles (Xe, Kr) are instantaneously released from the grain boundaries and from the gap between the cladding and the fuel. This occurs when the metal waste container and fuel cladding are first breached and is known as the instant release fraction. The second stage of dissolution involves the slow corrosion of the UO 2 matrix via the oxidation of U(IV) to U(VI) and the formation of higher oxide defect structures, such as U 4 O 9 clusters, at the surface and the grain boundaries. Secondary alteration products, such as coffinite (USiO 4 ) or uranyl (UO 2 2+ ) can be formed upon SNF matrix dissolution, under reducing or oxidising conditions, respectively. Since aqueous U(VI) species such as uranyl are highly soluble, and U(IV) compounds including UO 2 and coffinite are insoluble, it is preferable that reducing conditions are maintained in the engineered barrier during SNF disposal. The maintenance of reducing conditions is complicated by the production of radiolytic species in groundwater through radioactive decay of actinide daughter products in the SNF. Alpha particles emitted from the surface of fuel can travel a distance of ~40 μm in water, so only a thin film of water needs to be present on the surface of the SNF to be affected by α-radiolysis. Both oxidising (e.g. HO,HO 2, H 2 O 2 ) and reducing (e (aq), H,H 2 ) radiolytic species are produced; the presence of 12

14 Figure 7. Schematic of the key processes involved in the dissolution of SNF. oxidising species could lead to the oxidation of SNF and the subsequent release of radionuclides to the engineered barrier and biosphere. However, the presence of reducing species, especially H 2, can counteract the oxidising species and keep the conditions favourably reducing. A further complication is the radiolysis of species commonly found in groundwater (e.g. carbonate), which may alter the balance of radiolytic species towards more oxidising; this is not yet well understood. The Swedish KBS-3 engineered barrier concept has been designed to minimise, as far as possible, the dissolution of SNF by groundwater. The SNF waste package will be surrounded by a bentonite (clay) backfill that will slow down the transport of groundwater towards the waste package. The SNF will be placed in a gas-tight, corrosion-resistant copper container, with a cast iron overpack. If the copper container is breached by groundwater, the overpack will provide reducing conditions through the production of hydrogen upon its corrosion. This will also help to counter the production of oxidising radiolytic species. It is expected that this design will be highly effective. As such, it has also been chosen for the Finnish geological disposal facility. Plutonium: waste or resource? A significant stockpile of plutonium has been created in the UK from reprocessing of civil SNF, projected to reach 140 metric tonnes by the end of reprocessing operations. Substantial stockpiles of separated civil plutonium are also held in Russia and France, and surplus military plutonium in Russia and the US. Such stockpiles constitute a proliferation and security concern due to risk of diversion or misappropriation for use in nuclear weapons. The motivation for separating plutonium from civil nuclear fuel, by reprocessing, was to use this fissile material to fuel fast-breeder reactors, with the production of 13

15 additional plutonium from a 238 U blanket. The economic rationale was that the scarcity and demand of natural uranium resources would outstrip supply and, against the backdrop of rising fossil-fuel prices, the conservation and reuse of fissile material was deemed essential. However, fossil fuel prices remained largely stable, known reserves of uranium increased and the anticipated expansion of global civil nuclear-energy programmes was not realised. Since economic deployment of fast reactors could not be demonstrated, and reliable operation proved challenging, fast reactor programmes were largely abandoned. Nevertheless, reprocessing of civil SNF continued, giving rise to the present substantial stockpiles. One option for managing current civil plutonium stockpiles is to reuse this material as mixed oxide (MOX) fuel, composed of (U,Pu)O 2, in light water reactors. MOX fabrication involves a number of processes that include blending of PuO 2, mixing with UO 2, milling, pelletising and sintering, followed by loading of MOX fuel pellets into zircaloy fuel cladding to make fuel assemblies for light water reactors. All of these complex processes must be performed remotely, with automated operation due to the high radioactivity associated with PuO 2. Successful deployment of MOX fuel has been achieved in several countries, most notably France, Japan, the US and Russia. The prospect for MOX uptake in the UK remains weak, however: no reactor operator has expressed an interest in utilising MOX fuel, and neither the Sizewell B nor future Hinkley Point C pressurised water reactors are yet licensed to do so. The commercial and technical viability of industrial scale MOX production is also unclear. The MOX Fuel Fabrication Facility (MFFF) under construction in the US is far behind schedule, over budget and is now scheduled for cancellation. A further challenge in MOX fuel fabrication is the ingrowth of 241 Am from 241 Pu (with a half-life of 14 years). Since 241 Am is a strong γ-emitter, blending of the plutonium feedstock and adequate shielding of production facilities are required to reduce worker dose uptake. Alternatively, chemical extraction of 241 Am may be considered. However, this has not been demonstrated at industrial scale, and would add both cost and hazard to the overall fuel production process. Immobilisation options for plutonium The economic and technical challenges of commercial MOX fuel manufacture have led the UK to consider the immobilisation of its plutonium stockpile for ultimate disposal, since long-term managed storage is not a sustainable or cost-effective strategy. In any case, a fraction of the UK plutonium stockpile will not be suitable for reuse as MOX fuel and will require suitable treatment and disposal. Technology developed for this purpose could also be utilised to treat plutonium declared surplus to military requirements in the US and elsewhere. Several immobilisation options for plutonium have been evaluated. Encapsulation in cement is not considered a viable option; the achievable incorporation rate, based on safety and criticality concerns, would afford a prohibitively large volume of waste and life time management cost. Moreover, illicit recovery of plutonium from a cement waste package would not be as technically challenging as for glass and ceramic wasteforms. Immobilisation of plutonium by vitrification is 14

16 considered a viable option, using special glass formulations. However, plutonium accountancy (i.e. verification of material amounts) in vitrification processes, with reuse of a melter crucible, is a concern. The most promising materials for immobilisation of plutonium are tailored ceramic materials, in which plutonium is accommodated by solid solution in the crystal structure, with appropriate charge compensation. The batch-wise nature of ceramic production technology offers an inherent advantage in terms of plutonium accountancy, criticality and safeguarding considerations. The technology necessary for both ceramic and vitrification options requires considerable development and maturation. Research into crystalline ceramic wasteforms for actinides has been underway since the 1970s, when a multi-mineral phase titanate ceramic material called Synroc was conceived to immobilise radionuclides from reprocessed SNF. Synroc is an assemblage of hollandite, perovskite, zirconolite and rutile minerals, where actinides typically partition into the crystal structure of the zirconolite phase. Thus, zirconolite prototypically CaZrTi 2 O 7 could also be utilised as a dedicated ceramic wasteform for plutonium. Indeed, naturally occurring zirconolites incorporate a considerable inventory of the actinides uranium and thorium, which have been retained, under sustained self-radiation damage, for millions of years. Such natural analogues (which are lacking for glass materials), provide strong evidence of the long-term efficacy of ceramic materials for plutonium retention and can be used to parameterise and test predictive models of long-term performance in geological disposal systems. The chemical composition of synthetic zirconolite analogues can be engineered to accommodate the desired plutonium content, with the incorporation of neutron absorbers such as gadolinium and hafnium as an additional safeguard against criticality. Table 1 shows the range of mineral-ceramic phases that may fulfil the key requirements of a wasteform for plutonium. Each host phase has advantages or disadvantages with respect to durability, waste loading, chemical flexibility (i.e. ability to incorporate impurities and contaminants), processing compatibility (i.e. sintering temperature, pressure, compatibility with the radioactive environment), volume swelling (arising from α-decay that may also impact the aqueous durability) and natural analogues (to build confidence with respect to wasteform longevity). The choice of wasteform for immobilisation depends on the relative importance of each of these factors. In the US, significant research efforts have identified pyrochlore and zirconolite as the most suitable immobilisation hosts for plutonium unsuitable for MOX, while recent UK efforts (for example, at the University of Sheffield) have focused on the development of zirconolite glass-ceramics a mixture of a ceramic host phase to partition plutonium, distributed within a sodium aluminoborosilicate glass matrix to immobilise contaminants such as chloride for plutonium residues. One technology under development for ceramic immobilisation of plutonium at industrial scale is hot isostatic pressing (HIPing). Although there are currently no full-scale radioactive HIP facilities for nuclear waste immobilisation, HIP technology has been successfully demonstrated at full scale with inactive simulants, and is widely used in the fabrication of metals and ceramics on an industrial scale. The HIP 15

17 Table 1. Properties of mineral-ceramic phases identified as potential plutonium immobilisation hosts. Ln = lanthanide element (for fission products or Gd), An = actinide (U or Pu); green = high, yellow = intermediate, red = low. process simultaneously applies heat and pressure to consolidate and sinter materials. The potential benefits of using HIP technology for nuclear waste immobilisation include: The use of isostatic pressure, applied using an inert gas, promotes densification and eliminates internal porosity. This is advantageous for promoting good durability (lower porosity results in lower surface area for dissolution) and achieves volume reduction (up to 60%), leading to significant lifetime waste management cost savings. The waste is processed in hermetically sealed canisters so that no radioactive off-gas is produced, precluding the use of expensive off-gas treatment systems and generation of secondary wastes. The production of a sealed canister is also beneficial for disposal purposes; this can form part of the engineered barrier in a geological disposal facility. There is no requirement to pour the discharge product, as in a melter system, improving material accountability, important for fissile materials like plutonium. The combination of tailored ceramic wasteforms and HIP technology for their processing is also under consideration for 99 Tc-bearing wastes arising from medical isotope production (Australia). In future fuel cycles, where advanced separations technologies could be used to separate radionuclides according to their chemical properties and half-life, each radionuclide could be immobilised in an individual, 16

18 tailored crystalline ceramic phase. Until political decisions regarding the fate of the UK and US plutonium stockpiles are made, this immobilisation route will continue to be researched. Since such decisions may be far in the future, immobilisation may become an essential alternative to MOX fabrication. 4 Outlook Although nuclear waste management has been practised and researched for more than 40 years, we are still further than desired from achieving the goal of safe disposal of nuclear waste. This should be a necessary ethical prerequisite for the construction of new nuclear power stations and is required to fulfil the principle of intergenerational equity. The reasons for this lie behind the long timescales involved; political decisions are typically made with an outlook of only four to eight years, not hundreds of thousands. However, some countries are setting an example to the rest of the world; Sweden and Finland will have operational geological disposal facilities for SNF within the next few decades and other European nations will follow within this century. For those countries that have struggled to gain public (or political) acceptance for nuclear waste disposal (notably the UK and US), an alternative disposal option may be the solution. Deep borehole disposal where HLW and excess plutonium could be disposed of, permanently, in a small number of boreholes drilled several kilometres deep into the Earth s crust may be viewed more favourably than the shallower facilities currently planned. No such facility yet exists, but technological developments in oil-well drilling may facilitate future progress. As old nuclear power stations are retired from service, new materials and technologies will be required to immobilise decommissioning wastes, incorporating materials ranging from contaminated soil to pipework and masonry used in nuclear facilities. Recent advances have been made to use an approach of size reduction and vitrification for such wastes, but the melter technologies are yet to be used on radioactive materials at full scale. Much recent focus has been on developing decommissioning and immobilisation strategies for the Fukushima Daiichi and Chernobyl nuclear power plants. In these accident scenarios, SNF was melted with fuel cladding and fuel reactor components, generating a highly radioactive and heterogeneous material known as corium, which remains poorly understood. Advances in radiation-tolerant robotics are required to retrieve these materials, which must be characterised to understand how best they should be immobilised. In the future, it may be possible to significantly reduce the volume of HLW by a process known as transmutation. This would not eradicate HLW completely and the reduced volumes would still require disposal, but if it were possible to separate the long-lived actinides and fission products from SNF, they could be transmuted to shorter-lived radionuclides by irradiation in a reactor or an accelerator. The development of nuclear fusion as an alternative to nuclear fission would inevitably cause the demise of electricity generation by fission. The nuclear waste generated through fusion will arise from neutron activation of reactor materials (which are yet to be selected). However, since this will produce very short-lived 17

19 radionuclides, these wastes may only need to be managed for 100 years before they no longer pose a risk to human health. Additional resources Bodansky D (ed) 2004 Nuclear Energy: Principles Practices and Prospects (New York: Springer). An introduction to the nuclear fuel cycle. Choppin G, Liljenzin J O, Rydberg J and Ekberg C 2013 Radiochemistry and Nuclear Chemistry (Oxford: Elsevier). An introduction to basic radiation physics and radiochemistry. Chapman N and McCombie C (ed) 2003 Principles and Standards for the Disposal of Long-lived Radioactive Wastes (Oxford: Elsevier). An introduction to geological disposal of nuclear waste. Donald I W 2010 Waste Immobilisation in Glass and Ceramic Based Hosts (Oxford: Wiley). Further details on glass and ceramic materials for nuclear waste immobilisation. Konings R (ed) 2012 Comprehensive Nuclear Materials (Oxford: Elsevier). An introduction to radiation damage in nuclear materials, including nuclear waste. Lee W E, Ojovan M I and Jantzen C (ed) 2013 Radioactive Waste Management and Contaminated Site Clean-Up (Cambridge: Woodhead Publishing). An overview of radioactive waste management. Ojovan M I and Lee W E 2005 An Introduction to Nuclear Waste Immobilisation (Oxford: Elsevier). An overview of radioactive waste management, focusing largely on UK practices. Ojovan M I (ed) 2011 Handbook of Advanced Radioactive Waste Conditioning Technologies (Cambridge: Woodhead Publishing). An overview of the technologies used to immobilise nuclear waste. Yardley B W D, Ewing R C and Whittleston R A 2016 Deep-mined geological disposal of radioactive waste Elements A collection of articles on different geological disposal concepts and site selection. Hyatt N C 2017 Plutonium management policy in the United Kingdom: The need for a dual track strategy Energy Policy Further details on the Pu stockpile crisis. 18