Another version of the UREX+ process was tested in miniature centrifugal contactors at Argonne National lab in 2004 using the same spent fuel feed as previously described (Pereira 2005). Changes to the process included the coextraction of U, Tc, Np and Pu from the rest of the waste, see Fig. 6.6, routing the high active waste to further Cs/Sr removal. The benefit of this methodology, compared to the previous UREX+ flowsheet, is that the low specific activity of U, Tc, Np and Pu would be separated from the rest of the high active waste early in the process. This would make handling of the U/Tc + Np/Pu product streams much easier considering dose rates and shielding required. Another improvement is that less organic solvent is needed since only one extraction section is required followed by re-extrac­tion of uranium and technetium and reductive stripping of Np and Pu. Furthermore, no feed adjustment is needed for the second extraction of neptunium and plutonium where AHA is destroyed. The flowsheet for creating both U/Tc + Np/Pu product streams (Fig. 6.6) can be compared to only the upper part of the flowsheet of Fig. 6.5, for the UREX process. As in the first UREX+ test, the organic phase was not recycled. The recovery of neptunium and plutonium in the Np/Pu product was 87.2 and 99.8%, respectively. As before, some neptunium was lost to the raffinate but was recovered downstream. The largest impurity in the Np/Pu product was

zirconium, although this was claimed to relate to hydraulic problems in the Zr scrub section rather than chemical behavior.

The U/Tc separation was accomplished by anion exchange where the pertechnetate was retained by the resin and the uranium was effectively purified to a high degree, reaching all goals set for disposal as low level waste. The use of ion exchange as part of the PUREX process for Tc sepa­ration has been investigated separately (Dileep 2008) where anion exchange was used as a purification step for obtaining a pure plutonium product after the U/Pu split.

A number of flowsheets of the same general setup as those shown in Figs 6.5 and 6.6 have been tested in the UK since 2003 and are summarized in a publication by Birkett et al. (Birkett 2007). The U+Pu or U+Np product streams were first coextracted from the high active waste and transported by the loaded solvent to a separate stage for U/Np or U/Pu separation with the aid of AHA. Different tests were carried out to investigate the feasibil­ity of this process for different concentrations of Pu in the feed, and simu­lated 1) thermal MOX-fuel (7 wt.% Pu); 2) fuel from fast reactors (20 wt.% Pu); and 3) treatment of exotic Pu legacy waste (40 wt.% Pu). In all tests pure uranium product could be obtained. The purity of the plutonium product could potentially be controlled to accommodate proliferation resistance.

Historical perspective and current trends

Interest in the separation of fission-product cesium and strontium dates back to the early days of reprocessing over half a century ago. This interest has been reflected in a large body of scientific research and process devel­opment that is growing unabated to this day. Motivation has been threefold: advanced reprocessing, waste management, and industrial applications.25,26 Separations of 137Cs and 90Sr being considered for advanced reprocessing today are driven by the expected increase in repository capacity upon reducing the heat load that would otherwise result from inclusion of these two fission products.27-29 Given the large cost of adding unit operations to a reprocessing plant, however, separation may not be preferable to decay — storage,30 at least using currently available separation technologies. Nevertheless, research has continued to the point of demonstration and even multimegacurie separations have been done,31 thereby establishing a baseline of mature technology options for future consideration. Major use of 137Cs and 90Sr separations technology occurs in the area of waste manage — ment.32-37 Wastes include a great variety of stored radioactive materials but are dominated in volume by >108 gal of legacy defense reprocessing wastes stored in the United States and Russia. Strategies generally aim to concen­trate radionuclides into a compact volume for disposal with concomitant high decontamination of the bulk volume of the waste stream. Environmental risk is thereby reduced, and cost savings accrue by reducing the volume of high-activity waste destined for expensive storage facilities. Radiation sources containing 137Cs or 90Sr have been in demand for multiple industrial uses dating back to the 1950s, motivating considerable research and mega­curie recovery operations.25,26,37’38 Historically, legacy wastes, such as the separated radiocesium and radiostrontium currently stored at Hanford, have been viewed as a resource in this regard.

The status of technology development to separate cesium and strontium from both acidic and alkaline solutions was reviewed comprehensively approximately two decades ago,25,26 describing the considerable progress that had been made in previous decades and laying out the needs and challenges for future processing. A need was singled out therein for new solvent-extraction technologies for cesium and strontium removal from acidic streams. The potential of crown ethers, recognized in the radiochemi­cal community since the 1970s,39-42 had not yet led to a practical system, and promising dicarbollide chemistry. Although promising dicarbollide chemis­try had gained the attention of Czech researchers as early as 1974,43 it had not yet emerged from dependence on toxic diluents like nitrobenzene. Several ion-exchange and precipitation systems had been in use on the acid side, but all were seen as having drawbacks,25 such as difficult recovery of the radionuclide from the reagent, low capacity, or instability. A number of separations on the alkaline side, pertaining mainly to cesium, were available or being implemented, such as the In-Tank Precipitation (ITP) process using tetraphenylborate for cesium removal from tank waste at the Savannah River Site (SRS),44 phenol-formaldehyde resin functionalized with sulfonic acid groups (Duolite ARC-359), zeolites, and several solvent-extraction systems.25,26,33 Strontium, largely insoluble as hydroxide or carbonate salts or as co-precipitated with other metals, is not as important as cesium for separation on the alkaline side, but monosodium titanate had been demonstrated to be effective in sorbing traces from tank waste.44 Megacurie recovery of strontium from acidified tank sludge had been carried out previously.25

In the electrorefiner, LFP in the molten salt gradually builds up during processing of the spent fuels. In order to limit the decay heat content and the contamination of the cathode products, the molten salts in the electro­refiner are periodically treated in the ‘counter current extraction’ process and ‘zeolite column’ process to decrease LFP content. As shown in Fig. 10.5, feed salt from the electrorefiner comes in contact with liquid cadmium at approximately 450 °C to extract actinides into liquid cadmium (as a

liquid metal solvent). The treated salt (raffinate) is additionally decontami­nated during the ‘zeolite column’ process, and the decontaminated salt comes in contact with the cadmium (extract) from the first stage of the counter current extraction, to strip actinides from the molten salt to be recycled in the electrorefiner. The number (N) of stages necessary depends on the separation requirements and distribution characteristics. Figure 10.6 shows the measured distribution data that rules separation efficiency at each stage (Koyama, 1992). In a liquid chloride salt and cadmium metal system, equilibrium among the elements is achieved by redox reaction between cations in the salt and metal atoms in the cadmium. The chlorine anions remain in the salt and are not oxidized. Thus, the equilibrium among
two cations can be represented as an exchange reaction between the pair of elements. Take, for example

nUCl3 + 3M = nU + 3MCln 10.6

The equilibrium constant, Ke, for the reaction is

where [M], [MCln], AG°, R and T denotes activity of M in the cadmium phase, activity of M in the salt phase, standard free energy of formation, gas constant and temperature in Kelvin, respectively.

Here the distribution coefficient of an element, M, in a molten salt and liquid metal system is defined as

Y

Dm = Y“, 10.8

XM

where YM and XM denotes mole fraction of M in salt, and atom fraction of M in metal, respectively.

The separation factor of element M relative to uranium is defined as

SFm = — 10.9

M Du

According to the thermodynamic relationship described in equation [10.7], the separation factor is described as

SFM = Ke1/3DU(n-3)/3f-^^ YYuCl^ 1 10.10

VY MCln А Y и )

where у denotes the activity coefficient.

As shown in Fig. 10.6, the distribution coefficients of actinides and lan­thanides have a linear relationship with that of uranium. Agreement of the slopes suggest that these chlorides are the same, actually in trivalences. Hence, equation [10.10] can be simplified as

SFm = Ke f^-lfYuCl 1. 10.11

VYMCln A Yи )

This equation clearly shows the nature of the separation factor in this system. It is not just a chemical technological value but a function of ther­modynamic properties that depends only on the temperature. In the counter current extraction process, therefore, each stage has different distribution coefficients but the same separation factor for each pair of elements. As

shown in Table 10.2, the measured separation factors of trivalent lanthanides are more than ten times larger than those of trivalent actinides, suggesting that sufficient separation between actinides and lanthanides can be expected for the counter current extraction of several stages.

In nitric acid solutions, such as PUREX raffinates (where [HNO3] > 3 mol. L-1), the 4f lanthanide metallic cations (Ln) and the 5f americium and curium metallic cations (An) predominantly show the same oxidation state, +III, and many similar physical and chemical properties (Nash, 1993, 1994, Beitz, 1994, Morss, 1994, Marcus, 1997):

• They are considered as ‘hard acids’ in Pearson’s theory (Pearson, 1963).

• The 4f and 5f orbitals have a rather small radial extension and are more or less protected by the saturated lower electron orbitals, respectively

the 5s2-5p6 for the lanthanides and the 6s2-6p6 for the actinides. Thus, the nf electrons scarcely interact with electrons of neighbouring ligands and their electronic properties are only slightly affected by their environments.

• The ionic radius shortens along the 4f and 5f series as the atomic number increases. Thus, it is easy to predict the higher electrostatic reactivity (formalized by the ionic potential closely linked to the charge density) of an element of higher atomic number, Z, compared to that of an element of lower atomic number in the periodic table.

• Since Ln(III) and An(III) have the same positive charge (+3), their discrimination through solvent extraction involving ‘hard bases’ (e. g., ligands bearing oxygen donor atoms in their structures) will mainly be due to geometric and/or steric hindrance reasons: the better the fitting of a metallic cation radius with the cavity size of the complexing/extract — ing agent or its coordinating site, the better the discrimination. However, the separation of the two series of trivalent elements will not be com­plete because of the similarities in the ionic radii among 4f and 5f elements.

• Ln(III) and An(III) are highly hydrated in aqueous media: 8 to 9 water molecules can be numbered in their inner-coordination spheres, as com­pared to 4 to 5 only in the case of penta- and hexavalent actinides. It is, however, admitted, although difficult to demonstrate by a structural proof, that an outer-coordination sphere of water molecules exists and interacts with the water molecules present in the inner-coordination spheres of the metallic cations through hydrogen bonds.

• As for other metallic cations, hydration of the 4f and 5f trivalent ele­ments is of capital importance in their extraction mechanisms, since they can be partly or completely dehydrated while being extracted in organic solvents.

• The coordination numbers in complexes of trivalent lanthanides and actinides vary from 6 to 12, depending on the bonding chemical system involved.

However, a slight chemical behaviour difference does exist between the two series of trivalent elements: the 4f orbitals of the lanthanides are slightly more localized around their nuclei than the 5f orbitals of the actinides, which can consequently interact more easily with their electronic environ­ments than the corresponding lanthanides (Beitz, 1994, Morss, 1994). Unlike trivalent lanthanides, trivalent actinides create stronger chemical bonds with ligands bearing ‘softer’ donor atoms than oxygen, such as for instance sulphur or nitrogen (Musikas et al., 1983, Musikas, 1984). The drawback of hydrophilic and/or lipophilic compounds containing sulphur and/or nitro­gen atoms is their usually strong affinity for protons in acidic media.

Although more rational (considering the inventory of elements present), the direct and selective extraction of An(III) from PUREX raffinates has been the most challenging of the unresolved research topics radiochemists have addressed for the past 50 years throughout the world. This is why, except for specific single-step processes, such as SETFICS (Nakahara et al., 2007) or DIAMEX-SANEX (Madic et al., 2002) processes which will not be covered by this chapter, most of the strategies adopted to selectively recover trivalent minor actinides from PUREX raffinates show the same two-step process approach:

1. The co-extraction of the An(III) together with the Ln(III) in a front — head process, such as the TRUEX, DIAMEX, or TODGA processes, which make use of oxygen donor extractants (‘hard’ bases), such as carbamoyl phosphonate/phosphine oxide or diamide compounds, and specific scrubbings with hydrophilic masking agents to achieve the sep­aration of An(III) and Ln(III) from the rest of the fission products (FP).

2. The partition of An(III) from Ln(III) in a second cycle process, either through the selective stripping of the An(III) thanks to a hydrophilic highly selective ligand, or through the selective extraction of the An(III) thanks to a lipophilic highly selective extractant (both types of com­pounds bearing ‘soft base’ electron-donor atoms, the use of which is made possible by the lower acidity of the feeds coming from the front- head processes than that of the PUREX raffinates).

In high temperature gas-cooled reactors (HTGR), also known as fast reac­tors, graphite is utilized as the moderator of the nuclear reaction. The graphite is either used as part of the structural materials for the reactor

Steps SF ILW LLW Tailings Comments Mining and milling 65,000 In terms of radiation doses and numbers of people affected, uranium mining has been one of the most hazardous steps in the nuclear fuel chain, disproportionately impacting indigenous communities. Conversion — — 32-112 — Besides airborne and waterborne uranium, hazards include chemicals such as hydrofluoric acid, nitric acid, and fluorine gas. Enrichment 3-40 Typically buried at dump sites with a high risk of leaching radionuclides into the groundwater. Waste is contaminated with polychlorinated biphenyls (PCBs), chlorine, ammonia, nitrates, zinc and arsenic. Fuel fabrication " " 3-9 " Because fuel fabrication does not involve the production of liquid waste, its effects are mainly restricted to workers and are on the same order as for workers in the reprocessing sector. Reprocessing and vitrification not applicable not applicable not applicable not applicable Wastes from reprocessing, together with spent fuel, contain more radioactivity than any other waste in the fuel cycle. Phenolic and chlorinated compounds are produced in large amounts due to the use of decontamination reagents such as CCI4 together with phenolic tar (Gad Allah, 2008). Reactor operations — 22-33 86-130 — Boiling water reactors have considerable emissions of radioactive noble gases. Spent fuel storage and encapsulation 2 0.2 Considerable quantities of "low-level" waste are created due to fission products leaking into the spent fuel pools from cracks in the fuel cladding (Choi eta/., 1997). Spent fuel final disposal 26 — — — Insufficient treatment can cause continued exposure to environment and local population. Decommissioning — 9 333 — Most of the radioactivity from reactor decommissioning waste is in a relatively small volume of intensely radioactive material. Totals 26 33-44 457-624 65,000

Steps SF HFW IFW FEW Tailings Comments Mining and milling 50,060 Mill tailings account for over 95% of the total volume of the radioactive waste from MOX-OT processing cycle. This does not include mine wastes. Many tailings sites all over the world remain unremediated and/or neglected and pollute ground and surface water with radioactive and non-radioactive toxic substances. Conversion — — — 25-86 — Besides airborne and waterborne uranium, hazards include chemicals such as hydrofluoric acid, nitric acid, and fluorine gas. Enrichment 3-25 Typically buried at dump sites with a high risk of leaching radionuclides into the groundwater. Waste is contaminated with polychlorinated biphenyls (PCBs), chlorine, ammonia, nitrates, zinc and arsenic. Fuel fabrication 13 7.4-12.5 Because fuel fabrication does not involve the production of liquid waste, its effects are mainly restricted to workers and are on the same order as for workers in the reprocessing sector. Reprocessing and vitrification 2-4 17-39 8016-8037 As in FUE-OT system, wastes from reprocessing, together with spent fuel, contain result in the highest risk. The waste is high inorganic content. There is a particularly high risk of further contamination through accidents of storage facilities at the reprocessing plant. Reactor operations — — 22-33 86-130 — Boiling water reactors have considerable emissions of radioactive noble gases. Spent fuel storage and encapsulation 0.3 0.03 As in the FUE-OT system, large quantities of "low-level" waste are created due to fission products leaking into the spent fuel pools from cracks in the fuel cladding. Fisson products are trapped in resins in filters, which then become "low-level" waste in the United States and intermediate level waste in Europe. Spent fuel final disposal 26 — — — — Insufficient treatment can cause continued exposure to environment and local population. Decommissioning — — 10.1 315 — Most of the radioactivity from reactor decommissioning waste is in a relatively small volume of intensely radioactive material. Totals 26 2-4 62-95 8452-8615 50,060

core vessel or as fuel containment elements in the form of pebbles. The graphite used from natural sources contains non-carbon impurities within the carbon matrix. Among these impurities are oxygen and nitrogen from entrapped air, cobalt, chromium, calcium, iron, and sulfur (Khripunov et al, 2006). Upon exposure to high neutron flux, most of the impregnated impuri­ties are expected to transmute to unstable radioactive forms. For example, experimental exposure of graphite in nuclear reactors have shown that the stable forms of oxygen, nitrogen, and C-12 are converted to radiocarbon-14 (C-14) as shown in Table 15.3.

The radioactive fission products are created within the fuel grains and migrate through grain boundaries and then through microscopic cracks in the graphic matrix (Fig. 15.2). Most of the fission products are entrained in the matrix — a small proportion escapes through the outer layers into the gas phase. The challenge of reprocessing involves the separation of the metallic radionuclides from the graphite matrix and reducing the amount of C-14. Impurities in the fuel itself include: (1) metallic fission products (Mo, Tc, Ru, Rh, and Pd) which occur in the grain boundaries as immiscible micron to nanometre-sized metallic precipitates (e-particles); (2) fission products that occur as oxide precipitates of Rb, Cs, Ba, and Zr, and (3) fission products that form solid complexes with the UO2 fuel matrix, such as Sr, Zr, Nb, and the rare earth elements (Kleykamp, 1985; Shoesmith 2000; Buck et al., 2004; Bruno and Ewing, 2006).

As the 21st century unfolds, energy has become a theme underlying many of the challenges facing mankind. There is little doubt that standard of living closely correlates with the availability of energy resources. Raising the global standard of living will require providing power to those who cur­rently do not have it. It is estimated that two billion people do not have access to electricity.[1] This represents 30% of the world population. Providing power to these people, while continuing to satisfy the needs of the developed world, represents an enormous challenge, especially in the face of the threat of global climate change and decreasing availability of “clean” fossil fuels. Fission-based nuclear power will play a vital role in meeting this challenge. Nuclear power provides reliable base-load power with virtually no greenhouse gas emissions (other than those associated with mining and construction operations). This power production technol­ogy is well established and has proven to be safe and reliable, although it will be essential to maintain the improvements that have been made in plant operational safety and efficiency over the past several decades. One aspect of nuclear power that still provides significant technical challenges is the management of irradiated nuclear fuel.

For the most part, the United States has pursued a “once through” fuel cycle policy in which the uranium passes once through the power reactor and the irradiated fuel is emplaced intact in a geologic repository. However, no geologic repository has been licensed and this policy is presently being re-evaluated. At present, the fate of irradiated fuel from commercial power production reactors in the US is uncertain. Partial recycling of the irradiated fuel has been practiced in a number of countries (e. g., France, the United Kingdom, Russia, and Japan). These operations allow recycling of fissile uranium and plutonium back into the fuel cycle as mixed oxide (MOx) fuel, but they still result in problematic long-lived transuranic elements in the high-level waste that must be disposed of in a repository. The presence of the transuranic elements in waste placed in a repository requires engineer­ing of the repository to ensure safe performance for hundreds of thousands of years. Ensuring repository performance over such a large span of time is beyond human experience. Because of this, there has been growing inter­est in closing the nuclear fuel cycle in a manner that allows long-lived radionuclides to be recycled back into the fuel cycle and thus to have little effect on repository design or performance. Under such scenarios, the integ­rity of the repository need only be ensured for hundreds of years, a time frame well within the horizon of recorded history.

The subject matter of this book is the separation science and technology underpinning the efforts to manage irradiated nuclear fuel, including an examination of progress towards achieving a closed fuel cycle. The book is organized into three parts. The first part is aimed at providing fundamental scientific and engineering information related to nuclear fuel cycle separa­tions. The second part is devoted to describing standard and advanced technological solutions for application in nuclear fuel cycle separations. The third part reviews emerging and innovative separation and extraction tech­niques that are being pursued in the development of advanced nuclear fuel cycles.

Part I opens with two chapters summarizing fundamental actinide chem­istry as it relates to the separation of these elements and also their behavior in the environment, and the physical and chemical properties of actinides in particular, i. e. the most critical type of element of concern in nuclear fuel reprocessing. This is followed by a chapter describing the nuclear engineer­ing principles as they relate to aqueous separations. Chapters 4 and 5 discuss issues related to monitoring material flow through nuclear separation plants both for the purpose of process monitoring and control, and for safeguard­ing of special nuclear materials.

Part II of the book begins with Chapter 6 describing well-established technology for separating uranium and plutonium from dissolved irradiated fuel (i. e., the PUREX process). This chapter not only discusses the estab­lished PUREX methods, but also describes recent enhancements such as the recovery of a mixed uranium/plutonium product (the COEXTM process) and options for managing neptunium. This is followed by a chapter describ­ing the recent work performed in the US on advanced alternatives to the PUREX process. Chapter 8 describes recent developments in the separa­tion of fission products such as 137Cs and 90Sr from the dissolved irradiated fuel matrix. Taking this a step further, Chapter 9 discusses efforts to develop a single process that can extract and separate a number of radionuclides (actinides and fission products) together.

Part III of the book opens with Chapter 10 describing pyrochemical/ electrochemical separations and engineering. Chapter 11 details the quest to design new separation materials that are highly specific for selected fuel components, coverage that is enhanced by Chapter 12, which reviews developments in the partitioning and transmutation of radioactive wastes.

The development of such highly selective separations media would greatly simplify implementation of the separations required for closing the fuel cycle. Finally, the book closes with three chapters that discuss separation methods that are somewhat different from the traditional liquid-liquid extraction methods. Firstly, Chapter 13 discusses the possibility of using solid-phase extraction methods in nuclear fuel separations; secondly, Chapter 14 explores the use of supercritical fluid extraction and ionic liquids in advanced fuel cycle separations; and thirdly, Chapter 15 details biological treatment and bioremediation processes of use in separations science and for the recovery of useful materials from radioactive wastes.

It is our hope that this book will provide a useful reference to scientists and engineers working in the field of nuclear fuel cycle separations. But perhaps more importantly, we hope that it will provide a starting point for the young scientists and engineers who will rise to meet the challenge of safely managing the nuclear fuel cycle in a sustainable manner, enabling safe expansion of this low-carbon means of electrical production to raise mankind’s standard of living worldwide in the 21st century.

Finally, the editors and publisher would like to make a special dedication to Dr Troy Tranter, formerly of Idaho National Laboratory, USA, and author of Chapter 13 of this book, who sadly passed away in December 2010.

Reference

1. Rhodes, R.; Beller, D. The Need for Nuclear Power, Foreign Affairs, 2000, 79,

30-44.

Gregg J. Lumetta

Pacific Northwest National Laboratory P. O. Box 999, MSIN P7-22 Richland, WA 99352 USA

gregg. lumetta@pnl. gov (email)

Kenneth L. Nash

Washington State University P. O. Box 644630 Pullman, WA 99164-4630 USA

knash@wsu. edu (email)

Centrifuges are typically used to separate undissolved fission product particles and fuel cladding debris from the dissolved nuclear fuel prior to reprocessing. A typical unit, used in the Thermal Oxide Reprocessing Plant (THORP) at Sellafield in the UK, and similar to the units used at the La Hague nuclear site in France, is shown in Fig. 3.4. This has a rotating inner bowl that spins at several thousand rpm contained within an outer container. As was described in Section 3.2 on the Type 3 PSC, the outer

container is permanently welded into the in-cell pipework, while the inner bowl is suspended on a shaft from the cell top and can be withdrawn through a cell roof plug into a flask if this is ever needed. The motor and gearbox are located outside the cell and so can readily be maintained without breaking the cell containment.

This unit is operated “batch continuously” with the dissolved nuclear fuel solution (“dissolver liquor”) being fed into the inner bowl for an extended period, with the clarified product overflowing from the top of the bowl, over a weir inside the outer container, so that it then flows from the bottom of the outer container to a receipt tank.

The removed solids progressively build up on the inner bowl wall and, when this layer reaches a set thickness, the feed of dissolved fuel is stopped and the centrifuge changed to solids recovery mode. In this mode, high pressure water jets from a spray lance inside the inner bowl remove the solids “cake” and slurry it from the bottom of the bowl into a separate receiving tank. Feed of the dissolved fuel is then restarted and the cycle is repeated. Two such centrifuges are usually operated together to provide continuous solids removal capability.

An example of a safeguards application to a pyrochemical separation plant is found at the Fuel Conditioning Facility (FCF) located at Idaho National Laboratory (INL). INL, owned by the Department of Energy (DOE) and operated by the Battelle Energy Alliance (BEA), must comply with specific DOE orders regarding nuclear material control and accountability (DOE 2007). This order establishes a program for the control and accountability of nuclear materials at DOE-owned and DOE-leased facilities and DOE — owned nuclear materials at other facilities that are exempt from licensing by the Nuclear Regulatory Commission (NRC). Nuclear material control and accountability must be integrated with the Safeguards and Security Program for two reasons. First, it provides a mechanism for detecting a potential loss of nuclear material for safeguards and security. Second, it provides a periodic check of inventories to ensure that processes and mate­rials are within control limits (Vaden et al. 1996).

The basic FCF mission is to support the Fuel Cycle Research and Development Program (the successor to the Advanced Fuel Cycle Initiative (AFCI)) by treating sodium-bonded metal fuel and producing interim storage products and final waste forms. FCF has two adjacent hot cells known as the “air cell” (that contains a purified air atmosphere) and the

“argon cell” (that contains a purified argon atmosphere). Both hot cells house the remotely operated equipment used to electrochemically treat spent metallic nuclear fuel and are considered one Material Balance Area (MBA). Figure 5.1 shows a block diagram of the basic process steps. Simply stated, the operations performed in FCF consist of: 1) receipt of nuclear material and preparation into a form suitable for electrochemical process­ing; 2) electrochemical processing to remove nuclear material from fission products, bond sodium, etc.; 3) production of a low-enriched product for storage; and 4) preparation of waste forms. The process has been described in greater detail by Ackerman (1991) and Mariani (1993).

Because nuclear materials processed in FCF are contained in many material types and forms, material accountability requires measuring the nuclear material content of all flow streams entering and exiting the MBA and the physical inventory of nuclear material within the MBA. The inven­tory difference (ID) is defined as the difference between the measured inventory and what is expected to be in the inventory based on the previous
inventory and measured flows into and out of the process. The ID is calcu­lated via the following equation (DOE 1995).

ID = (BI + TI) — (EI + TO)

In this equation the summation of the ending inventory (EI) and transfers of nuclear material out of the MBA (TO) are subtracted from the summa­tion of the beginning inventory (BI) and transfers of nuclear material into the MBA (TI). Because measurement errors will occur, the actual amount of material measured will differ somewhat from the expected quantity, most likely resulting in a non-zero ID. The probability of detecting the loss of a given quantity of material (the loss detection capability) depends upon the uncertainty associated with the determination of the ID. In the FCF, mate­rial control and accountability uses twice the standard deviation of the inventory difference for the limit of error, which is propagated from all measurement and sampling uncertainties in an operation. The limit of error means the true ID has a 95% probability of being within two sigma of the measured ID. The true ID is zero if all materials have been measured and accounted for.

Near-real-time accountancy (NRTA) in FCF is accomplished by a com­bination of neutronics calculations, process models, physical measurements, and a computer-based mass tracking (MTG) system. The MTG system tracks the location and masses, by element and isotope, of nuclear material — containing items in near real time. Items may be storage containers, process­ing equipment, and fuel elements and assemblies. The masses are determined using in-cell balances with isotopic and elemental compositions determined by neutronics calculations, by previous measurements, or by computations based on process models. The neutronics calculations and process models are established by measurements applied to a particular process step. Mass values derived from process models are updated when measurement results are available. The MTG system provides a model of discrete accountable items distributed in space and time and constitutes a complete historical record (Adams et al. 1996). Figure 5.2 is a photograph of some of the track­ing stations in FCF that assist in the NRTA.

With this database, material accountancy over the whole facility can be calculated for any specified time interval and space. Material accountancy uses the item weights and compositions and associated uncertainties. The system uses the best available information, which may either include measured or model weights. The Materials Accounting with Sequential Testing (MAWST) computer code is typically used for propagating the errors and establishing the inventory difference and limit of error (Picard and Hafer 1991).

Moving beyond the Fuel Conditioning Facility, safeguarding any process­ing facility requires periodic material balances (Li et al. 2002). This neces­sitates measuring the input streams and output streams. In addition, an accounting of all materials in the physical inventory is also necessary to close the material balance. Two approaches are proposed, a clean out of the physical inventory or measuring the physical inventory. A complete cleanout of the physical inventory is very disruptive to the process and may not be possible in some cases. Measuring the physical inventory is difficult and may result in large uncertainties in the measurement and material balance. As stated in the FCF example, measuring the physical inventory allows for NRTA. Research has just barely begun on non-destructive assay (NDA) of the input stream. In-situ measurements of the physical inventory have not been fully developed. FCF relies heavily on process modeling for its NRTA. In conclusion, a major research and development effort is necessary to measure the input and output streams and the physical inven­tory to establish NRTA for safeguarding a dry processing facility.

Unlike other components of a closed fuel cycle, the feedstock to the recy­cling facility has a great deal of variation whether its origin is SNF, or other high — or low-level wastes (legacy defense, medical, industrial, etc.). In addi­tion, the final product and/or waste form needs to meet very tight product specifications or waste acceptance criteria. Failure to do so results at best in materials that need to be reprocessed, or at worst in an orphan waste form with no disposition path. The need to design a robust process that can accommodate variations in the feed and tight product and waste specifica­tions requires bracketing the acceptable feed compositions and testing the process at a wide range of conditions. It is unfeasible to do so experimen­tally, as not all possible feed combinations can be reproduced and it is prohibitively costly. The use of models to design flowsheets allows the design of a facility for a wide range of feed compositions and an optimized process that ensures product and waste specifications are meet.

Investigation of conditions for extraction of radionuclides from salt-rich solutions

Laboratory studies of the possibility of simultaneous recovery of long-lived radionuclides from HLW in the framework of a single process have resulted

in the development of a flowsheet for waste treatment on the basis of the CCD, CMPO and PEG in diluent FS-13 extraction system. Since HLW, as a rule, has a complicated chemical composition, the influence of different HLW components on the extraction of long-lived radionuclides needs to be investigated to create an effective technology.

In particular, owing to the rather low concentration of CMPO in the extraction mixture, its extraction properties with regard to trivalent TPE and REE depend strongly on the concentration and state of such extract­able impurities such as zirconium and iron, which may compete with TPE and REE. Tables 9.10 and 9.11 show the effect of zirconium and iron on the distribution coefficients of europium from 1 M HNO3, using an extract­ant of 0.08 M CCD + 0.024 M CMPO + 1.4% PEG-400 in FS-13.

Tables 9.10 and 9.11 show that zirconium and iron exert a suppressing effect on europium extraction. Increasing the aqueous solution acidity to 1.5 M reduces the iron effect drastically, and thus special precautions should be taken to suppress zirconium extraction.

The extraction of hindering impurities and thus their effect can be sup­pressed with the use of different complexones, such as fluorine-ion or citric acid. As citric acid deteriorates the properties of cement (in the case of cementing LLW-raffinate) after the extraction of radionuclides, preference was given to fluorine-ion. Table 9.12 shows the effect of fluorine-ion on europium extraction at various concentrations of zirconium.

The data given in Table 9.12 illustrate the possibility of establishing condi­tions for the application of fluorine-ion to suppress zirconium extraction. As to the iron effect, the table shows that, even at the molar ratio 1 : 1 between iron and fluorine-ion, iron extraction is effectively suppressed.

244 Advanced separation techniques for nuclear fuel reprocessing Variant 3

• stripping of Cs, Sr, An and REE by a solution containing 1 M guanidine carbonate and 20 g/l DTPA.

On the basis of experimental results concerning the processes of extrac­tion and stripping, the flowsheets using the UNEX-extractant have been developed which involve the fraction or combined separation of long-lived radionuclides from HLW.