Because minor actinide bearing fuel is still a new field, a tremendous amount of excellent research has been accomplished to date [7]. It appears that the greatest emphasis has been placed on heterogeneous target fuel concepts with a high content of minor actinides only (i. e., no plutonium). This trend may have been driven by the desire to set the performance envelope, namely to understand and prove performance characteristics in extreme cases. However, this approach may overlook some practical approaches amenable to commercial-scale deployment in practice.

The physicochemical properties of materials involved in the fuel design and fabrication of minor actinide fuels are needed to understand and improve performance of fuel elements exposed to high temperatures, cor­rosive coolants and a radioactive environment. The objectives for develop­ment of a physicochemical properties database are to assemble a database on fuel properties and performance sufficient to support a safety/licensing case, and to develop a fabrication and quality assurance process that will enable effective and economic fuel fabrication. The required physicochemi­cal properties include a range of thermodynamic (enthalpy and heat capac­ity, melting temperature, enthalpy of fusion, vapour pressure, vaporization, thermal expansion, density and surface tension), transport (thermal con­ductivity and diffusivity, emissivity and optical constants) and mechanical (hardness creep, thermal shock and swelling) properties.

These properties (Table 12.4) are needed to understand fuel performance and are also required for the modelling of fuel behaviour. Relevant experi­mental data for all minor actinide elements are not available for all proper­ties, and may include datasets for a mixture of actinides. The database is also sparse for metal, oxide, nitride and other potential fuel forms of the minor actinides. A database is available for a few Np compounds, while data for Am and Cm are sparse. The collection of property data for minor acti­nides is an ongoing effort.

The irradiation behaviour of minor actinide fuels may vary from that of conventional fuels in several ways. Most prominent is the increased fuel pin helium gas inventory due to capture and decay sequences associated with 241Am and a significant amount of 242Cm, which decays by a emission (half­life of 169 days) to 238Pu. The additional helium gas inventory can lead to higher fuel swelling rates and is an additional source term for fuel pin over­pressurization. An additional interesting phenomenon relates to the evolu­tion of isotopic mixtures in fuel with high 241Am and 237Np loadings, and the effect this evolution has on pin power. As plutonium isotopes are bred in fuel from neutron capture during irradiation, particularly in the thermal spectrum, fuel rod power increases as a function of irradiation time.

Other issues to be studied include fission product behaviour, and optimi­zation of the oxygen to metal ratio in oxide fuels. Due to the shift in isotopic composition of the starting fuel, the isotopic and chemical distribution of fission products also shifts relative to that of standard MOX fuel.

Many fuel types, encompassing different matrices and chemical forms, have been considered for minor actinide bearing fuels. Several unconven­tional fabrication techniques have been explored as well. The in-reactor irradiation tests conducted to date are all very promising. As post­irradiation examination results become available, they will provide valua­ble information to guide additional future irradiation tests. Obviously, more R&D and in-reactor irradiation tests are required to qualify minor actinide bearing fuels. The incorporation of MA has some impact on the physicochemical properties of fuel material. Some results are available for the incorporation of MA in MOX fuel (for example, lower melting tem­perature, influence of stoichiometry on thermal conductivity, redistribu­tion of Am). But additional data are needed to guarantee the safe operation of reactor and fuel cycle facilities (fuel fabrication and reprocessing).

In the US, AFC fuel test hardware was designed to simulate fast reactor test conditions in Idaho National Laboratory’s advanced test reactor ATR: the AFC-1 irradiation test series was designed to evaluate the feasibility of actinide-bearing fuel forms in sodium cooled fast reactors for the transmu­tation of actinides from spent nuclear fuel [8]. The fuel rods had the same diameter as EBR-II fuel, but were reduced in length. All fuels, both nitride and metal, were sodium bonded inside stainless steel Type 421 (HT-9) clad­ding with an inert plenum gas. AFC-1B, AFC-1F, and AFC-1^ irradiation test capsules provided irradiation performance data at intermediate burn ups of 4 to 8 at.% on non fertile and fertile actinide transmutation fuel forms containing plutonium, neptunium and americium isotopes. AFC-1D, AFC-1G, and AFC-1H capsules extended fuel performance with satisfac­tory results, in terms of gas release and microstructural data compared to U-xPu-10Zr fuel when correlated with fission density.

In Japan, the “Am-1” programme has been conducted in order to inves­tigate the irradiation behaviour of Americium containing MOX fuel in the experimental fast reactor Joyo [9, 10]. Two irradiation experiments were conducted in the Joyo MK-III 3rd operational cycle to research early thermal behaviour of MA-MOX fuel. Six prepared fuel pins included MOX fuel containing 3% or 5% americium (Am-MOX), MOX fuel containing 2% americium and 2% neptunium (Np/Am-MOX), and reference MOX fuel. Ceramography results showed that structural changes such as lenticu­lar pores and a central void occurred early, within the brief 10 min of irra­diation. The results of electron probe microanalysis revealed that the concentration of Am increased in the vicinity of the central void. Post irra­diation examination of these pins did confirm fuel melting and local con­centration evolutions under irradiation of NpO2-x or AmO2-x in the (U, Pu) O2-x fuel. These test results are expected to reduce uncertainties in the design margin for the design of MA-MOX fuels.

The SUPERFACT irradiation in Phenix (1986-1988) represents the main body of existing knowledge on in pile behaviour of MOX fuel loaded with MA [11]. This project demonstrated the feasibility of Am or Np incorpora­tion of up to 2% in MOX fuels. The main constraint in introducing minor actinides into the core (homogeneous recycling mode) is linked to their impact via core reactivity and kinetic factors.

For heterogeneous recycling and in particular in the case of blankets loaded with MAs, high MA content raises the question of managing the large quantity of He produced. Hence, a specific transmutation fuel micro­structure must be developed, which requires envisioning several innovative steps in irradiation systems.

Most fast reactor irradiation tests, like the SUPERFACT experiment, were done in the French Phenix SFR, which was finally shut down in 2009 [12]. Since there are only a handful of fast reactors still in operation that can provide prototypic irradiation test environments in fast spectrum, an international collaboration to expand test capabilities and optimize the limited availability of irradiation test facilities is desirable. For example, under the framework of the GEN-IV SFR programme [13], an international collaboration project called GACID (Global Actinide Cycle International Demonstration) is being conducted (2007-2016) with participation of French CEA, US DOE and Japanese JAEA. The objective of the project is to demonstrate the transmutation of minor actinides in a 20% Pu MOX fuel with the Monju SFR located in Tsuruga, Japan. The fuel pins will be manufactured at the French CEA Atalante hot cells in Marcoule, using USA MA feedstocks. Data obtained from the GACID irradiation project will provide a feasibility assessment of MOX fuel matrix for transmutation.

Partitioning and transmutation of radioactive waste 375

The biosorption process is designed to remove metallic species, especially those originating from the nuclear reaction process in power plants, for further processing and recovery. The majority of these are the products of nuclear fission of uranium to form lighter elements. During the first several hundred years after the fuel is removed from a reactor, fission products are considered the most hazardous elements to living organisms in the environ­ment. Among these, radiostrontium (Sr-90) is one of the most abundant radioactive components of nuclear waste (Watson et al., 1989). Strontium can be highly mobile in both soils and groundwater systems (Dewiere et al., 2004) and it has a half-life of 28 years.

Due to its chemical similarity with calcium, it is easily incorporated into bone material in mammals. When incorporated in the organisms in this manner, it continues to irradiate localized tissues with the eventual develop­ment of bone sarcoma and leukaemia (Chen, 1997). The main disadvantages of using conventional adsorbents, such as zeolites and synthetic organic ion exchangers, for strontium removal from radioactive waste are: their decreas­ing efficiency at higher pH, and inhibited performance at high sodium concentration (Chaalal and Islam, 2001). Microbial adsorbents, on the other hand, have been shown to possess high capacities for the selective uptake of a range of metals and radionuclides from dilute metal-bearing solutions (Beveridge, 1989; Mullen et al., 1989; Chubar et al., 2008).

In the following sub-sections, the biologically mediated reactions at the surface of cells are explained and the operational conditions for the bioad­sorbents and effects of co-occurrence with other cationic species are exam­ined. Sulfate reducing bacteria are used as an example due to their demonstrated ability in adsorbing a range of metals including palladium (II) and calcium (II).

In the operation of solvent extraction systems in the laboratory, where new reagents and processes are typically developed, conditions are typically maintained in a moderately idealized state. In particular, the effects of “loading”, i. e., contacting aqueous solutions containing high concentrations of metal ions with extractant solutions targeting those metal ions, are typically not investigated until an advanced stage of process development is approached. In this limit, deviations from ideality are often observed. For example, a UO22+ extraction reaction having the following ideal stoichiometry,

UO22+ + 2 NO3- + 2 Ex = UO2(NO3)2Ex2

might adhere to this ideal stoichiometry through some range of concentra­tions, or more appropriately up to a solubility limit in the organic phase. At

higher concentrations of metal ions, the extraction stoichiometry might drop to

UO22+ + 2 NO3- + 1 Ex = UO2(NO3)2Ex

While the metal ion continues to partition, the possibility solute-molecule reorganization in the organic phase can occur leading to precipitation or to the separation of a third liquid phase (third phase formation).

In this metastable (though typically reversible) condition, the aqueous phase retains much of its essential character, while the organic phase will split into solute-poor (mainly diluent) and solute-rich (mainly extracted metal complex and excess extractant molecules) phases. This condition is at best an annoyance as a source of process upset and at worst a serious safety hazard in the operation of a reprocessing facility, as the unintentional formation of a critical mass of fissile material can occur with potentially disastrous consequences. Such phenomena are well known in solvent extrac­tion of metal ions and mineral acids. The minimization of the potential for such an event to occur is addressed through appropriate alteration of diluent structures and mixtures of diluents, effective extractant design or by carefully designing process monitoring and control procedures to prevent overloading of extractant solutions. Recent research being conducted in the US and in particular in Europe have provided unique insights into the nature and driving force for such interactions. In the design of any facility or process that incorporates fissile materials, it is quite important to always consider these possibilities.

Centrifugal contactors are classed as stage-wise or equilibrium contactors because, like mixer-settlers, they manifest a step concentration profile as the aqueous and solvent phases pass from stage to stage. Also, like mixer- settlers, they consist of mixing and phase separation compartments. However, phase separation occurs under the action of centrifugal, rather than gravitational forces, which leads to more compact processing units than mixer-settlers.

Various designs of centrifugal contactor have been developed and they are used extensively for small-scale process development tests (e. g. Leonard, 1997). Centrifugal contactors were used successfully in the first UNF repro­cessing plant at the Savannah River Site but since then they have seen limited use. However, their compact size makes them ideal for processing solutions containing high concentrations of fissile material and Baron (2008) has described their recent industrial application to plutonium product purification in the UP2 UNF recycling plant at La Hague. In addition, their compact size also makes them amenable to modular processing unit con­cepts, such as the Modular Caustic Side Solvent Extraction Unit currently being used to separate cesium from radioactive wastes stored at the Savannah River Site (Poirier, 2008 and Geeting, 2008). Centrifugal contac­tors are being considered for future thermal and fast reactor UNF recycling plants (e. g. Balasubramanian, 1992, Washiya, 2004, Duan, 2005 and Law, 2006).

The centrifugal contactors receiving most recent attention have integral mixing and phase separation compartments and are more accurately referred to as annular centrifugal contactors. As shown in Fig. 3.15, the mixing compartment consists of an annulus bound by the outer housing and an internal rotor bowl that spins at several thousand revolutions per minute. The two liquid phases enter the annulus and are mixed as a result of the shear forces set up by the rotating rotor and the bottom vanes. The mixed phase is drawn from the annulus, between the rotor base and housing and into the rotor bowl. The phases separate as a result of the centrifugal forces and exit the rotor bowl over their respective weirs. Banks of centrifugal contactors are connected together in a similar manner as mixer-settlers to permit counter-current flow (Fig. 3.16).

Annular centrifugal contactor size is usually specified by the rotor diam­eter. Laboratory-scale units used for testing flowsheets typically have 2-cm diameter rotors with a throughput capacity of 40 mL/minute (Leonard, 1997). Units with 5-cm diameter rotors and throughputs up to 5 L/minute are used for engineering scale tests (e. g. Law, 2006). Industrial scale units with rotors of diameters from 12.5 cm are available. Clean-in-place capabil­ity has been incorporated into 12.5 cm units by adding to the rotor a hollow central shaft with spray nozzles that can effectively flush out solids (Garn, 2008).

Residence time in the mixing annulus is very short, of the order of seconds. However, the turbulent energy is sufficiently high that very small drops (fractions of a millimeter) are formed with a high interfacial area that provides for efficient mass transfer. Nonetheless, centrifugal contactors are generally not ideal for liquid-liquid systems exhibiting slow interfacial mass

3.15 Schematic representation of an annular centrifugal contactor.

3.16 Bank of four centrifugal contactors (two in view) with Perspex housing showing inter-connections. Source: Nuclear Decommissioning Authority ("NDA"), copyright: Nuclear Decommissioning Authority ("NDA").

transfer. Wrong-phase entrainment is again a measure of hydraulic perfor­mance and Arm (2006) has shown entrainment first decreases and then increases with increasing rotor speed. This phenomenon is indicative of how mixing intensity and phase separation performance are related by the rotor speed. The larger drops formed at low rotor speeds are easier to separate, but this is counteracted by the fact that phase separation improves with increasing rotor speed. As the rotor speed increases, the drops become smaller and less easy to separate. Wrong-phase entrainment also increases with increasing flow rate because the mixed phase has less time to separate. Therefore, there is an optimum rotor speed and flow rate to minimize wrong-phase entrainment.

Some important chemical engineering attributes of annular centrifugal contactors are provided in Table 3.4.

The organic phase exiting the codecontamination cycle contains most of the actinides (U, Pu, and Np) initially present in the dissolved fuel. Some resid­ual fission products, notably Tc and Zr, are also present in the organic phase. The next step of the process (referring to Fig. 6.2) is partitioning, which involves the selective stripping of Pu and U from the loaded organic in two complementary steps. The first is Pu partitioning and the later is U stripping.

Plutonium partitioning

The first operation in the partitioning cycle involves selective plutonium back-extraction or stripping from the loaded organic phase. This is accom­plished by the reduction of extractable Pu(IV) to the inextractable Pu(III) oxidation state. A variety of reducing agents have historically been used for the reduction and concomitant partitioning of plutonium, the most promi­nent being: (a) uranous cation, U(IV), (b) hydroxylamine nitrate (HAN), or (c) ferrous sulphamate, Fe(NH2SO3)2. Of these reducing agents, the later, ferrous sulphamate, is no longer used in modern PUREX plants and will not be further discussed. Hydroxlyamine nitrate is used in some instances in the downstream Pu purification cycle and will be further described sub­sequently. The reductant universally used in modern facilities for Pu parti­tioning is U(IV). However, to explain the logic behind the process diagram, it is once again convenient to regress to a brief review of the pertinent process chemistry.

The process was demonstrated in 2003 and 2007. Different separations modules were tested for the recovery of Cs/Sr and Am/Cm products. In 2003, CCD-PEG and Cyanex 301 were tested. In 2007, FPEX and TALSPEAK were tested. In addition, in 2003, Tc was separated from uranium by selective stripping while in 2007 an ion exchange resin was used to make the process more amendable to an industrial application.

In 2003, the Np fraction was lower than the target value and a high lan­thanide content was observed in the Am/Cm product which deemed the Cyanex 301 separation module unacceptable. Process adjustments in NPEX resulted in an excellent recovery of Np in 2007. In addition, the replacement of the Cyanex 301 separation module with TALSPEAK resulted in excel­lent recovery of Am/Cm from the Ln fission products. Both CCD-PEG and the FPEX separations modules demonstrated high Cs/Sr recoveries. Results are given in Table 7.5.

To assess the possibility of solidification of the low-level raffinate, with the aim of its near-surface storage, the traditional LLW cementing technique was tested.

Test results indicated that, during cementing, the solidification process proceeded uniformly, and the cement blocks produced did not crack on discharge from the mold or in subsequent storage. Although the cement bulk per volume of raffinate being solidified was rather high (1.5 m3 per 1 m3 raffinate), the cost of the cementing process should not be very high, because the low content of radionuclides in the raffinate (LLW) allows the solidification process to be undertaken with the use of routine non­radiochemical equipment.

C. HILL, CEA, France

Abstract: This chapter discusses the methodology deployed in the European partitioning strategy to design highly selective extractants for long-lived radionuclide separation: calix[4]arenes for caesium, malonamides for the co-extraction of trivalent minor actinides (Am, Cm) and lanthanides (Ln(III)), and nitrogen-donor ligands, such as bis-triazinyl-pyridines, for the separation of trivalent minor actinides from Ln(III).

Key words: selectivity, calixarene-crowns, diamides, bis-triazinyl-pyridines.

11.1 Introduction

In the quest to develop mature technologies capable of partitioning long — lived radionuclides (LLRN) from spent nuclear fuels, hydrometallurgy has consensually become a reference route, probably because of the successes of PUREX1 process industrial implementations in Europe and elsewhere since the 1970s (Benedict et al., 1981, Schulz et al., 1990, Birkett et al., 2005). Furthermore, solvent extraction allows high selectivity and recovery yields to be reached without generating excessive volumes of secondary waste; its flexibility in adapting to various spent nuclear fuel characteristics and fuel cycle options, as envisaged in the GenIV initiative, appears industrially attractive. Solvent extraction has been a subject of intensive nuclear research activities in Europe, Russia, China, Japan, India, and the United States since the early 1960s (Marcus and Kertes, 1969, Sekine and Hasegawa, 1977, Musikas, 1986, Rydberg et al., 1992, Danesi, 2004, Rydberg et al., 2004).

After a brief introduction on the advantages of the Partitioning and Transmutation (P&T) strategy, this chapter identifies the LLRN targeted in the European P&T policy. The methodology deployed to design highly 1 PUREX process, for Plutonium URanium EXtraction selective lipophilic compounds is then illustrated by three examples: (i) calix[4]arenes for caesium separation, (ii) diamides for the co-extraction of trivalent minor actinides and lanthanides from PUREX raffinates, and

(iii) nitrogen-donor ligands for the discrimination between trivalent minor actinides and lanthanides. Finally, the main progress achieved in these respective processes is described.

Advances in the field of solid-phase extraction, along with substantial improvements in automated control systems and on-line monitoring, suggest that the technology merits further consideration for nuclear fuel reprocess­ing applications. The technology is probably more suited to MA and Ln separations subsequent to the removal of U and Pu by other means, or for polishing applications in any number of processing and/or raffinate streams. Issues with high U loading will likely preclude the use of solid-phase extrac­tion materials for primary actinide separations, but this decision is ulti­mately a function of the stability of the given material and economics of the process under consideration. It is noteworthy that many researchers have recognized the need for more information regarding the robustness of the solid-phase extraction composites. This is evidenced by attempts to increase substrate stability via the addition of inorganic materials as well as bench-top studies to assess extractant losses and radiation effects. Still, more data is needed to evaluate the issues mentioned in the previous section. Future work should include comprehensive assessments of resin performance under the multiple cycles of loading and stripping correspond­ing to those typical of an actual process. Some testing at larger scales, and smaller tests using actual reprocessing solutions, will also be needed to accurately assess the viability of the technology. The added complexity of these studies will certainly drive up cost and likely require some type of programmatic support. This is typically the issue when transferring any technology from small scale to large scale applications. The information to be gained is, however, essential for performing adequate engineering and cost-benefit analyses of the technology.

13.5 Sources of further information and advice

The works by Cortina et al., (1994a, b), Braun and Ghersini (1975), and the numerous journal publications by E. P. Horwitz and coworkers are excellent sources for more detailed information on solid-phase extraction.

13.6 Acknowledgment

I would like to thank Mitchell Greenhalgh of the Idaho National Laboratory for fruitful discussions and immense help in gathering information for this work.