Category Archives: Advanced separation techniques for nuclear fuel reprocessing and radioactive waste treatment

Future trends

The problem of radiation pollution in the environment is expected to increase as industries in the world experience pressure to reduce their carbon atmospheric contributions. The dilemma is that production of goods will only increase with population growth. Most countries are considering nuclear energy as the interim solution to the growing energy crisis. This will require environmental engineers and scientists to delve deeper into research on environmentally friendly processes to counter the pollution effects to avoid leaving a pollution legacy for the future generations. Advanced microbial cultures will be sought to treat a wider variety of recalcitrant pollutants. Discussions on the possibility of genetically engineering special­ized cultures for the purpose are not new in the environmental engineering fraternity. However, the application of ideas bears a large ethical burden as it is forbidden in almost all countries in the world to introduce genetically engineered organisms into the environment. Less aggressive methods for dealing with the problem without violating ethics include in situ bioaug­mentation and molecular bioaugmentation to a certain extent.

In situ bioaugmentation

This entails identifying indigenous species of bacteria within the vicinity of the contaminated site and determining the critical carbon sources and nutri­ents that could be supplied to encourage the growth of the target species. When the selected nutrients are introduced into the environment, either by injection into boreholes or by spreading on the ground, the target species will out-compete other species and will be able to degrade the contami­nants. The potential problem with this form of bioaugmentation is that the nutrients may be viewed as pollutants in their own right especially at the beginning of the bioaugmentation process when microbial loading is very low. The nutrients such as NO3- and SO42- have undesirable pollution effects on receiving water bodies such as eutrophication of streams receiving the base flow from the remediated areas.

Molecular bioaugmentation

The molecular bioaugmentation process utilizes genetic carriers such as transposons and plasmids to shuttle genetic information for toxic metal remediation into native species in the environment or species already adapted to the target environment. Several species of bacteria are capable of picking up and retaining circular fragments of DNA called Broad-Host — Range Plasmids which may be engineered to carry specific genes for the degradation of xenobiotic compounds and transformation of toxic metals (Weightman et al, 1984; Vincze and Bowra, 2006). The same process can be applied using genetically engineered linear DNA called transposons. Although studies have been conducted using these techniques in laboratory microcosms, the application in actual environments has not been attempted (Hill et al., 1994). In the future, it is foreseeable that these methods will find wide application for the new pollutant varieties that may be untreatable by conventional methods.

Biofractionation and bioseparation of elements

A very little understood application of bioseparation involves using micro­organisms to discriminate radioisotopes by size. So far, this application has remained conceptual due to limited understanding on the structure and function of organisms that are suspected to achieve biofractionation (Molokwane and Chirwa, 2009). In the latter study, Molokwane and Chirwa observed with a modest degree of certainty that microbial cells previously isolated from a high radiation-exposed facility accumulated C-14 while growing on a C-14/C-12 carbon matrix from powdered nuclear graphite. The experiment was conducted in a closed loop chemostat system equipped with biofilters for collection of suspended matter for analysis. The observed metabolic activity in the cells indicated that the process was possible under very low dissolved oxygen, suggesting that the microorganisms preferred inorganic forms of carbon as the primary carbon source. Bacteria that utilize inorganic carbon sources such as CO2 and HCO3- as primary carbon sources — known as autotrophic organisms — favour anaerobic conditions for growth. However, in this preliminary study, the amount of C-14 remain­ing in solution which could be required to draw a mass balance on C-14 in the system was not measured. These preliminary results on C-12/C-14 bioseparation hold promise for future HTGR nuclear reactors that produce large amounts of low radiation level waste as expired nuclear graphite. Success in the above process is also important for the decontamination and recovery of nuclear graphite from decommissioned plants for reuse in new reactors.

15.9 Sources of further information and advice

Chow B. G. and Jones G. S. (1999) Managing Wastes With and Without Plutonium Separation (Santa Monica, Calif.: RAND, 1999). The reprocess­ing LLW figure also uses data from Groupe Radioecologie Nord Cotentin, Inventaire des rejets radioactifs des installations nucleates, vol. 1, July 1999, p. 19.

International Union of Pure and Applied Chemistry (IUPAC), 1984. Makhijani A., Hu H. and Yih K. (1995). Nuclear Wastelands: A Global Guide to Nuclear Weapons Production and Its Health and Environmental Effects, Cambridge, Mass.: MIT Press, Massachusetts USA.

United States Nuclear Regulatory Commission (http://www. nrc. gov/waste).

[1] Estimating required heights and diameters for the columns. The height determines how much of the resin capacity is utilized before the con­taminant concentration in the effluent becomes unacceptable. The diam­eter mainly determines the liquid throughput for a given utilization of the resin capacity.

• Channeling of the liquid through the ion exchange bed. Larger diameter columns can suffer from channeling if the feed is not adequately distrib­uted across the column cross-section. Channeling can be exacerbated by fissures in resin beds caused by it swelling and shrinking.

• Feed distributor design. A well-designed feed distribution device can mitigate channeling at least to some degree in a large diameter column.

• Deciding on upward or downward flow of liquid through the bed. Up-flow velocities are limited by the need to prevent fluidization of

[2] A mixer compartment where the two liquids are mixed to form a disper­sion. Mixing is typically achieved by a mechanical impeller with the motor and gearbox mounted outside the biological shielding for ease of maintenance. Mass transfer is primarily achieved in the mixing compartment.

• A settler compartment where the two liquids are separated by gravity and exit through a system of weirs. The weir system can be designed for the separated liquids to flow to the succeeding contactor under gravity without the need for inter-stage pumping.

Mixer-settlers’ continued application in nuclear facilities is warranted on the basis of their mechanical simplicity and industrial pedigree and their important chemical engineering attributes are summarized in Table 3.2. Mixer-settlers are well-established equipment for liquid-liquid extraction unit operations for processing radioactive material and are employed in currently operating commercial UNF recycling plants (e. g. THORP, Sellafield, UK; UP3, Cap La Hague, France; and the Rokkasho Reprocessing Plant, RRP, Rokkasho Mura, Japan) mainly for purifying uranium product streams in which the fissile material content is sufficiently low that nuclear criticality can never be reached even within the non-ever-safe geometry of these contactors.

[3] Solvent extraction cycles are dedicated to virtually complete recovery of the uranium and plutonium products with a high degree of purifica­tion from the other constituents of used nuclear fuels, particularly with

[4] Dissolution of the solid used nuclear fuels to produce liquid feed to the separation (solvent extraction) operations. This portion of the process, typically termed headend operations, is an important consideration in determining and controlling the underlying chemistry of the subsequent separations processes. Obviously, the characteristics of irradiated fuel vary widely with a variety of interrelated factors, e. g. fuel type (metallic, UO2, MOx, etc.); reactor type (PWR, LWR, BWR, Fast, etc.); burnup; and enrichment.

Notes: (1) The repository addressed by the GNEP program was Yucca Mountain.

[6] Transuranic elements (of concern): Pu — plutonium, Np — neptunium, Am — americium and Cm — curium.

[7] Non-TRU: Waste containing no more than 3700 becquerels (100 nanocuries) of alpha-emitting transuranic isotopes per gram of waste and half-lives greater than 20 years.

[8] the need to decrease HLW acidity resulting in an increase of volumes of feed solutions;

• the use of complexones and salts of organic acids for stripping of actinides and lanthanides, which makes further handling of the strip product more difficult;

• insufficient mutual purification of Cs-Sr and An-Ln fractions.

Additionally, the use of this system is complicated by the fact that the syn­thesis and purification of phosphorylated PEGs is a rather sophisticated

[9] The reduction potentials of elements in LiCl-KCl eutectic salt into liquid cadmium alloy were derived when the concentrations of the elements in the salt and in liquid cadmium were the same (0.001 mole fraction).

[10] SFCs/Na = DCs/DNa, where DM is the distribution ratio of the extracted alkali cation M.

[11] A DIAMEX raffinate is the genuine highly active solution issued from the imple­mentation of a DIAMEX process, co-extracting trivalent actinides and lanthanides from a PUREX raffinate (issued from the reprocessing of a dissolver liquor by TBP solvent to recover uranium and plutonium).

[12] surrogate, spiked, or genuine PUREX raffinates, as well as well as PUREX concentrates (Serrano-Purroy et al., 2005a, b, Modolo et al., 2007a), as the feeds;

• DMDOHEMA dissolved at 0.65 M in HTP, as the solvent;

• mixer-settlers, centrifugal extractors, rotating ‘Couette-Taylor’ effect columns, and pulsed columns, as the laboratory scale contactors.

Oxalic acid and A-(2-hydroxyethyl)ethylenediamine-A, A’,A’-triacetic acid (HEDTA) were added to the feeds and the scrubbing solutions to limit the extraction of Mo and Zr on the one hand, and Pd on the other. The DIAMEX flowsheet tested in 2005 at the CEA Marcoule (France) on a genuine PUREX raffinate (31 litres, from the reprocessing of 15 kg of a 52 GWd/t UOX type spent nuclear fuel: 6.5 years of cooling), in three

[13] SANEX, for Separation of ActiNides by Extraction. The objectives of the SANEX process are to recover more than 99.9% of the An(III) in a purified product stream containing less than 5 wt.% of Ln(III).

[14] Dry impregnation — this is the most often used method wherein the extractant, or extractant diluted with an appropriate organic diluent, is

Boundary conditions

The outer and inner boundary conditions for dissolved species u and biomass x are defined by:

ju = kLu*(uB(t) — u/s(t,

z = L/, outer boundary


jx = :K(U)■XfL/,

z = L/, outer boundary


ju = 0,

z = 0, inner boundary


jx = 0

z = 0, inner boundary


where kLU = DwU/Lw, is the mass transfer rate coefficient (L2T^), and u/ = dissolved species concentration at the liquid/biofilm interface (ML-3).

Simulation and parameter optimization

The system described above is an oversimplification of the actual biofilm processes in nature. However, the example serves to illustrate how complex the solution for such a simplified version could be. In the old days before fast and efficient computers, solving such problems manually was unthink­able. Lately, computer speed has increased exponentially and memory is no longer a limiting factor. Innovative tools for simulation of the PDE system and heuristic approaches for estimating parameters are now available. For example, Nkhalambayausi-Chirwa and Wang (2005) applied a custom PDE solver to solve the model equations for simultaneous Cr+6 reduction and phenol degradation in a dual species biofilm reactor. The PDE solver uti­lized the (fourth-order) Runge-Kutta method with spatial discretization using the (second-order) Crank-Nicholson and Backward Euler finite dif­ference methods for the biofilm spatial profiles. The solution of the biofilm PDE system of equations is shown as a solid line in the effluent from a biofilm reactor showing the prediction of the metal reduction process in the biofilm (Figs 15.12 and 15.13).

Parameter optimization

The parameters in the biofilm model in the example above were estimated using a heuristic procedure — Genetic Search Algorithm (GSA). The first version of this algorithm was implemented in the C programming language using subroutines adopted from Hunter (1998). The GSA uses the inverse of the mean residual sum of squares (MRSS) computed as the global vari­ance (a2) as a fitness function during parameter optimization using the principles of evolution and natural selection (KrishnaKumar, 1993). The global variance is the main objective function for GSA computed as:

G2 =—— Y((obs — y)2 15.16

n — qt!

where a = average deviation of the model from measured values, yobs = observed variables, y = simulated variables, n = number of observations, and q = degrees of freedom representing number of parameters being evaluated.

15.12 Simulation of the reduced metal Mn+ (Cr6+) in a dual species biofilm culture under a range of hydraulic loading conditions: 24 h HRT (Phase I-VI); 11.7 h HRT (Phase VII-X); 6 h HRT (Phase XI-XIV); 17.9 h HRT (Phase XV-XVIII).

15.13 Simulation of the carbon source (P) and metabolite (U) concentration in a dual species biofilm culture under a range of hydraulic loading conditions: 24 h HRT (Phase I-VI); 11.7 h HRT (Phase VII-X); 6 h HRT (Phase XI-XIV); 17.9 h HRT (Phase XV-XVIII).

In Figs 15.12 and 15.13, parameters were estimated from the data obtained from the operation of the reactor at 24 hours hydraulic retention time (HRT) (Phase I-VI). The rest of the phases (VII-XVIII) were simulated using the optimized parameters. The results show high confidence in the optimization routine as the model accurately tracked the trends in effluent concentrations for both the electron donor (P) and the electron sink (M).

Application of biofilm process

The applications of mathematical models to complex biofilm process resembling natural systems are very rare. Most of the reports are based on laboratory-scale pure cultures. Black-box approaches are normally used to evaluate performance of actual systems. The above quoted example by Nkhalambayausi-Chirwa and Wang (2005) was one of the few efforts to mechanistically model a mixed culture system.

The efforts to understand the fundamental nature of biofilm systems are worthwhile since these offer unique solutions to contaminant treatment. For instance, cells growing in biofilm cultures have been observed to perform better than the same species suspended in medium (Semprini and McCarty, 1981). One reason offered for the better performance of cells in the biofilm environment is the effect of shielding from high toxicity levels. It is esti­mated in most biofilm systems that the bulk liquid concentration is much higher than the concentration in the deeper layers of the biofilm. Other complex interactions are also known to exist within the biofilm system. Cultures grown in a cooperative system where some species of microorgan­isms require a product from other species for survival (Nkhalambayausi — Chirwa and Wang, 2005). The mass transport conditions and balanced retention of substrates are required to sustain a culture for a specific purpose.

Biofilm systems have demonstrated the potential to package complex treatment systems into a small space to achieve removal of multiple sub­strates. These systems are especially useful when relatively sensitive species of bacteria are used to treat toxic waste. It is envisaged in this report that biofilm systems will in future form a significant part of the treatment and recovery regime for nuclear and radioactive waste.

Structure and strength of complexes

As has been noted above, actinide ions in their common solution oxidation states (3+ to 6+) are all hard Lewis acids, and their bonds with aqueous ligands are predominantly ionic. Several decades ago it was observed that for a given ligand, the strength of actinide complexes increased with the “effective” cationic electrostatic charge of the actinide ions (Rao, Choppin 1984):

AnO2+ < An3+ < AnO22+ < An4+ 2.5

The effective cationic charges of both the actinyl(V) and actinyl(VI) ions, larger than their overall, formal charge, suggest that the oxygen atoms of both O = An5/6 = O cations retain a partial negative charge and the bonds between the actinide cations and the ligands in the equatorial plane are

Table 2.6 Nominal, overall and effective cationic charge of actinides

Cationic charge


Am3*, Pu3+




Pu4+, U4+






Overall (formal)

1 +









considerably stronger than would be indicated by their formal charge of +1/+2. The values of the effective charge, determined experimentally (Choppin, Rao, 1984) have been confirmed by theoretical calculations (Walch, Ellis, 1976; Matsika, Pitzer, 2000), providing theoretical foundations for the observed behavior of actinides (Choppin, Jensen, 2006) (Table 2.6).

The thermodynamic bond strengths of actinide-ligand complexes are determined primarily by electrostatic attraction of metal and ligand modi­fied by steric constraints; aside from the dioxocations, directed valence effects are generally not evident in actinide coordination compounds. The electrostatic attraction between an actinide cation and a ligand is propor­tional to the product of the effective charges of the metal and ligand divided by the actinide-ligand distance (Choppin, Jensen, 2006). The steric con­straints may arise from the properties of the actinide cation (ion size and presence or absence of actinyl oxygen atoms) or of the ligand (number and spatial relationship of donor atoms, size of the chelate rings, and flexibility of ligand conformations).

If the interactions of ligands with both the actinide cation and proton are governed by the same (electrostatic) interactions, some properties of their metal complexes, such as the stability constant, stoichiometry or structure can often be predicted from the chemistry of their chemical analogs (related ligands or other metal ions of similar properties). Linear free energy cor­relations of structurally-similar complexes can be used to understand dif­ferences in coordination geometries for actinide complexes in solutions (Choppin, 1996; Choppin, Jensen, 2006; Paulenova, Clark, 2009). One can predict the relative strength of the metal complexes by comparison of the ligand affinity for a proton and metal cation. Gibbs free energy of a reaction is proportional to its equilibrium constant K:

AG = -2.3RTlogK 2.6

where R is the universal gas constant and T is the absolute temperature. The logarithmic constants for complex stability constant and the proton­ation of ligand can be used for correlating and interpreting observed values. Obviously, the ligand protonation constant can be expressed as a constant


of the reverse process, ligand dissociation, and log kH will be replaced with pKa.

The next three thermodynamic correlations (Figs 2.3-2.5) are based on the assumption that the interactions between the ligand and cation are primarily electrostatic in nature. When the coordination chemistry of com­plexes is the same, linear correlations between stability constants are appar­ent. A comprehensive database of ligand pKa values (Smith, Martell, 2006)



2.5 Correlation between log pi and the sum of the protonation constants (XpKA) for a variety of complexes with the uranyl cation.

was used in correlations below. Thermodynamic data for gluconate and oxalate ligands were measured at WSU (Paulenova, Clark, 2009).

A nearly ideal correlation between the 1 : 1 stability constants for the neptunyl and uranyl cations with a variety of organic and inorganic ligands confirms the same coordination geometry for both the di-oxo-linear struc­tures of f-elements, and can be used (with caution) to predict the stability constants if one of constant for this pair is known.

Though the different structures of the An3+/4+ cations and An5+/6+dioxocations should impose restrictions on the correlation of thermodynamic data, it is noteworthy that data for the actinyl cation (U(VI), Np(V)) complexes generally correlates well with corresponding data for complexes of amino — carboxylates with the simple, spherical cations (Paulenova, Clark, 2009). Although the coordination geometry for the dioxocations differs from the spherical cations, the correlation between the first stability constants (log) for NTA (nitrilotriacetic) and EDTA (ethylenediamine-N, N,N’,N’-tetra acetic) acids with a variety of metal cations is consistent for both NTA and EDTA. As displayed in Fig. 2.4., the formally monovalent cation NpO2+ lies half-way between the monovalent and divalent spherical cations, and the correlation is satisfactory also for uranyl cation.

The correlation between log p1 and the sum of the protonation constants (XpKa) for neptunyl complexes with different carboxylate ligands is dis­played in Fig. 2.5. The hydroxy-monocarboxylates are correlated worse than aminocarboxylates, and lie between dicarboxylate (oxalate) and other carboxylate ligands. This is presumed to result from the strong donor effect of the oxygens present in the molecules of hydroxycarboxylates.

Spectroscopic demonstration using commercial fuel

Samples of commercial fuel were taken from a high-burnup ATM-109 fuel (Vaidyanathan 1997). The ATM-109 fuel consists of 6 rods that were also produced by General Electric (GE). The rods were initially irradiated in the Quad Cities I reactor beginning in February 1979 until September 1987 in a normal assembly amassing an average exposure of 43 MWd/kgU. The rods were then moved to a carrier assembly and irradiated from November 1989 until September 1992. When removed from the reactor, the average exposure reached 79 MWd/kgU. The rods spent a total of 3508 on-power days in the reactor. Post-irradiation examinations were performed at GE’s Vallecitos Nuclear Center. The fuel sample dissolved for the spectroscopic demonstration had a burnup of 70 MWd/kgU and an initial enrichment of approximately 3%.