An Evaluation of the Electrometallurgical Approach for Treatment of Excess Weapons Plutonium (1996)

Two general “waste” forms are under investigation in conjunction with electrometallurgical processing of spent fuel and the technology's adaptation to disposal of excess weapons plutonium. The first waste form is a glass-bonded synthetic Linde Type A (LTA) zeolite that incorporates radionuclides, including the fission products and possibly transuranic components, into structural cavities (α- and β-cages) within the zeolite via ion exchange with molten salt from the electrorefining process. The second general waste form is a phase-separated metallic waste composed predominantly of the cladding material into which the noble metal fission products from the spent fuel are incorporated via solid solution. Currently, both waste forms are being evaluated as to their suitability for incorporation of excess weapons plutonium, including their ability to comply with the “spent fuel standard” for the final disposition of plutonium.

As discussed in Chapter 3, a molten chloride stream containing chlorides of plutonium, residual uranium, fission products other than noble metals, and transuranic elements would be produced by the simultaneous electrometallurgical treatment of excess weapons plutonium, DOE spent nuclear fuel, and in some cases additional ¹³⁷Cs now in storage at Hanford. The resulting molten chloride stream would be passed though successive columns of salt-loaded synthetic LTA zeolite for essentially complete (>99.9%) removal of plutonium, uranium, fission products, transuranic elements, and added ¹³⁷Cs. Various radioactive cations would transfer into open structural cavities of the zeolite and would be charge balanced by the negative charges on the zeolite framework and occluded chloride ions. Radioactive anions would exchange with chloride ions occluded in the zeolite.

The resulting loaded zeolite, along with associated processing salts (LiCl, KCl), would be size reduced and tumbled with additional salt-free zeolite at elevated temperature (above the melting point of the salt but below the temperature at which structural changes occur in the zeolite) to produce a free-flowing zeolite mixture. A borosilicate glass binder frit would then be added, and the final waste form would be produced by hot isostatic pressing of the mixture. The resulting material is described as a glass-bonded zeolite (GBZ) waste form, also termed “mineral waste.” The GBZ waste form has yet to be produced with actual waste components using this column-loading and hot isostatic pressing procedure. A few column-loading tests and hot isostatic pressing tests with separate material were completed by ANL using simulated waste during late 1995. However, questions regarding mineralogical structure of the processed zeolite, loading capacity, compositional homogeneity, radioelement selectivity, and competing ion effects have not been examined and resolved.

These and other questions are discussed below under the categories of

• zeolite column loading,

• glass-bonded zeolite production, and

• waste form testing and acceptability.

To date, consideration has been given largely to use of the electrometallurgical technique for disposition of excess weapons plutonium as an add-on to an existing spent nuclear fuel treatment process. For this reason, there is considerable uncertainty about concentrations of fission products, transuranic elements, residual uranium, and added ¹³⁷Cs in the molten salt feed to the zeolite columns.

Zeolite sorption of molten chloride salts has been under development at ANL for more than 5 years. The chemistry of ion exchange and salt occlusion in the zeolite has been studied in the absence of large quantities of radioactive materials, but the technology has not been demonstrated on a large scale. The current development scale is with batch sizes of about 1 kilogram.

Work has begun on ion exchange column tests using simulated waste. These studies must ultimately provide sufficient engineering data to determine operable ranges of column configuration, ion exchange materials, and ion exchange particle dimensions, as well as effective operating conditions. The tests must also establish the thermal, chemical, mechanical, and radiological stabilities under expected column loading conditions.

Results to date have shown that ions are removed selectively by the zeolite, dependent largely on ionic charge, with higher-valent ions displacing lower-valent ions. Additional important questions remain concerning how plutonium and the other radioelement ions will load spatially on the zeolite columns, the capacity of the columns, and the ability to maintain sufficiently uniform loadings of plutonium and fission products. Information also is needed on the effects of temperature distribution in the zeolite columns on loading selectivity and sorption kinetics of the various radioelements. Finally, some of these effects will introduce complications into criticality control and engineering optimization of zeolite column sizing and performance, as well as the ultimate attainment of “spent fuel equivalency” of the GBZ.

Fabrication of the GBZ waste is being studied at the 10- to 100-gram batch scale. Screening tests are under way to select the best glass frit to produce an immobilization form having the necessary leach resistance and mechanical properties. Much of the development equipment became available only in late 1995. The committee is concerned with the potentially long development period that may be necessary for the iterative testing and optimization of the GBZ fabrication process.

The GBZ characteristics will determine the actual mechanism (e.g., ion exchange, dissolution) for release of radioelements from the zeolite waste form over the long time periods relevant to geologic disposal. The committee has several areas of concern regarding the progress and adequacy of the characterization and performance testing of the GBZ waste form. In addition, changes in the zeolite structure could alter the suitability of GBZ as a long-term waste form.

In the area of radiologic stability, work to date on radiation damage has been done largely, if not exclusively, via gamma irradiation of laboratory specimens. The committee notes that there are additional important issues arising from unexamined radioactive decay effects, including charge-balance discrepancies arising from radioactive decay of incorporated radionuclides (e.g., radioactive decay of ¹³⁷Cs⁺ to ¹³⁷Ba²⁺) and formation of colloidal metal from gamma irradiation of metal chlorides.

Questions also remain concerning chemical stability, including the effects of impurities such as fluoride ion on zeolite stability and on the stability of the borosilicate glass phase. Rates and mechanisms for alteration of the zeolite by the interaction with water vapor and liquid water are not well established. The committee believes that further study of thermal effects is necessary. Evaluations of heat generation rate and of the potential criticality of plutonium-loaded zeolite that meets the “spent fuel standard” are lacking at this time. Similarly, there is no information on the expected thermal stability of GBZ over the extended elevated temperature range postulated for the Yucca Mountain repository for high-level waste disposal.

Waste form testing has been conducted on a simulated mineral waste form composed of a batch-processed (in contrast to the intended column pass-through process) zeolite A that is physically embedded within a borosilicate glass binder. Preliminary tests reportedly show dissolution rates of ~0.6 g/m²/day for non-radioactive matrix components; these rates are similar in magnitude to those observed in analogous tests of glass waste forms.

These results, however, cannot be used to infer (1) that a zeolite-based waste form would meet repository acceptance criteria, or (2) that a zeolite waste form could be qualified using test protocols developed for other waste forms. For example, the presence of a waste-free glass surrounding the zeolite particles may provide short-term physical protection of the embedded zeolite that could not be relied on for long-term (>100 years) storage under the conditions of a geologic repository. Tests on the stability and release rate of included radioelements should be conducted directly on the “as-produced” zeolite waste form as well as on the subsequent GBZ waste form. It has been shown that the presence of other engineered barriers may radically change the release rate of radionuclides from waste forms, either by chemical interactions or imposition of mass-transport constraints.¹ Acceptability of a waste form must be predicated on the performance of an integrated system of isolation barriers, which includes the waste form as the initial location and source of radionuclides.

For the assessment of long-term waste form performance, the combined cumulative releases and doses have not been calculated. The degree of risk associated with this proposed waste form is unknown relative to the technical risk associated with storage of commercial spent nuclear fuel and vitrified defense wastes, because the GBZ waste form has not yet been developed and tested.

Criticality concerns related to the long-term post-closure repository performance of the GBZ waste form cannot be addressed readily, because the exact composition of the waste forms is still undetermined. The resulting technical concerns are exacerbated by the expected concentration of plutonium at the front end of the loading columns during processing. Some consideration has been given by ANL to introduction of neutron absorbers, but few data are available to support further evaluation of this approach.

The development of the GBZ waste form is viewed by the committee as less mature than that of glass or ceramic waste forms and represents a major technical and programmatic uncertainty. As noted by the committee in its earlier report,²

The time and cost for qualifying any new waste form are expected to be large, and the qualification process is fraught with technical and political pitfalls. To date, no waste forms have been licensed or qualified for geologic disposal, although a large body of knowledge has been accumulated on borosilicate glass, which is the leading candidate waste form for high-level waste and is favored over other waste form types.

The committee concludes that for disposition of excess WPu, issues related to waste form acceptability are of the highest level of importance relative to all other aspects of development of the electrometallurgical technique. However, the committee is not convinced that a sufficiently aggressive program is being pursued to demonstrate waste form performance in a timely manner.³

Cladding metals (zircaloy and stainless steels) are obtained from the electrorefining process as a separate waste stream. Because of their similar electrochemical properties, noble metal fission product elements are expected to be incorporated in this separated waste stream. This waste stream would be melted and cast into disposal canisters. Two subtypes of metal waste form could be produced: either a phase-separated matrix of predominant Zr-rich and subdominant Fe-rich phases or a phase-separated matrix of predominant Fe-rich and subdominant Zr-rich phases, for zircaloy or stainless steel cladding, respectively.

As with the GBZ waste form, a number of questions remain about the production and properties of metal matrices. These questions concern the number, degree of compositional uniformity, and spatial scale of separated phases, which depend on processing conditions, bulk composition, and possibly on the presence of minor alloying components. These characteristics may affect the waste form's performance and suitability. Preliminary evidence suggests that plutonium is relatively enriched in the Zr-rich phase, but this has not been quantified and confirmed. With increasing concentrations of plutonium and other radioelements (e.g., TRU, Cs) necessary for compliance with the “spent fuel standard,” the need for additional phase separations of radioelements cannot be ruled out. Even at lower concentrations, the partitioning behavior of radioelements other than plutonium has not been determined. Evaluation of heat generation rates and of potential criticality is also lacking at this time for plutonium-loaded metal matrices.

Preliminary, short-term waste-form tests reportedly show a corrosion rate of ~0.02 g/m²/day for the cladding metal waste forms at ambient temperature in contact with deionized water. This result is not unexpected, given the comparative insolubility of metals and the relatively benign conditions of such tests. The appropriateness of existing waste-form test protocols, developed for glassy and crystalline aluminosilicate/oxide systems has not been established for application to other waste forms, however. Furthermore, the issue of how to test phase-separated waste forms has not been addressed. For example, the observed preferential fractionation of plutonium into the Zr-phase, the possible separation of an undetermined radioelement-enriched phase, or the possible preferential degradation of the Zr- or Fe-rich

phase make it probable that a new test protocol will be needed to address the unique characteristics of these metal matrices and the effects of these properties on the long-term acceptability of the waste form. As with zeolite waste forms, the acceptability of the metal waste form should depend on the performance of an integrated system of isolation barriers that includes the waste form as the initial location and source of radionuclides.

In the area of radioelement loading of zeolite columns, particular emphasis should be given to

• establishing the range of parameters, with respect to zeolite type, configuration, and operating conditions, that give satisfactory column performance, and

• determining the thermal, chemical, mechanical, and radiological stability of zeolite under expected column loading conditions.

For testing and evaluation of waste forms, the committee recommends the following:

• A schedule should be developed and implemented for demonstrating waste form performance over a time period commensurate with DOE' s plans for treatment of spent nuclear fuel and conversion of WPu to a form suitable for ultimate disposal. Evaluation of waste form performance is of equal concern for application of the electrometallurgical technique to treatment of DOE SNF, although the latter application is governed by a different schedule.

• Waste-form testing should be conducted on the “as-produced” zeolite host phase for radionuclides, as well as on the GBZ waste form.