Electrometallurgical Techniques for DOE Spent Fuel Treatment: Final Report (2000)

The DOE is preparing an environmental impact statement (EIS) for the treatment and management of the remaining sodium-bonded spent nuclear fuel in its inventory using the EMT process. Table 5.1 lists the DOE sodium-bonded fuels by category, quantity, characteristics, and storage location. The total sodium-bonded fuel in the DOE inventory is about 60 metric tons heavy metal (MTHM). The EIS preparation process began in February 1999 with the publishing of the Notice of Intent; a Record of Decision is expected in early 2000.

If the EMT process is to be used successfully to treat sodium-bonded fuels, or any other spent fuels,¹ a number of activities must be carried out after the EMT demonstration, which concluded in June 1999. First and foremost, ANL must complete all the activities required to qualify both the metal and ceramic waste baseline forms for repository disposal. Although not formally part of the demonstration, at the urging of the committee ANL developed test plans and began acquiring waste form performance data prior to the end of the demonstration.² The status of waste form development and qualification was reviewed in this committee’s ninth report.³

In addition to efforts to address these waste form issues, at least two other post-demonstration efforts are essential if the remaining sodium-bonded fuel is to be treated successfully. First, ANL-E must provide ongoing technical support to operations at ANL-W, and ANL-W must complete the required facility modifications and qualify the new, larger-scale equipment needed to handle the increased volume of fuel. These constitute a minimum set of post-demonstration activities.

Other activities that were either abandoned or considerably reduced in scope, or were started too late during the demonstration project, could be considered for continued R&D in the post-demonstration period. These include the following:

• Determine the feasibility of pressureless sintering for producing a ceramic waste form and if feasible, qualify the waste form;

• Develop a high-throughput electrorefiner (HTER);

• Complete the development of the zeolite column to separate plutonium and fission products from the salt; and, finally,

• Complete the development of the lithium oxide reduction step as a front end-process to treat oxide fuel.

Of these post-demonstration activities, pressureless sintering and HTER and zeolite column development are required only if it could be shown that their implementation would significantly reduce the cost or the time required to treat the remaining sodium-bonded fuel. The remaining activity, development of the lithium oxide reduction step, is required if a decision is made by DOE to treat that fraction of the sodium-bonded EBR-II fuel, referred to as “disrupted” EBR-II fuel, for which the cladding has been breached and deterioration by oxide formation has occurred, or if DOE wishes to develop EMT for processing oxide fuels. Alternatively, the lithium oxide reduction work should continue if DOE wants an electrometallurgical process development that can treat oxide fuels.

In the remainder of this chapter, the committee discusses the post-demonstration activities planned by ANL and offers related recommendations.

The specific data required for waste form qualification are determined by the need to ensure the long-term safety of a deep geologic repository containing such waste forms. DOE-RW is preparing, but has not yet finalized, acceptance criteria for DOE spent nuclear fuel and high level waste.

The technical basis for such acceptance criteria has been addressed in previously published safety assessments for the proposed repository at Yucca Mountain, Nevada. The DOE-RW Yucca Mountain Project has conducted several system studies on repository safety. The total system performance assessment study in 1995 (TSPA-95),⁴ in particular, reviews the technical basis for data needs with respect to waste-form and repository performance. The TSPA-95 report makes clear that there are several aspects to waste-form performance that assure safe levels for radionuclide releases from a repository.

The first aspect is the waste-matrix dissolution rate, also called the alteration or leach rate, that controls the long-term release of soluble radioactive elements. The second aspect is the solubility-limited concentration for a given radioactive element, imposed either by equilibrium between groundwater and a stable waste form matrix or by equilibrium between groundwater and alteration products that form because they are more stable than the dissolving waste form matrix. These aspects correspond exactly to those identified in a previous NRC study (the

Waste Isolation System Panel—WISP—report).⁵ A third aspect is the potential formation of radionuclide-bearing colloids arising from waste form dissolution.

Each of these aspects will be influenced by environmental and design factors of the repository system. For example, dissolution rate will be a function of groundwater composition, temperature, and surface area contacted by groundwater over time. Solubility constraints, while also a function of groundwater composition and temperature, are intensive parameters that are not surface-area dependent.

Post-demonstration qualification testing of EMT-produced waste forms must therefore focus on the following:

• Long-term dissolution rate of the initial waste-form matrix (including the effects of fractures),

• Formation of radioactive element solubility-limiting solids, and

• Potential formation of radionuclide-bearing colloids.

These aspects of waste form qualification as they apply specifically to EMT waste forms are discussed in the following sections.

The waste acceptance product specifications (WAPS) include short-term product consistency tests (PCTs) to determine the initial dissolution rate (normalized leach rate) of the EMT waste forms. In Chapter 4, the committee expresses reservations about whether the current WAPS, based on WAPS developed for monolithic borosilicate glass, are appropriate for multiphase EMT waste forms. In particular, normalized leach rates of stable elements (e.g., Na, Si) that are common to two phases in the CWF, namely sodalite and borosilicate binder, are likely to be difficult to interpret unambiguously. More importantly, the radionuclide release behavior of other phases in the CWF that do not contain these stable components, such as U-Pu oxide, actinide silicate, and oxychlorides, cannot be determined by a conventional PCT devised for monolithic borosilicate glass.

A previous NRC report criticized “leach rate” as a measure of the long-term performance of waste forms under expected repository conditions.⁶ The key concern is that short-term tests on waste forms under closed-system conditions fail to incorporate mass-transfer constraints (e.g., diffusive or diffusive-advective transport). Consideration of mass-transfer constraints shows that ultimately two factors control the long-term release of most radionuclides from a multibarrier repository. These are radioelement solubility limits, which are imposed either by the solubility of a stable waste matrix or formation of more stable alteration products, and fracturing, which can expose more surface area that contains the high-alpha-activity phases.^7,8

Especially germane to the charge of this committee, however, is the recognition that solubility limits that determine long-term repository performance are strongly dependent on “the compositional parameters of the concentration-limiting solid phase.”⁹ Slight changes in chemical factors (e.g., waste-form composition, groundwater composition) can lead to different solubility-limiting solids with extreme differences (by several orders of magnitude) in predicted solubility concentrations for radioelements.¹⁰

The post-demonstration evaluation of the long-term performance of EMT waste forms, especially the CWF, under repository conditions must address this aspect of solubility limits for radioelements. Such an effort would address the formation and characterization of alteration phases under appropriate conditions, as well as measurement of the solubility-limited concentrations of radioelements. ANL cites the American Society for Testing and Materials (ASTM) Standard-C1174-97¹¹ as a basis for its test program for qualifying waste forms. With respect to predicting the long-term behavior of waste forms, the 1997 ASTM standard makes no reference to the perspectives and conclusions of the WISP report¹² regarding the formation of more stable, solubility-limiting solids. Nor does the ASTM standard reference the TSPA-95 safety assessment report¹³ or equivalent system-level safety analyses that affirm the importance of solubility-limiting phases. A possible reason is that the ASTM standard restricts itself solely to the “alteration” of initial waste forms and other barrier materials.¹⁴ As previously noted,¹⁵ the alteration rate (leach rate) of waste forms is an irrelevant factor in the overall release rate of most radionuclides from waste forms emplaced in deep geologic repositories.

The committee notes that ANL has used a bulk dissolution rate (leach) test method developed for a single-phase nuclear waste form (borosilicate glass). It is now clear, however, that the CWF and the MWF are multiphase waste forms and that each phase has a very different radionuclide composition. Such different phases will likely experience different rates of dissolution under repository conditions. Hence, the present use of bulk leach rate tests has extremely limited value in assessing any meaningful measure of the performance of multiphase EMT waste forms in a geologic repository.

There is considerable concern regarding the potential for rapid migration of significant quantities of radionuclides, especially Pu, at Yucca Mountain via colloidal transport.¹⁷ The as-produced specimens of the CWF contain separate U-Pu oxide particles on the order of 10 nm in radius. Such separate U-Pu oxide phases could conceivably contribute to colloidal transport upon waste form dissolution under repository conditions.

The committee notes that assessment of repository safety will be a function of the performance of all waste emplaced in such a geologic disposal system. The number of containers and the total DOE radionuclide inventory of EMT waste forms are extremely minor in comparison with the volume of commercial spent nuclear fuel and defense waste processing facility borosilicate glass intended for co-disposal at Yucca Mountain. It is conceivable that the uncertainties in radionuclide inventory and release-rate performance of these dominant waste forms may have a far greater impact on meeting a total-system safety standard than would conservative bounding assumptions made regarding EMT waste forms.

Alternatively, an arbitrary 10,000-year cutoff may be applied by the Environmental Protection Agency in its eventual safety standard for Yucca Mountain. In such a case, isolation strategies involving high-integrity containers designed to physically isolate HLW for more than 10,000 years could obviate the need for any long-term performance testing of waste forms.¹⁸ As noted in this committee’s fourth report, the 1983 WISP report argued that extrapolations from leach tests may not be “applicable to predicting performance of waste packages in geologic repositories.”^19,20 Furthermore, even if a base case of extended containment were to be accepted, the potential for early failures in containers, arising from fabrication or emplacement operations, would still require a database on waste form performance.

Should DOE decide to use electrometallurgical technology to further process spent nuclear fuel in its inventory, most of the equipment used for the EMT demonstration process would be used as is during these post-demonstration operations. However, in addition to replacement of the driver-element chopper, the waste processing equipment will have to be resized to meet the increased throughput. For example, for the ceramic waste form a new V-mixer with double the capacity of the existing one, and a larger hot isostatic press (HIP) capable of processing a 51-cm-diameter can will be required. Additionally, a new furnace dedicated to producing the metal waste form is needed. This new furnace is intended to both distill the salt from the cladding and cast the cladding into an ingot. Design changes to the existing cathode processor are also necessary so that the crucible load capacity can be increased from 54 to 75 kg. During inventory operations, the cathode processor and casting furnace used in post- demonstration activities will be dedicated to uranium processing.

The facility will be modified to allow repair of equipment in the hot cell in such a way as to not disrupt ongoing operations. In addition to decreasing downtime due to equipment malfunctions, some parts of the infra-structure such as the cell purification and refrigeration systems will be upgraded.

ANL is also investigating alternative methods to use of HIP for ceramic waste form fabrication. Such a method is “pressureless sintering.”

The preparation steps for pressureless sintering are essentially identical to those for the HIP process. The process for pressureless sintering is discussed in detail in Chapter 3. Studies are being conducted to establish optimal fabrication procedures, and to establish whether pressureless sintering can produce a suitable waste form. The short-term leach characteristics of waste forms from HIP and pressureless sintering are also essentially identical, although the committee has previously noted that such short-term tests may not be fully indicative of long-term release-rate performance of waste forms under expected repository conditions. Furthermore, the heterogeneous nature of the multi-phase ceramic waste form mandates examination of the microstructure and phase composition of as-produced waste forms.

Pressureless sintering may provide advantages relative to the HIP process during fabrication. The main technical advantage of the pressureless sintering over the HIP process would be its higher throughput per square foot of process space. It should be noted, however, that although the initial steps in the HIP process (i.e., loading, evacuating, and sealing of the HIP can) are carried out in the Ar cell, the actual HIP is carried out in air. Pressureless sintering would have to be carried out completely in the Ar cell. Also, the HIP process is inherently a batch process (load, bring to temperature and pressure and hold, and finally return to ambient temperature and pressure), whereas pressureless sintering, which depends only on maintaining a sample at a specified temperature for a specified length of time, can be operated in practice as a continuous process.

In the post-demonstration period, a HTER, particularly if it could be cost-effectively developed and implemented in a timely fashion, could offer the advantage of considerably reducing the time required to treat the remaining sodium-bonded fuel. Currently, the proposed schedule calls for the remaining sodium-bonded spent fuel (approximately 58 metric tons of heavy metal-MTHM) to be treated at a rate of approximately 5 MTHM/year over a period of 12 to 13 years. The 5 MTHM/year rate is based on estimated rates of 0.6 MTHM/year and 4.4 MTHM/year for the Mark-IV and Mark-V, respectively. To achieve the estimated rate, the Mark-V will operate anode-cathode modules (ACMs) in three of the four ports simultaneously at a deposition rate of approximately 400 g/hour per ACM and a 50% duty cycle.

A processing rate of greater than 5 MTHM/year could potentially be achieved by further development of the 25-inch HTER at ANL-E. The committee notes that the rate of 5 MTHM/year rate could be doubled simply by adding a second Mark-V to the Ar cell at ANL-W. Since the development work on the Mark-V is essentially over and its capabilities have been demonstrated, this alternative appears to offer distinct cost advantages. Adding a second Mark-V will double the rate only if the electrorefining step is the rate-limiting step.

Initially the EMT process included both a multistage pyrocontactor and a zeolite column to treat (recycle) the spent electrorefiner salt. The purpose of the pyrocontactor was to remove the residual transuranic elements and the purpose of the zeolite column was to extract and immobilize the fission products. After this treatment, the salt was to be returned to the electrorefiner, and the waste-bearing zeolite from the column was to be converted to a solid monolithic form. This plan was later modified after the revised environmental assessment for the demonstration project prohibited using a cadmium cathode to separate out plutonium. As a result, the plutonium remained with the salt and the pyrocontactor became obsolete. However, work continued at ANL-E to develop the zeolite column. The plan was to process spent fuel so as to raise the level of fission product contamination to a point that as a result of running the salt through the zeolite column, the loading of the salt would be about 3 wt %. Because of this revised plan, radioactive waste treatment operations could not be started until late in the process, and

characterization and qualification of a waste-load waste form would have to come even later. As a result, radioactive waste form production and characterization were not part of the original EMT demonstration project.

However, part way through the demonstration project, ANL changed directions and decided not to pursue salt recycle at that time. Instead, ANL implemented the so-called throw-away salt option, in which a portion of the plutonium- and fission-product-contaminated salt would be removed from the electrorefiner, and in a batch process, mixed directly with small zeolite and glass particles. This mixture would then be subjected to the HIP process to produce the desired waste form. The decision to implement the throw-away salt option resulted in essentially abandoning further radioactive zeolite column development. However at ANL-E, scale-up, remote design, and “cold” column operation development continued, although at a reduced level, in parallel with the EBR-II demonstration.

If the zeolite column work is to be resumed in the post-demonstration period, important questions remain such as how Pu and the other radioactive elements will load spatially on the column, the capacity of the columns, and the ability to maintain sufficiently uniform loading of Pu and fission products. Information is also needed on the effects of temperature and of potential temperature gradients on selectivity and sorption kinetics, as well as particle size and flow rate. Resolving the uncertainties related to these questions and issues would require a reasonable amount of R&D, but the problems should be ones that can be solved. However, the loss of water in the column in the early phases of elution may prove to be an intractable problem that would prevent the use of the column process. Small amounts of residual water on the zeolite are required to react with the metal oxychlorides to convert them to the desired oxide.²²

For EMT to be used to treat oxide fuels, a head-end step is required to convert the oxide to metal. ANL-E has been pursuing the use of lithium metal as a reducing agent in molten LiCl salts to effect this conversion. The lithium metal is regenerated by electrolysis of the resulting lithium oxide so that the lithium and LiCl can be recycled. The interface between the reduction step and the electrorefining step is critical because any lithium or Li₂O carried over in the reduction could interact with UCl₃ in the electrorefiner. However, ANL has demonstrated the technical feasibility of coupling the reduction and electrorefining steps using existing technology.²³ However,

some work remains to be done to complete the development of the Li reduction step. Some of the issues that remain are the extent of reduction of Pu₂O_3, the optimal design and materials of construction for the electrowinning cathode, the kinetics of UO₂ reduction, and methods for handling metallic lithium. Continued R&D is required, but the committee believes there are no “show stoppers.”

This page intentionally left blank.