Electrometallurgical Techniques for DOE Spent Fuel Treatment: An Assessment of Waste Form Development and Characterization (1999)

The U.S. Department of Energy (DOE), through its Office of Civilian Radioactive Waste Management (RW), is assessing the viability of permanent disposal of spent nuclear fuel (SNF) and high-level waste (HLW) in a deep geologic repository at Yucca Mountain, Nevada.¹ Long-term isolation of these nuclear materials from the biosphere is planned through a “defense-in-depth” approach of multiple natural and engineered barriers. The performance and compatibility of the Argonne National Laboratory (ANL) waste forms must be assessed within a system context of overall repository safety. The committee recognizes that while general specifications have been identified for acceptance of DOE spent fuel and waste forms into the repository program,² explicit criteria for evaluating such waste forms under relevant repository conditions are not yet available. Furthermore, the committee recognizes that it is not waste-form performance per se, but rather performance of the integrated system of engineered and natural barriers that must be demonstrated for safety. Nevertheless, DOE asked the committee to evaluate ANL's progress in taking appropriate steps that would be necessary for obtaining the necessary regulatory approvals in the future for the waste forms produced by the electrometallurgical process.

DOE-RW is charged with responsibility for determining overall waste-acceptance product specifications.³ However, at this time RW has not yet promulgated waste-acceptance product specifications for any waste form destined for placement in a geologic repository. As a result, any evaluation of ANL's testing program for its ceramic waste form (CWF) and metal waste form (MWF) must be performed within the context of this lack of criteria. The committee does not intend to evaluate whether these waste forms will eventually be accepted in a geologic repository, but rather whether ANL's present testing protocol is adequate to characterize the CWF and MWF.

The actual safety standards for a geologic repository have yet to be promulgated by the U.S. Environmental Protection Agency (EPA). There has been considerable debate concerning an appropriate safety standard, a topic on which the National Research Council has published a special study.⁴ The recent study on Yucca Mountain by RW,⁵ for example, indicates that if an arbitrary 10,000-year cutoff is assumed for dose modeling, an isolation strategy based on extended containment (i.e., preventing groundwater from contacting waste forms) would greatly mitigate the importance of waste forms for repository safety. Until EPA finalizes a safety standard, it is difficult for

the U.S. Nuclear Regulatory Commission and DOE to determine exactly what data will be required for safety assessment in support of repository licensing.

The MWF and the CWF are HLW forms intended for final disposition in a geologic repository. The electrometallurgical technology (EMT) Program has developed a waste-qualification program that is patterned after the protocols used for the waste qualification of Defense HLW (DHLW) borosilicate glass.⁶ The early phases of waste-form acceptance modeling and data collection activities by the EMT Program are being conducted during ANL's demonstration project for the EMT process. To support a final waste-acceptance decision, however, major qualification and characterization activities will continue beyond the end of the demonstration project.

To date, both commercial spent fuel and vitrified defense HLW have been subjected to detailed characterizations conducted with respect to their performance in a geologic repository.⁷ Such characterization data for borosilicate glass have been used to guide isolation strategies as well as initial design of engineered barrier systems to assess the viability of a geologic repository. ⁸ At present, only DOE SNF is being grouped by RW with respect to common characteristics. For the EMT waste forms, preliminary evaluation is being performed, and no final decision on the EMT waste form has been made by RW. This preliminary evaluation is using bounding data on these waste forms.⁹

To assure coordination between RW and DOE's Office of Environmental Management (EM), the “Memorandum of Agreement (MOA) for the Acceptance of DOE Spent Nuclear Fuel and High-Level Radioactive Waste” was issued.¹⁰ This MOA establishes the terms and conditions under which RW will make available disposal services to EM for DOE SNF and HLW. The responsibilities of RW and EM relative to data collection, transportation, storage (if needed), safeguards, characterization, and final acceptance for disposal of these materials are identified. The responsibility to treat the EBR fuel rests with DOE's Office of Nuclear Energy, Science, and Technology (NE), but the ultimate disposition of this treated EBR fuel and any HLW waste forms generated is the responsibility of EM. Hence the waste-acceptance activities of the EMT Program will be guided by the MOA.

Figure 2 in Chapter 2 shows a flow diagram of the interrelated waste characterization and verification activities of both RW and the producers of DOE HLW (EM and NE).¹¹ From these activities, documents are to be produced that will support the decision for acceptance of EMT HLW forms by RW. This flow diagram is broadly similar to flow diagrams presented in “Appendix C: Subagreement on the DOE SNF and HLW Technical Baseline” in the MOA and the Savannah River Laboratory Defense

Glass Program.¹² In particular, Appendix C of the MOA addresses issues related to the development, concurrence, distribution, compliance, and conformance verification of acceptance criteria for DOE SNF and HLW.

The specific waste qualification requirements for waste-form producers (NE and EM) by the repository operator (RW) are shown in Figure 2 of Chapter 2 and are defined in Appendix C of the MOA¹³ as follows:

Acceptance: The transfer of responsibility, custody, and physical possession of DOE SNF or HLW from EM to RW at the EM site.

Acceptance Criteria: All technical and programmatic requirements that must be satisfied by DOE SNF and HLW for the repository program to meet regulatory requirements. RW is currently preparing a “Civilian Radioactive Waste Management System (CRWMS) Acceptance Criteria” document.

Waste Acceptance Product Specifications (WAPS): The documentation by a HLW producer that identifies the technical specifications for the HLW waste forms.

Waste Form Compliance Plan (WCP): The documentation prepared by a HLW producer describing planned analyses, tests, and engineering development work to be undertaken, as well as information to be included in waste-form production records to demonstrate compliance of the proposed waste form with waste acceptance specifications.

Waste Form Qualification Report (WQR): The documentation prepared by a HLW producer which describes results of analyses, tests, and engineering development work performed to demonstrate waste-form compliance with acceptance specifications.

Broadly, two types of data requirements can be envisioned within the waste-form acceptance activities shown in Figure 2 of Chapter 2. The first type of data is collected to verify that the as-produced HLW waste forms consistently conform to acceptance specifications for disposal, including topics such as particulates, pyrophoricity, dimensions, radionuclide inventories, and heat-generation rate.

The second type of data are those more directly connected to the long-term (10,000 years or more) performance characteristics of such HLW forms under expected repository conditions. The “Waste Testing and Qualification” activity shown in Figure 2 of Chapter 2 is conducted by the waste-form producer to provide specific input to RW's repository performance assessment task. The MOA's section VII “Acceptance Criteria”

notes that RW shall be responsible for the long-term performance characterization of HLW starting in fiscal year 1999.¹⁴

Issues remain that impact the ability of the EMT Program to fully document its plans and schedule for achieving a future waste-acceptance decision. First, the EMT Program is concluding a directed demonstration phase that supports issuance of an Environmental Impact Statement (EIS) regarding continued application of the EMT process to the remaining inventory of EBR-II spent fuel. As previously noted, the EMT Program must orient its current activities to provide evidence of successful compliance with demonstration criteria.¹⁵

A final consideration is that the initial draft of RW's Acceptance Criteria document should be issued for review in 1999. This new document may modify the actual waste-acceptance strategies and waste-acceptance criteria that the EMT Program is currently following.

Finding: From interactions with RW, ANL has developed a strategy appropriately based on RW's waste acceptance criteria for the characterization of its MFW and CFW for eventual acceptance by RW.

This strategy encompasses its characterization protocols, including short-term test procedures, for its ceramic and metal waste forms.

Conclusion: Continued interaction between ANL and RW will become even more important in the postdemonstration phase.

Conclusion: There remains uncertainty regarding which DOE organization will be charged with the ultimate responsibility for performance-confirmation testing of waste forms suitable to support a repository licensing decision. As this uncertainty in responsibility could lead to costly duplication of effort and lack of consensus among DOE organizations regarding data supporting future decisions, DOE should take the lead in achieving a documented resolution to this issue.

The electrometallurgical treatment of spent EBR-II reactor fuel involves a set of operations designed to disassemble driver and blanket fuel pins, to refine and recover the uranium metal contained therein, and to segregate the radioactive waste components. The radioactive waste components are consolidated into two forms, CWF and MWF. The CWF includes transuranic elements and fission products in a glass-ceramic matrix, whereas the MWF contains noble metal fission products in a fuel-cladding matrix. The

MWF may contain up to 4 weight % noble metal fission products and up to 11 wt % uranium. The long-term corrosion behavior of this type of alloy is not known and needs, therefore, to be determined in the corrosion tests carried out at Argonne National Laboratory-East (ANL-E). The EBR-II driver fuel is primarily a uranium-10 weight % zirconium alloy with type 316, D9, or HT9 stainless steel cladding. Because zirconium is a principal component of the driver fuel, zirconium will be a significant component of the metal waste stream. For the entire EBR-II spent fuel inventory, the base metal waste stream composition is stainless steel containing approximately 15 weight % zirconium, labeled SS-15Zr by ANL personnel.¹⁶ Thus, MWF testing at ANL-W has focused primarily on this and similar alloys.

The MWF is obtained by melting at 1,600 °C in an inert atmosphere the cladding residue that remains from the electrorefiner step. The molten residue is adjusted to contain 15 weight % zirconium and then cast into ingots. Corrosion resistance and noble metal fission product retention are the principal requirements for the safe application of the MWF. Waste-form qualification involves experimental testing and model development. Research at ANL-E has evaluated alloy metallurgy and alloy properties, including mechanical properties, thermophysical properties, and corrosion behavior.¹⁷ The corrosion resistance of SS-15Zr alloys has been determined using immersion tests, electrochemical tests, and accelerated corrosion tests (vapor hydration, high temperature immersion, and product consistency tests).

Plans for qualification testing beyond June 1999 and testing highlights have been presented at the committee's recent meetings.^18,19 The MWF test plan consists of attribute tests, characterization tests, accelerated tests, and service condition tests. The attribute tests, as defined by ANL, are designed to provide material property information using electron microscopy, x-ray analysis, and neutron diffraction. Good progress seems to have been achieved in the identification of the various phases of SS-15Zr-type materials. Noble metal-rich precipitates have not been observed.

The characterization tests consist of immersion testing in sealed Teflon™ vessels at 90 °C in J-13 (simulated Yucca Mountain well water) and deionized water. The tests that have been terminated showed either no corrosive attack or only minor tarnish. The test solutions have been submitted for elemental analysis. Justification for whether the planned total of 856 tests is necessary to achieve the goals of the project has not been provided to the committee.

The accelerated tests are designed to shorten the test period and consist of immersion in deionized water in a titanium vessel at 200 °C for 28 days. Six alloy compositions were tested. Corrosion rates were very low and no correlation of elemental leaching with alloy composition was found.

Electrochemical corrosion testing is based on the polarization resistance technique (ASTM G59),²⁰ which is used to measure instantaneous corrosion rates.²¹ Electrochemical corrosion testing is used to screen out alloy compositions that may not be suitable for repository disposal. An example was given for an SS-1Ag-2Nb-1Pd-1Ru alloy that did not contain zirconium and had high corrosion rates at pH 2. Corrosion rates for alloys that contained from 15 to 20 weight % zirconium were low. Corrosion rates of the MWF alloys in J-13 and in solutions of pH = 2, 4, and 10 were low and similar to those of SS316 and alloy C22. These results are not surprising considering that the solutions tested did not contain chloride ions that could have initiated localized corrosion. Corrosion rate data for MWF materials were also compared to those of copper and mild steel.

The results from pulsed-flow immersion tests of SS-15Zr alloys containing Nb, Pd, Rh, Ru, and Tc in J-13 water at 90 °C for up to 275 days showed a sudden increase in Tc release rates after about 150 days; however, the overall release rate remained relatively small. The cause of this behavior is under investigation. The results of the immersion tests, which have shown that only small amounts of fission and actinides are dissolved in the test solution, suggest that corrosion of the SS-15Zr MWF is not a dominant release mechanism. Corrosion appears to be retarded by the formation of a passivating oxide layer that may trap the fission products and actinides, limiting their release.

Finding: Some of the corrosion products, which may sequester radionuclides, might remain on the sample surface and might not be detected by solution analysis.

Vapor hydration tests have been performed in sealed SS vessels for 56 and 182 days. It was found that the corrosion rates were greatly accelerated by exposure to steam. The corrosion product layer for SS-15Zr has been estimated to have a thickness of about 1 µm after 56 days based on visual observations. Samples containing less than 5 weight % zirconium (or no zirconium) were heavily rusted and contained numerous pits. The chemical nature of the corrosion products is under investigation.

No data have been presented to the committee thus far for “standard” SS-15Zr samples. ANL personnel did discuss corrosion testing of SS-15Zr MWF samples at a 1998 meeting, concluding that “SS-Zr waste forms are very resistant to the normal corrosion conditions envisioned at the proposed Yucca Mountain geologic repository.”²²

Conclusion: Corrosion behavior data for the SS-15Zr MWF standard need to be obtained.

The effect of radiation on corrosion behavior has been discussed only briefly. Calculations carried out at ANL seem to suggest that radiation levels in the MWF are too low to affect the corrosion resistance.

The toxicity characteristic leaching procedure (TCLP) test data suggest that the MWF passes the TCLP test. The results for the release of Ag, As, Ba, Cd, Cr, Hg, Pb, and Se were below the detection limits of acceptable methods.

Recommendation: Surface analysis by x-ray photoelectron spectroscopy (XPS) or Auger electron spectroscopy (AES) should be performed for selected samples to determine the chemical composition of passivation filings and/or corrosion products. Because a large number of samples to be tested differ only slightly in minor alloying elements, it is recommended that only a few of these samples be subjected to full characterization. These samples should be selected using a statistical analysis approach.

Based on the results from the various corrosion tests, ANL concluded that SS-15Zr shows corrosion behavior similar to that of stainless steels such as SS316. High corrosion rates were observed in electrochemical and vapor hydration tests for alloys with less than 5 weight % zirconium. In the immersion tests, high release of silver was observed for an alloy that did not contain zirconium. The significance of these results is not clear because the alloy composition was below the lower limit of the zirconium specification. The corrosion resistance of the MWF appears to be dominated by the passivation behavior of the alloy, and dissolution is not considered to be a dominant release mechanism of the radionuclides.

Finding: Results from corrosion testing of the MWF in rather benign environments suggest that the corrosion behavior of the MWF is similar to that of stainless steel.

Finding: At the present time, ANL has not indicated how it plans to conduct crevice corrosion studies.

The tests to be performed after June 1999 have not been finalized. Electrochemical tests are to be performed at elevated temperatures in order to shorten the test period. It was suggested by the committee that ANL concentrate on a few key samples, expose them at higher temperatures, and obtain electrochemical and surface analysis data. Tests are also to be conducted in chloride solutions with concentrations of up to 10,000 ppm, which are credible conditions that might be encountered in the repository.

Finding: ANL has carried out a number of corrosion tests using mild solutions. Under these conditions, significant corrosion damage to the MWF is not expected.

Recommendations: Instead of continuing to conduct a large number of corrosion tests using mild conditions, it would be better to subject a few carefully selected samples to additional evaluation by surface analysis to determine the chemical composition of the corrosion products. It may be better to concentrate on a few key

samples, expose them at higher temperatures, and then obtain electrochemical and surface analysis data.

Guidance for carrying out the needed pitting scans can be obtained from ASTM G61.²³ Corrosion rate data could also be obtained from such measurements. The proposed study of crevice corrosion needs careful design of an artificial crevice with consideration of the proposed application of these alloys. ASTM G48,²⁴ which describes a multiple crevice assembly, could serve as guidance.

Two of the Success Criteria Goals for ANL's demonstration project relate to MWF specifically. These criteria and associated work to date are as follows:

Criterion Goal 2-2. Develop metal waste specifications that are based on performance characterization results of small samples with variations in the principal constituents: zirconium, uranium, technetium, plutonium, neptunium, and noble metals. Determine performance characterization with electrochemical techniques, corrosion tests, vapor hydration tests, and attribute tests.

Accomplishments as of June 15, 1999. The analysis of 110 samples, 80 of which were spiked with radioactive constituents, is 99% complete and will be summarized in the Metal Waste Form Handbook. Analyses were performed with a wide variety of techniques and procedures, including scanning electron microscopy (SEM) and transmission electron microscopy (TEM), immersion tests, vapor hydration tests, thermal aging tests, crevice corrosion tests, neutron diffraction analysis, bulk material tests, electrochemical corrosion tests, and TCLP. These tests provide the basis for performance of the MWF with reasonable variations in composition of the major components (zirconium and SS) and minor components (U, Tc, Pu, Np, and noble metals). Because Pu and Np are trace constituents in the MWF, the metallurgical behavior of these elements was investigated with small ingot samples that were spiked with up to 6 weight % Pu and up to 2 weight % Np.

Criterion Goal 2-3. Develop metal waste process specifications for major process variables: operating temperatures, hold time, and cooling rate.

Accomplishments as of June 15, 1999. Three ingots have been cast with irradiated cladding hulls from the electrorefiner and have shown acceptable casting parameters at 1600 °C and a 2-hour hold time. Characterization of these ingots has shown that they are similar to the metal waste test matrix samples. The operating parameters are summarized in the Metal Waste Form Qualification Plan, which is being compiled.

In addition to the tests that have been completed as part of the demonstration project, there are a number of long-term tests on the metal waste form whose results will not be known until after then end of the project. As a result, full characterization of the metal waste form may not be complete by the end of the demonstration.

As described in Chapter 2, the reference CWF for the demonstration is glass-bonded sodalite formed from the hot isostatic pressing (HIP) process.²⁵ Sodalite is the thermolysis product formed during the HIP of the salt-loaded zeolite 4A. As shown in Figure 1 in Chapter 2, the salt from the electorefiner containing transuranic elements and fission products is blended with dried zeolite and is heated to occlude the salt into the zeolite. The salt-loaded zeolite is then densified into the CWF by HIP. By appropriate choice of temperature and pressure, the zeolite is converted into sodalite during the HIP process. Scale-up problems previously encountered in the heating of the salt/zeolite blend, and particle size mismatch, were successfully resolved.²⁶

ANL reported²⁷ on the detailed testing to support CWF qualification using “scoping” tests.²⁸ These scoping tests include studies of solution exchange with the CWF; product consistency tests in which the waste form, which is crushed and sieved to achieve suitable particle sizes and washed to remove fines, is leached; the Material Characterization Center Test-1 (MCC-1), a static leach test that uses a monolithic sample; pH stat tests; accessible free salt measurements; and vapor hydration. ²⁹ These tests are carried out over a range of environmental conditions but do not provide data on the long-term release-rate performance of these waste forms with respect to relatively insoluble radioelements.

Limited tests indicate that dissolution of the sodalite matrix, rather than ion exchange, controls the release to solution of radioelements such as cesium and

strontium.³⁰ The MCC-1 test reflects early dissolution rates, and the PCT tests reflect somewhat longer time dissolution. Neither test, however, reflects the open-system, mass-transport conditions that govern the actual release rates of most radionuclides from the waste package system for disposal.^31,32 The release of highly soluble radioelements (e.g., iodine and cesium), however, probably will be limited by the dissolution rate of the encapsulating phase. Tests to date show that the CWF dissolves at a rate equal to or less than reference defense high-level waste (DHLW) borosilicate glass. This anticipates that the CWF repository performance will be comparable to that of the reference borosilicate glass.

Finding: ANL's tests of several months' duration indicate that the CWF dissolves at a rate equal to or less than that of the reference DHLW borosilicate glass.

Conclusion: If dissolution remains the dominant release mechanism under actual repository conditions, then the release performance of the CWF will be at least comparable to that of DHLW borosilicate glass.

As stated previously, the performance and compatibility of the CWF must be assessed within a system context of overall repository safety. However, until that assessment has been completed, it is significant that the dissolution rate of the CWF is lower than that of reference DHLW borosilicate glass when the two waste forms are subjected to comparable test conditions. The committee is aware that long-term test results could alter this observation regarding the relative dissolution rates of the two waste forms.

ANL reported33 on modeling for waste form lifetime in the repository. The model attempts to predict long-term behavior based on short-term data. The model is based on a transition state theory approach for the rate of silicate mineral dissolution.34,35 The model incorporates two effects. The first is the forward reaction rate in the absence of dissolved silicic acid. The forward rate is both temperature and pH dependent. The second effect takes into account the relative degree of saturation of silicic acid with respect to the solubility product of the dissolving solid. In previous applications to borosilicate glass, the solubility product of a proxy solid phase (for example, amorphous silica) is used. This approach has been used to model dissolution rates of high-level and low-level borosilicate waste glass for performance assessments. ANL's use of the model is aimed at predicting degradation and radionuclide release for the CWF. The model assumes congruent dissolution of the sodalite—i.e., aluminum and silicon (and presumably other

minor and trace components) are released to solution at rates whose ratio equals their stoichiometric proportions in the sodalite phase. ^36,37 This assumption should be validated. Input data to the model is based on MCC-1 (forward rate) and PCT data (relative saturation) on both simulated and actual CWF. ANL has yet to demonstrate that its resulting model adequately represents measured data on the rate of general corrosion of the crystalline CWF.

ANL³⁸ discussed accelerated alpha damage testing on simulated CWF doped with 0.2 to 2.5 weight % ²³⁸Pu or ²³⁹Pu. The ²³⁸Pu is shorter lived, and hence generates a larger alpha flux than would be found in the CWF.³⁹ Early results to date at low doses indicate limited cumulative damage. Results were from x-ray diffraction lattice parameters, PCTs, and some SEM and density examination. SEM indicated that plutonium oxide nanocrystals were formed and clustered along grain boundaries. There was some discussion by the committee that the grain size of the plutonium oxide, might lead to increased solubility and has the potential for colloidal particle formation.⁴⁰

Finding: During the conduct of the alpha-decay tests, plutonium oxide was observed as nanocrystals in the grain boundaries.

Conclusion: Plutonium may not be in the sodalite phase.⁴¹ Its presence in potentially colloid-sized products may have implications for the long-term release behavior of plutonium and any other radionuclides that also segregate into such colloid-sized phases.

1. The CWF is a multi-phase, nonhomogeneous material that introduces several issues. SEM images showed the presence, for example, of separate phases of cerium and neodymium oxide, as well as separate nonoccluded salt phases. Many actinides might preferentially partition into such rare-earth oxide phases, and fission products might partition into the free-salt phases. The EMT Program's waste-form qualification program is based on adaptation of models and test protocols developed for DHLW borosilicate glass.

Recommendation: The EMT Program should continue to evaluate and demonstrate that test protocols and conceptual models developed for monolithic single-phase borosilicate glass can adequately represent the behavior of the nonhomogeneous multiphase EMT CWF.

2. There is to date little or no direct evidence quantifying the distribution of radionuclides among the phases. This is an area where sufficient information is needed to ensure adequate understanding of the material performance. SEM has shown that cerium, neodymium, and plutonium oxides exist as separate phases.

Finding: The material characterization center test (MMC-1) and product consistency test (PCT) designed to model the release behavior of inert, major components of the CWF may be irrelevant with respect to evaluating the release of plutonium and other actinides partitioned into separate oxide phases.

3. The committee has noted the importance of understanding the effects of alpha-recoil damage on the CWF as well as the effects of fission product decay on the stability of sodalite- and zeolite-based CWF. ANL has prepared CWF samples containing ²³⁸Pu for evaluating the effects of alpha-recoil damage to the CWF. During the operation of the electrometallurgical process, fission product elements that are thermodynamically less noble than uranium accumulate in the molten salt bath from the electrorefiner. These fission product elements then batch equilibrate dry Type A zeolite with the molten salt to produce a CWF that contains fission products. These samples will be evaluated subsequently to determine the effect of fission product loading on the CWF as well as effects of radiation damage to the CWF.

The observations from SEM and TEM of the simulated CWF used for alpha-decay damage studies show that the plutonium is present as nanocrystals in a separate oxide phase, primarily at the sodalite grain boundaries. Because the plutonium exists as an oxide in the grain boundaries, it is problematical whether the effects of alpha recoil damage from plutonium decay would be observed by x-ray diffraction measurements of the unit cell dimensions (percent swelling) of the bulk sodalite phase. X-ray diffraction measurements may not be meaningful if the plutonium is not in the sodalite phase because the penetration distance of the alpha-recoil particles would occur predominantly at the grain boundary region. The current x-ray diffraction test would supply damage information only if the plutonium were distributed evenly in the sodalite lattice. Furthermore, the formation of nanocrystals of plutonium oxide after fabrication of the CWF raises questions as to whether plutonium might be released through a geologic repository system as plutonium oxide colloids rather than as dissolved plutonium.

HIP has been adopted as the reference technology for densification during the demonstration program. To date, tests have been successful using 4.5-inch cans (a limited number of tests with 8-inch cans have also been performed). A cooperative research and development agreement (CRADA) between ANL and the Australian Nuclear Science and Technology Organisation is developing an 8-inch HIP can

technology. Initial results appear promising. However, for production in the postdemonstration period, a scale-up to an 18-inch can is planned. ANL is seeking an outside contractor to develop the 18-inch can but, as of this writing, has not located a willing and able vendor. The planned scale-up in canister size, combined with remote-handled hot cell operation, may introduce an unresolved safety question. ⁴²

It has been ANL's intent to use a zeolite column rather than the current batch equilibration method for processing the remaining EBR-II fuel in the postdemonstration period. Column operation is under development but is not part of the demonstration. In the batch mode, traces of water in the zeolite prevent attack of the zeolite by chlorides of uranium and plutonium. However, with column operation, water will not be present and therefore this protection will not occur. For column operation to be viable, scaled-up column performance needs to be evaluated.

Two of the Success Criteria Goals for ANL's demonstration project relate to CWF specifically. These criteria and associated work to date are as follows:

Criterion Goal 2-4. Develop ceramic waste specifications that are based on performance characterization results of samples with principal constituent variations: glass fission products, uranium and plutonium. Determine performance characteristics with attribute, characterization, accelerated, and service-condition tests.

Accomplishments as of June 15, 1999. Approximately 900 laboratory-scale and 83 demonstration-scale samples were produced under different conditions. The results from the characterization of these samples were used to establish glass-bonded sodalite as the reference CWF composition. The present status of the work is documented in the report entitled “Preliminary Report on Glass-Bonded Sodalite Properties and Behavior.”

Criterion Goal 2-5. Develop ceramic waste process specifications for major process variables: free chloride, zeolite moisture content, and chloride per unit cell.

Accomplishments as of June 15, 1999. Tests of the CWF have examined the importance of free chloride, zeolite moisture content, and chloride per unit cell. The process specifications for these variables have been defined.^43,44

Analysis of the CWF samples and evaluation of the resulting data may not be complete at the end of the demonstration program. Hence, meeting the above two success criteria goals may be delayed beyond the end of the demonstration period until analysis and data evaluation can be completed.

Advanced Fabrication. Fabrication technology alternatives to HIP have been considered by ANL. These consist of hot uniaxial pressing (HUP), where the amount of glass is increased to about 25 mass %, and pressureless sintering (PLS), where the amount of glass is increased to about 50 mass %. ANL has stated that it is only using HUP to prepare simulated CWF and is not considering it further. However, pressureless sintering is under development as an alternative to HIP for the postdemonstration period and appears to be an attractive alternative to HIP. PLS would avoid concerns about scale-up in canister size and unresolved safety issues associated with remote HIP in a hot cell.

Conclusion: The committee believes that ANL is taking appropriate steps to coordinate its waste qualification program with the DOE-RW repository program. It remains undemonstrated, however, that direct adaptation of test procedures and models developed for measuring the rate of general corrosion of the matrix of homogeneous, vitrified HLW forms are sufficient for evaluating the performance of the heterogeneous, crystalline CWF under long-term repository conditions.

It is significant that the tests conducted so far by ANL indicate that the rate of corrosion of the CWF is comparable to that of the reference borosilicate glass. It remains to be seen whether this behavior will continue in long-term testing.

Conclusion: These continuing concerns are not expected to jeopardize the timely completion of the EBR-II demonstration project in 1999, but attention should be devoted to their resolution prior to extending the EMT process past the demonstration.

When criteria for acceptance of waste forms for geologic repository placement are adopted by RW, test procedures for the waste forms produced by the electrometallurgical process may require modification. The committee believes, however, that the test procedures used for the MWF and CWF are appropriate for the completion of ANL's demonstration project.