This report provides a summary of the results obtained for a limited variability study for glasses containing Crystalline Silicotitanate (CST), Monosodium Titanate (MST), and either simulated Purex or HM sludge. Twenty-two glasses containing Purex sludge and three glasses containing HM sludge were fabricated and tested. The fabricated glasses were tested for durability using the 7-day Product Consistency Test (PCT) and characterized by measuring the viscosity at 1,150 C and by determining an approximate, bounding liquidus temperature. The current models used by Defense Waste Processing Facility (DWPF) for predicting durability, viscosity, and liquidus temperature were applied to all 25 glasses. The goal of this work was to identify any major problems from a glass perspective, within the scope of this effort, which could potentially preclude the use of CST at DWPF.
This long-standing symposia series has become the premier, international forum for scientific and engineering issues related to all levels and types of radioactive wastes and their management. Topics include: fuel cladding and spent nuclear fuel; container fabrication and corrosion; performance assessment; repository performance; radionuclide sorption and transport; cement-based materials and waste containment; corrosion of ceramic wasteforms; structure and characterization of ceramics; radiation effects; natural analogs; wasteform characterization and processing; and corrosion and characterization of glass wasteforms.
The viscosity of glass melts containing four simulated sludge types and two frit candidates (Frits 18 and 21) was measured over the temperature range 750 to 1200°C. The viscosity of melts made with either frit was reduced by the addition of high iron sludge, unchanged by average sludge, and increased by composite and high aluminum sludge. High aluminium sludge greatly increased the viscosity. Frit 21 (containing 4 wt % Li2O substituted for 4 wt % Na2O in Frit 18) was clearly better than Frit 18 in terms of its low viscosity. However, further reductions in viscosity are desirable, especially for glasses containing high aluminum sludge. Changing any frit component by 1 wt % did not significantly affect the viscosity of the glasses. Therefore, variability of 1 wt % in any frit component can be tolerated.
Glass produced during the Purex 4 campaigns of the Integrated DWPF Melter System (IDMS) and the 774 Research Melter contained a lower fraction of sludge components than targeted by the Product Composition Control System (PCCS). Purex 4 glass was more durable than the benchmark (EA) glass, but was less durable than most other simulated SRS high-level waste glasses. Further, the measured durability of Purex 4 glass was not as well correlated with the durability predicted from the DWPF process control algorithm, probably because the algorithm was developed to predict the durability of SRS high-level waste glasses with higher sludge content than Purex 4. A melter run, designated Purex 4 Remediation, was performed using the 774 Research Melter to determine if the initial PCCS target composition determined for Purex 4 would produce acceptable glass whose durability could be accurately modeled by the DWPF glass durability algorithm. Reagent grade oxides and carbonates were added to Purex 4 melter feed stock to simulate a higher sludge loading. Each canister of glass produced was sampled and the glass durability was determined by the Product Consistency Test method. This document details the durability data and subsequent analysis.
The objective of this study was to experimentally measure the properties and performance of a series of glasses with compositions that could represent high level waste Sludge Batch 5 (SB5) as vitrified at the Savannah River Site Defense Waste Processing Facility. These data were used to guide frit optimization efforts as the SB5 composition was finalized. Glass compositions for this study were developed by combining a series of SB5 composition projections with a group of candidate frits. The study glasses were fabricated using depleted uranium and their chemical compositions, crystalline contents and chemical durabilities were characterized. Trevorite was the only crystalline phase that was identified in a few of the study glasses after slow cooling, and is not of concern as spinels have been shown to have little impact on the durability of high level waste glasses. Chemical durability was quantified using the Product Consistency Test (PCT). All of the glasses had very acceptable durability performance. The results of this study indicate that a frit composition can be identified that will provide a processable and durable glass when combined with SB5.
This report is the fourth in a series of studies of the impacts of the addition of Crystalline Silicotitanate (CST) and Monosodium Titanate (MST) from the Small Column Ion Exchange (SCIX) process on the Defense Waste Processing Facility (DWPF) glass waste form and the applicability of the DWPF process control models. MST from the Salt Waste Processing Facility (SWPF) is also considered in the study. The KT08-series of glasses was designed to evaluate any impacts of the inclusion of uranium and thorium in glasses containing the SCIX components. The KT09-series of glasses was designed to study the effect of increasing Al2O3 and K2O concentrations on the propensity for crystallization of titanium containing phases in high TiO2 concentration glasses. Earlier work on the KT05-series glasses recommended that the impact of these two components be studied further. Increased Al2O3 concentrations have been shown to improve the properties and performance of high waste loading glasses, and K2O has been reported to improve the retention of TiO2 in silicate glasses. The KT10-series of compositions was designed to evaluate any impacts of the SCIX components at concentrations 50% higher than currently projected.a The glasses were fabricated in the laboratory and characterized to identify crystallization, to verify chemical compositions, to measure viscosity, and to measure durability. Liquidus temperature measurements for the KT10-series glasses are underway and will be reported separately. All but one of the KT08-series glasses were found to be amorphous by X-ray diffraction (XRD). One of the slowly cooled glasses contained a small amount of trevorite, which had no practical impact on the durability of the glass and is typically found in DWPF-type glasses. The measured Product Consistency Test (PCT) responses for the KT08-series glasses are well predicted by the DWPF models. The viscosities of the KT08-series glasses were generally well predicted by the DWPF model. No unexpected issues were encountered when uranium and thorium were added to the glasses with SCIX components. Increased Al2O3 concentrations were not successful in preventing the formation of iron titanate crystals in the KT09-series glasses. Increased K2O concentrations were successful in hindering the formation of iron titanates in some of the glasses after the canister centerline cooled (CCC) heat treatment. However, this result did not apply to all of the CCC versions of the glasses, indicating a compositional dependence of this effect. In addition, high concentrations of K2O have been shown to hinder the ability of the DWPF durability and viscosity models to predict the performance of these glasses. The usefulness of increased K2O concentrations in preventing the formation of iron titanates may therefore be limited. Further characterization was not performed for the KT09-series glasses since the type of crystallization formed was the characteristic of interest for these compositions. All of the KT10-series glasses were XRD amorphous, regardless of heat treatment. Chemical composition measurements showed that the glasses met the targeted concentrations for each oxide. In general, the measured PCT responses of the KT10-series glasses were well predicted by the DWPF models. The measured, normalized release values for silicon for some of the glasses fell above the 95% confidence interval for the predicted values; however, the PCT responses for these glasses remain considerably lower than that of the benchmark Environmental Assessment (EA) glass. The viscosities of the KT10-series glasses were generally well predicted by the DWPF model. The next step in this study will be to compile all of the data developed and further compare the measured properties and performance with those predicted by the current DWPF Product Composition Control System (PCCS) models. Recommendations will then be made as to which models, if any, may need to be modified in order to accommodate the material from SCIX into DWPF glass production.
The Defense Waste Processing Facility (DWPF) is about to process High Level Waste (HLW) Sludge Batch 4 (SB4). This sludge batch is high in alumina and nepheline can crystallize readily depending on the glass composition. Large concentrations of crystallized nepheline can have an adverse effect on HLW glass durability. Several studies have been performed to study the potential for nepheline formation in SB4. The Phase 3 Nepheline Formation study of SB4 glasses examined sixteen different glasses made with four different frits. Melt rate experiments were performed by the Process Science and Engineering Section (PS & E) of the Savannah River National Laboratory (SRNL) using the four frits from the Phase 3 work, plus additional high B2O3/high Fe2O3 frits. Preliminary results from these tests showed the potential for significant improvements in melt rate for SB4 glasses using a higher B2O3-containing frit, particularly Frit 503. The main objective of this study was to investigate the durability of SB4 glasses produced with a high B2O3 frit likely to be recommended for SB4 processing. In addition, a range of waste loadings (WLs) was selected to continue to assess the effectiveness of a nepheline discriminator in predicting concentrations of nepheline crystallization that would be sufficient to influence the durability response of the glass. Five glasses were selected for this study, covering a WL range of 30 to 50 wt% in 5 wt% increments. The Frit 503 glasses were batched and melted. Specimens of each glass were heat-treated to simulate cooling along the centerline of a DWPF-type canister (ccc) to gauge the effects of thermal history on product performance. Visual observations on both quenched and ccc glasses were documented. A representative sample from each glass was submitted to the SRNL Process Science Analytical Laboratory (PSAL) for chemical analysis to confirm that the as-fabricated glasses corresponded to the defined target compositions. The Product Consistency Test (PCT, ASTM C1285) was performed in triplicate on each Frit 503 quenched and ccc glass to assess chemical durability. The experimental test matrix also included the Environmental Assessment (EA) glass and the Approved Reference Material (ARM-1) glass. Representative samples of all the ccc glasses were examined for homogeneity visually and by X-ray diffraction (XRD) analysis. Chemical composition measurements indicated that the experimental glasses were close to their target compositions. PCT results showed that all of the Fit 503 quenched glasses had an acceptable durability compared to the EA benchmark glass. The durability of one of the ccc glasses, NEPHB-04, was statistically greater than its quenched counterpart. However, this was shown to be of little practical significance, as the durability of the NEPHB-04 ccc glass was acceptable when compared to the durability of the EA benchmark glass. Visual observations and PCT results indicated that all of the Frit 503 quenched glasses were free of any crystallization that impacts durability. For the ccc glasses, XRD results indicated that the lower WL glasses (30 to 40 wt%) were amorphous, which was consistent with visual observations and PCT responses. The higher WL glasses (45 and 50 wt%) were shown by XRD to contain spinel (trevorite, NiFe2O4). It is possible that some of the other high WL glasses also contained some nepheline, but that the amount of nepheline crystallization was below the detection limit (0.5 vol%) associated with XRD. The results indicate that Frit 503 is a good candidate for SB4 processing, based on chemical durability of homogeneous and devitrified glasses over a WL range of 30 - 50%. It should be noted that the higher WL glasses would not be fit for processing in DWPF as they exceed other process related criteria (such as liquidus temperature). However, this is only one of many factors influencing the frit selection. Melt rate and the final SB4 composition are also important factors in frit selection. Additional melt rate studies are currently underway, and the final composition projection for SB4 is expected shortly.
