Choose an item. EWG-RPT-023 Rev 0.0 PNNL-29023 X-Ray Diffraction and Product Consistency Test Results for the Phase 6 Study of Nepheline Formation in High-Level Waste Glasses September 2019 CE Lonergan JO Kroll CH Skidmore ZJ Nelson JD Vienna Prepared for the U.S. Department of Energy under Contract DE-AC05-76RL01830
80
Embed
X-Ray Diffraction and Product Consistency Test Results for the Phase 6 … · 2020. 11. 25. · EWG-RPT-023 Rev 0.0 PNNL-29023 X-Ray Diffraction and Product Consistency Test Results
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Choose an item.
EWG-RPT-023 Rev 0.0 PNNL-29023
X-Ray Diffraction and Product Consistency Test Results for the Phase 6 Study of Nepheline Formation in High-Level Waste Glasses September 2019
CE Lonergan JO Kroll CH Skidmore ZJ Nelson JD Vienna
Prepared for the U.S. Department of Energy under Contract DE-AC05-76RL01830
Choose an item.
DISCLAIMER
This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor Battelle Memorial Institute, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof, or Battelle Memorial Institute. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.
PACIFIC NORTHWEST NATIONAL LABORATORY operated by BATTELLE
for the UNITED STATES DEPARTMENT OF ENERGY
under Contract DE-AC05-76RL01830
Printed in the United States of America
Available to DOE and DOE contractors from the Office of Scientific and Technical Information,
X-Ray Diffraction and Product Consistency Test Results for the Phase 6 Study of Nepheline Formation in Hanford High-Level Waste Glasses
September 2019 CE Lonergan JO Kroll CH Skidmore ZJ Nelson JD Vienna Prepared for the U.S. Department of Energy under Contract DE-AC05-76RL01830 Pacific Northwest National Laboratory Richland, Washington 99354
EWG-RPT-023 Rev 0.0 PNNL-29023
Summary ii
Summary
Hanford high-level waste (HLW) contains relatively high concentrations of Al2O3. A major constraint limiting the waste loading of high-Al2O3 glasses is nepheline (nominally NaAlSiO4) formation upon slow cooling of HLW glasses after melts are poured into steel canisters. The model currently planned to be used at the Hanford Waste Treatment and Immobilization Plant (WTP) for avoiding nepheline formation is unnecessarily conservative and drastically limits waste loading.1
To increase operational flexibility and effectively operate the WTP, the effects of glass composition on glass properties must be determined and glass property-composition models must be developed.2 Determining the impacts of glass composition on nepheline formation and the effects of nepheline on Product Consistency Test (PCT) response is an important part of this effort. During this work data related to the Hanford high-Al2O3 HLW composition region was generated to supplement existing data for model development. Twenty glasses were fabricated, heat treated, and analyzed for crystallinity and PCT response. The heat treatment was designed to mimic the canister centerline cooling (CCC) profile of Hanford HLW glass canisters.3
It was found that only 6 of the 20 prepared glasses (#4, #7, #8, #10, #19, and #20) formed nepheline upon heat treatment. All glasses except for one (#19) precipitated spinel upon quenching and no glasses contained nepheline upon quenching. Of the 20 glasses, the PCT-normalized boron release of the quenched glasses and all but five of the CCC glasses satisfied the environmental assessment glass benchmark value (16.695 g/L).4 The CCC glass samples that exceeded the limit were #8 (48.21 g/L), #10 (19.77 g/L), #13 (20.17 g/L), #19 (60.85 g/L), and #20 (61.08 g/L). Those glasses coincide with the highest amounts of nepheline found in the CCC glasses, except for #13, which did not form nepheline after CCC.
1 Vienna JD, D-S Kim, DC Skorski, and J Matyas. 2013. Glass Property Models and Constraints for Estimating the Glass to Be Produced at Hanford by Implementing Current Advanced Glass Formulation Efforts. PNNL-22631 (ORP-58289), Rev. 1, Pacific Northwest National Laboratory, Richland, Washington. Available at http://www.pnnl.gov/main/publications/external/technical_reports/PNNL-22631Rev1.pdf. 2 Peeler DK, JD Vienna, MJ Schweiger, and KM Fox. 2015. Advanced High-Level Waste Glass Research and Development Plan. PNNL-24450, Pacific Northwest National Laboratory Richland, Washington. Available at http://www.pnnl.gov/main/publications/external/technical_reports/PNNL-24450.pdf. 3 Petkus L. October 19, 2003. “Canister Centerline Cooling Data, 24590-PADC-F00029 Rev 1.” Memorandum to C Musick. River Protection Project, Waste Treatment Plant, Richland, Washington. 4 Jantzen CM, NE Bibler, DC Beam, CL Crawford, and MA Pickett. 1993. Characterization of the Defense Waste Processing Facility (DWPF) Environmental Assessment (EA) Glass Standard Reference Material. WSRC-TR-92-346, Rev. 1, Westinghouse Savannah River Company, Aiken, South Carolina.
EWG-RPT-023 Rev 0.0 PNNL-29023
Acknowledgments iii
Acknowledgments
The authors gratefully acknowledge the financial support provided by the U.S. Department of Energy Office of River Protection Waste Treatment and Immobilization Plant Project, managed by Tom Fletcher, with technical oversight by Albert Kruger.
The authors thank Kevin Fox of Savannah River National Laboratory for his help in the analysis and testing of the glasses. We also thank Dong-Sang Kim (PNNL) for his technical review, Matt Wilburn of (PNNL) for the editorial review, and Hans Brandal and Veronica Perez (PNNL) for programmatic support during the conduct of this work.
JMP 13 JMP® version 13.0.0 (SAS Institute Inc. 2016)
KH potassium hydroxide fusion
LM lithium metaborate fusion
ND nepheline discriminator
NN neural network
NR normalized release
OB optical basicity
ORP Office of River Protection
PCT Product Consistency Test
PF sodium peroxide fusion
PNNL Pacific Northwest National Laboratory
QA quality assurance
R&D research and development
SEM scanning electron microscopy
SRNL Savannah River National Laboratory
TanH hyperbolic tangent
XRD X-ray diffraction
WTP Waste Treatment and Immobilization Plant
EWG-RPT-023 Rev 0.0 PNNL-29023
Contents v
Contents
Summary ....................................................................................................................................................... ii
Acknowledgments ........................................................................................................................................ iii
Acronyms and Abbreviations ...................................................................................................................... iv
Contents ........................................................................................................................................................ v
Appendix A – Measured and Targeted Glass Compositions .................................................................... A.1
Appendix B – X-Ray Diffraction Data ......................................................................................................B.1
EWG-RPT-023 Rev 0.0 PNNL-29023
Contents vi
Figures
Figure 1. Predicted probability of nepheline formation versus nepheline concentration, as determined by XRD. The highest concentrations of nepheline are found in the following CCC samples. X = NP6-20, □ – NP6-19, and ▼ = NP6-08. (Glasses in red precipitated nepheline upon CCC heat treatment while glasses represented by black circles did not.) ........................................................................................................ 14
Figure 2. SEM images, in secondary mode, of various quenched glasses. ................................................. 15
Figure 3. SEM images of several glasses after CCC heat treatments. ........................................................ 16
Figure 4. Boron ln(NR) rates for quenched glasses versus CCC glasses. Glasses in red precipitated nepheline upon CCC heat treatment while the black circles did not. X = NP6-20, □ – NP6-19, and ▼ = NP6-08. .................................................................... 18
Figure 5. A plot of the difference in crystal concentrations (CCC minus quenched values) vs. the difference in normalized B release values for CCC glasses minus quenched glasses. Glasses in red precipitated nepheline upon CCC heat treatment and glasses presented in black did not. X = NP6-20, □ – NP6-19, and ▼ = NP6-08. ............ 19
Tables
Table 1. Composition of “Others” mixture (in mass fraction). ..................................................................... 6
Table 2. Upper and lower bounds for components (in mass fraction). ......................................................... 7
Table 3. Compositions of 20 statistically designed glasses. ......................................................................... 8
Table 4. Preparation and measurement methods used in reporting the concentrations of the analytes in this study (Fox et al. 2019). .............................................................................. 9
Table 6. Crystal identification and quantification in as-quenched glasses. ................................................ 12
Table 7. Crystal identification and quantification results using XRD for the glasses in this study post-CCC treatment. ......................................................................................................... 13
Table 8. PCT method A release rates for B, Li, Na, and Si normalized to target compositions. Values below are averaged from the triplicate measurements (< = below detection limit). ................................................................................................................. 17
EWG-RPT-023 Rev 0.0 PNNL-29023
Introduction 1
1.0 Introduction
The U.S. Department of Energy (DOE) manages approximately 210,000 m3 of radioactive waste in
underground tanks on the Hanford Site. The Hanford Waste Treatment and Immobilization Plant (WTP) is currently being constructed to vitrify the waste in borosilicate glasses. Prior to processing the low-activity waste (LAW) fraction will be removed from the high-level waste (HLW) and processed separately. Each fraction will be blended with glass-forming chemicals, melted at ~1150°C, and then the molten glass will be poured into stainless steel canisters to cool and solidify (DOE 2000).
According to the 2008 HLW feed vector used by Vienna et al. (2013), many HLW compositions contain high concentrations of Al2O3. The Al2O3 fraction is projected to range from roughly 10 to 70 mass% on a calcined oxide basis after caustic leaching. The loading of high-Al2O3 HLW may be drastically reduced by the constraint(s) applied to reduce the probability of forming nepheline (nominally NaAlSiO4) upon slow cooling of glasses. Nepheline precipitation from an HLW glass during cooling is a major concern because it can reduce the durability of the resulting glass by removing three moles of glass forming oxides (one mole of Al2O3 and two moles of SiO2) for every mole of Na2O (Kim et al. 1995). When nepheline is present in a waste glass, it is difficult to predict the Product Consistency Test (PCT) response. It has also been documented that eucryptite (LiAlSiO4) has a similar effect on glass durability (McCloy and Vienna 2010). To meet disposal requirements, nepheline formation must be avoided, or the amount of nepheline formed and its impact on the PCT response must be predicted. The ability to predict the amount of nepheline formed and the PCT response would provide a basis for specifying a constraint to avoid HLW glass compositions that would yield unacceptable PCT responses.
The current constraint used to avoid nepheline formation in glasses produced by WTP is unnecessarily conservative and limits waste loading (Vienna et al. 2016). A nepheline discriminator (ND) was developed to reduce the risk of nepheline precipitation during canister centerline cooling (CCC) heat treatment (Li et al. 1997). This approach is based on limiting the normalized SiO2 concentration (NSi) using the following:
0.62 (1)
where gi is the mass fraction of the ith component in the glass. A better discriminator is needed as ND screens out glass compositions that often do not precipitate nepheline, while allowing glass compositions that have been shown to crystallize nepheline. In an effort to reduce some of the conservatism in the ND, optical basicity (OB) was proposed as a constraint (Rodriguez et al. 2011; McCloy et al. 2011). The OB of a glass (Λglass) can be calculated from a glass composition using the following equation:
Ʌ∑ Ʌ∑
(2)
where qi = the number of oxygen atoms in the ith component oxide xi = the mole fraction of the ith component oxide in glass Λi = the molar basicity of the ith component oxide
The revised constraint allows glasses with NSi < 0.62 if the OB of the melt is less than 0.55. This approach did reduce some of the conservatism, but still limits the waste loading of high-Al2O3 glasses (Vienna et al. 2016).
EWG-RPT-023 Rev 0.0 PNNL-29023
Introduction 2
A neural network (NN) model has also been proposed to estimate the probability of nepheline formation for a specific glass composition (Vienna et al. 2013). This approach was selected because it can account for highly non-linear effects of components. The NN model consists of three nodes using the hyperbolic tangent (TanH) transfer function. The output from the three nodes was compared to a probability cutoff value to assign a binary response (i.e., whether nepheline formed or not) for a glass composition. However, the NN method involves complex calculations that require much more data than is available for typical glass properties, and it is difficult to determine the uncertainties of predictions made with the model (Vienna et al. 2016).
To optimize waste loading, a new approach is needed to limit nepheline precipitation during slow cooling of HLW canisters. It has been demonstrated that B2O3, CaO, Fe2O3, K2O, and Li2O affect nepheline formation (Li et al. 1997). A ternary sub-mixture model that takes advantage of the success of the ND while considering the effects of other components in the melt has been developed (Vienna et al. 2017).
The current nepheline formation models are based on a dataset composed of 747 glasses (Vienna et al. 2017; Stanfill et al. 2019). This dataset was gathered from multiple studies under different methods at different labs. The glasses that are currently part of this dataset do not adequately cover high-Al2O3 regions of interest as defined by the predicted waste compositions from the Hanford tanks. The objective of this subtask is to develop data that will enable the current nepheline model to be refined and validated and eventually to develop an improved nepheline model with good predictability in the high-alumina glass composition region for Hanford HLW glasses. This is the first step in creating a model that can support plant operation and waste form qualification activities. This objective will be achieved by varying glass compositions in the region of interest and measuring both the concentration of crystalline phases formed and the PCT responses of samples heat treated according to the simulated Hanford HLW CCC temperature schedule.
The first phase of this study deployed a strategy of varying seven components of interest (Al2O3, B2O3, CaO, Fe2O3, Li2O, Na2O, SiO2) one-at-a-time around a baseline glass to formulate a total of 15 glass compositions. Only one glass from this study formed nepheline crystals during CCC heat treatment. This demonstrated that the HLW glass composition region for which nepheline-free glasses can be formulated was larger than expected but yielded little data useful for modeling. The Phase 2 nepheline study deployed the same one-component-at-a-time approach, but varied Al2O3, B2O3, Li2O, Na2O, and SiO2 around a revised baseline glass (BL3) that was adjusted so that roughly 10 to 15 mass% nepheline would form during CCC heat treatment. A total of 12 of the 14 unique glasses developed during Phase 2 precipitated nepheline upon CCC, and one of the two remaining glasses contained eucryptite.
The data from the first two phases showed that the composition region for which nepheline precipitated was not easily described by first order effects. In Phase 3, a statistical design surrounding the composition region generated two-at-a-time (2X) and three-at-a-time (3X) changes in Al2O3, B2O3, Li2O, Na2O, and SiO2 using the revised baseline glass (BL3) and lower and upper bounds of each component from the Phase 2 study. The 2X and 3X changes in all possible combinations of pairs or triples of components were made by a statistical program, JMP® version 13.0.0 [SAS Institute Inc. 2016] (JMP 13), and then the “net change” was distributed proportionally among the remaining components that were not involved in the 2X or 3X changes. JMP 13 chose six clusters of glasses from all the possible 2X and 3X combinations computed, from which one cluster of 16 glasses was selected to be tested. Results from Phases 1 through 3 are reported by Kroll et al. (2016).
Phase 4 of the study consisted of a matrix of 45 glasses that was statistically designed to cover the Hanford high-alumina composition region as efficiently as possible using an extreme vertices mixture design by varying Al2O3, B2O3, Bi2O3, CaO, Cr2O3, Fe2O3, Li2O, MgO, MnO, Na2O, P2O5, SiO2, and ZrO2 (Vienna et al. 2014). A total of 14 of the 45 glasses precipitated nepheline after CCC treatment.
EWG-RPT-023 Rev 0.0 PNNL-29023
Introduction 3
The Phase 5 study aimed at further expanding the dataset of glasses susceptible to nepheline formation, consisting of a set of 26 glasses developed by varying Al2O3, B2O3, CaO, Fe2O3, Li2O, Na2O, P2O5, SiO2, and an “Others” mixture all at the same time. JMP 13 was used to generate the compositions using a mixture design and space filling technique. Single- and multi-component constraints were used to set bounds on component concentrations and limit melt viscosity. A replicate of BL3 was also tested in this study. A total of 11 of the 26 glasses were found to precipitate nepheline upon CCC treatment. A more detailed description of the results from the Phase 5 study is given by Kroll et al. (2018).
The current study (Phase 6) also used JMP 13 statistical software to generate a matrix of 20 glasses using a space filling technique. The aim of this study was to refine the logistic regression sub-mixture model (see Stanfill et al. 2019) by filling in gaps of predicted nepheline formation probabilities. This report summarizes the results of the Phase 6 study, which primarily involves PCT and X-ray diffraction (XRD) analysis for quenched and CCC heat-treated glasses.
1.1 Quality Assurance
This work emphasized demonstrating proof of principle with the intent of solving a specific problem or meeting a practical need. The information associated with this report may be used to support design input.
1.1.1 PNNL QA Program
The Pacific Northwest National Laboratory (PNNL) Quality Assurance (QA) Program is based upon the requirements as defined in the DOE Order 414.1D, Quality Assurance, and 10 CFR 830, Energy/Nuclear Safety Management, Subpart A, “Quality Assurance Requirements” (a.k.a., the Quality Rule). PNNL has chosen to implement the following consensus standards in a graded approach:
ASME NQA-1-2000, Quality Assurance Requirements for Nuclear Facility Applications, Part I, “Requirements for Quality Assurance Programs for Nuclear Facilities”
ASME NQA-1-2000, Part II, Subpart 2.7, “Quality Assurance Requirements for Computer Software for Nuclear Facility Applications,” including problem reporting and corrective action
ASME NQA-1-2000, Part IV, Subpart 4.2, “Guidance on Graded Application of Quality Assurance (QA) for Nuclear-Related Research and Development.”
The PNNL Quality Assurance Program Description/Quality Management M&O Program Description describes the Laboratory-level QA program that applies to all work performed by PNNL. Laboratory-level procedures for implementing the QA requirements described in the standards identified above are deployed through PNNL’s web-based “How Do I…?” (HDI) system, a standards-based system for managing and deploying requirements and procedures to PNNL staff. The HDI procedures (called Workflows and Work Controls) provide detailed guidance for performing tasks, such as protecting classified information and procuring items and services, as well as general guidelines for performing research-related tasks, such as preparing and reviewing calculations and calibrating and controlling measuring and test equipment.
1.1.2 ORP Glass Support Work QA Program
The DOE Office of River Protection (ORP) Glass Support Work project is performed under the PNNL Nuclear Quality Assurance Program (NQAP); this program establishes an ASME NQA-1-2012, 10 CFR 830, Subpart A DOE Order (O) 414.1D compliant QA program for use by research and development (R&D) projects and programs that are compatible with the editions of ASME NQA-1 2000 through 2012.
EWG-RPT-023 Rev 0.0 PNNL-29023
Introduction 4
The work of this report was performed to the QA level of applied research:
Applied research work activities (or deliverables) apply to nuclear and non-nuclear R&D (work activities or deliverables) that are processes initiated-with-the-intent of solving a specific problem or meeting a practical need. For applied research activities, grading is minimal and largely contingent upon the complexity of the research and the ability to duplicate the research if data were lost. The elements of QA grading, including the level of documentation, were applied to the program-, project-, and task-levels.
EWG-RPT-023 Rev 0.0 PNNL-29023
Experimental Methods 5
2.0 Experimental Methods
The glasses formulated by statistical design were prepared by standard melt/quench protocols as described in Section 2.2. After preparation, the glasses were subjected to heat treatments and PCT, described in Sections 2.4 and 2.7, respectively, to understand crystal formation and its impacts on leaching response. Additionally, compositions were analyzed to ensure no errors in preparation occurred (procedures described in Section 2.3).
2.1 Composition Matrix Design The aim of this study was to refine the logistic regression sub-mixture model (see Stanfill et al. 2019) by filling in gaps of predicted nepheline formation probabilities. A matrix of 20 glasses was developed using a space-filling design technique with JMP™ version 13 statistical software. An in-depth description of the design effort was published in Piepel et al. (2018). Eight components, Al2O3, B2O3, CaO, Fe2O3, Li2O, Na2O, SiO2, and an “Others” mixture, were varied all-at-a-time. The composition of the Others mixture is listed in mass fraction in Table 1. Multi-component constraints were used to limit melt viscosity to the range of 4 to 10 Pa·s and the probability of nepheline formation as predicted by the logistic regression sub-mixture model to the range of 0.20 to 0.65. The lower and upper limits of the single-component constraints in mass fraction are shown in
EWG-RPT-023 Rev 0.0 PNNL-29023
Experimental Methods 6
Table 2. The 20 glass compositions tested in this study are presented in Table 3.
Table 1. Composition of “Others” mixture (in mass fraction).
Component Mass Fraction
Bi2O3 0.1498
Cr2O3 0.1835
F 0.0586
MgO 0.0234
MnO 0.2375
NiO 0.0215
P2O5 0.1727
PbO 0.0108
RuO2 0.0067
SO3 0.0540
SrO 0.0107
ZrO2 0.0705
EWG-RPT-023 Rev 0.0 PNNL-29023
Experimental Methods 7
Table 2. Upper and lower bounds for components (in mass fraction).
Component Lower Limit Upper Limit
Al2O3 0.2150 0.3200
B2O3 0.1400 0.2400
CaO 0.0000 0.0500
Fe2O3 0.0200 0.0500
Li2O 0.0000 0.0600
Na2O 0.0700 0.1600
SiO2 0.2250 0.3350
Others 0.0300 0.1200
EWG-RPT-023 Rev 0.0 PNNL-29023
Experimental Methods 8
Table 3. Compositions of 20 statistically designed glasses.
Glass ID Al2O3 B2O3 CaO Fe2O3 Li2O Na2O SiO2 Bi2O3 Cr2O3 F MgO MnO NiO P2O5 PbO RuO2 SO3 SrO ZrO2
Glasses were prepared in 200-g batches with oxides (Al Bi2O3, Cr2O3, Fe2O3, MnO, NiO, PbO, SiO2, and ZrO2) and carbonates (CaCO3, Li2CO3, Na2CO3, and SrCO3) for components shown above in Table 3. NaF, NaPO3, and Na2SO4 were used as the sources of F, P2O5, and SO3, respectively. Boric acid was used as the B2O3 source, and Al(OH)3 was chosen as the aluminum additive. RuO2 was added as 1.5% RuNO(NO3)3 solution dripped onto the necessary amount of SiO2 and placed in an oven for at least 2 h at 90 °C. Once dried, the SiO2 + RuO2 mixture was placed in a container with the rest of the chemicals and homogenized in a vibratory mill with an agate milling chamber for 4 min in two to four separate additions. Once mixed, the batch was added into a Pt-10%Rh crucible in two to four separate additions and melted at 1150 °C for 1 h. The melt was quenched by pouring onto a stainless-steel plate. After quenching, the glass was ground in the vibratory mill for 4 min, using a tungsten carbide chamber, and melted again at 1150°C for a second melt. After preparation, samples were either sent for chemical analysis (discussed further in Section 2.3) or prepared for heat treatment according to the HLW CCC profile (discussed further in Section 2.4)
2.3 Glass Composition Analysis
Glasses were analyzed by Savannah River National Laboratory (SRNL) and the results were published in Fox et al. (2019). Quenched samples were sent for inductively coupled plasma-atomic emission spectroscopy (ICP-AES) and ion chromatography (IC). The low-level reference material was included in the analysis as a reference to ensure the proper operation of the ICP-AES and IC. Glasses were dissolved at SRNL according to the techniques listed in Table 4; more information can be found in Fox et al. 2019.
Table 4. Preparation and measurement methods used in reporting the concentrations of the analytes in this study (Fox et al. 2019).
Analyte Preparation
Method Measurement
Method Al PF ICP-AES B PF ICP-AES Bi LM ICP-AES Ca PF ICP-AES Cr PF ICP-AES F KH IC Fe LM ICP-AES Li PH ICP-AES Mg LM ICP-AES Mn PF ICP-AES Na LM ICP-AES Ni PF ICP-AES P PF ICP-AES Pb LM ICP-AES Ru LM ICP-AES S LM ICP-AES Si PF ICP-AES Sr LM ICP-AES Zr PF ICP-AES PF = sodium peroxide fusion KH = potassium hydroxide fusion LM = lithium metaborate fusion
EWG-RPT-023 Rev 0.0 PNNL-29023
Experimental Methods 10
2.4 Canister Centerline Cooling Heat Treatment
To perform the CCC heat treatment, a crushed sample of quenched glass with particles sized roughly 5 mm in diameter was placed in either a 1.2 × 1.2 × 1.2 cm or a 2.54 × 2.54 × 2.54 cm Pt-10% Rh foil crucible with a tight-fitting Pt-10% Rh foil lid. The glass-loaded crucibles were placed in a furnace preheated to the melt temperature (1150°C) and allowed to dwell for 30 min. Then, the furnace temperature was quickly reduced to 1050°C (at an estimated rate of -12.5°C/min) and cooled as prescribed in Table 5.
Table 5. CCC heat treatment schedule.
Segment Start Temp
(°C) Stop Temp
(°C) Rate
(°C/min)
1(a) 1150 1150 0.000(a)
2(b) 1150 1050 Free fall(b)
3 1050 980 -1.556
4 980 930 -0.806
5 930 875 -0.591
6 875 825 -0.388
7 825 775 -0.253
8 775 725 -0.278
9 725 400 -0.304
(a) Segment 1 is a 30-min dwell at the glass melt temperature (1150°C).
(b) Segment 2 free fall is at an estimated rate of -12.5°C/min.
When the furnace temperature was below 400°C (below the glass transition temperature), the furnace power was turned off and the glass cooled naturally to room temperature. This profile is the temperature schedule of CCC treatment for Hanford HLW glasses planned for use at WTP.1
2.5 Scanning Electron Microscopy Analysis
A JEOL 5900 scanning electron microscopy (SEM) instrument was used for the SEM images. Samples after quenching and post-CCC treatment were placed in Allied Technologies quickset epoxy to be mounted for polishing. Once mounted, samples were polished using silicon carbide polishing paper in the following grits: 320, 400, 600, 800, and 1200. After the final polishing pad, diamond suspension water-based polishing slurries from Allied Technologies were used to polish the samples to a 1-micron finish.
Once polished, the samples were coated with a 2.5-nm layer of platinum, and a strip of carbon tape was applied, to mitigate electron charge buildup during SEM analysis. Images were taken in backscatter mode and secondary mode.
1 Petkus LL. 2003. “Canister Centerline Cooling Data, Revision 1,” to C.A. Musick, CCN: 074851, October 29, 2003, River Protection Project, Hanford Tank Waste Treatment and Immobilization Plant, Richland, Washington.
PCT was performed at SRNL on three replicate samples from both as-quenched and CCC glasses according to Method A of ASTM-C-1285-14. The results of which were published in Fox et al. 2019. Bulk samples were ground and sieved to a particle size of 0.149 to 0.074 mm (-100 +200 mesh) and tested in stainless steel vessels at 90± 2 °C for 7 days (+/-2%). Fifteen milliliters of Type-I ASTM water and 1.5 g of sieved glass were placed into the vessels prior to testing. Also included in the experimental test matrix was the Approved Reference Material (ARM-1) glass, the environmental assessment (EA) benchmark glass (Jantzen et al. 1993) and blanks from the vessel cleaning batch.
After 7 days, the vessels were removed from the oven and allowed to cool to room temperature. After cooling, pH was measured on a small aliquot of the solution and the remaining solution was used to perform ICP-AES analyses. The solution for chemical analysis was filtered and acidified by adding 4 mL of 0.4M HNO3 to 6 mL of leachate (2:3 volume dilution). The leachates for the EA glass were further diluted with de-ionized (DI) water at a factor of 1 mL leachate to 10 mL of DI water.
EWG-RPT-023 Rev 0.0 PNNL-29023
Results and Discussion 12
3.0 Results and Discussion
The 20 glass compositions were analyzed as described in Section 2.3, and there was no indication of a mis-batch or significant error. The sums of the measured compositions for all glasses varied between 95.53% and 100.72%. The measurements summed for the total values are shown in Appendix A, Table A-1. Overall the Al2O3, Li2O, and Na2O concentrations were low (~ 6 relative wt% or less) for some of the glasses while Fe2O3 was higher than targeted (by less than 5 relative wt% difference). F and SO3 were low for most glasses which may indicate volatility during melting. Further comparisons can be made using the data in Table A-1.
3.1 Crystal Fraction Analysis Upon Quenching and Post-Canister Centerline Cooling Treatment
After melting, the quenched glasses were characterized using XRD to determine if any crystals developed during melting or upon cooling. All glasses contained crystals upon cooling in concentrations of about 6.5 wt% or less. Spinel was measured in every glass except NP6-20, which precipitated an iron silicate. The results are summarized in Table 6 and the XRD data can be found in Appendix B.
Table 6. Crystal identification and quantification in as-quenched glasses.
All glasses were then subjected to an HLW CCC heat treatment, as described in Section 2.4. After treatment, the glass was characterized by XRD to identify and quantify the crystals present in the glass. Table 7 shows the results for all 20 glasses investigated in this study.
EWG-RPT-023 Rev 0.0 PNNL-29023
Results and Discussion 13
Nepheline was found in 6 of the 20 glasses (NP6-4, 7, 8, 10, 19, and 20) while spinel was found in all but one (NP6-19). Additional phases present include hematite, various silicates, fluorapatite, and others that were in concentrations of less than 0.5 wt%. It is interesting to note that spinel precipitated in NP6-19 upon quenching, but after CCC spinel was not found in the sample, which instead crystallized out nepheline and hematite. Also shown below in Table 7 is the nepheline predictor which was determined for the glasses using a nonlinear, multi-component model described by Piepel et al. (2018).
Table 7. Crystal identification and quantification results using XRD for the glasses in this study post-CCC treatment.
A key issue that this matrix was intended to address is the misclassification rate of glasses that will, or will not, crystallize nepheline. The nepheline predictor shows the estimated probability of nepheline formation for the glasses in the Phase 6 matrix and a plot of the probability versus the nepheline concentration after CCC is shown below.
EWG-RPT-023 Rev 0.0 PNNL-29023
Results and Discussion 14
Figure 1. Predicted probability of nepheline formation versus nepheline concentration, as determined by XRD. The highest concentrations of nepheline are found in the following CCC samples. X = NP6-20, □ – NP6-19, and ▼ = NP6-08. (Glasses in red precipitated nepheline upon CCC heat treatment while glasses represented by black circles did not.)
The prediction of nepheline behavior for this matrix is still not as accurate as needed, as seen by instances of no nepheline formation for glasses with high probabilities of predicted nepheline.
3.2 SEM Analysis
SEM images were taken of several quenched and CCC glasses to understand the morphology of the crystals present in the glass. Five quenched samples (NP6-05, 06, 08, 13, and 16) with a varying range of spinel concentrations, from 3.01 wt% to 7.99 wt%, were imaged with SEM; example images are provided in Figure 2. As noted in the XRD results, all the glasses mentioned above only precipitated spinel crystals upon cooling after melting.
EWG-RPT-023 Rev 0.0 PNNL-29023
Results and Discussion 15
Figure 2. SEM images, in secondary mode, of various quenched glasses.
All five glasses show similar spinel morphologies. Glass NP6-06 also contained a larger (10 µm × 20 µm), oval-shaped cluster of spinel crystals.
Six of the glasses subjected to the CCC treatment were imaged, and the results are shown in Figure 2. The images show spinel (small white features) formation along with nepheline (large gray features). The XRD of NP6-19 indicates no spinel was detected but hematite was present. NP6-8 and NP6-20 have darker
EWG-RPT-023 Rev 0.0 PNNL-29023
Results and Discussion 16
gray regions in the larger nepheline crystals, and it is unclear at this time what those features may be attributed to.
Figure 3. SEM images of several glasses after CCC heat treatments.
NP6-04-CCC
NP6-07-CCC
NP6-08-CCC
NP6-10-CCC
NP6-19-CCC
NP6-20-CCC
EWG-RPT-023 Rev 0.0 PNNL-29023
Results and Discussion 17
3.3 Product Consistency Testing Analysis
PCT was measured on quenched and heat-treated samples by SRNL as described in Section 2.7. Full experimental details and the testing results can be found in Fox et al. 2019.
The normalized release (NR) results, normalized to the target compositions, are shown in Table 8 for Li, B, Na, and Si for quenched and CCC samples of each glass.
Table 8. PCT method A release rates for B, Li, Na, and Si normalized to target compositions. Values below are averaged from the triplicate measurements (< = below detection limit).
Glass ID Status NRB (g/L) NRLi (g/L) NRNa (g/L) NRSi (g/L)
The highest boron releases for CCC samples were found for the following compositions, starting with the largest value: NP6-20 (61.1 g/L), NP6-19 (60.8 g/L), and NP6-08 (48.2 g/L), which coincide with the glasses that contained the largest amount of nepheline after the CCC heat treatment. The boron normalized release rates for quenched glasses are compared to the heat-treated glasses in Figure 4.
Figure 4. Boron ln(NR) rates for quenched glasses versus CCC glasses. Glasses in red precipitated nepheline upon CCC heat treatment while the black circles did not. X = NP6-20, □ – NP6-19, and ▼ = NP6-08.
There appears to be no clear correlation between normalized release for quenched vs. CCC glasses. In general, the nepheline-containing glasses (in red) have significantly higher CCC NR values compared to the quenched glasses (Figure 5).
EWG-RPT-023 Rev 0.0 PNNL-29023
Results and Discussion 19
Figure 5. A plot of the difference in crystal concentrations (CCC minus quenched values) vs. the difference in normalized B release values for CCC glasses minus quenched glasses. Glasses in red precipitated nepheline upon CCC heat treatment and glasses presented in black did not. X = NP6-20, □ – NP6-19, and ▼ = NP6-08.
There is a clear indication that higher levels of crystals tend to have a higher boron release in the CCC PCT analysis, particularly as compared to the quenched NRB values. As seen in Figure 5, the glasses with the highest releases are NP6-08, NP6-19, and NP6-20, which also have the highest nepheline/crystal concentrations.
It is interesting to note that only five of the CCC glasses exceeded the EA benchmark value (16.695 g/L) for normalized boron release. All the glasses with elevated boron release, NP6-(8, 10, 13, 19, and 20)-CCC, contained nepheline upon CCC treatment except for NP6-13, which only precipitated spinel in the as-quenched and post-CCC sample. This is noteworthy as there are other glasses that precipitated nepheline during CCC but did not exceed 16.695 g/L for boron response, and NP6-13 contained no nepheline and exhibited a high normalized release value (NRB = 20.17 g/L).
The glasses with higher releases had relatively high concentrations of Na2O. Additionally, the glasses had less than 6 wt% spinel after CCC as well as less than 3 wt% spinel upon quenching, but other glasses with lower release rates exhibited similar behavior. Further work is recommended to understand this response.
EWG-RPT-023 Rev 0.0 PNNL-29023
Summary 20
4.0 Summary
Glasses in a matrix developed via a space-filling design were subjected to CCC heat treatments and analyzed using XRD and SEM and tested for the PCT response. The nepheline predictor established that use of previous data did not predict the formation of nepheline well, as this was a compositional region in a highly misclassified compositional space.
XRD results showed that nepheline formed in only 6 of the 20 glasses after heat treatment, while all but one of the glasses contained spinel upon quenching and after heat treatment. SEM confirmed the presence of spinel upon quenching and the presence of spinel inside of the nepheline crystals after CCC.
Interestingly, only 5 of the 20 glasses exceeded the EA PCT response for boron release. While all six glasses that precipitated nepheline upon CCC had relatively higher release rates compared to their quenched counterparts, one of the five glasses exceeding the EA response did not contain nepheline. This result should be investigated further as it would be expected that the presence of nepheline would result in poor performance for PCT.
EWG-RPT-023 Rev 0.0 PNNL-29023
References 21
5.0 References
ASTM. 2014. Standard Test Methods for Determining Chemical Durability of Nuclear Waste Glasses: The Product Consistency Test (PCT). ASTM C-1285, ASTM International. West Conshohocken, Pennsylvania.
DOE 2000. Design, Construction, and Commissioning of the Hanford Tank Waste Treatment and Immobilization Plant. U.S. Department of Energy, Office of River Protection, Richland, Washington.
Fox KM, TB Edwards, MC Hsieh, and WT Riley. 2019. Chemical Composition Analysis and Product Consistency Tests of the ORP Phase 6 Nepheline Study Glasses. SRNL-STI-2018-00645, Rev. 0, Savannah River National Laboratory, Aiken, South Carolina.
Jantzen CM, NE Bibler, DC Beam, CL Crawford, and MA Pickett. 1993. Characterization of the Defense Waste Processing Facility (DWPF) Environmental Assessment (EA) Glass Standard Reference Material. WSRC-TR-92-346, Rev. 1, Westinghouse Savannah River Company, Aiken, South Carolina.
JMP™ Pro, Ver. 13.0, Computer Software, SAS Institute Inc., Cary, NC (2016).
Kim D-S, D Peeler, and P Hrma. 1995. “Effect of Crystallization on the Chemical Durability of Simulated Nuclear Waste Glasses.” Ceramic Transactions 61:177-185.
Kroll JO, JD Vienna, MJ Schweiger, GF Piepel, and SK Cooley. 2016. Results from Phase 1, 2, and 3 Studies on Nepheline Formation in High-Level Waste Glasses Containing High Concentrations of Alumina. PNNL-26057, Rev. 0.0; EWG-RPT-011, Rev. 0.0, Pacific Northwest National Laboratory, Richland, Washington.
Kroll JO, JD Vienna, ZJ Nelson, and CH Skidmore. 2018. Results from Phase 5 Study on Nepheline Formation in High-Level Waste Glasses Containing High Concentrations of Alumina. PNNL-27555, Rev. 0.0; EWG-RPT-017, Rev0.0, Pacific Northwest National Laboratory, Richland, Washington.
Li H, JD Vienna, P Hrma, DE Smith, and MJ Schweiger. 1997. “Nepheline Precipitation in High-Level Waste Glasses: Compositional Effects and Impact on the Waste Form Acceptability.” Scientific Basis for Nuclear Waste Management XX 465:261-268. Materials Research Society, Pittsburgh, Pennsylvania.
McCloy JS and JD Vienna. 2010. “Glass Composition Constraint Recommendations for Use in Life-Cycle Mission Modeling.” PNNL-19372. Pacific Northwest National Laboratory, Richland, Washington. Available at https://www.pnnl.gov/main/publications/external/technical_reports/PNNL-19372.pdf.
McCloy JS, MJ Schweiger, CP Rodriguez, and JD Vienna. 2011. “Nepheline Crystallization in Nuclear Waste Glasses: Progress toward Acceptance of High-Alumina Formulations.” International Journal of Applied Glass Science 2(3):201-214. Available at http://onlinelibrary.wiley.com/doi/10.1111/j.2041-1294.2011.00055.x/full.
Piepel GF, BA Stanfill, SK Cooley, B Jones, JO Kroll, and JD Vienna. 2018. “Developing a space-filling mixture experiment design when the components are subject to linear and nonlinear constraints” Quality Engineering. https://doi.org/10.1080/08982112.2018.1517887
Rodriguez CP, JS McCloy, MJ Schweiger, JV Crum, and A Winschell. 2011. Optical Basicity and Nepheline Crystallization in High Alumina Glasses. PNNL-20184. Pacific Northwest National
EWG-RPT-023 Rev 0.0 PNNL-29023
References 22
Laboratory, Richland, Washington. Available at http://www.pnnl.gov/main/publications/external/technical_reports/PNNL-20184.pdf.
Stanfill BA, GF Piepel, JD Vienna, and SK Cooley. 2019. “Nonlinear logistic regression mixture experiment modelling for binary data using dimensionally-reduced components.” Quality and Reliability Engineering International. Submitted.
Vienna JD, D-S Kim, DC Skorski, and J Matyas. 2013. Glass Property Models and Constraints for Estimating the Glass to Be Produced at Hanford by Implementing Current Advanced Glass Formulation Efforts. PNNL-22631, Rev. 1; ORP-58289, Pacific Northwest National Laboratory, Richland, Washington. Available at http://www.pnnl.gov/main/publications/external/technical_reports/PNNL-22631Rev1.pdf.
Vienna JD, D-S Kim, MJ Schweiger, GF Piepel, JO Kroll, and AA Kruger. 2014. Glass Formulation and Testing for U.S. High-Level Tank Wastes Project 17210 Year 1 Status Report: October 15, 2014. PNNL-SA-84872, Pacific Northwest National Laboratory, Richland, Washington. Available at https://www.hanford.gov/files.cfm/Glass_Formulation_Year_1_Report_Final_10-15-14,_PNNL-SA-84872.pdf.
Vienna JD, GF Piepel, DS Kim, JV Crum, CE Lonergan, BA Stanfill, BJ Riley, SK Cooley, and T Jin. 2016. Update of Hanford Glass Property Models and Constraints for Use in Estimating the Glass Mass to be Produced at Hanford by Implementing Current Enhanced Glass Formulation Efforts. PNNL-25835, Pacific Northwest National Laboratory, Richland, Washington. Available at http://www.pnnl.gov/main/publications/external/technical_reports/PNNL-25835.pdf.
Vienna JD, JO Kroll, PR Hrma, JB Lang, and JV Crum. 2017. “Submixture Model to Predict Nepheline Precipitation in Waste Glasses.” International Journal of Applied Glass Science 8(2):143-157. DOI: 10.1111/ijag.12207
EWG-RPT-023 Rev 0.0 PNNL-29023
Appendix A A.1
Appendix A – Measured and Targeted Glass Compositions
This appendix contains the compositional data measured and reported by SRNL in SRNL-STI-2018-00645 (Fox et al. 2019). BDL = below detection limit.
Table A-1: Comparison of targeted and measured glass compositions