Glass Design Using Machine Learning Property Models with Prediction Uncertainties: Nuclear Waste Glass Formulation

Lu, Xiaonan; Weller, Zachary D.; Gervasio, Vivianaluxa; Vienna, John D.

doi:10.1016/j.jnoncrysol.2024.122907

Glass Design Using Machine Learning Property Models with Prediction Uncertainties: Nuclear Waste Glass Formulation

Journal Article · 01 May 2024 · Journal of Non-crystalline Solids

DOI:https://doi.org/10.1016/j.jnoncrysol.2024.122907· OSTI ID:2409376

^[1]; Weller, Zachary D. ^[1]; Gervasio, Vivianaluxa ^[1]; Vienna, John D. ^[1]

BATTELLE (PACIFIC NW LAB)

The United States Department of Energy is responsible for managing the legacy nuclear waste stored in underground tanks at the Hanford Site. The waste will be separately vitrified as low-activity waste and high-level waste fractions. Waste glass formulation algorithms have been traditionally developed using partial quadratic mixture property-composition models. Recently, machine learning (ML) techniques have been used to predict glass properties and discover new glass materials for nuclear waste vitrification, and these advancements can be utilized to improve waste glass composition design. In this proof-of-principle study, ML algorithms such as Gaussian process regression (GPR) were used to interpolate glass properties (e.g., viscosity, electrical conductivity, chemical durability). After selecting appropriate sets of GPR hyper-parameters for each property, an optimization program was developed to formulate glass compositions to maximize waste loading while simultaneously satisfying property within constraints. The results of the ML-based waste loadings and glass compositions were compared to those obtained using the traditional methods. Comparing to the previous glass design framework, the ML-based optimization methods offer improved glass designs and a streamlined approach to generation of optimally designed data and near real-time updates.

Research Organization:: Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)

Sponsoring Organization:: USDOE

DOE Contract Number:: AC05-76RL01830

OSTI ID:: 2409376

Report Number(s):: PNNL-SA-193529

Journal Information:: Journal of Non-crystalline Solids, Journal Name: Journal of Non-crystalline Solids Vol. 631

Country of Publication:: United States

Language:: English

References (31)

Machine learning as a tool to design glasses with controlled dissolution for healthcare applications Han, Taihao; Stone-Weiss, Nicholas; Huang, Jie Acta Biomaterialia, Vol. 107 https://doi.org/10.1016/j.actbio.2020.02.037	journal	April 2020
Effect of composition on peraluminous glass properties: An application to HLW containment Piovesan, V.; Bardez-Giboire, I.; Perret, D. Journal of Nuclear Materials, Vol. 483 https://doi.org/10.1016/j.jnucmat.2016.11.008	journal	January 2017
Forty years of durability assessment of nuclear waste glass by standard methods Thorpe, Clare L.; Neeway, James J.; Pearce, Carolyn I. npj Materials Degradation, Vol. 5, Issue 1 https://doi.org/10.1038/s41529-021-00210-4	journal	December 2021
Array programming with NumPy Harris, Charles R.; Millman, K. Jarrod; van der Walt, Stéfan J. Nature, Vol. 585, Issue 7825 https://doi.org/10.1038/s41586-020-2649-2	journal	September 2020
Machine Learning with Gradient-Based Optimization of Nuclear Waste Vitrification with Uncertainties and Constraints Gunnell, LaGrande Lowell; Manwaring, Kyle; Lu, Xiaonan Processes, Vol. 10, Issue 11 https://doi.org/10.3390/pr10112365	journal	November 2022
Optimizing the Prediction Accuracy of Concrete Compressive Strength Based on a Comparison of Data-Mining Techniques Chou, Jui-Sheng; Chiu, Chien-Kuo; Farfoura, Mahmoud Journal of Computing in Civil Engineering, Vol. 25, Issue 3 https://doi.org/10.1061/(ASCE)CP.1943-5487.0000088	journal	May 2011
A 25-year laboratory experiment on French SON68 nuclear glass leached in a granitic environment – First investigations Guittonneau, C.; Gin, S.; Godon, N. Journal of Nuclear Materials, Vol. 408, Issue 1 https://doi.org/10.1016/j.jnucmat.2010.10.075	journal	January 2011
Predicting radioactive waste glass dissolution with machine learning Lillington, Joseph N. P.; Goût, Thomas L.; Harrison, Mike T. Journal of Non-Crystalline Solids, Vol. 533 https://doi.org/10.1016/j.jnoncrysol.2019.119852	journal	April 2020
Compositional Models of Glass/Melt Properties and their Use for Glass Formulation Vienna, John D. Procedia Materials Science, Vol. 7 https://doi.org/10.1016/j.mspro.2014.10.020	journal	January 2014
Hanford low-activity waste vitrification: A review Marcial, José; Riley, Brian J.; Kruger, Albert A. Journal of Hazardous Materials, Vol. 461 https://doi.org/10.1016/j.jhazmat.2023.132437	journal	January 2024
Matplotlib: A 2D Graphics Environment Hunter, John D. Computing in Science & Engineering, Vol. 9, Issue 3 https://doi.org/10.1109/MCSE.2007.55	journal	January 2007
Glass Property Models, Constraints, and Formulation Approaches for Vitrification of High-Level Nuclear Wastes at the US Hanford Site Kim, Dongsang Journal of the Korean Ceramic Society, Vol. 52, Issue 2 https://doi.org/10.4191/kcers.2015.52.2.92	journal	March 2015
Impacts of constraints and uncertainties on projected amount of Hanford low-activity waste glasses Lu, Xiaonan; Kim, Dong-Sang; Vienna, John D. Nuclear Engineering and Design, Vol. 385 https://doi.org/10.1016/j.nucengdes.2021.111543	journal	December 2021
Prediction of Compressive Strength of Concrete: Critical Comparison of Performance of a Hybrid Machine Learning Model with Standalone Models Cook, Rachel; Lapeyre, Jonathan; Ma, Hongyan Journal of Materials in Civil Engineering, Vol. 31, Issue 11 https://doi.org/10.1061/(ASCE)MT.1943-5533.0002902	journal	November 2019
Machine learning unveils composition-property relationships in chalcogenide glasses Mastelini, Saulo Martiello; Cassar, Daniel R.; Alcobaça, Edesio Acta Materialia, Vol. 240 https://doi.org/10.1016/j.actamat.2022.118302	journal	November 2022
SciPy 1.0: fundamental algorithms for scientific computing in Python Virtanen, Pauli; Gommers, Ralf; Oliphant, Travis E. Nature Methods https://doi.org/10.1038/s41592-019-0686-2	journal	February 2020
A crucible salt saturation method for determining sulfur solubility in glass melt Jin, Tongan; Kim, Dongsang; Darnell, Lori P. International Journal of Applied Glass Science, Vol. 10, Issue 1 https://doi.org/10.1111/ijag.12366	journal	June 2018
Predicting nepheline precipitation in waste glasses using ternary submixture model and machine learning Lu, Xiaonan; Sargin, Irmak; Vienna, John D. Journal of the American Ceramic Society, Vol. 104, Issue 11 https://doi.org/10.1111/jace.17983	journal	July 2021
Product Consistency Test and Vapor Hydration Test Comparisons of a Radioactive Hanford Waste Glass with its Non-Radioactive Simulant Glass Reiser, Joelle T.; Neeway, James; Parruzot, Benjamin P. https://doi.org/10.2172/1998870	report	April 2023
GEKKO Optimization Suite Beal, Logan; Hill, Daniel; Martin, R. Processes, Vol. 6, Issue 8 https://doi.org/10.3390/pr6080106	journal	July 2018
Glassy Wasteforms for Nuclear Waste Immobilization Ojovan, Michael I.; Lee, William E. Metallurgical and Materials Transactions A, Vol. 42, Issue 4 https://doi.org/10.1007/s11661-010-0525-7	journal	November 2010
Data-driven machine learning for disposal of high-level nuclear waste: A review Hu, Guang; Pfingsten, Wilfried Annals of Nuclear Energy, Vol. 180 https://doi.org/10.1016/j.anucene.2022.109452	journal	January 2023
An international initiative on long-term behavior of high-level nuclear waste glass Gin, S.; Abdelouas, A.; Criscenti, L. J. Materials Today, Vol. 16, Issue 6 https://doi.org/10.1016/j.mattod.2013.06.008	journal	June 2013
Nuclear Waste Vitrification in the United States: Recent Developments and Future Options: Nuclear Waste Vitrification in the United States. Vienna, John D. International Journal of Applied Glass Science, Vol. 1, Issue 3 https://doi.org/10.1111/j.2041-1294.2010.00023.x	journal	September 2010
Can the compressive strength of concrete be estimated from knowledge of the mixture proportions?: New insights from statistical analysis and machine learning methods Young, Benjamin A.; Hall, Alex; Pilon, Laurent Cement and Concrete Research, Vol. 115 https://doi.org/10.1016/j.cemconres.2018.09.006	journal	January 2019
A history of Hanford tank waste, implications for waste treatment, and disposal Colburn, Heather A.; Peterson, Reid A. Environmental Progress & Sustainable Energy, Vol. 40, Issue 1 https://doi.org/10.1002/ep.13567	journal	December 2020
The immobilization of high level radioactive wastes using ceramics and glasses Donald, I. W.; Metcalfe, B. L.; Taylor, R. N. J. Journal of Materials Science, Vol. 32, Issue 22, p. 5851-5887 https://doi.org/10.1023/A:1018646507438	journal	January 1997
Machine learning in concrete strength simulations: Multi-nation data analytics Chou, Jui-Sheng; Tsai, Chih-Fong; Pham, Anh-Duc Construction and Building Materials, Vol. 73 https://doi.org/10.1016/j.conbuildmat.2014.09.054	journal	December 2014
Vitrification of High Level Waste in the UK Harrison, Mike T. Procedia Materials Science, Vol. 7 https://doi.org/10.1016/j.mspro.2014.10.003	journal	January 2014
Data‐driven glass/ceramic science research: Insights from the glass and ceramic and data science/informatics communities De Guire, Eileen; Bartolo, Laura; Brindle, Ross Journal of the American Ceramic Society, Vol. 102, Issue 11 https://doi.org/10.1111/jace.16677	journal	May 2019
Data Structures for Statistical Computing in Python McKinney, Wes Proceedings of the Python in Science Conference https://doi.org/10.25080/Majora-92bf1922-00a	conference	January 2010

Similar Records

Glass Property Models, Constraints, and Formulation Approaches for Vitrification of High-Level Nuclear Wastes at the US Hanford Site

Journal Article · 2015 · Journal of the Korean Ceramic Society · OSTI ID:1244810

Compositional Models of Glass/Melt Properties and their Use for Glass Formulation

Journal Article · 2014 · Procedia Materials Science · OSTI ID:1209064

Preliminary Enhanced LAW Glass Formulation Algorithm

Technical Report · 2020 · OSTI ID:1592112

Related Subjects

glass formulation
machine learning
constraints
prediction uncertainty
artificial intelligence

Glass Design Using Machine Learning Property Models with Prediction Uncertainties: Nuclear Waste Glass Formulation

Citation Formats

References (31)

Similar Records

Related Subjects