“`html
As nations globally witness a revival in nuclear energy initiatives, the issues concerning the disposal of nuclear waste continue to be as politically sensitive as ever. The United States, for example, has indefinitely postponed its sole long-term underground nuclear waste repository. Researchers are employing both modeling and experimental approaches to examine the impacts of subterranean nuclear waste disposal and ultimately aim to enhance public confidence in the decision-making process.
Recent findings from researchers at MIT, Lawrence Berkeley National Laboratory, and the University of Orléans demonstrate progress in this area. The research reveals that simulations of underground nuclear waste interactions, produced by innovative, high-performance-computing software, correspond closely with experimental outcomes from a research facility in Switzerland.
The study, co-authored by MIT doctoral candidate Dauren Sarsenbayev and Assistant Professor Haruko Wainwright, alongside Christophe Tournassat and Carl Steefel, is published in the journal PNAS.
“These advanced new computational resources, together with real-life experiments such as those at the Mont Terri research site in Switzerland, aid us in comprehending how radionuclides will move in interconnected underground systems,” states Sarsenbayev, the lead author of the study.
The authors aspire that the findings will bolster assurance among lawmakers and the public regarding the long-term safety of subterranean nuclear waste disposal.
“This investigation — integrating both computational methods and experiments — is crucial to enhance our confidence in waste disposal safety evaluations,” remarks Wainwright. “With nuclear energy re-emerging as a crucial solution for addressing climate change and ensuring energy security, it is essential to validate disposal routes.”
Contrasting simulations with experiments
Currently, disposing of nuclear waste in deep geological formations is deemed the safest long-term approach for managing high-level radioactive waste. Consequently, significant efforts have been directed towards exploring the migration behaviors of radionuclides from nuclear waste within various natural and engineered geological materials.
Since its establishment in 1996, the Mont Terri research site in northern Switzerland has acted as a vital testing ground for an international group of researchers interested in exploring materials such as Opalinus clay — a dense, impermeable claystone prevalent in the tunneled areas of the mountain.
“It is widely acknowledged as one of the most significant real-world experimental sites due to the decades of data available on the interactions between cement and clay, which are the key materials proposed for use by countries worldwide in engineered barrier systems and geological repositories for nuclear waste,” clarifies Sarsenbayev.
For their research, Sarsenbayev and Wainwright teamed up with co-authors Tournassat and Steefel, who have developed high-performance computing software to enhance modeling of interactions between nuclear waste and both engineered and natural materials.
To date, multiple challenges have restricted scientists’ understanding of how nuclear waste interacts with cement-clay barriers. For instance, the barriers consist of irregularly mixed materials located deep underground. Moreover, the existing category of models commonly utilized to simulate radionuclide interactions with cement-clay does not account for electrostatic effects related to the negatively charged clay minerals in the barriers.
The new software by Tournassat and Steefel considers electrostatic effects, making it the sole program capable of simulating those interactions in three-dimensional space. The software, named CrunchODiTi, was developed from the established framework CrunchFlow and received its latest update this year. It is designed to operate concurrently on multiple high-performance computers.
For the research, the scientists examined a 13-year-old experiment, initially centered on the interactions between cement and clay rock. Recently, a combination of both negatively and positively charged ions was introduced to the borehole located near the center of the cement placed in the formation. The researchers concentrated on a 1-centimeter-thick area between the radionuclides and cement-clay referred to as the “skin.” They juxtaposed their experimental findings against the software simulation, discovering that the two datasets were consistent.
“The findings are quite significant because prior to this, these models did not align well with field data,” remarks Sarsenbayev. “It’s intriguing how fine-scale phenomena at the ‘skin’ between cement and clay, whose physical and chemical attributes change over time, could be utilized to reconcile the experimental and simulation data.”
The experimental outcomes demonstrated that the model successfully addressed electrostatic effects associated with the clay-rich formation and the interactions between materials in Mont Terri over time.
“This is all driven by decades of effort to comprehend what transpires at these interfaces,” Sarsenbayev states. “It has been hypothesized that mineral precipitation and porosity blockage occur at this interface, and our findings strongly indicate that.”
“This application necessitates millions of degrees of freedom since these multibarrier systems require high resolution and substantial computational capacity,” Sarsenbayev mentions. “This software is ideally suited for the Mont Terri experiment.”
Evaluating waste disposal strategies
The novel model could now supersede older models that have been applied to perform safety and performance evaluations of underground geological repositories.
“If the U.S. ultimately chooses to dispose of nuclear waste in a geological repository, these models could prescribe the most suitable materials to utilize,” Sarsenbayev explains. “For example, currently, clay is considered an appropriate storage medium, but salt formations represent another potential option that could be used. These models enable us to track the fate of radionuclides over millennia. We can utilize them to understand interactions over timescales ranging from months to years to numerous millions of years.”
Sarsenbayev mentions that the model is relatively accessible to other researchers and that future endeavors may explore employing machine learning to develop less computationally intensive surrogate models.
Additional data from the experiment will be released later this month. The team intends to compare those findings with further simulations.
“Our collaborators will effectively receive a block of cement and clay, enabling them to conduct experiments to determine the exact thickness of the skin along with all the minerals and processes present at this interface,” Sarsenbayev states. “It’s a significant project and takes time, but we wanted to share preliminary data and this software as soon as possible.”
For the time being, the researchers aim for their study to contribute to a long-term solution for nuclear waste storage that policymakers and the public can endorse.
“This is a multidisciplinary study that involves real-world experiments demonstrating our capability to predict the fate of radionuclides in the subsurface,” asserts Sarsenbayev. “The motto of MIT’s Department of Nuclear Science and Engineering is ‘Science. Systems. Society.’ I believe this integrates all three areas.”
“`