Modeling Nuclear Fuel Behavior for Enhanced Reactor Performance and Safety

By BRENT LEWIS, KHALED SHAHEEN, AND MICHAEL J. WELLAND

Nuclear energy offers a green solution for electric power generation. Researchers at Royal Military College in Canada show how to get the most out of nuclear fuel.

Schematic of a CANDU plant. Used with permission from A. El-Jaby.

Figure 1. Schematic of a CANDU plant. Used with permission from A. El-Jaby.

Nuclear energy is gaining attention across the globe as a practical alternative- energy solution. In comparison to thermal sources where fossil fuels are burned to create energy, nuclear power can be considered a clean form of energy because there is no combustion during the nuclear reaction. Hence, nuclear energy does not emit greenhouse gases, acidic gases, or particulates, all issues related to global warming and environmental damage. Additionally, the highenergy density of nuclear fission reduces the use of natural resources, not to mention the impact of extracting those resources. Finally, nuclear fuel can be reprocessed to extract even more of the unused energy from the fuel, making it recyclable. All reasons why research into how to efficiently and safely get the most out of nuclear fuel can be considered of great importance.

One country that has been utilizing nuclear power generation for some time is Canada, where nuclear energy generates 15% of Canada’s electric power. To help further Canada’s nuclear technology domestically and throughout the world, nuclear energy research is supported at the university level, including at the Royal Military College (RMC) of Canada in Ontario. According to Brent Lewis, Ph.D., Industrial Research Chair (IRC) in Nuclear Fuel at RMC, the university’s work will help support the CANDU (CANada Deuterium Uranium) nuclear industry for improved reactor operational support, fuel performance prediction, and reactor safety code analysis. “This research is of particular importance in light of the need for increased capacity and the realization of next-generation reactors, which can offset the greenhouse gas footprint,” said Dr. Lewis.

Summary of coupled phenomena and properties governing nuclear fuel performance.

Figure 2. Summary of coupled phenomena and properties governing nuclear fuel performance.

The Research Team

Since nuclear fuel is a key enabling technology for the enhancement of reactor safety, performance, and economics, Dr. Lewis and his team at RMC are conducting research on nuclear fuel behavior. Dr. Lewis explained how the main goals of the IRC in nuclear fuel are to better understand nuclear fuel performance during normal and reactor accident conditions, including the behavior of advanced and next-generation fuel designs.

Because of its multiphysics capabilities, Dr. Lewis chose COMSOL for his team to model the complexities of nuclear fuel behavior (Figure 2). “COMSOL allows us to quickly focus on the model development and given application with its multiphysics capability and its ability for numerical solution of the governing (coupled) partial differential equations for heat and mass transport,” said Dr. Lewis. “Since COMSOL is used and tested by many people in the world, we trust its solver and therefore can specifi cally focus on the model development.”

Accurately Modeling the Fuel Element

One of Dr. Lewis’s Ph.D. students, Khaled Shaheen, M.A.Sc., is currently modeling the behavior of a nuclear fuel element — intact or defective — during reactor operation. “There are many considerations that must be taken into account in order to simulate the reactor temperature under varying circumstances — material composition and properties that are changing over time, and a geometry that is evolving in response to the physics of the system,” said Mr. Shaheen. For example, Dr. Lewis explained how, “The Structural Mechanics Module allows us to develop mechanistically-based models for describing the thermo-mechanical behavior of nuclear fuel. The multiphysics platform allows us to couple governing partial and ordinary differential equations.” It is these sophisticated models that Mr. Shaheen relies on in order to accurately predict all of these coupled phenomena together.

One specific challenge Mr. Shaheen discussed about his work was how geometry and fi ssion gas behavior affect the heat transport within the fuel rod. The heat transport simultaneously affects the physical deformation of the rod, and the production and release of fission gases, which makes modeling fuel performance a highly-coupled problem. “The multiphysics capability of COMSOL, and the ability to implement two-way coupling between the different physics interfaces and the moving mesh interface, are key to being able to solve this problem. Adding new effects to a model is straightforward, which means that we can always add new and more detailed physics to increase the accuracy, sophistication, and applicability of our code,” said Mr. Shaheen. “Also significant is the ability to control meshing parameters. As we expand our models to multiple dimensions, the different scales make it very important to deal with aspect ratios, meaning that we need the flexibility in customizing meshes,” he added.

Temperature profi le of oxidized
defective fuel. Fuel element temperature
slice plot in degrees Kelvin.

Figure 3. Temperature profile of oxidized defective fuel. Fuel element temperature slice plot in degrees Kelvin.

Combining Models

While Mr. Shaheen was working on nuclear fuel behavior during normal reactor conditions, his colleague Michael J. Welland, Ph.D. — currently a research fellow at the JRC- Institute for Transuranium Elements in Karlsruhe, Germany — was completing his Ph.D. thesis at RMC concentrating on fuel during reactor accident conditions, specifi cally fuel melting. “The fuel melting model work was conducted to improve the understanding of the melting process and what might be expected should it occur. As such, it contributes to the safe and effi cient operation of the modern reactors and may help in the design of the next-generation reactors,” stated Dr. Welland.

If a fuel rod becomes defective, the coolant could make contact with the fuel and oxidize it, explained Dr. Welland. Oxidized fuel has a reduced thermal conductivity and a lower initial melting temperature. This has the effect of increasing the temperature in the center of the fuel, and the potential for centerline melting under upset conditions (Figure 3). In reactor operation, the melting of the fuel is forbidden as it can challenge the structural integrity of the reactor core. A limit is therefore imposed on the power at which the reactor can be safely operated.

The team at RMC had a fuel oxidation model that was used to simulate ten cases of defective fuel oxidation that had occurred during real power reactor operation. “It was possible to take features from the model used to predict oxidation and integrate them, with some modification, into the fuel-melting model,” said Mr. Shaheen.

Dr. Welland pointed out that while the initial model for fuel oxidation was developed for normal operating conditions and therefore required some modifying, this required little more than copying and pasting the appropriate expressions. “As both models were in COMSOL, we were able to truly combine them, rather than coupling them as one would have to do were they developed as separate modules or standalone codes. The result was that this could happen quickly and easily and that the resulting model is of good quality,” he said.

Ongoing Efforts

The team at RMC is also modeling Zircaloy hydriding, iodine-induced stress corrosion cracking, and fi ssion product transport in defective fuel and to plan and design an instrumented out-reactor fuel test to provide data for model validation. “The COMSOL numerical platform can be easily learned by students and the programs passed on to other students for ongoing and continued research development,” said Dr. Lewis. “ This software allows us to quickly develop and test models in a number of graduate theses within a reasonable timeframe. These models can then be used by the industry after they have been conceptually developed and tested at the university.”

The modeling team (left to right) Brent Lewis, Khaled Shaheen, and
Michael J. Welland.

The modeling team (left to right) Brent Lewis, Khaled Shaheen, and Michael J. Welland.