CASL simulations add insight into operating nuclear reactor cores
Reactor core simulations: (Top) a slice through the full core showing power generation (red signifies more power, blue less); (middle) a close-up view of several fuel sub-assemblies showing fuel, control pins, and water channels; (bottom) the lower right corner of a full core view with control rods (blue) inserted.
Powerful computers and great minds from national laboratories, academia and industry are already notching successes that could help commercial light water nuclear reactors operate more efficiently and reliably for decades.
Through the Consortium for Advanced Simulation of Light Water Reactors, the first Energy Innovation Hub established by the US Department of Energy just three years ago, researchers recently performed their first fullscale simulation of a reactor during startup. While that's a significant milestone, it represents just the beginning for this partnership of 10 core players, says CASL Director Doug Kothe, who noted that leadership computers such as ORNL's Titan—the fastest in the US— are providing unprecedented modeling and simulation capabilities.
"With CASL tools using Titan, our scientists are seeking to gain a better understanding of what's happening with, for example, tens of thousands of fuel rods in a reactor core," Kothe says. "The amount of information we can see is not only unprecedented, but also revealing." This new window into a reactor core, while initially made possible through CASL modeling and simulation technology that effectively utilizes high-performance computing platforms like Titan, can then be used to increase the capabilities of industrial-class computers in common use today.
The recent simulation used CASL's new Virtual Environment for Reactor Applications. The VERA simulations were directly compared against operational data taken at Tennessee Valley Authority's Watts Bar Nuclear Plant, and the favorable comparisons showed that the software environment is both accurate and useful. Through VERA and simulations to come, researchers will gain a better understanding of reactor performance with much greater fidelity than provided by methods of the past.
While this initial VERA simulation focused on the startup cycle, future simulations will examine full power operations of the TVA reactor, which will utilize current VERA development that is integrating the thermal hydraulics behavior, fuel performance and surface chemistry. These additional capabilities will allow not only a greater understanding of operating reactors, but also spark insights that Kothe and colleagues are confident can stimulate advances in reactor operations.
"Our vision is to predict with confidence the safe, reliable and economically competitive performance of nuclear reactors through science-based modeling and simulation technologies," Kothe says. "These predictive technologies can then be deployed on common computers used broadly throughout the nuclear energy industry."
CASL is headquartered at ORNL, and its core partners include: the Electric Power Research Institute, Idaho National Laboratory, Los Alamos National Laboratory, Massachusetts Institute of Technology, North Carolina State University, Sandia National Laboratories, TVA, University of Michigan, Westinghouse Electric Company and ORNL.
Industry partners counting on CASL
From the nuclear power plant business' perspective, maximizing the life span and performance of the more than 100 reactors in operation across the nation are of critical importance. With 100,000 megawatts of power generation capacity, today's reactors are supplying nearly 20 percent of US electricity.
While a new generation of reactors emerges, many years of useful life remain for these proven sources of power, which represent billions of dollars of investment.
To accomplish their goal of helping industry continue to optimize performance, CASL researchers focus on the in-vessel reactor core phenomena of pressurized water reactors, the most common type of light water reactor in the US. They are studying, for example, the behavior of nuclear fuel during all operating conditions, while looking for potential modifications to enhance safety and efficiency.
Kothe also notes that performance expectations for first-generation nuclear power reactors needed to be conservative in order to guarantee safety with what was then new technology. Fifty years later, modern tools and supercomputers are allowing the CASL team to gain a deeper understanding of underlying processes such as thermal hydraulics, fuel rod performance, neutronics, surface chemistry and corrosion, and structural dynamics.
"For example, our new capabilities will allow us to look closely at reactor core models operating with 193 fuel assemblies, nearly 51,000 fuel rods and about 18 million fuel pellets," Kothe says. "These elements operate in a high-temperature, high-pressure, high-radiation environment for three to five years. Our software is evolving to simulate these conditions and predict performance, providing industry with some of the information it needs to meet its goals."
CASL core partners provide a unique perspective based on decades of working with industry members. Some partners envision CASL research leading to nuclear power plants that are more flexible in operation, using, for example, "gray" control rods, an approach that helps a reactor match power output more closely to demand. Opportunities may also exist for more accident-tolerant fuels, or for improvements to reactor core designs that allow more fuel burnup and consequently less waste to be produced.
For others a state-of-the-art understanding of nuclear power plant phenomena and performance can benefit from advanced modeling and simulation, exactly the role filled by CASL. A rigorous simulation could address scenarios in which power is limited by certain behaviors in the core—with insight potentially leading to reduced fuel costs or increased power ratings.
Meanwhile, other team members are interested in modeling and simulation work that would result in a clearer understanding of reactor safety margins. During the industry's first generation of light water reactor development, safety margins were set very carefully in order to ensure public safety with the new technology. Now, with more than 50 years of operating experience and increased understanding of the science gained through tools such as VERA, there may be some room for margins to be adjusted and performance increased, while still preserving the highest levels of public safety.
VERA in action
Through what the CASL team calls "test stands," scientists are providing early deployment of CASL technologies into active nuclear design and engineering environments.
"Our focus is on enhancing the performance of light water reactors with advanced nuclear reactor modeling and simulation technology," says ORNL's Jess Gehin, a nuclear engineer and member of the CASL team. "Through the test stand effort, we're able to address issues that are important to the nuclear industry. We're also able to receive highly constructive feedback to help us continuously improve CASL's simulation capabilities."
CASL test stands offer flexibility, allowing for siting at a CASL industry partner, council member or collaborator site, and use of ORNL-based computing assets or local computing assets. "In a sense a CASL test stand is similar to a rocket test stand," Kothe says. "We use a test stand to determine performance characteristics of VERA, but I'll add that we also allow and want our industry partners and industry council members to be the ones doing the testing. And we want them to test VERA in the process of trying to do 'real work.'"
Kothe noted that working with Westinghouse, for example, CASL deployed its first test stand at a Cranberry, Penn., site this summer. It is allowing Westinghouse and CASL to apply and test VERA on core physics analysis of the AP1000 Pressurized Water Reactor and its advanced first core design. Simulating the advanced core design provides a challenging scenario to test the VERA tools' prediction capabilities. In return CASL gets valuable and candid feedback on whether VERA is useful and usable. "This feedback can be in any number of forms—examining feature needs, quality, robustness or computational performance," Kothe says. "The possibilities are endless, but getting this feedback via test stands now is invaluable because it allows our active development to address any problems and issues."
Current industry analysis techniques are effective at helping scientists and engineers understand and predict the performance of materials, components and subsystems of nuclear power plants. Often, however, they are based on a collection of simplifications and calibrations that ultimately limit the applicability and predictability of the technique.
By pressing beyond the techniques of the past and applying contemporary methods, the CASL team is pursuing a deeper understanding of phenomena that are taking place every day in nuclear reactors across the US.
Armed with this collection of capabilities, CASL takes on "challenge problems" that encompass phenomena limiting the performance of some pressurized reactors. These problems are what drive the development of higher-fidelity tools that combine multiple complex physical and chemical processes taking place simultaneously. By demonstrating the application of these tools to existing issues, immediate insights can be delivered to the commercial nuclear power industry. As Gehin sums it up, "With CASL, we are developing the next generation of reactor simulation tools that offer huge potential for improving our abilities to simulate reactor operation and performance."
CASL Core Partners
CASL Contributing Partners