BYU engineering research finds key to quicker nuclear power: artificial intelligence

🌈 Abstract

The article discusses how a BYU professor is using artificial intelligence (AI) to significantly reduce the time and cost required to design and license modern nuclear reactors in the United States.

🙋 Q&A

[01] Leveraging AI for Nuclear Reactor Design

1. How can AI help speed up the nuclear reactor design and licensing process?

• AI can reduce the heavy computational burden and time required for the complex multi-scale, multi-physics simulations needed to design a nuclear reactor.
• By replacing some of the traditional thermal hydraulic and neutronics simulations with trained machine learning models, the design process can be optimized much faster.
• This allows the design space to be explored more efficiently, focusing on the optimal design rather than combing through a wide range of possibilities.
• While humans still make the final design decisions and safety assessments, the AI-assisted approach can save a significant amount of time and cost upfront.

2. What are the key benefits of using AI in nuclear reactor design?

• Reduces the typical 20-year timeline and $1 billion cost to license a new nuclear reactor design by potentially cutting a decade or more off the process.
• Enables faster deployment of new nuclear power capacity to meet growing electricity demands, while keeping power costs down for consumers.
• Helps make nuclear power a more viable and environmentally friendly option to meet future energy needs.

3. What are the specific technical details of the AI-based approach?

• The research involves building machine learning algorithms to predict temperature profiles based on variable reactor parameters.
• This allows the design elements to be geometrically optimized much faster than traditional methods.
• For example, the AI algorithm was able to find an optimal nuclear shield design in just 2 days, compared to 6 months using traditional methods.

[02] Collaboration and Impact

1. Who are the researchers involved in this project?

• The lead researcher is Matt Memmott, a chemical engineering professor at BYU.
• The BYU research team includes Andrew Larsen, Ross Lee, Braden Clayton, Edwards Mercado, Ethan Wright, Brent Edgerton, and Brian Gonda, as well as chemical engineering professor John Hedengren.
• Collaborators from Alpha Tech Research Corp, Caden Wilson and John Benson, also contributed to the research.

2. What is the potential impact of this AI-powered approach to nuclear reactor design?

• It could significantly reduce the time and cost barriers to deploying new nuclear power capacity, making it a more viable option to meet growing electricity demands.
• This could help nuclear power play a larger role in providing emissions-free, baseload electricity to the grid and address future energy needs.
• Ultimately, the goal is to make nuclear power a more affordable and accessible option for environmentally-friendly power generation.