A professor at Brigham Young University has discovered a method to significantly reduce the time and cost involved in designing and licensing modern nuclear reactors by integrating artificial intelligence (AI) into the process.

Typically, licensing a new nuclear reactor design in the United States takes around 20 years and costs about $1 billion. Building the reactor adds another five years and costs between $5 and $30 billion. However, chemical engineering professor Matt Memmott believes that using AI in the computational design process could cut a decade or more off this timeline, saving millions of dollars. This advancement is crucial given the increasing energy demands.

Engineers face complex challenges, from quantum-scale neutron behavior to macro-scale coolant flow and heat transfer. Memmott explains that these elements are tightly interconnected, making the design process lengthy and data-intensive. AI can alleviate this burden, leading to faster power production and lower costs for consumers.

Memmott's research demonstrates that replacing some thermal hydraulic and neutronics simulations with a trained machine learning model can predict temperature profiles based on variable geometric reactor parameters. This approach optimizes reactor design at a fraction of the traditional computational expense. Memmott and his colleagues developed and tested multiple machine learning algorithms, identifying the most effective one for optimizing nuclear shield design.

Their findings, published in the journal Nuclear Engineering and Design, show that their AI model can optimize design elements much faster than traditional methods. In just two days, the AI algorithm determined an optimal nuclear-reactor shield design, a task that previously took a molten salt reactor company six months.

While humans still make the final design decisions and conduct safety assessments, AI significantly reduces the initial time investment. Memmott emphasizes the importance of this development, stating that as electricity demand soars, nuclear power remains the only emissions-free option capable of meeting gigawatt-scale needs.