Researchers at Idaho National Laboratory (INL) recently performed their first digital twin test of a simulated microreactor. The successful demonstration builds on advancements in remote monitoring, autonomous control, and predictive capabilities that can help lower operating costs of microreactor technologies and enhance their safety.
Researchers built a virtual model of the Microreactor Agile Non-nuclear Experimental Testbed (MAGNET) using sensor data and open-source technologies to create a consistent flow of information and real-time data sharing. This digital twin allowed researchers to test, evaluate, and predict microreactor behaviors under different operating conditions.
Through integrated machine learning, the digital twin successfully predicted future heat pipe temperatures and detected trends toward unfavorable threshold temperatures. The virtual model then autonomously controlled the heat pipe by adjusting its temperature to avoid potential complications. Researchers used a separate computer system to capture a 3D model of the heat pipe, along with sensor temperatures.