NUCLEAR POWERCutting Nuclear Power Plant Costs: Argonne Develops Framework for Smarter Maintenance
Merge a multiphysics simulation with real nuclear reactor inspection data and the result is a revolutionizing tool that predicts component failure before it happens. The study combines advanced simulations with real-world testing to predict how feedwater heater tubes, which preheat water before it enters a nuclear reactor, break down over time. The powerful new tool improves maintenance strategies and extend component lifespan.
Merge a multiphysics simulation with real nuclear reactor inspection data and the result is a revolutionizing tool that predicts component failure before it happens.
Researchers at the U.S. Department of Energy’s (DOE) Argonne National Laboratory have developed an innovative framework to improve maintenance schedules for critical components in nuclear power plants. This breakthrough could save millions of dollars on operating costs while keeping power reliable.
The study, published in Engineering Failure Analysis, combines advanced simulations with real-world testing to predict how feedwater heater (FWH) tubes break down over time. This information offers nuclear power plant operators a powerful tool to improve maintenance strategies and extend component lifespan.
FWHs preheat water before it enters the reactor in a nuclear power plant. When a plant goes from a cooldown state to full operational power, the water changes temperature from cold to hot. This causes fatigue on FWH tubes and other components.
A Cost-Saving Solution for the Nuclear Industry
Maintenance of FWHs is challenging. Over time and after a set number of cycles, plant operators manually inspect the FWH tubes to see if they need to be repaired. Current inspection methods are infrequent, labor-intensive and expensive, leaving gaps in system health monitoring.
Existing approaches, like run-to-failure methods, lack the precision needed to detect early signs of degradation or identify specific areas of concern.
The failure of a single FWH tube could cost $1.2 million each year. Plants commonly contain 30 of these FWHs. Collectively, across all 94 U.S. nuclear reactors, there are over one thousand FWHs that need be maintained.
“This is exactly what the industry was looking for,” said Richard Vilim, senior nuclear engineer and an author of the paper. “Power plants operate on a fixed maintenance schedule, and unplanned maintenance is costly. They approached us with this challenge, asking if we could develop a method to predict when maintenance should occur. There’s no existing technology that provides this critical information.”
Predicting Component Lifespan with Advanced Simulation Modeling
To find the optimal time for maintenance, the research team developed a multiphysics simulation model. The model integrated fluid-mechanics, condensation, solid mechanics and cumulative damage theory.
The researchers found it challenging to combine multiple physics disciplines for the simulation.
“The physics and geometry of FWHs are complex,” said Akshay Dave, manager of the intelligent systems group and co-author of the paper. “So, to model it, we needed to set up a complicated computation simulation. Then we had to ask ourselves, how do we translate that into something a utility can use? How do we translate that into something that can inform their maintenance schedule?”
The Key to Validating Simulations Is Experimental Data
The researchers built a framework, linking simulation results with real inspection data to accurately predict the remaining useful life of FWH tubes. The model predicted tube lifespan to be approximately 29 years, which closely matches the actual replacement history. It also accurately identified areas of stress along the tube.
“The most difficult part of complex simulations is validating it,” said Yeni Li, Argonne scholar and lead author on the paper. “The key to validating simulations is with experimental data. Only at Argonne can we find the concentration of skills in one area to be able to do that.”
The framework was validated against experimental inspection data and replacement history from an operating power plant. It showed degradation patterns and lifespan predictions to be consistent with what the researchers modeled.
Scalable Solutions for Nuclear Resilience
There are many practical implications for this framework. For nuclear power plant operators, the framework enables predictive maintenance, which reduces operational costs and minimizes downtime. It also improves system reliability and safety by identifying critical degradation areas before they become a problem. Economically, the framework could save nuclear power plants millions annually.
Finally, the framework offers a scalable solution for other heat exchangers and critical components in the energy industry. It also supports the design of next-generation advanced reactors by providing insights into component degradation.
Going forward, the researchers plan to expand the framework by incorporating additional degradation mechanisms, such as corrosion and wear.
“As we aim to extend our capabilities across the entire utility industry, collaborating with industry partners to cross-validate our data is essential,” said Dave. “This effort holds significant promise for next-generation advanced reactors currently in the design phase, where our insights could be invaluable.”
By combining advanced simulations with experimental validation, this research offers a practical innovation to enhance efficiency, reliability and safety in the utilities sector.
Marguerite Huber is Communications Coordinator at Argonne National Laboratory. The article was originally posted to the website of Argonne National Laboratory.
