Simulation Outruns Conservative Nuclear Plant Designs

Simulate large nuclear structures, like this 2 million node mechanical surface treatment reactor vessel component design. Courtesy of Areva.

A real strength of simulation is its ability to assess phenomena which are difficult, expensive, or impossible to test in real life. Few industries can appreciate these difficulties as much as the nuclear power industry. And considering the potential consequences of failure, using simulation to prove the safety of a power plant is key to the wellbeing of society and the project’s profitability.

“You need to justify that your design matches the design regulations,” said Dr Tomasz Kisielewicz, ESI Group’s Executive VP for Services Operations. “You must specify the level of stress to ensure safety based on these three core safety rules: confinement, reactor cooling, and the cooling of spent fuel.”

Kisielewicz described that a decade ago the International Atomic Energy Agency (IAEA) changed the nuclear safety specifications to reflect the growing power of simulation. Using a conservative design approach isn’t interesting when you can simulate a realistic scenario. Simulation enabled the precise modeling of earthquakes, explosions, tsunami, and other climatic events, used as loading conditions recreated to assess the nuclear plant.

“In the 70’s, design regulations specified safety factors due to unknowns. Therefore, these safety margins were based on conservatism applied to standard loading conditions. ESI’s Computer-Aided Engineering (CAE) software like SYSTUS and SYSWELD, for example, can accurately visualize and assess reality. Instead of using set factors, the codes are assessing how the plant is really affected by the extreme load conditions. Unfortunately, though this approach allows plant engineers to deliver safer designs, you cannot totally eliminate risks,” explained Kisielewicz.

Designing Nuclear Power Plants through Simulation

A seismic analysis of nuclear plant auxiliaries. Courtesy of Areva.

“Using a CFD solver coupled with ESI’s software Virtual Performance Solution you can recreate a natural disaster like the one that hit Fukushima. This explicit code can handle non-linear behaviors efficiently. You can also simulate the manufacturing of the parts and see how that may affect the model and load conditions. You can really get a realistic model using ESI solutions,” expressed Kisielewicz.

This invites the question: If the effects of Fukushima can now be simulated, why did the disaster cause so much suffering? The answer is in the loading conditions. The nuclear plant at Fukushima met the regulations that were drawn up at the time of design. Even if the safety of a plant is designed taking into account a catastrophic event of a magnitude expected only once in a million years, the time of that occurrence remains unknown.

“The earthquake that hit the plant was thousands of times more powerful than the event expected in safety standards and the tsunami wave twice as high. The standards have now been changed to include an ultimate safety system that will be able to survive the extreme loads with an annual probability of 10^-8 (once in 100,000,000 years),” said Kisielewicz.

To assess the power plant’s design properly, engineers must account for the stress, fractures, and even for the manufacturing of the plant. Some typical assessments look into the effects of seismic activities, drops, fatigue, distortions, electro-magnetism, and more. To perform these tasks, the engineer must have a broad CAE tool belt to simulate the loads. Various ESI solutions can tackle these challenges, including:

ProCAST: simulates casting processes, predicts porosity & defects
Virtual Performance Solution: end-to end performance analysis considering the manufacturing properties, multi-domain performance including crash/ drop tests, durability, fatigue and more
SYSTUS: stress, fracture & electromagnetics
SYSWELD: welding, distortion, & heat treatment
ACE+ Suite: advanced Computational Fluid Dynamics & multiphysics

Extending a Nuclear Plant’s Life

Elasto-plastic study of steam generator tube support plates. Courtesy of Areva.

Due to politics, economics, and public opinion, many societies are looking to extend the life of nuclear facilities as opposed to producing new ones. However, one cannot just decide to keep a nuclear system going.

“Operators and licensing authorities must justify the extension of life for a plant. Many are looking to extend from 40 to 60 years, but in the US, many plants are looking to 80 years. It can be done if the plant is still safe with respect to design rules. But, the fatigue of metals, wear and tear is real, and metals’ microstructures change drastically under radiation,” said Kisielewicz.

Doubling the lifespan of a plant is no small feat. Fatigue and irradiation can make metals brittle by transitioning their microstructure to a glassier state. As such, the catastrophic loads assessed in design cannot be applied to a plant model with updated stress characteristics and stress reports. These stress reports use a large portion of the ESI library such as mechanical analysis, elastic-plastic studies, and non-linear material properties.

“By using simulation, it must be justified that the core safety functions are maintained for this extended life. Will the radioactivity be contained? Will the shut down and cooling of the reactor be effective? Will the cooling, handling, and storage of the spent fuel be safe or preprocessed properly? The tests are similar to the design of the plant but the initial conditions have changed,” explained Kisielewicz.

Keeping up with Nuclear Power Plant Maintenance

A prediction of a 40 year long crack propagation. Courtesy of Areva.

Maintenance involves simulations of more basic life and use of the plant. Typical maintenance assessments include component fatigue, crack propagation and residual stress. Every year, the plant must pass an inspection period and every ten years an in-depth maintenance is planned. The latter is typically performed during major shutdowns where plant repairs take place.

“During the yearly inspections, you identify each defect in critical components, and see through simulation if they are acceptable for the long term. Simulation can predict how the defect can affect the life of the plant. Based on that prediction, during an upcoming ten-year in-depth maintenance, you can either replace, repair, plug, or weld the component. However, it is very costly to shut down the plant in order to plug a component like a steam generator tube, especially when it has to be done under emergency conditions. At the end of the day, you need to generate revenue. Here is where simulation is key as it can get you back in production faster without impairing safety and it can help in planning major outages rather than reacting to unforeseen critical events,” said Kisielewicz.

End of Life Decommissioning and Dismantling

Decommissioning and dismantling a nuclear power plant are two separate processes. During the decommissioning process, the plant stops producing power but continues some of its functions to maintain reactor and fuel cooling. In a process that could take years, fuel is removed from the plant as it cools.

“After stopping the reactor you still have the fuel producing heat. You will need to maintain the operations of the plant to cool that fuel. During this time, you will need to maintain the same catastrophic external load simulations of the plant that you did during the design and operations of the plant,” explained Kisielewicz.

In other words, the fuel and irradiated plant still pose a threat and must therefore be protected from the same potential catastrophic events as any plant in operation. These safety assessments must also be performed on the storage and reprocessing facilities.

“Once the fuel is removed you can start dismantling the plant. At this stage, the engineer must ensure that the plant will not collapse as it is being taken apart. This can take decades to remove all the irradiated materials. These materials must be stored for a long term so storage must be assessed through simulation. Though not as severe as the assessments during decommissioning, you must still protect the environment.”

San Onofre Fluid Elastic Instability

San Onofre generating station. Courtesy of Jelson25.

Another mainstay of simulation is its ability to assess new phenomena. One example was the fluid elastic instability witnessed at the San Onofre power plant in Southern California.

Within the heat exchange tubes, an unstable vibration was created when the cross flow reached a certain velocity. This vibration was severe enough to risk the collisions and cracking of the tubes.

“The tubes are part of the confinement boundary,” expressed Kisielewicz. “The water in the tubes is radioactive so any leak would break the confinement rule.”

This vibration was a big risk to the plant as the phenomena seemed random at first. “We modeled the steam generator to help predict the phenomena in the computer. Now that we can predict the phenomena we can work to solve it.”

Unfortunately, the vibrations were not the only things affecting the life of the plant. The study was shut down as San Onofre moved into its decommissioning stage. However, with this phenomenon now reproducible in simulation, other plants might be saved.

Furthermore, with the success of the fluid elastic phenomena simulation, I wouldn’t be surprised if San Onofre engineers continued to use ESI solutions during the decommissioning and dismantling of the plant.

ESI Group has sponsored promotion of their nuclear simulation software suite on ENGINEERING.com. They have no editorial input to this post - all opinions are mine. Shawn Wasserman