The fast-paced depletion of fossil fuels and environmental concerns have intensified interest in renewable energies, with dispatchable solar energy emerging as a key alternative. Concentrated solar power (CSP) technology, utilizing thermal energy storage (TES) systems with molten salts at 560 °C, offers significant potential for large-scale energy generation. However, these extreme conditions pose challenges related to material corrosion, which is critical for maintaining the efficiency and lifespan of CSP. This research modeled the corrosion process of 347H stainless steel in molten solar salt (60% NaNO₃ + 40% KNO₃) melted at 400 °C using a cellular automaton (CA) algorithm. The CA model simulated oxide growth under pilot-plant conditions in a lattice of 400 × 400 cells. SEM-EDS imaging compared the model with a mean squared error of 2%, corresponding to a corrosion layer of 4.25 µm after 168 h. The simulation applied von Neumann and Margolus neighborhoods for the ion movement and reaction rules, achieving a cell size of 0.125 µm and 10.08 s per iteration. This study demonstrates the CA model’s effectiveness in replicating corrosion processes, offering a tool to optimize material performance in CSP systems. Additionally, continuing this investigation could contribute to the development of industrial applications, enabling the design of preventive strategies for large-scale operations.

The fast-paced depletion of fossil fuels and environmental concerns have intensified interest in renewable energies, with dispatchable solar energy emerging as a key alternative. Concentrated solar power (CSP) technology, utilizing thermal energy storage (TES) systems with molten salts at 560°C, offers significant potential for large-scale energy generation. However, these extreme conditions pose challenges related to material corrosion, critical for maintaining CSP efficiency and lifespan. This research models the corrosion process of 347H stainless steel in molten solar salt (60% NaNO3 + 40% KNO3) melted at 400°C using a cellular automaton (CA) algorithm. The CA model simulates oxide growth under pilot plant conditions in a lattice of 400x400 cells. SEM-EDS imaging compared the model with a mean squared error of 2%, corresponding to a corrosion layer of 4.25 µm after 168 hours. The simulation applied von Neumann and Margolus neighborhoods for ion movement and reaction rules, achieving a cell size of 0.125 µm and 10.08 seconds per iteration. This study demonstrates the CA model’s effectiveness in replicating corrosion processes, offering a tool to optimize material performance in CSP systems.