Uncertainty in CFD simulation of reactors and approaches to improve the confidence of simulation results

Computational Fluid Dynamics (CFD) plays an important role in the design and optimization of reactor thermal-hydraulic systems and serves as an essential component of the ”virtual reactor” projects globally. However, the credibility of CFD simulations is often questioned due to various sources of error and uncertainty, limiting their broader applications in engineering practice. A rigorous process of Verification, Validation, and Uncertainty Quantification (V&V&UQ) is widely acknowledged as essential for assessing the credibility of CFD simulation results. Nevertheless, existing guidelines and standards remain largely theoretical, lacking practical implementation strategies, which prevents a comprehensive understanding of V&V&UQ among researchers. With the enhancement of computational resources, there is potential to improve the credibility of CFD simulation of reactors by reducing the introduction of errors and uncertainties while enhancing V&V&UQ tools and data accessibility. This paper identifies key sources of error and uncertainty across different stages of CFD simulation of reactors from a software developer’s perspective. Furthermore, it explores approaches to improve the credibility of CFD simulation of reactors, emphasizing four critical areas: minimizing model uncertainty, reducing numerical uncertainty, establishing a robust mesh quality evaluation framework, and advancing supportive tools and datasets. By clarifying sources of error and uncertainty and providing guidance for future developments, this paper aims to contribute to the enhanced credibility of CFD simulation of reactors in the future. • Reviews the research on V&V&UQ in reactor CFD simulations. • Identifies key sources of error and uncertainty in reactor CFD simulations. • Explores approaches to improve the credibility of reactor CFD simulations.