Speaker
Description
In the rapidly expanding global nuclear decommissioning market, the efficacious decontamination of metallic waste is pivotal for enhancing waste management economics and ensuring operational safety. Recent pilot-scale demonstrations at the Kori Unit 1 site have validated a synergistic process—integrating ultrasonic cavitation with inorganic salt solutions—which successfully reduced surface contamination of highly active specimens to background (BKG) levels, satisfying stringent clearance criteria. However, inherent mechanical limitations in equipment design often lead to non-uniform decontamination and process inefficiencies when dealing with complex geometries and diverse surface characteristics. Furthermore, conventional manual operations, characterized by lengthy process cycles of approximately nine hours, significantly limit throughput and necessitate technical advancements to mitigate radiological risks to personnel in high-risk environments.
To address these challenges, this study presents the development of an AI-based intelligent real-time hybrid decontamination system featuring multidimensional pre-recognition and autonomous process optimization.
First, the system employs advanced sensor fusion and vision algorithms to precisely characterize the material properties, complex geometry, surface roughness, and initial radioactivity distribution of the waste prior to processing.
Second, an AI-driven predictive model, trained on empirical performance data, dynamically optimizes critical operational parameters—such as chemical concentration, ultrasonic power intensity, frequency, and duration—to establish ideal decontamination conditions in real-time.
Third, a real-time feedback loop monitors the decontamination progression to eliminate redundant cycles, automatically terminating the process upon reaching target levels to maximize throughput.
The proposed intelligent framework actively compensates for mechanical variables through AI-driven adaptability, maximizing decontamination efficiency per unit time. Moreover, the implementation of an automated system minimizes manual intervention, providing inherent safety for workers. This AI-integrated hybrid solution establishes a new technological standard for next-generation decontamination, ensuring both economic viability and safety in the global nuclear decommissioning industry.