A computer-based simulation methodology of the predetermined maintenance scheme of an irradiation facility

Abstract

This study presents a computational framework for maintenance modelling aimed at addressing the complexities of the test cell (TC) within the IFMIF DONES, a complex industrial facility in the realm of fusion materials irradiation and testing. The proposed framework, which provides an insight into the maintenance process, is based on high-level Petri nets and captures maintenance tasks, duration, delays, and interactions among components. The proposed method employs global sensitivity analysis (GSA) to gain a better understanding of the influence of the numerous parameters on the system’s performance. Three maintenance scenarios are explored, taking into account factors such as workforce assumptions, shift expenses, delays, and task completion timings. Monte Carlo simulations evaluate the probabilistic behaviour of the considered maintenance scenarios to allow for a thorough examination of the impact of uncertainties on maintenance operations. Examination of shift data yields insights into optimizing maintenance strategies minimizing downtime, and enhancing cost effectiveness. The results underscore the significance of implementing a night shift to improve facility availability. Furthermore, the proposed maintenance model is compared against a regression metamodel, which is validated by comparing Sobol indices derived from the Petri net. The novelty of this research is shown through combining shift data analysis, global sensitivity analysis, and innovative hybrid modelling aimed at enhancing maintenance planning within an irradiation facility.