Safety-driven design of automation systems in nuclear facilities, the SPES use case

by Dr Giordano Lilli (LNL)

Wednesday 22 May 2024, 16:00 → 18:00 Europe/Rome

C. Villi meeting room

Automation technologies employed in critical tasks within nuclear facilities provide clear advantages in reducing staff exposure, but they also involve reliability challenges and safety implications connected with potential failure scenarios during operation. Nuclear laboratories and industrial automation sectors exhibit quite distinct approaches to safety assessment and harmonization. This talk aims to showcase how the early integration of safety in the design process might be advantageous for both reliability enhancement and risk reduction. The study takes advantage of the remote handling infrastructure that is currently being developed for the transport and storage of radioactive Target Ion Source (TIS) units within the Selective Production of Exotic Species (SPES) nuclear research facility. A semi-quantitative Probabilistic Risk Assessment (PRA) has been developed to assess severe failure scenarios that might occur during remote handling procedures. A hybrid methodology combining HAZard and Operability analysis (HAZOP) and Layer Of Protection Analysis (LOPA) systematically investigated the various nodes, determining the likelihood of failure scenarios, and evaluating their consequences. Following the identification of criticalities, the PRA proposed a number of safeguards, recommendations, and design upgrades that would increase the robustness and maintainability of key components. The evaluation and optimization of maintenance activities have been recognized as key weaknesses. To face this shortcoming, some key essential Front-End assemblies experienced a thorough redesign leading to an improved maintenance and the introduction of backup actuation features. In addition, the most critical maintenance tasks have been evaluated in an extensive experimental campaign that allowed to optimize the interventions in accordance with the As Low As Reasonably Achievable (ALARA) principles and to estimate the time required for each specific activity. In the last section, safety of automation software is discussed. Overall, the adoption of the described techniques resulted in a significant increase in the level of safety of the facility’s automation. The proposed approach can be easily extended to the design of safety-critical systems in other contexts.”