Speaker
Description
Reliable machine protection is critical for large-scale particle accelerators, as critical failures can cause severe equipment damage and long operational delays. For the LHC, an unprotected magnet quench could cause months of downtime. Ensuring safe operation while maintaining high machine availability requires a dedicated reliability assurance process. This includes qualitative hazard analysis to identify failure modes, their system-level impacts, and eliminate design flaws such as single points of failure. Quantitative reliability models, incorporating component failure rates, demand rates, and planned maintenance intervals, are then used to verify that protection functions meet defined reliability targets over the intended operational lifetime.
To demonstrate the applicability to accelerator machine protection systems, this process is applied to a part of the LHC Quench Detection System, responsible for protecting thousands of superconducting magnets. Two candidate architectures with different levels of power-supply redundancy are benchmarked. Single points of failure are eliminated using the qualitative analysis. Adequate monitoring and inspection policies to meet target reliability metrics are derived with the quantitative analysis, establishing a relationship between power-supply redundancy and required inspection interval. The presented methodology can be generalised to a variety of applications in reliability assurance, spanning protection systems both within and beyond the accelerator domain.