Politiques de Tests Partiels \& Syst\`emes de S\'ecurit\'e

TL;DR

This paper introduces formulas for assessing the failure probability of MooN safety systems under partial and full testing, optimizing test strategies to minimize system failure risk.

Contribution

It presents a novel approach to evaluate and optimize partial and full testing policies for MooN architectures, improving reliability assessment.

Findings

Formulas for PFD assessment of MooN systems are proposed.

Performance estimation methods for test policies are demonstrated.

Optimization of partial test distribution reduces average PFD.

Abstract

A set of general formulas is proposed for the probability of failure on demand (PFD) assessment of MooN architecture (i.e. k-out-of-n) systems subject to partial and full tests. Partial tests (e.g. visual inspections, imperfect testing) may detect only some failures, whereas owing to a full test, the system is restored to an as good as new condition. Following the proposed approach and according to an example, performance estimations of the system and test policies are presented, by using the feedback from partial and full tests. An optimization of the partial test distribution is also proposed, which allows reducing the average probability of system failure on demand (PFDavg).