Safety devices represent one of the most important issues requiring serious attention by maintenance personnel. As a rule, when they fail, they fail silently and invisibly, with potentially disastrous results. Hence, the question facing reliability engineers is “How often should the safety device be tested for functionality?” That is, what is the “Failure Finding Interval” (FFI) that will provide the desired functional availability?
Where μprotective device is the mean time between failure of a backup function or safety device and μmulitple failure is the specified mean time between multiple failure (i.e. a failure of the protected funciton at a moment of unavailability of the protective device). Moubray’s formulas for FFI, such as the one above for a single device protecting a single system are based on reasonable approximations and assumptions (in most cases). If applied appropriately within the maintenance plan, they will achieve an acceptable, documented risk of “multiple” failure (the failure of a protected function at a moment when its protective function is unavailable).
Applying these formula within a “real” maintenance context, however, challenges us beyond the limit of conventional maintenance information management practice. Living RCM (LRCM) provides the practical way to check and update assumptions as part of our normal work order procedures. By completing a work order according to LRCM principles (i.e. event type, and failure mode RCM reference), software will assist us to update the assumed values of µ (failure rate) and their underlying respective failure probability distributions and apply appropriate formula for updating FFIs in the maintenance plan.
© 2011, Murray Wiseman. All rights reserved.