Gary''s Student <GarysStud...@discussions.microsoft.com> wrote...

...

That's for a memoryless process, i.e., one in which an item's mean
time till failure is independent of it's age. If mean time till
failure and age are anticorrelated (older items more likely to fail
sooner than newer items), this won't be the case. However, the OP
hasn't provided sufficient detail to provide an answer if that's the
case. For instance, handling items that are older than their life
expectancy depends on the hazard function.