and functional safety concerns and it is pointless to concern oneself with the third significant figure
(or even possibly the second). I would suggest that only if an identified failure rate changes by a factor
two or more would it be appropriate to consider revisiting PFD/PFH calculations, useful life
assumptions, or equipment selection options. The failure rate of modern equipment is so low that
random failures will be relatively infrequent unless there is a large population. In practical integrity
terms, the dominant issues are likely to be systematic influences, and management and maintenance
practices. Nevertheless, it is appropriate to establish a practicable basis for identifying possible end-
of-useful-life or misplaced initial assumptions and hopefully the above will help in this regard.