ETL 1110-2-560
30 Jun 01
1/ n
n
∗ (%)
= ∏ λi
λ
(C-1)
merged
i=1
where
λmerged = a summary failure rate derived from several constituent data source individual failure rates
n = number of records having failures
λi = individual failure rate from each source having failures. Individual failure rates were
calculated by using a single failure divided by the total number of operating hours for that
component
% = percentage of total operating hours associated with data entries having failures to the total
operating hours of entries without failures. Percentage for Corps analysis was taken to be
100 percent since all data entries had failures and information related to the overall
population and total number of operating hours for the population is unknown
The merged failure data are shown in Table C-1.
CAUTION: The resulting merged failure rates have inherently high variability and more closely
represent worst-case failure rates. Real failure rates will be less than those presented. As more data
are gathered from occurred failures, the merged failure rates will more closely approximate the real
failure data.
C-4