Approximate queueing formulae are often employed for the
practical evaluation of manufacturing system performance.
Common approximations do not fully address practical
issues such as idle tools with work in queue, travel time
between stages of production, removal of lots from queue
pending process issue resolution and the tendency of lots to
defect from a failed server in favor of an equivalent available
server. In this paper, approximate queueing formulae
are proposed which extend popular existing formulae. To
test the quality of the proposed approximations, data from
production toolsets in IBM’s 200mm semiconductor manufacturing
fabricator is considered. It is demonstrated that
the approximations perform well on the toolsets studied.