A lower confidence limit for reliability of a coherent system with independent component via the CHA algorithm

In this paper, we solve the problem of finding interval estimate for system reliability via
the CHA algorithm (Chaudhuri et al., 2001) following the Easterling (1972) approach. We
consider a coherent system composed of independent components. No distributional assumption
is made for the component life times. A closed form expression for the standard
error of the system reliability, for a given mission of duration, is obtained. The method of
calculating the 100(1-\alpha) % lower confidence limit for the system reliability is illustrated
for a simple low-pressure coolant injection system (LPCI) with two pumps (Blischke and
Murthy, 2000). Both the CHA algorithm and the usual variance method are used for calculations.
Some simulation results are also reported. This paper basically extends the results
of Easterling to any coherent system.