Block censoring scheme with two-parameter exponential distribution

This paper deals with a new censoring scheme, called ‘Block Censoring’ which reduces considerably the total time on test in the life testing experiments with respect to the common used experimental tests such as rightly censored data. This new scheme is analysed when the lifetimes of products follow the two-parameter exponential distribution. Specially, it is proved that the respective spacings are independently distributed exponential. The problem of estimating parameters is investigated in details.A Monte Carlo simulation is conducted for obtaining the optimal block censoring scheme in the sense of the shortest expected test time. Finally, a real data set on times to breakdown of an insulating fluid between electrodes from Nelson [Applied life data analysis. New York: Wiley; 1982. p.105] is analysed.