Expected Bayesian credible limit of reliability parameters in the case of zero-failure data for exponential distribution

E-Bayesian credible limit estimations of failure rate, mean life and reliability based on four different prior distributions of hyper parameter are given for zero-failure data from products with exponential distribution. Relations among E-Bayesian credible limits and classical confidence limit are discussed. The results show that E-Bayesian credible limits are superior to the classical confidence limit. Finally, calculation is performed according to practical problem, and the analysis of estimation result is carried on.