Considering confidence levels in reliability simulation

The proposed procedure is able to include confidence bounds of the design project's input parameters into simulation. The input is broken down into its descriptive parameters whose population is calculated. Random discrete values are drawn and combined again. A number of random samples are drawn. This is repeated for all input parameters of interest. The samples are combined and integrated into an adequate lifetime model resulting in a multitude of lifetime curves which can be interpreted with confidence bounds. Thus, the variance of the Bq-life can be expressed by means of simulation with respect to its scattering input parameters. The procedure allows the combination of additional information from different sources (e.g. testing, expert knowledge etc.) on the kind of distribution, the distributions' parameters etc.