Statistical methods giving evidence of a connection between fuel design concept and fuel element failure rate

The proof that changes in the fuel element concept must not result in an increased fuel element failure rate can only be furnished in respect of operating power reactors. Suitable strategies of respective investigations can be derived from methods of mathematical experiment planning. The model of binomial distribution represents an appropriate basis. Data examples are given for estimates concerning confidence limits and the extent of necessary random samples. A sequential stepwise procedure has proved to be advantageous. A modification of the sequential analysis by Wals provides a simple statistical approximation to this effect. (orig.)