A New cross section adjustment method of removing systematic errors in fast reactors

Takeda, Toshikazu*; Yokoyama, Kenji   ; Sugino, Kazuteru

A new cross section adjustment method has been derived in which systematic errors in measured data and calculated results of neutronics characteristics are estimated and removed in the adjustment. Bias factors which are the ratio between measured data and calculated results are used to estimate systematic errors. The difference of the bias factors from unity is caused generally by systematic errors and stochastic errors. Therefore by determining whether the difference is within the total stochastic errors of measurements and calculations, systematic errors are estimated. Since stochastic errors are determined for individual confidence levels, systematic errors are also dependent to the confidence levels. The method has been applied to cross section adjustments using 589 measured data obtained from fast critical assemblies and fast reactors. The adjustments results are compared with those of the conventional adjustment method. Also the effect of the confidence level to the adjusted cross sections is discussed.