Application of the tentative guide to the expression of uncertainties to the reevaluation of a few french critical experiments.
G. Poullot, J. Anno, D. Doutriaux, ICSBEP 2000, 26-30/06/2000, St-Petersbourg (Russia).
When using a reevaluation reported in the ICSBEP Handbook, problems of different types can be encountered:
- the mixture of uncertainties on physical parameters either coming from a large number of measurements or given as tolerances,
- the lack of information about the level of confidence and degrees of freedom corresponding to reported uncertainties,
- inconsistencies in the reported level of confidence for statistical uncertainties corresponding to 1, 2 or 3 s,
·- when the uncertainty on a physical parameter is converted in reactivity equivalent uncertainty, the total experimental uncertainty is given without mentioning the level of confidence.
A former presentation (Reference 1) at the ICSBEP meeting gave general guidelines on the uncertainty treatment in the reevaluation process by the participants of the ICSBEP working group. This paper applies these guidelines to a few uncertainty calculations performed with the French Calculation Scheme named CRISTAL used at the Institute of Protection and Nuclear Safety, and gives the names of codes currently used.
Appendix A presents a glossary of metrology and associated statistics terms currently used in this report, selected from References 2 and 3.
Since most of the lattice experiments used the same type of fuel, the Appendix B is devoted to the derivation of uncertainties for UO2 fuel rods.
Other Appendices give numerical applications of uncertainty calculation for a selected set of experiments.