2010 International Conference on Nuclear Data for Science and Technology(ND2010), Jeju, Korea, April 26-30, 2010
A new Monte Carlo (MC) formulation aimed at quantifying effects of both statistical and nuclear
cross section uncertainties on neutronics design parameters on nuclear systems is presented. The
effectiveness of the formulation is examined through computing the contribution of the cross section
uncertainties to the variance of the MC estimate on keff of two fast spectrum criticality benchmark
problems; GODIVA and JEZEBEL using McCARD as a MC neutronics code and JENDL 3.3
for needed cross section and covariance data. It is shown that the MC formulation is capable of
identifying the nuclide and the type of the cross sections that contribute significantly to, and to
elicit the necessary improvement in the existing covariance data to reduce, the uncertainties of the
criticality calculations.
