Keff uncertainty due to uncertainties in non-isotopic parameters of fissionable systems

In the criticality safety design of fissionable systems using computer codes, uncertainties in the non-isotopic parameters other than the isotopic data should be also considered to assure there is an adequate minimum margin of subcriticality (MMS). Uncertainties in the non-isotopic parameters and isotopic data can produce an uncertainty in the neutron distribution, and thus in the calculated value of neutron multiplication factor (keff). However, most studies have been limited to effects of uncertainties in isotopic data (e.g. cross sections) on calculated keff values whereas the keff uncertainty due to uncertainties in non-isotopic parameters (e.g. spatial distribution of isotopes in the system) remains largely unstudied. Here we discuss the Monte Carlo based methods for the analysis of keff uncertainty due to uncertainties in non-isotopic parameters. An uncertainty analysis on 30B UF6 cylinder storage array is performed to illustrate the independent forward Monte Carlo method. In the independent forward Monte Carlo method, the non-isotopic parameters (such as material densities and compositions, enrichment, geometries and etc.) are sampled from given probability density functions of parameters. The resulting uncertainty in keff is determined by the explicit uncertainty propagation through the forward Monte Carlo neutron transport calculation. Unlike other Monte Carlo keff uncertainty analysis methods, the independent forward Monte Carlo method is not limited to the degree of changes in non-isotopic parameters and can be used for both small perturbations and large variations. The illustration example shows that keff uncertainties due to uncertainties in non-isotopic parameters can be comparable to those due to uncertainties in cross sections. Therefore, an appropriate MMS should take into account keff uncertainties due to uncertainties in non-isotopic parameters to the best available knowledge when more realistic modeling of fissionable systems or processes is needed. (authors)