Program - Brookhaven National Laboratory

Development of the Fission-Product Yield Library for BROND-3
A.B. Koldobsky, V.M. Zhivun
The National Research Nuclear University “MEPhI,” Moscow, Russia
A.I. Blokhin
A.I. Leypunsky Institute of Physics and Power Engineering, Bondarenko sq., 1, 249033 Obninsk, Russia
The importance of the fission-product yield data induced by neutrons and photons was widely stressed.
Practically all national evaluated data libraries contain the sublibraries with these data. In this report
we present the development of the fission-product yield files for BROND-3/FPYLD. BROND-3/FPYLD
library contains independent and cumulative fission product yields for 31 fissile nuclei (Th-228 - Cm-248)
at more than 10 neutron incident energies and for 3 fissile nuclei (U-235, U-238, Pu-239) at 20 photon
incident energies. The intercomparison of the independent fission-product yield data from the ENDF/B-
VII.0, JEFF-3.1.1 evaluated nuclear data libraries with the our data were performed.
Corresponding author: A.I. Blokhin
Session KD Fission Yields
Wednesday March 6, 2013
Room: Empire West at 1:30 PM
KD 1 1:30 PM
Uncertainty Propagation of Fission Product Yield Data in Spent Fuel Inventory
Calculations
Robert W. Mills
UK National Nuclear Laboratory
Uncertainty propagation in complex computational problems can be carried out by the random perturbation
of input data sampled from a probability distribution derived from the evaluated mean and standard
deviation of each datum input into the computer code with subsequent analysis of the results giving the
distribution of the calculated parameter of interest about its mean. However, this basic approach ignores
the correlations between the data. For neutron cross-sections it is currently possible to use a ”Total Monte
Carlo” approach to consider those correlations resulting from the processes of evaluation, processing and
application solution [1], but for correlations resulting from experimental measurements and for other types
of data, whose evaluation method is not yet fully automated, alternative methods are required. The ACAB
inventory code [2] is being developed to handle such correlations but require the data covariance matrices.
This work considers the generation of such covariance matrices for fission product yield data. Spent fuel
inventory calculations use independent fission product yields as an input. These yields are derived from the
large data set of published measurements including mostly chain and cumulative yields with relatively few
independent yield measurements. The independent yields and their uncertainties in evaluations are thus
determined from semi-empirical models with parameters fitted from the available data [3]. However, the
cumulative yields can be calculated from the independent yields and the respective nuclides’ decay data.
As more cumulative yield measurements are available than independent yields and typically have smaller
uncertainties, it is possible to use these measurements to produce covariance matrices to better estimate
the uncertainties of fission product nuclide number densities during fuel irradiation and cooling. A method
to determine these independent fission product yield covariance matrices is presented. This paper studies
the use of random perturbation of nuclear data. Firstly, the effect of different probability distribution
functions on a simple decay chain is shown and important conclusions high-lighted. Then the method is
demonstrated for a fission pulse decay heat calculation using the FISPIN code [4] with and without the
155