Improved criticality search techniques for low- and high-enriched systems

The purpose of this work is to develop a parameter search capability for use in criticality safety analyses of nuclear fuel shipping containers. This new search capability is being developed for use in the SCALE computer code system. SCALE is being developed for licensing evaluation of various transportation package designs. It can perform criticality safety analysis, shielding analysis, and heat transfer analysis on systems described in multi-dimensional geometry.

The work described in this report is specifically concerned with the criticality safety analysis capability of SCALE, and particularly, with two sequences that will provide SCALE users with the option to perform criticality searches on systems described in either one-dimensional or three-dimensional geometries.

This search package includes a method for an optimization search (maxima or minima) and a method for a fixed value search. The optimization search is a search for the most reactive state of a system. It searches for the value of a parameter, such as fuel assembly lattice pitch or fuel concentration, that corresponds to the maximum K-effective of the system. Similarly, the fixed value search is a search for the value of a parameter that corresponds to a fixed value of K-effective. Both searches rely on least squares curve fitting to obtain the information needed to make an estimate. Specifically, the data points are least squares fitted to a cubic polynomial which is solved analytically for parameter values that correspond to maximum or fixed values of K-effective.

iii

iv This new search capability has been implemented into a FORTRAN computer routine named OpTMIZ which is used to perform the parameter modifications in SCALE's two criticality search options. Results obtained from OPTMIz show that fixed value criticality searching is greatly improved with regard to accuracy and computing cost relative to the existing search capability available in KENO-IV. Also, results show that the new optimum search capability available in OPTMIz is an accurate and reliable search for locating parameter values that correspond to maximum values of K-effective.