Accuracy and Computational Time of a Hierarchy of Growth Rate Definitions for Breeder Reactor Fuel

For a hierarchy of four logically different definitions for calculating the asymptotic growth of fast breeder reactor fuel, an investigation is performed concerning the comparative accuracy and computational effort associated with each definition. The definition based on detailed calculation of the accumulating fuel in an expanding park of reactors asymptotically yields the most accurate value of the infinite time growth rate, γ^∞, which is used as a reference value. The computational effort involved with the park definition is very large. The definition based on the single reactor calculation of the equilibrium surplus production rate and fuel inventory gives a value for γ^∞ of comparable accuracy to the park definition and uses significantly less central processor unit (CPU) time. The third definition is based on a continuous treatment of the reactor fuel cycle for a single reactor and gives a value for γ∞ that accurately approximates the second definition. The continuous definition requires very little CPU time. The fourth definition employs the isotopic breeding worths, w_i*, for a projection of the asymptotic growth rate. The CPU time involved in this definition is practically nil if its calculation is based on the few-cycle depletion calculation normally performed for core design and critical enrichment evaluations. The small inaccuracy (≃1%) of the breeding-worth-based definition is well within the inaccuracy range that results unavoidably from other sources such as nuclear cross sections, group constants, and flux calculations. This fully justifies the use of this approach in routine calculations.