Target accuracy assessment for China Experimental Fast Reactor based on subspace method

Abstract

According to the design uncertainty limit of reactor physics response calculation, target accuracy assessment solves a inequality-nonlinear-constrained optimization problem and provides the priorities of nuclear data uncertainty requirements, which is beneficial to the improvement of safety and economy of nuclear reactors. Due to the ultra-large amount of nuclear data, solving the optimization problem in full-space is improbable. Subspace method could reduce calculation space dimensions effectively and improve the stability of numerical solution, in the meantime the high-dimensional information is mostly retained. Studies on nuclear data target accuracy assessment for the first core of China Experimental Fast Reactor (CEFR) at BOC and EOC based on subspace method shows that, within the 0.3% target accuracy of effective multiplication factor, computational dimension can be reduced from over 1000 to less than 100. The most affected reactions are the scattering and radiation capture of ²³Na, ⁵²Cr, ⁵⁶Fe and ²³⁵U. The results of this paper provide a comparable target accuracy assessment for similar fast reactors utilizing enriched uranium dioxide fuel, and are helpful for communicating across pipeline from evaluator to end user.