Optimization Control of Void Effect in Axially Heterogeneous Sodium-cooled Fast Reactors Based on Non-dominated Sorting Genetic Algorithm

The sodium-cooled fast reactors (SFRs) represent a significant option among Generation Ⅳ reactors, offering potential benefits such as nuclear fuel breeding, improved uranium utilization, and the transmutation of high-level radioactive waste. In SFRs, sodium voiding decreases neutron absorption and hardens the neutron energy spectrum, resulting in positive reactivity. At the same time, the increased neutron mean free path leads to greater neutron leakage, which contributes negative reactivity. The net sodium void effect can be controlled by balancing these competing mechanisms. Consequently, axially heterogeneous core designs play a crucial role in influencing neutron leakage and are key to mitigating positive sodium void reactivity. Furthermore, sustainable SFR development must address challenges related to nuclear fuel fabrication and waste management, requiring flexible breeding capabilities at various stages. The breeding ratio directly impacts fuel utilization efficiency and neutron economy, making its optimization essential for the long-term viability of these reactors. In SFRs, a trade-off exists between controlling sodium void worth and adjusting the breeding ratio, both of which are critical objectives. This study develops an optimization method coupling neutron transport calculations with a genetic algorithm to design axially heterogeneous core configurations, applied to a 1 000 MWth metal-fueled SFR. The sodium void worth and breeding ratio were selected as target parameters for optimization. The NSGA-Ⅱ algorithm, implemented via Python’s Geatpy library, was employed to encode the geometric parameters of the core fuel zone. A neutronics-genetic algorithm framework was established to explore the design space, evaluating core physics performance based on breeding ratio and sodium void worth. The genetic algorithm iteratively searched for Pareto-optimal solutions within the MET-1000 benchmark core geometry, aiming to regulate sodium void effects while enabling flexible tuning of breeding performance. The Pareto front, representing optimal solutions where no objective can be improved without compromising another, was successfully derived with rapid convergence. Quantitative analysis of the Pareto front elucidated the inherent trade-off between sodium void worth and breeding ratio. Axial heterogeneity enabled modulation of sodium void worth (−119.7-2144.1 pcm) and breeding ratio (1.14-1.42). Spatial analysis reveals that the reduction in sodium void worth primarily results from increased neutron leakage near the sodium plenum, necessitating the positioning of fissile zones closer to these regions. Breeding performance is regulated through the axial arrangement of fertile and fissile zones. However, higher breeding ratios require increased fuel enrichment, which negatively impacts the doubling time. Compared to the MET-3000, MET-1000 exhibits narrower breeding ratio optimization ranges due to core size limitations, necessitating further design refinements to enhance breeding capability.