Development of a design method for burnable poisons in small modular fast reactors

• Develop a fast prediction method for burnable poison assemblies to evaluate the compensation ability of different designs. • Establish the Monte Carlo two-step method, which has good accuracy in the calculation of burnup reactivity loss. • Use Genetic algorithm coupled with Monte Carlo program to optimize burnable poison assembly arrangements in core. Long-cycle small modular fast reactors have high excess reactivity. The use of burnable poison can compensate for excess reactivity and improve core safety. Burnable poison design of fast reactors is a complicated problem, the empirical design method is time-consuming and does not guarantee optimal compensation. In this study, a design method for burnable poison in fast reactors is developed. Firstly, a fast prediction method of compensation ability is formed to optimize poison assembly, the evaluation of optimal concentration in a specific geometry and compensation in different geometries can be advanced. Secondly, a multi-group Monte Carlo depletion method is formed to accelerate calculation, and the optimization of poison arrangement was carried out based on a genetic algorithm. Optimization reduced core burnup reactivity loss from 10,043 pcm to 3841 pcm in 3.5 years, and the residual penalty is 301 pcm. The design method can provide a reference for the subsequent design of burnable poison in a fast reactor.