A neutronic model to find promising candidates as burnable poisons in fast reactors

Small modular liquid-metal-cooled fast reactors may require dedicated provisions to reach very long core lifetime, or to reduce the reactivity swing if minimizing control requirements is needed. In such cases, the use of burnable poisons may be a necessary route. Unfortunately, to date very few materials are known that may be used as burnable poisons in fast reactors. Goal of this paper is therefore to propose a model by which the expected neutronic behavior of a candidate burnable poison material in fast spectrum can be estimated without the need for extensive neutronic calculations, for which a detailed description of the system in terms of materials and geometry would be required. The neutronic behavior of a material is estimated by considering separately its performance in terms of poisonousness, defined as the ability to provide enough negative reactivity to compensate for the initial excess reactivity of the fuel, from its performance in terms of burnability, defined as the ability to balance in time the reactivity change due to fuel depletion. Although the values calculated by the model equations do not have physical meaning per se, they can be used to compare different materials with each other in order to make a preliminary assessment of their neutronic behavior. The model was first tested with seven candidates (Eu 2 O 3 , Gd 2 O 3 , Dy 2 O 3 , Er 2 O 3 , NpO 2 , AmO 2 and B 4 C), and then assessed against the results obtained from simulations with MCNP6.1 for a lead fast reactor fuel assembly with UO 2 enriched in 235 U at 19.75 wt% in which, one at a time, the various candidates have been added. Despite some small differences from the simulations especially at low burnups, the comparison confirmed the prediction capability of the model up to a poison content of 10 at.%. • Burnable poisons in small liquid-metal-cooled fast reactors may be a necessary route. • Need to screen neutronically promising materials as burnable poisons in fast spectrum. • A neutronic model is developed that does not require extensive calculations by codes. • Material performance estimated analyzing separately poisonousness and burnability. • Model prediction capability confirmed by comparison with MCNP6.1 simulations.