Instabilities of burnup calculations in a SERPENT/OPENFOAM coupling

Nowadays, different multi-physics approaches are available, in order to improve the reactor safety analysis and the design of different types of nuclear systems. Moreover, these methods allow the investigation of physical effects at different scales of time and space. On the other hand, multi-physics simulations could improve the fuel management and estimate the amount of long-lived radionuclides in spent nuclear fuel for different types of nuclear reactors. In this context, the development of burnup analysis with the thermal-hydraulics feedback in Monte Carlo codes, that assures accuracy and flexible implementation, is a challenging task. In particular, burnup calculations could present numerical instabilities and their evaluation is a mandatory for the reliability of the burnup solution. In this paper, we discuss the effects of numerical instabilities on fuel depletion in the Serpent code, for simulations both with uniform fields and density/temperature distributions obtained by an external solver in OpenFOAM. We carry out a burnup analysis for 10 axial zones, simulating the entire fuel cycle of a simplified fuel cell, composed by an UO2 pin surrounded by water. Particularly, we will show the instabilities on neutron fluxes and nuclide densities obtained by the predictor-corrector scheme. The issue is fixed by the adoption of the Stochastic Implicit Euler (SIE) method, that is unconditionally stable.