High-fidelity multiphysics modeling of pulsed reactor heat generation in the Annular Core Research Reactor fuel using Serpent 2

Abstract

Sandia National Laboratories’ Annular Core Research Reactor is a unique pulsed reactor using UO${}_2$-BeO fuel. Unlike TRIGA reactors, which often operate in steady-state with infrequent pulsing, the ACRR operates almost exclusively in pulsed mode. Throughout the last 15 years, computational reactor physics analyses of the ACRR have involved either kinetic simulations with reduced physics approaches to the neutron distribution or a detailed Monte Carlo criticality simulation in steady-state. This paper presents an effort to perform time-dependent Monte Carlo modeling of pulsed operation of the ACRR for the purpose of determining the heat generation in the fuel. This time and space-dependent volumetric heat generation will serve as a source term for future analysis of the physical characteristics of the fuel after many pulses to inform decisions about ACRR operation and the future design of ACRR-like reactors. The Serpent Monte Carlo code is coupled to a simple Python script providing updated fuel temperatures after each time step, allowing on-the-fly adjustment of fuel cross sections. Beginning with a simple pin of UO${}_2$-BeO fuel, models increase in complexity to the full model of the ACRR. Results from the full model of the ACRR are compared to experimental results, and computational efficiency and heat generation plots are discussed.