Monte Carlo based equilibrium cycle analysis of one-dimensional breed and burn benchmark problem

As part of initial neutronics codes verification and validation (V and V) efforts for Traveling Wave Reactors (TWR) at TerraPower, the equilibrium cycle analysis of the one-dimensional breed and burn benchmark problem was performed using Monte Carlo codes (MCNP-5.1.60 and MCNP-XT), the depletion code, ORIGEN-2.2, and a set of driver Python scripts. Several goals of this work are: 1) to establish the reference solution for the benchmark problem by series of sensitivity calculations verifying both the equilibrium cycle modeling methodology and base modeling input parameters; 2) to verify the MCNP-XT transport solver against MCNP-5 for depletion problems; 3) to demonstrate the reactor shutdown length effect. The computational benchmark problem consists of 100 one-dimensional fuel slabs of 5 cm thickness each. Two fuel shuffling sequences examined are the inward convergent shuffling and the convergent divergent shuffling. The equilibrium cycle modeling methodology was developed by iterating flux calculations and depletion calculations. Investigated sensitivities include the number of depletion slices in a fuel slab, the number of depletion steps per cycle, and the depletion step setting, i.e., fluxes and reaction rates at beginning of step (BOS), end of step (EOS), or average of BOS and EOS. Furthermore, the reactivity effect of the reactor shutdown length is found to be about +100 pcm affecting only beginning of equilibrium cycle (BOEC), which is caused by the holdup of Np239 inventory in the core. (author)