A comparison between depletion calculations of a fuel assembly design by stochastic codes and deterministic codes is presented in this thesis. The calculations are based on a graphite-filled MOX fuel assembly design. The infinite multiplication factors and isotope inventory changes as the function of burnup obtained by MONTEBURNS, SCALE5.1 module TRITON/KENO V.a - stochastic method and SCALE5.1 module TRITON/NEWT - deterministic method are compared with those obtained by the HELIOS code. The calculation using MONTEBURNS is carried out with continuous energy cross section library based on ENDF/B VI, while SCALE5.1 module TRITON uses 238-group ENDF/B VI cross section library, whereas HELIOS uses 89-group ENDF/B V cross section library. The MONTEBURNS results show that the average absolute difference as the function of burnup is less than 0.26% in the eigenvalue with respect to SCALE5.1 module TRITON/KENO V.a and less than 0.32% in the eigenvalue with respect to SCALE5.1 module TRITON/NEWT. The uranium and plutonium depletion rates calculated by MONTEBURNS and SCALE module TRITON/KENO V.a and SCALE module TRITON/NEWT have quite good agreement. However, the absolute difference in the initial multiplication factor between MONTEBURNS results and HELIOS results is quite large, around 1.45% and the isotope inventory changes showed quite differently at the later burnup steps. The cross section library difference (continuous and multi-group energy cross section) and the different decay chains are found to be the reasons which cause that discrepancy between stochastic and deterministic methods.
Additionally, the Monte Carlo depletion calculation with leakage corrected critical spectrum was performed and its results are compared with those of deterministic depletion calculation with critical buckling search. Although critical spectra are obtained using two different methods, the isotope inventory changes from MONTEBURNS leakage correction calculation and SCALE module TR...