Biodiesel is a complex mixture of long-chain methyl esters. Accurate numerical study of biodiesel combustion requires detailed chemistry of large biodiesel surrogates representing realistic biodiesel fuels. However, the detailed kinetic mechanisms for large biodiesel surrogates involve large number of species and reactions, leading to extremely expensive or even infeasible computation when incorporated in engine CFD (computational fluid dynamics) calculations. In this study, the scheme of on-the-fly mechanism reduction incorporated with engine CFD code KIVA-3V is first extended to two large biodiesel surrogate mechanisms, namely methyl decanoate and methyl-9-decenoate. Detailed combustion and engine performance characterization for biodiesel fuels are enabled, and the computational intensity is significantly reduced with satisfactory accuracy. In the simulations, lower CO and NO emissions and lower engine power are observed for biodiesel surrogates. Combustion features such as early oxidation of ester groups are also well captured. This work provides the insight to study combustion and engine operation of complex fuels on the basis of detailed chemistry with, efficient mechanism reduction technique.