Residence time and mixing time predictions were investigated using two hybrid RANS-LES models (SBES and SST-SAS) and compared with experimental data and URANS approaches for the case study of a laboratory-scale draft tube reactor. The URANS models are found to over-predict the mixing time by a factor of ~2× and to poorly predict the exit residence time behaviour. In contrast, the two scale-resolving models are found to produce similar predictions which are in good agreement with the experimental data for the mixing time and the shape of the average exit residence time curve. The use of multiple tracers in the scale-resolving simulations shows that both models can also capture the range of residence time behaviours observed in the experiment. The results achieved are considered applicable to a broad range of flows and the simulation times required to achieve good predictions with the scale-resolving models are considered practical for industrial simulation.