Determining residence time distributions in oscillatory baffled reactors: A comparison between experiments and CFD-simulations
Matthias Adams, Georgios D. Stefanidis, Tom Van Gerven
Abstract
Oscillatory flow reactors are a process intensification technology that aims at enabling plug flow-like operation for inherently slow processes. However, the plug flow regime is usually challenging to obtain in practice. To that end, this work focuses on simulating residence time distributions in oscillatory baffled reactors with for the first time an experimental validation of the simulated results. Four different physical models are implemented to simulate residence time distributions: 2D-axisymmetric laminar, 2D-axisymmetric turbulent κ - ϵ , 3D laminar and 3D turbulent κ - ϵ . Different flow conditions are tested ranging between 50 and 250 for the net Reynolds number and 50 and 300 for the oscillatory Reynolds number. The comparison between experiments and simulations is done qualitatively and quantitatively. The quantitative parameters include: the mean residence time, the root mean square error and the number of ideal tanks in series. The results show that in almost all tested flow conditions 3D laminar physics are necessary to predict the experimental residence time distribution. However, for the tested flow parameters, literature advises in general to use 2D-axisymmetric laminar physics to model local flow patterns. This demonstrates the need for new guidelines to model global parameters in oscillatory baffled reactors and the importance of experimental validation. • 3D laminar physics simulate residence time distribution in OBR for R e O < 300 , R e n e t < 150 . • Jetting occurs with dominant net flow increasing the effects of experimental imperfections. • The mean residence time is not a reliable parameter for experimental validation.