Flocculation is an operation of significant industrial relevance commonly encountered in many processes, including water and wastewater treatment. The physico-chemical phenomena of this process is strongly affected by the magnitude of the velocity gradients generated, typically through agitation, in rapid mix devices and flocculation vessels.

In this work the fluid dynamic characteristics of mechanically agitated systems, namely three different types of stirred tanks, which can be used as flocculation vessels, were studied. Both the mean and fluctuating velocities in all three directions were measured by using a Laser-Doppler Velocimeter (LDV). The velocity distribution, fluctuating velocities, power consumption and local velocity gradient were numerically predicted with FLUENT, a computational fluid dynamic (CFD) software, using k-e model, algebraic stress model (ASM), or Reynolds Stress Model (RSM) to simulate turbulence effect. The experimentally obtained mean velocities and turbulent kinetic energies on the top and bottom horizontal surfaces of the region swept by the impeller were used as boundary conditions in the simulations.

Significant agreement between the experimental data and the numerical predictions of the three dimensional velocities and turbulent kinetic energies was obtained in all cases.

A novel approach to numerically calculate the local velocity gradient (G) in turbulent flocculation tanks was developed. The distribution of local G values in three mixing systems was mapped through two new methods: the complete definition of local velocity gradient method and the local energy dissipation method. Results show that both methods can provide similar information about the local G value distribution. The trajectory of a solid particle with physical properties similar to those to a floc particle moving in three different mixing systems was numerically determined. The G value experienced by the particle as a function of time was also determined. A new parameter, the velocity gradient-time integral along a particle trajectory, was proposed and calculated. It is expected that the approach developed in this study will provide the foundations for a more accurate characterization of the fluid dynamics of flocculation systems.