Abstract
Accurate measurement of power numbers on lab scale reactors can provide useful information about the needed energy to reduce mass transfer problems. Next, this knowledge can be used to avoid scale-up problems (Paul et al., 2004). In this study, the accuracy and applicability of four methods to determine the power number in a 1 L reactor are tested: calorimetry, torque measurement, computational fluid dynamics (CFD) and the Furukawa et al. correlation (Furukawa et al., 2012). Experiments with water and 2-octanol are performed with a 4-bladed 45 ° pitched blade turbine (4PBT) and 6-bladed Rushton turbine in baffled and unbaffled conditions. At low rotational speed, experimental techniques record higher power numbers compared to the theoretical techniques. The calorimetric method is only accurate at rotational speeds above 300 RPM due to insufficient heat flow at lower settings. Torque measurement is most accurate at low rotational speed, before the vortex reaches the stirrer and creates cavities that lead to inaccurate results. At rotational speeds higher than 300 RPM, differences between the four techniques are less than 30 %, confirming the accuracy of all methods. Therefore, literature correlations can be used for a quick estimation of the power number in lab scale reactor experiments at high rotational speed. However, experimental techniques are recommended for accurate power number measurements at low rotational speed and for specific reactor set-ups which are not described in literature.
