Power calculation for mixing tank

Hu

I want to calculate the power input for the liquid in the mixing tank after using CFX. From the literature, I have found three different ways to calculate the power input.

1) CFX tutorial examples: Calculating the torques on the blades, then multiplied by the rotational speed. 2) Bird et al. Transfer Phenomena: Calculating the torques on the baffles and stationary walls, then multiplied by the rotational speed. 3) Some published papers: Integrating the turbulence eddy dissipation over the whole volume then multiplied by the liquid density.

Obviously I got different results from the three different methods. The first one the values are much higher than the other two. The second one is the lowest one, as the torque on one side of baffle is positive but on the other side is negative, the total torque on the baffles is much lower than that on the blades. I am quite confused which one I should go for.

Many thanks.