SFuller
Chemical
- Jul 7, 2003
- 8
Hi,
I am working on scaling up a process from a research lab to pilot plant scale. The process involves continuously separating two liquid phases (aqueous and organic) that are fed as a mixed feed to a centrifugal separator. The separation has been proven to work in a lab unit.
My question is: how does a liquid centrifugal separation scale up? The local vendor maintains that I should keep the product of the g-force at the wall of the machine and the residence time in the machine constant for scale-up. This does not take into account the increased distance that liquid drops in the machine have to travel however so I would expect that I should divide by the radius of the machine to take that into account. So for scale up
(G1 . T1 / R1 ) = (G2 . T2 / R2) where
G is G-force, T is residence time and R is bowl radius. The vendor maintains that the radius is not a factor. Essentially I am using the angular velocity (rad/s) instead of the G-force as a scale-up factor.
Has anyone else experience of this?
Regards,
Steve.
I am working on scaling up a process from a research lab to pilot plant scale. The process involves continuously separating two liquid phases (aqueous and organic) that are fed as a mixed feed to a centrifugal separator. The separation has been proven to work in a lab unit.
My question is: how does a liquid centrifugal separation scale up? The local vendor maintains that I should keep the product of the g-force at the wall of the machine and the residence time in the machine constant for scale-up. This does not take into account the increased distance that liquid drops in the machine have to travel however so I would expect that I should divide by the radius of the machine to take that into account. So for scale up
(G1 . T1 / R1 ) = (G2 . T2 / R2) where
G is G-force, T is residence time and R is bowl radius. The vendor maintains that the radius is not a factor. Essentially I am using the angular velocity (rad/s) instead of the G-force as a scale-up factor.
Has anyone else experience of this?
Regards,
Steve.