Three methods to seperate water out of oil
1) Vacuum Dehydration
2) Coalescers
3) Cetrifuge
Each of this has its own advantages and disadvantages. The main disadvantage of a VD is the energy consumption and throughput. The main advantage of VD is ultimate dryness level. Levels below saturation limit of the oil are possible which would be virtually unattainable using a coalescer or centrifuge. One thing to realise here is the fact that a VD's performance
(both eff. and energy consuption/unit time) is greater than other mechnical means (centrifuge and coalescers) only when the water ingression rate is low. Some companies experimented with a combination of Coalescer and VD by splitting the stream into two. I am not sure whether any work has been done to check the stability of the additive package at the temperatures and pressures encountered in a VD. I know that many studies have been done in the field of coalescing to look at the effect on additive package using a coalescer.