Decanter Design

[tusk334]

Can anyone please advise on how to design a decanter? The mole fractions of the three components(ethyl acetate, toluene, water) are known. However, the appropriate steps needed to determine the two liquid phases inside the decanter is not known.

 

[hduncan]

Sounds like your designing a liquid/liquid separator. You don't have to know a whole lot to do this, but you need to know how each phase is going to split to get your physical properties. Chemical Process Equipment by Walas is a good book that covers what you need. You will have to figure out how much residence time you want to allow to make sure that you get a good separation. You will also need to calculate the reynolds number for each phase.

 

[25362]

If you'd expect to have two phases with EA partitioned between water and toluene (lighter phase) the typical approach is to use a horizontal drum.
Sometimes electrostatic precipitation is used for coalescing of water being a polar molecule.

The volume of the decanter would be indicated, not only to provide sufficient time for gravity settling to take place but also for sufficient surge volume to remove uncertainties when feeding other units from this drum.

For water settling out of toluene the settling rate may be 1 ft/min. If a drop of water has to fall through a layer of hydrocarbon 3 ft deep and coalesce into the bulk of the water phase, it might take about 3-6 min.

A reasonable time for water is 5 min. The settling rate will depend on the viscosity of the hydrocarbon layer (affected by temperature), the lower the viscosity the quicker the settling time. The formula normally used for the settling time, t, in minutes is
t = 3 visc/sp.gr difference
visc measured in cP (for the the predominant continuous liquid phase) and sp. gr. difference in densities in g/cm[sup]3[/sup] between the two phases.

Since water would settle and run along the bottom of the drum, its removal is sometimes helped out by using a draw-off boot, especially when the this phase is in relative small quantities.

The total time for sizing is the sum of both, the settling time and the surge time.

For sizing of a drum containing two liquid phases, you may find useful information in More on design of reflux drums by B. Sigales on the Chem. Engineering issue of Sept. 29, 1975, page 87+.