I strongly recomend you to perform a process simulation of your system, using any of the available tools in the market, in order that you could have a better understanding of operational safety ranges before you go through an equipment sizing. Entrainer recovery wil always be impacting your operation direct cost if not proper recovered. First you'll be facing a problem of solubilty to ensure proper separation temperature at overheads composition, proper temperature should be achieved before decanting. Second you'll be facing separation problem based on gravity, density differential, and as far as I remember also surface tension.

Similar mixtures, as yours described, can be decanted by conventional gravity settling vessels, where separation can be improved by colaescer devices such as internal packing mesh, or candle type mesh. Desicion will be a matter of separation efficiency and economics. A small scale lab testing is recommended is you're in the intention of selecting internals to promote separation, in order that you do not get short on design. Important consideration on design is also incondensables presence in the stream.

After decanting you'll be in the requirement to recover the portion of the entrainer(butanol/butyl acetate) that is carried on heavy phase (water) in a separated column, and also you'll be in the requirement on purging a portion of the light phase stream to remove butyl acrylate or lighter components, that eventually will accumulate on overheads and break azeotrope composition at column, before it is refluxed. If incondensables are present you´ll be in the requirement of another cooler and a scrubber prior to venting to your fume collector.

A couple of points.

There is a good write up on the design of continuous decanters in the Handbook of Separation Techniques.

Basically, you want to slow the velocity of each phase down to a Re less than the transition region.  This allows for the dispersed phase to migrate to the bulk.  The parameters that matter include all those needed to calculate a Re, settling velocity and coalescence time.  Typically you pick a droplet size to suit your needs and perform the calc.

You have a genuine opportunity to understand the rudimentary factors of this design.  Work out the details on paper or on a spreadsheet then confirm with a simulation.  

Calculations of settling velocity upon which vessel size is based assume laminar flow in the settling region, a premise which is usually far from accurate.  Even when Re calculated from vessel dimensions, liquid depth, and flow rates is <2000, reality is that the geometry with which the 3-phase condensate stream enters the vessel can have the effect of an outboard motor on the liquid in the vessel.  Key element to success is an inlet distributor properly designed for disengagement and stilling to minimize eddies in the bulk liquid.  If the fluid system produces finely-dispersed droplets of one phase in the other, a coalescer pad is helpful.  Another trick is to provide a pack of plates parallel to the flow direction and tilted at 45 degrees so that to disengage, a droplet need rise or sink only the distance to the next higher or lower plate, where it coalesces with other droplets of that phase and slides up or down the plate to the edge until it reaches the layer of the same phase.  Plates are preferably made of a material that is wetted by the phase whose droplets are the limit on separation.  By reducing the float or sink distance necessary, the plate pack can reduce the vessel size required for a target separation by a factor of several.