Water / Organic separation

[Homayun]

Hi all,

I have a question regarding separating a high hydrocarbon-alcohol (C9OH) and water mixture. The water concentration is very low, 0.3 wt%.
We have in our installation a liquid-liquid separator that works on basis of density difference of these two liquids. But it's not working that satisfactory. Our aim is to modify this vessel in order to perform better. One option is to install horizontal baffles that enhance the separation and collect the water in a boot at the bottom of the vessel. But we are not sure whether this is a good option.

I have been told that there are certian packing materials that are suitable for separaing low quantities of water from an organic liquid.
Does anybody have any experience with such a packing or internal system?

 

[lattar]

At least in small scale molecular sieves (zeolites) are used to separate small amounts of water from alcohols. Regeneration of zeolites is done simply by heating.
I think suitable zeolites are for example acidic zeolites.

You should also consider centrifugation and membane processes as possible solutions.

 

[Homayun]

Thanks Lattar,
But remember I am talking about an exisitng L/L separator that should retrofitted in order to function better. THere are no Heating mediums available and we do't want to install them.
Centrifgation will also not help.

 

[StoneCold]

Homayun
I wonder what the mixture looks like and if you are even getting phase separation at these levels. You may want to pull some samples or make some in the lab at different water concentrations. That may also give you a better feel for how to redesign the liquid liquid separator. Most alcohols are very soluble in water (vise versa) and the solubility is very temperature dependant. In my own opinion I would say that if you are getting down to 0.3wt% water that you are doing great and if you want it drier then you need to finish the job with sieves or a membrane as Latter suggested.
My two cents.

StoneCold

 

[sshep]

Hey Homayun,

We work with alcohol in this range. Water and C9OH basically don't mix, so a simple visual settling test should tell the story on whether the physical size and density of your water drops can be successful. I don't have direct experience with your system because we do an LLE mix and settle a wide range of alcohols/water/solids and then distill so any entrained aqueous phase is captured in the distillate. We do add ammonia to reduce emulsions and thereby assist phase seperation.

In general insure a horizontal settler has effective residence time to work by making sure your inlet velocity is low and not pointed to the outlet (i.e. elbow 180deg away, use deflection plate, etc). You want a nice uniform horizontal laminar flow using the whole vessel cross-section rather than local currents which can short-circuit volume in the settler. Coalescing media at the inlet will help if your settling tests show that the droplets are too small. This page by a media vendor seemed like a useful background.

http://www.acsseparations.com/coalescer_design.htm

best wishes, sshep