merox caustic carryover

[rogered]

Hello everybody

My question is the following:

We have a merox process for treating naphtha. It has a coalex demister on top of the extractor column to remove caustic from the hydrocarbon stream. Upstream of the column is a naphtha splitter reflux drum and pump. Downstream is storage and export facilities. Currently there's no caustic carryover equipment to deal with a blip in the hydrocarbon flowrate. It seems like a good idea to put in some equipment to deal with this transient condition. Looking at previous projects it seems that a caustic holdup drum (but no sand settler)is typically installed. There's a range of holdups of 3 minutes and 20 minutes.

Has anyone any experience of the requirement for such equipment and if so which of the 2 (3mins or 20mins) looks more reasonable, and what is the basis for this holdup?

many thanks

 

[SeanB]

Actually what you are referring to is retention time and not hold up - they are two different things. It is quite common to have a caustic settler and sometimes even a water wash drum to remove any residual caustic. You need to know the average particle diameter of your caustic droplets to determine your terminal velcity. From there you calculate the diameter of the vessel required to give you the required cross sectional velocity. From there and your liquid flowrate you have your retention time.

 

[rogered]

Sean B

Many thanks for post - I take your point about retention time.
However the size of particles, by which I mean the size of the water particles in the separator/holdup drum outlet, is unknown. Any ideas of what to take for this?

A further complication/question - In the licensor (UOP) proposal they suggest (on our insistence because they didn't propose additional equipment items) holdup drum plus sand settler. I assume that crude separation occurs in the holdup drum and remaining water droplets are coalesced in the downstream sand settler. I also assume (have asked the licensor this question) that the first drum takes slugs of caustic but has no continuous level control- this assumes that slugs occur but that the operators intervene and regain control of the process (ie. stop the disturbance that caused the blip in the first place).