5 replies to this topic

[Fr3dd]

Hello Everyone...

Currently i'm working in the design of an Amine Regeneration facility of a Refinery, and i'm having a hard time trying to design an Amine flash drum.

Here´s the thing... this amine flash drum will be able to handle about 870,000 lb/h of a rich amine solution (the lean amine solution proportion is 60% amine 40% water). This drum will be horizontal (bucket and weir type). Obviously, the rate of hydrocarbons expected in normal operating conditions is very low, but, in order to complete the design an asumption of 0.1% wt. of hydrocarbon is suposed (Several international practices support this assumption).

I have read some material about the design of this kind of equipment, and, as usual in oil/water separation settlers, the Stoke's Law is very important for the design. The drum must be large enough to guarantee the rise of the oil droplets thus allowing and efficient Oil/solution separation. Also, an oil droplet diameter of 0.005 inches is suggested for Stoke's Law calculations if no other information about this diameter is available.

However, after the sizing under these considerations, the resulting equipment is much larger than expected. I've noticed that the assumed oil drop diameter has a huge influence in the drum size. I've been searching on the web for technical papers or other information about this matter, but I haven't found anything yet.

I'll apreciate all the help you can bring me on this. If you have any questions, or need more specific information, just ask me.

Thanks in advance.

[Zauberberg]

Usually, the mixed liquid phase residence time of 5 minutes will be sufficient for proper hydrocarbon/amine separation. GPSA Databook recommends 3-10 minutes, based on my experience in three plants 5 minute is good enough. Some more rigorous standards recommend maximum axial velocity of liquid phase inside the vessel as 15 mm/sec.

Also, looking from a practical perspective, why would you need a perfect separation inside the flash drum? It doesn't result in the optimum process/economical design solution. If you look at the operating facts, there is no harm of having some liquid hydrocarbons entering regenerator (the focus is on word "some" - not "significant quantity") as they would usually end up with the overhead acid gas. No big deal.

[Dacs]

Can't say much at the moment but in my experience, I've seen three-phase separators with some sort of coalescer inside the flash drum.

Maybe you should look into it.

[Erwin APRIANDI]

Dear Fr33d,

Based on my experience in designing amine sweetening unit,

The size of the flash drum is determined by two criteria:
• To provide residence time for droplets of heavy hydrocarbons to settle under gravity from
the rich amine.
• To provide buffer storage of rich amine prior to entering the regenerator.

While steps should be taken to reduce the ingress of hydrocarbon into the rich amine, even in
well run systems it is inevitable that the rich amine will become contaminated with some
hydrocarbons. The rich amine flash drum relies on gravity settling to achieve phase
separation. This takes place in accordance with Stokes Law for lamina flow. The following
graph illustrates the settling time needed for C5 hydrocarbon in 25% DEA for a range of
droplet sizes. It is considered that using a drop size of between 125 to 150 microns is a realistic
basis for good hydrocarbon/amine separation, for which a residence time of 30 minutes is
needed. Thirty minutes also provides adequate buffer storage to minimise fluctuations in
flowrate and composition of rich amine being fed to the regenerator.

But you have to make a consideration based on case to case basis and you may reduce it if you
installed any internal to the flash drum to accelerate the separation of rich amine solution and oil

It also should be noted that C5 is the lightest and lowest density hydrocarbon that will be present in
the liquid phase and that heavier hydrocarbons with a density closer to that of rich amine will
take longer to separate. When more expensive tertiary proprietary amines are used (such as
MDEA formulations) it is even more important to ensure adequate residence time, because any
loss of amine due to regenerator upsets will be more costly.

[Art Montemayor]

Erwin:

You have left out the first (if not, the most important) reason for having an amine flash drum in a high pressure amine acid gas removal unit: the need reduce the excessive high pressure that could enter a very low pressure (relatively atmospheric) amine stripper/reboiler as well as eliminating the hydrocarbon gases and liquids that are dissolved in the amine due to the relative high pressure in the acid gas absorber.

The fact that heavy and liquid hydrocarbons are separated in the Flash Drum as well and that the Drum serves as a surge and capacitance vessel are all a consequence of carrying out the first priority: removal of dissolved hydrocarbon gases that would impose a heavy and needless burden on the acid gas stripper.

I agree with your remaining explanation of the reasoning and design of the Flash Drum. My experience also agrees with your 30 minutes of residence time – I have used this figure in the past as my nominal time. My design basis is for a longer duration and, as you explain, may vary from case-to-case. Each design case should be studied and designed on its own merits and needs.

[Erwin APRIANDI]

Thank you Art,

Nice to have a reply from the real Expert as you,
I'll keep that noted in my notebook

cheers