Flash Drum calculations

[Ron74]

Hi Everyone,

I need to design a small scale flash drum for pilot plant, I need to remove methanol residues from paraffin wax. How to do this calculation when I cannot find any VLE graph for paraffin/Methanol.

Thanks,
Ron

 

[Latexman]

Hi Ron,

Paraffin wax is about a C₂₄H₅₀ alkane. Use ideal behavior. The huge difference in relative volatility should make it an easy separation.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[Ron74]

Thank you Latexman,
I know the separation should be easy, but I need numbers for the pilot design, and I am actually dealing with a mixture of C30-34, and some residue of C14-18.
The truth is that I am very rusty with those calculations, can you direct me to some example?
Thanks,
Ron

 

[Latexman]

Look in McCabe & Smith and C. Judson King. They have plenty of examples in them. Probably none matching your situation tho. Do you have access to a simulator, like Aspen? That would make child's play out of this for you.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[Ron74]

I wish I did, that what made me forget some of those basic calculations.
Another issue with ASPEN is it lacks the database for heavier hydrocarbons.

 

[georgeverghese]

Paraffins are basically straight chain alkanes, so look for alkanes in the C15-C30 range on your simulator's chemical component database.
You can make a good stab at this by manual calcs for this adiabatic flash, assuming methanol solubility in high mol wt alkanes is next to zero. All you need is the flash drum pressure, feed temp / press, and feed flow, and VLE-thermo props for methanol and Cp value for the paraffin mix.

 

[apetri]

you may adopt a ideal model (as suggested by others) or fit some exp value (in literature you may find VLE data up to methanol-C12 and a few points above) , in practice any EOS (or equivalent) model with a suitable alpha correlation should be able to do the work, note that std. alpha correlations (such as those in PR or SRK) may not work well with heavy HC,
for multi phase equilibria (including solids and reactions) I prefer Prode (see prode.com), the free version has a limited database including up to C15 (n-Pentadecane) but it is free and includes over 30 thermodynamic models plus a data regression utility etc. etc.
the commercial versions (price 200 euro) include 1650 (or 2200) chemicals, these versions include C30 and above...

 

[Ron74]

Thank you for your help, I have managed to calculate the flash properties, but I am having an issue with the sizing of the flush drum.
Liquid flow rate is 662 kg/h, gas flow rate 72 kg/h. Densities of 710 kg/m3 and 0.9.
I calculated the tank diameter with 2 methods, Watkin's correlation, i got vapor velocity of 2.54 m/s, and a diameter of 10 cm.
Using Souders-Brown, I assumed K=0.2 velocity of 5.5 m/s and diameter of 7 cm. Both are very low, and the tank will not be able to hold the liquid volume without having a very long tank.
Both calculation for vertical tanks.
Any suggestions?

Thanks,
Ron

 

[Latexman]

A larger diameter will have less droplet entrainment, so look at those as minimum diameters and make it a practical size.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.