Water condensation in Naphtha Stabilizers

[thiru4vino]

Dear All!

Can anybody help me to understand why we get water condensate in Naphtha stabilizer that operates at 8.5 kg/cm2 (g) and a top temperature of 85 Deg.C ?

To an extent, I can understand that water when present in small quantities, say 100-200 ppm in Naphtha, it will be carried by LPG vapours to the overhead, though the operating pressure is 8.5 kg/cm2(g) and the top temperature is less than 100 Deg.C.

Can i get a clear picture of this event which proves the theory wrong.? Can we simulate the condition in Hysys or Pro-II and if possible, how can we do it?? esp, giving the qty of water in the feed...

Thanks for everything you do..
cheers

atm

 

[owg]

I too would like to see an explanation of this although the question is not fully specified. My guess is that if there is not too much water in the feed it will come out the top and can be drawn off. If there is too much it will accumulate in the tower gradually blanking off trays until you get an upset. Cutting the reboiler will let the water out the bottom to the next guy and you start the cycle again.

HAZOP at

http://www.curryhydrocarbons.ca

 

[UmeshMathur]

chillboy and owg:

Water is sparingly soluble in hydrocarbons, so it has very high K-values in the column. Hence, it gets carried up along with the vapors to the overheads. Upon condensation, it separates into a second liquid phase as the solubility in the hydrocarbon phase is low and only a tiny fraction of the incoming amount of water can be dissolved in it.

Depending on how much water there is in the feed, you may or may not see a second liquid phase in the condenser.

By the way, estimating the solubility of water rigorously in hydrocarbons is a non-trivial matter. However, the major simulators use various approximations to model the K-values in the column itself so that, for simple columns with one liquid phase only, the water is generally predicted to be carried up the column as it should.

Similarly, the prediction of water solubility in the hydrocarbon phase (reflux plus distillate) is generally not done too accurately either.

If there's too much water in the feed, you probably should check for a 2-liquid phase feed. Also, there may now be a second liquid phase on many trays in the column and solving that simulation problem requires an entirely different algorithm, i.e., one capable of handling two liquid phases. Also, the thermo package used must also predict liquid phase splitting accurately.

 

[mbeychok]

I agree with UmeshMathur' response. As an approximate rule-of-thumb, a moderate amount of water present in a hydrocarbon distillation column behaves as if it had a relative volatility equivalent to propane.

Milton Beychok
(Visit me at www.air-dispersion.com)
.