9 replies to this topic

[AllyK]

Hey guys,

I have a question regarding my distillation column design. I have two components in my mixture, within 25 degrees celcius of each other in boiling point. I have therefore deduced an increase in pressure is vital in increasing this margin to ensure better separation.

I have used the Clausius Clapeyron equation, going up in increments of 1 atm, then taking note of new Temperatures (i assume these are new boiling points).

My main problem is that i have reached 15 atm and i have only gained a boiling point margin of 60 degrees.

Is 15 atm far too high for a distillation column? is there any other method i can use in conjunction with pressure, to further increase this margin?

I'd really appreciate the help!

[AllyK]

Just to clarify i have been using:

ln(P1/P2)= (enthalpy of vaporization)/R * (1/T2-1/T1)

i have used this equation for each component (IBAP and IBPE) and compared my new temperatures at a particular pressure, in the form of a graph (where on is superimposed on the other)

[katmar]

It is feasible to run a distillation column at 15 bar, but who told you that 25 deg C difference between the top and bottom temperatures is too little? I have seen plenty of columns working very successfully with less than that. I suspect that it would be more economic to just add a few extra trays and run it at atmospheric pressure. If your bottom temperature gets too high then normal heating methods (eg steam) might not work and the cost of installing a high temperature furnace just to heat a distillation column can be more than the column itself.

[Art Montemayor]

I kind of remember that the Boiling Point Curve is an indication of how pressure affects the ease of separating a binary mixture: a vacuum amplifies the distance between the vapor curve and the liquid curve; pressure decreases the distance between the two curves and this makes a distillation more difficult.

Your approach is doing just the opposite. Are you sure that applying more pressure will make the separation of a binary easier and not more difficult?

[pavanayi]

AllyK,
Echoing whats been mentioned already, an increase in pressure will decrease the relative volatility and hence increase the difficulty in separation between components.

But more importantly, the pressure you select for the distillation column is not primarily based on the degree of separation of boiling points. Its more to do with the column top temperature.

What components are being split in your distillation column? What is the previous unit operation to this distillation column and what pressure and temperature is the process stream at the outlet of that unit operation?

[AllyK]

Thanks guys. Being new to designing stuff i was a little thrown off.

Pavanayi, my components were Isobutyl phenyl ethanol (95% conversion from isobutylacetophenone) and unreacted isobutylacetophenone (i am manufacturing ibuprofen). I was thinking of doing this as a binary distillation as i have assumed by-products are in trace amounts.

Anyway, thanks a bunch for the help guys, am i right in just assuming my column is running at 1 atmosphere then? Katmar you said it is a reasonable difference but would you adjust it to a pressure just slightly higher than atmospheric pressure or would you just leave it at 1 atm?

Art Montemayor, thanks for that insight. I did not consider this and was thinking only of making my boiling points greater in difference. Would you too just keep this at 1 atm?

Cheers guys!

[katmar]

In what is termed at "atmospheric pressure" column the condenser would be vented to atmosphere (unless there are toxicity considerations) and as you go down the column the pressure would increase by the pressure drop across each plate. This makes very little difference (usually) to the vapor-liquid equilibrium in terms of x-y data, but it does affect the boiling points. There are many distillation columns still working out there that were designed before the software was available, and those columns would not have been corrected for pressure on each tray. It was just too much work to do by hand. Nowadays the simulation software can take into account the tray pressure drops and automatically correct the boiling point and even the x-y data for the increased pressure.

[Zubair Exclaim]

I kind of remember that the Boiling Point Curve is an indication of how pressure affects the ease of separating a binary mixture: a vacuum amplifies the distance between the vapor curve and the liquid curve; pressure decreases the distance between the two curves and this makes a distillation more difficult. Your approach is doing just the opposite. Are you sure that applying more pressure will make the separation of a binary easier and not more difficult?

Now this is intresting , i was forced to spend like a hour on this... guys do correct me if i am wrong --i have taken two similar components and used simulator to assess the boiling point difference behaviour of two components, It is intresting to note that the difference between boiling points of two components increases with increase in pressure. Now this is news to me too..... --though i am not too sure about the behaviour that a mixture would give on a bubble point and dew point curve but i think that may act in asimilar way . --i am investigating more into it like behaviour of polar non polar etc will get back to you guys. wont jump to conclusion yet see attached excel sheet also

Attached Files

•  boiling point elevation.xls   20.5KB   37 downloads

Edited by Art Montemayor, 26 March 2012 - 07:18 AM.

[AllyK]

Hey guys just another question. I have completed a feed tray calculation and i have found there are 5 stages above my feed tray (including condenser) However, i have 6 total steps including the reboiler. I then decided my location doing 6-5=1. Does this mean my feed is entering the reboiler? Is this even possible?

To clarify things i have used the Kirkbride equation:

log(Nr/Ns)=0.206*((B/D)*(xf,hk/xf,lk)*(xb,lk/xd,hk)^2)

and have arranged my formula in terms of Ns=6-Nr to calculate a value for Nr.

It says in Coulson and Richardson that xf, xb and xd are concentrations. am i right in assuming they are just the fractions of each component in my feed/distillate/bottoms (where each component is running in moles/hr)?

Any help will be greatly appreciated!

[AllyK]

I'll add that Nr is the number of stages above the feed tray and Ns, the number below. Also B=molar flow bottom product and D=molar flow distillate, where i simply added the molar flowrates of the components above and including the light key (for D) and where i added the molar flowrates below and including the heavy key (for B ).

Cheers

Edited by AllyK, 03 April 2012 - 08:39 AM.