9 replies to this topic

[ipai]

Hi,

 

I am revisiting parameter use in butyraldehyde distillation tower. The feed contains n-butyraldehyde and iso-butyraldehyde (41 ton/hr).

 

The main issue is the water content (moisture, current reading 1.5 wt%) in iso-butyraldehyde (output at the top overhead, 4ton/hr) is quite high compare to previous year and I cannot pin point any obvious parameter as the moisture increase.

 

The operating condition at the overhead are P=1.27 bar and T=70degC. And the operating condition at the overhead drum is T=63 degC, with P= assuming same with the overhead pressure.

 

What P and T parameter I should use to maintain the moisture below 1.2 wt%?

 

Or any reference or step guideline so I can check the suitable parameter for the tower.

[breizh]

IPAI

Without information about the feed composition ( % mol or mass ) including water difficult to say anything . No leak in the overhead condenser ?

my 2 cents
Breizh

[ipai]

Dear Breizh,

 

Feed 41t/h. Feed composition 11wt% - iBa, 87wt% - nBa, 0.2wt% - moisture.

 

The overhead condenser is fin fan, so leak might not cause to the moisture.

 

 

Any particular physical properties that I should look at?

[breizh]

IPAI ,

Probably good to check the parameters change over the years .
According to the boiling points of the material you should not see water with the distillate .

Any change on feed rate , Reflux , Heat duty , working Pressure and Temperature?

No issue with the packing or trays ? No change on column DP ?

All sensors are well maintained , calibrated.

Probably good to establish a photography (Before and After)

Sorry I'm not knowledgeable about your system .

Breizh

[ipai]

Breizh,

 

I have check and there some changes for the past years. Just recently, the moisture content keep increasing since last Dec, and we propose to change back to the original parameter where the moisture is low. Currently the moisture is maintain with consistent moisture content.

 

If I want to focus with the pressure, should I reduce it or increase it ? 

[PingPong]

Note that isobutyraldehyde and water form an azeotrope, at least at atmospheric pressure, with some 5 to 6 % water and a lower boiling point than isobutyraldehyde.

At different pressure that azeotrope may shift to a higher or lower water concentration.

 

I suspect that the feed preheat temperature of the column can also have an impact on quantity of water that ends up in the distillate.

May be worth trying to lower that feed preheat temperature. Disadvantage of that is of course a higher reboiler duty.

 

One would need to do simulation to see impact of all parameters, but that would require more data than given above.

[ipai]

PingPong,

 

Yes, I agree with you about the feed preheat temperature, it will increase the reboiler duty and our reboiler already at max capacity.

 

Where do I can find the azeotrope data/graph for isobutyraldehyde and water? Do you have any reference?

 

Btw, thank you for your input.

[PingPong]

I have no experience with this particular system, but I know that when water and hydrocarbons containing O-atoms are together there are often interactions resulting in both molecules having activity coefficients departing from 1. This indicates non-ideal behavior and sometimes azeotropes.

 

Googling isobutyraldehyde water azeotrope gave me a few hits that imply that an azeotrope exists, such as

http://toxnet.nlm.ni...term @DOCNO 614

stating: forms azeotrope with water containing 94% isobutyraldehyde; azeotrope BP: 59 deg C @ 760 mm Hg

thereby referring to the Merck Index.

 

Whether there can really be an azeotrope at the pressure in your column is not really that important. What matters is that because of the non-ideal behavior water will be more volatile in this system than if one would be distilling alkanes with boiling points of 64 and 75 ^oC.

 

Does the overhead drum (reflux drum) have a water boot to allow separation of free water?

 

Is it possible to operate with subcooled liquid in the overhead drum, so as to separate as much free water as possible, or is the column pressure determined by the temperature in the overhead drum? In other words: how is the column pressure controlled?

 

EDIT: I used PRO/II to check for an azeotrope in the IsoButyraldehyde - Water system. At atmospheric pressure it predicts an azeotrope with 79 mol% IB and 21 mol% water, as can be seen in graph below (at 1.27 bar(abs) it is more like 78.5/21.5 so practically the same).

 

 IsoButyraldehyde-Water Azeotrope.jpg   29.01KB   2 downloads

 

That differs from the 94%IB/6%W that can be found via internet, but it may be that that is not mol% but wt%......(who knows).

PRO/II is not really the most accurate tool for this kind of systems. ASPEN Plus would probably be more accurate.

 

To complicate matters even further: n-butyraldehyde and water also form an azeotrope. Some hits in google say at 12 wt% water at atmospheric pressure. However I find 14 mol% water using PRO/II which is quite different from that 12 wt%. But maybe that 12 wt% is meant to be 12 mol%.......(who knows).

 

To conclude, some good news: isobutyraldehyde and n-butyraldehyde do not form an azeotrope.

Edited by PingPong, 17 September 2014 - 12:56 PM.

[ipai]

Dear PingPong,

 

Thank you for such informative post for this discussion. With your information, I have work on to got some help and get information by using ASPEN plus. Here is what I have got;

 

Physical Property model: NRTL

Azetrope mixture at 1 bar: 60 degC

With;

Water 27mol%, 8.77wt%

iBA 72mol%, 91wt%

 

Azetrope mixture at 1.25 bar: 66 degC

With;

Water 29mol%, 9.37wt%

iBA 71mol%, 91wt%

 

 

Physical Property model: UNIQUAC

Azetrope mixture at 1 bar: 60 degC

With;

Water 25mol%, 7.61wt%

iBA 75mol%, 92wt%

Azetrope mixture at 1.25 bar: 66 degC

With;

Water 26mol%, 8.07wt%

iBA 74mol%, 92wt%

 

 

 

Physical Property model: PENG-ROB

Azetrope mixture at 1 bar: 53 degC

With;

Water 48mol%, 19wt%

iBA 52mol%, 81wt%

Azetrope mixture at 1.25 bar: 60 degC

With;

Water 49mol%, 19wt%

iBA 51mol%, 80wt%

 

 

 

And to answer your question;

 

Question:

Does the overhead drum (reflux drum) have a water boot to allow separation of free water?

 

Answer:

The reflux drum does not have a water boot to separate water. I believe the original design does not expect high moisture from the feed. The plant has been running since 9 years ago, until recently we have high moisture.

 

 

Question:

1. Is it possible to operate with subcooled liquid in the overhead drum, so as to separate as much free water as possible, or

2. is the column pressure determined by the temperature in the overhead drum? In other words: how is the column pressure controlled?

 

Answer:

1. No it is possible to operate with subcooled liquid in the overhead as I mentioned earlier the overhead piping use fin fan cooler.

2. Could you explain more on this "column pressure determined by the temperature in the overhead drum". Column pressure controller: In my system, the pressure reading is come from the overhead column piping before going to fin fan cooler. The pressure reading will send signal to 4 valve (split range) according to the split range reading.

 

 

I still do not have any solid theory to change any parameter for the moisture, and the moisture still maintain in range 1.45 wt% to 1.5 wt%.

 

I appreciate for any idea. and thank you in advance.

[PingPong]

Column pressure controller: In my system, the pressure reading is come from the overhead column piping before going to fin fan cooler. The pressure reading will send signal to 4 valve (split range) according to the split range reading.

That still does not explain the mechanism by which the pressure in the column and the reflux drum are maintained.

Please upload part of P&ID showing column top, condensor, reflux drum and pressure controls.

 

 

 

I still do not have any solid theory to change any parameter for the moisture, and the moisture still maintain in range 1.45 wt% to 1.5 wt%.

Due to the highly non-ideal behavior of water in this system it is almost impossible to predict what should be done to reduce water content of distillate, using only gut feeling.

If you have access to Aspen Plus then you should model the whole column, and tweak the tray efficiencies above and below feed such that the model reproduces actual operation.

After that you can vary operating parameters in the model to see what change improves the distillate product.

It may very well be that the only solution is to reduce water content of column feed, as most of that water seems to end up in the distillate.

Edited by PingPong, 23 September 2014 - 05:54 AM.