9 replies to this topic

[Anup Paul]

Problem:

We would like to increase the feed to our distillation column.
Currently overall DP across the column is 320 mbar

As per tray datasheet, maximum allowable pressure drop is 245 mbar
And % flood is 80%

I generally take the (maximum allowable pressure drop) / (% flood ) to get the flooding DP.

In our case = 245/ 0.8 = 306 mbar

So I assume that this column already in flooding condition.

Now I want your suggestion to tell me the method I used here is correct or wrong.

If my method is wrong then what is the right method to identify the flooding limits in distillation column

[Hansson]

Dear Aelias,

As a rule of thumb, the maximum pressure drop per tray when you are at the werge of flooding is about 10 mbar. Multiply your total number of trays with 10 mbar and you get a rough estimate of maximum pressure drop before flood.

Now I do not know much about your column, but the reported pressure drop is not that far away from the design pressure drop, so I doubt that the entire tower is flooding.

I would suggest that you increase reboiler duty (or stripping medium) in small incremental steps (about 2-5% of the start value per step). Record the pressure drop at each step. Then plot either pressure drop versus reboiler duty or pressure drop versus reflux+distillate rate.

If you have sieve trays, your curve will be continuosly increasing pressure drop with reboiler duty. If you have valve trays, the curve will have a somewhat flat region followed by a part where the pressure drop increases with reboiler duty. As you pass the flood point, pressure drop will normally skyrocket.

Try this and show your result on an exxell sheet and I will have a look at it and try to help you find the flood point. Depending on where it is compared to todays service, we will be able to tell if the wanted through-put increase is attainable.

Regards
Hansson

[Zauberberg]

Hansson has given very good guidelines.

Another thing to consider is to conduct a performance test of the column, with feed rate increased gradually, and you keep an eye on column performance while maintaining product quality specs. It is possible that you will hit some other constraint (e.g. reboiler duty, condenser duty, reflux pump capacity etc.) before you actually reach the flooding point of the column. Performance test is one of the best ways to benchmark equipment operation, and to see up to what limits it can be stretched.

[Anup Paul]

Dear Hansson and Zauberberg,

Thanks a lot for your suggestions.

Total no of trays = 28
based on your suggestion, that we can assume 10 mbar dp per tray at flooding condition, then the total DP = 280 mbar

As per tray datasheet, maximum allowable pressure drop is 245 mbar
And % flood is 80%

and now we are operating at 320 mbar so are we are at flooding condtions???

This tower receives only vapor feed so there is no reboiler but we have condenser.

Pls. find the attachment for plot of dp vs (reflux + distillate).

Valve details: Glitsch V-1 style

Thanks
Aelias

Attached Files

•  Tower DP.xlsx   45.4KB   255 downloads

[DB Shah]

Dear Aelias.
When column floods before a process engg knows, the panel operator will shout. If your column is flooding you will notice-
1. Very high DP & fluctuations in pressure drop.
2. Unstable column profile, high temperature profile and no level in the bottom & suddenly you see the bottom full of liquid with column temp profile dropping.
3. Poor top/bottom quality

So as far as your case I do not see the urgency in your query and hence can assume your column is still not flooding. You have given your column profile, what is the time interval of the readings (ie every second or minute reading)

Secondly I would be worried for the max DP constraints given by Koch Glitsch. Higher DP in column than recommended may lead to internal tray damage, lifting of the trays. Confirm the operation with KG, give your operating data to them & they will gladly simulate your column and give you necessary recommendations.

[Anup Paul]

Dear Shah,

Thanks a lot for your suggestions.

We want to increase our plant load, but we suspect that after increasing the feed to this tower it may get flooded. Now we want to know how far we are reaching this flooding limit. Based on my calculation (maximum allowable pressure drop from datasheet) / (% flood ) I suspect that we are exactly on the verge of flooding. Now I need your help to validate this.

[Hansson]

Dear Aelias,

Thank you for your excel sheet. I am sad to say that I am not sure if you are flooding or not.
At first glance, it appears that you are not. However, as DB Shah mentioned, the time-line for your test could possibly help to clearify this. You see, pressure drop increases with time as you pass the flood point. The absoulte value of your DP is on the high side, but I was wondering how it is measured? Do you have a DP cell or a calculated DP from two separate pressure measurements? There is often a DP even as there is no flow, due to liquid accumulation in the instrument legs, so in reality the DP can be higher or lower then measured.

I am sorry, but I think I need some more data. How high weirs are you using?
How did you manage to increase reflux? Do you control feed inlet temperature or what?
I think we will be able to give you a answer, but I am still not sure.

To be continued
Regards
Hansson

[Hansson]

Dear Aelias,

I update two examples of pressure vs. gas flow for two sections that indead are flooding. They are both for valve trays. One figure is for a rectifying section and one is for a stripping section. From these charts, the flood point is easily noticed.

Regards
Hansson

Attached Files

•  DP.doc   137KB   174 downloads

[pavanayi]

Aelias,
You have already received good recomendations. I might suggest an alternate way to roughly calculate how far you are from flooding.

From Lieberman's 'A working guide to process equipment', the total height of clear liquid in the downcomer is formed by four parameters
1. Liquid backup due to escape velocity from downcomer on to the tray below
2. Weir height
3. Crest height of liquid overflowing the outlet weir
4. Pressure drop of the vapour flowing through the tray (converted to liquid head)


To calculate the height of liquid in the downcomer due to liquid flowing through the downcomer clearence,

∆H = 0.6 V²
where
∆H is inches of clear liquid backup in the downcomer due to head loss under the downcomer
V is the horizontal component of liquid velocity in ft/s as the liquid escapes from the downcomer.

You can make an assumption that the liquid flowrate through the column is constant, and with the tray information you should have, you should be able to calculate the area available for liquid in the downcomer to escape onto the tray. That should give you the horizontal velocity component and thus ∆H.

Since you already have an average tray pressue drop, you can convert that to liquid head.
for ex, if your tray pr drop is 320/28 = 11.43 mbar is approx equal to 4.6 inches of water. Convert that to inches of the liquid.

We actually have a foam instead of clear liquid within the downcomer. Liebermann suggests an aeration factor of 0.5 can be used. That means, if from the above calculations, you compute 10 inches of clear liquid, in reality, it will be 20 inches of actual froth.

If you know the height of the downcomer, that would enable you to have an educated idea on how far you are towards flooding.

If you can get your hands on that book, you can read the assumptions for the above calculation as well, so that you are confident on the numbers you arrive at.

Edited by pavanayi, 11 November 2011 - 02:44 PM.

[sheiko]

Lieberman proposes in the above mentionned book several ways to evaluate tray operating conditions.

To me, the simpler one consists in comparing the pressure drop per tray (expressed in inches of liquid) to the weir height (in inches):
- if DP > 3 times weir height, then flooding
- if DP < weir height, then weeping or dumping

You will find the weir height in the trays data sheet, but it should be close to 2-3 inches.

Edited by sheiko, 12 November 2011 - 09:55 PM.