10 replies to this topic

[USR]

Dear Forum members,
In conditions when the ambient temperatures are high and even after using the maximum available condenser capacity if it is not possible to produce enough reflux how could one effectively run distillation columns?

[Padmakar Katre]

Dear Forum members,
In conditions when the ambient temperatures are high and even after using the maximum available condenser capacity if it is not possible to produce enough reflux how could one effectively run distillation columns?

Hi,
Could you please elaborate the issue further giving us the background of the system. This is general issues water cooled exchangers when the local wet bulb temperatures are higher resulting the cooling water supply temperatures high or in case of air coolers demand of required area increases due to lower LMTD. I am giving in general reply. We can reply in detail and specifc once you share the background of the system in detail.

[Hansson]

Dear USR,

One way to keep fractionation OK when reflux is limited is to decreased the column feed pre-heat. Of course, this option only applies where feed pre-heat is controlled. If feed pre-heat is reduced, you reduce the internal traffic in the rectifying section of the tower. This allows you to increase reboiler duty until you once again reach your reflux limitation. This improves the overall fractionation in your tower.

Another possibility you can try is to increase the tower pressure. This increases bubble and dew point temperatures, giving you a greater temperature difference between your cooling media and the process media. Thus, the cooling duty increases and you can get higher reflux rates. A drawback with higher pressures is that most towers will give a lower relative volatility between the key components with increasing pressure. Often, the overall effect of increasing the pressure will be increased fractionation. You simply have to try for each column.

Kind regards
Hansson

[USR]

Dear Hansson,
I believe you have good hands on experience with distillation columns.I suppose you are in operations.The first suggestion you have given sounds very practical and effective.
Raising tower pressure may not be that effective.I believe it has more drawbacks,guess we reach a trade off here.Increased pressure will reduce the volumetric flow rate of the vapours leading to frequent dumping,pressure fluctuations.

Edited by USR, 20 August 2011 - 04:56 AM.

[Hansson]

Dear Hansson,
I believe you have good hands on experience with distillation columns.I suppose you are in operations.The first suggestion you have given sounds very practical and effective.
Raising tower pressure may not be that effective.I believe it has more drawbacks,guess we reach a trade off here.Increased pressure will reduce the volumetric flow rate of the vapours leading to frequent dumping,pressure fluctuations.

Dear USR,
You are quite right, I currently work in operations. The recomendation about reducing pre-heat always work as long as you do not load up the stripping section until you flood it. I do not think that you should role out my second recomendation about increasing tower pressure though. It does not always work, but for more then 50% of the towers where I have tried this, the overall effect has been beneficial. You are absolutely right that it will not work in some cases, but I believe in practical testing. If the test shows that it was a bad idea, it is not a big deal, you simply switch back to the previous operational mode. For a few hours you have done something less economical. However, if the fix works, you can run for years in a more economical mode. You are also quite right that you will get higher risks for dumping if you increase the tower pressure, but unless you are already operating close to the dump point (most sieve trays have turn-down to 50% of design load and valve trays even greater turn-down), dumping will rarely be a problem. On the other hand, at normal operation you are (or at least should) be closer to the flood point than the dump point, and increasing tower pressure will get you further from the flood point (both jet flood and downcomer back-up flood). If you indeed do try to increase tower pressure, it would be nice to here how it works out for you.

Kind regards
Hansson

[Padmakar Katre]

Dear USR,

One way to keep fractionation OK when reflux is limited is to decreased the column feed pre-heat. Of course, this option only applies where feed pre-heat is controlled. If feed pre-heat is reduced, you reduce the internal traffic in the rectifying section of the tower. This allows you to increase reboiler duty until you once again reach your reflux limitation. This improves the overall fractionation in your tower.

Hi,
What difference it makes in case of decresed feed preheat to compensate by increased reboiler duty. The problem is with your overhead condenser i.e. coolent medium i.e. increased supply temperature. The reboiler duty is sometimes split into the feed preheater and the bottom reboiler with due precautions to separate the vapor-liquid at feed stage. Waiting for you explaination on this issue.

[Hansson]

Dear USR,

One way to keep fractionation OK when reflux is limited is to decreased the column feed pre-heat. Of course, this option only applies where feed pre-heat is controlled. If feed pre-heat is reduced, you reduce the internal traffic in the rectifying section of the tower. This allows you to increase reboiler duty until you once again reach your reflux limitation. This improves the overall fractionation in your tower.

Hi,
What difference it makes in case of decresed feed preheat to compensate by increased reboiler duty. The problem is with your overhead condenser i.e. coolent medium i.e. increased supply temperature. The reboiler duty is sometimes split into the feed preheater and the bottom reboiler with due precautions to separate the vapor-liquid at feed stage. Waiting for you explaination on this issue.

Hi Padmakar S Katre,

It makes all the difference in the world.
Let us look at a simple example. A column is fed with 200 t/h of feed, half is taken as distillate and half as bottom product. Experience has shown that in order to get on spec. products, a reflux of 300 t/h is needed with normal OH-T and feed T. That means that the OH-condenser has to condense 100+300 t/h and possibly sub-cool it from the top tray temperature to the normal OH-T. Now then, lets say that the ambient temperature increases and fin fan cooler duty will go down or water cooler supply temperature goes up. If we no longer have enough utilities (for argues sake, let's say that we have fin fan coolers and all fans are switched on and are running on maximum speed and all louvers are open as much as possible). If this is not eneough to condense our 400 t/h, we have few options. One option is to decrease feed rate. This is typically not a very popular choise. If you try to keep feed rate and reboiler duty, the condenser duty is not enough to condense all gases coming to the OH-condenser, hence gas will come to the reflux drum. Preeusre will rise and eventuelly pop your SRV (safety relief valve). The second choise is to decrease reboiler duty so that the condenser duty will be able to condense all the OH gases so that the SRV will not pop and send your valuable product to the flare. Lets say that you have to decrease reboiler duty until your OH gas rate is 360 t/h before the SRV closes and you are back in stable operation. Now if you still have 100 t/h of distillate, you only have 260 t/h of reflux, while 300 t/h was needed to get you on spec. products. Here is where my suggestion comes in. If you lower feed pre-heat (or increase feed pre-cooling), the reflux rate will decline on constant reboiler duty. Decrease pre-heat until the reflux is 220 t/h. That means that you now have extra capacity in your OH-condenser. Increase reboiler duty until you once again have a reflux of 260 t/h. Your reflux ratio is as bad as once you started decreasing pre-heat, but your stripping ratio is back on where it used to be. Reduce feed pre-heat further and then increase reboiler duty by as much. Now, your reflux ratio is still worse than normal, but your stripping ratio is higher then normal. That way you can compensate for a lesser traffic in the rectifying section of the tower, by using the stripping section more then normal. If you are still not convinced, run a simple simulation, make a McCabe-Thiele diagram or test it on one of your own towers.

Regards
Hansson

Edited by Hansson, 26 August 2011 - 03:36 PM.

[Zauberberg]

Hansson is correct. Decreasing fead preheat unloads the rectifying section of the column while increasing load in the stripping section, due to higher reboiler heat input (for the same tower bottoms temperature as in the first case).

Apart from flooding issues, the degree of fractionation in the rectifying section will decrease as the feed preheat is reduced, resulting from reduced vapor/liquid traffic above the feed entry point.

[Root]

Reducing the vapor load in rectifying section will reduce the reflux ratio and over head product quality will effect and also reduce net over head draw.
Column pressure increasing idea will not work, as per him condenser duty problem. He can over come with this problem with varying the reflux flow (ratio) and keep monitoring over head product distillation range in optimum limits.

[Hansson]

Hi Padmakar S Katre,

It makes all the difference in the world.
Let us look at a simple example. A column is fed with 200 t/h of feed, half is taken as distillate and half as bottom product. Experience has shown that in order to get on spec. products, a reflux of 300 t/h is needed with normal OH-T and feed T. That means that the OH-condenser has to condense 100+300 t/h and possibly sub-cool it from the top tray temperature to the normal OH-T. Now then, lets say that the ambient temperature increases and fin fan cooler duty will go down or water cooler supply temperature goes up. If we no longer have enough utilities (for argues sake, let's say that we have fin fan coolers and all fans are switched on and are running on maximum speed and all louvers are open as much as possible). If this is not eneough to condense our 400 t/h, we have few options. One option is to decrease feed rate. This is typically not a very popular choise. If you try to keep feed rate and reboiler duty, the condenser duty is not enough to condense all gases coming to the OH-condenser, hence gas will come to the reflux drum. Preeusre will rise and eventuelly pop your SRV (safety relief valve). The second choise is to decrease reboiler duty so that the condenser duty will be able to condense all the OH gases so that the SRV will not pop and send your valuable product to the flare. Lets say that you have to decrease reboiler duty until your OH gas rate is 360 t/h before the SRV closes and you are back in stable operation. Now if you still have 100 t/h of distillate, you only have 260 t/h of reflux, while 300 t/h was needed to get you on spec. products. Here is where my suggestion comes in. If you lower feed pre-heat (or increase feed pre-cooling), the reflux rate will decline on constant reboiler duty. Decrease pre-heat until the reflux is 220 t/h. That means that you now have extra capacity in your OH-condenser. Increase reboiler duty until you once again have a reflux of 260 t/h. Your reflux ratio is as bad as once you started decreasing pre-heat, but your stripping ratio is back on where it used to be. Reduce feed pre-heat further and then increase reboiler duty by as much. Now, your reflux ratio is still worse than normal, but your stripping ratio is higher then normal. That way you can compensate for a lesser traffic in the rectifying section of the tower, by using the stripping section more then normal. If you are still not convinced, run a simple simulation, make a McCabe-Thiele diagram or test it on one of your own towers.

Regards
Hansson

[Elvin]

USR,

Hansson is right if you can control feed temperature. In many cases this is not possible.
Therefore, what I applied several times, was to reduce reboiler duty, in fact column bottom temperature with a 2-5 degrees in parallel with reflux flow, more and less to keep same reflux ratio and to avoid flooding. Keep this conditions for a while, hours, until you will get right composition of reflux.

USR, seems you are at LPG column, or a range of 30 to 60 centigrade at top of it, will work for sure...

That one with pressure, it is OK, works but you need to be a good DCS operator with process engineering background.

Good Luck,
Elvin