39 replies to this topic

[panoska]

Hello everyone,

 

I would like to ask you a simple question that I cannot really get it.

 

How do we create vacuum in a distillation column?

 

Thank you in advance!

[Art Montemayor]

Panagiotis:

 

What do you mean by your question?

 

• Are you trying to find out how to avoid creating a vacuum in a distillation column and system?
• Or do you want to find out how a distillation column is operated under a vacuum?

 

The answer to your question depends on what you are trying to achieve or to avoid.

[latexman]

With a vacuum pump or steam jets downstream of the condenser.

[breizh]

Hi,

Question is weird ! You need to be more specific.

Check for ejector or vacuum pump using your favorite search engine . 

Check also Graham website to get more input.

Technical Documents Library - Graham Corporation | Graham Corporation (graham-mfg.com)

Good luck

Breizh 

[katmar]

You will of course need some type of vacuum pump, but it is important to understand that the function of the vacuum pump is only to remove the incondensible gases. The main generator of the vacuum is the condenser and this needs to be supplied with sufficiently cold water to condense the vapors at the operating pressure that you want to achieve. The pressure in the column will be controlled by the temperature and flow rate of the condenser cooling water, and the area and HTC of the condenser.

[panoska]

First of all, thank you all for your replies.

 

Art, indeed my question is not clear at all. My actual question is your 2nd question: "Do you want to find out how a distillation column is operated under a vacuum?"

To be more precise I am wondering how vacuum is created inside the column? 

 

Breiz, I will go through your document. I believe it will help me. Thanks for providing me this document.

 

Katmar, thank you for your explanation.  In a vacuum distillation of a refinery, the inlet feed to the column is at atmospheric conditions and then the vacuum is created inside the column with the aid of the steam ejectors? Why do you mention that the condenser creates the vacuum inside the column?

[katmar]

panagiotis, I suggest you review the concepts of vapor pressure, vapor liquid equilibrium and partial pressure.

[SilverShaded]

In a refinery the feed to a vacuum column is not at atmospheric conditions.

 

You create a deep vacuum at the top of the column, typical 12mm Hg or thereabouts either by steam ejectors or a combination of steam ejectors, vacuum pumps and pre-condensers.  Typically pre-condensers are not used anymore as there are limits on how much vaccum you can pull and the trend has been for deeper vacuums.

 

Ok so the top of the column as at roughly 12mm Hg, you then have pressure drop down the column unitl you get the flash zone, which is probably ~20 mm Hg higher pressure than the top of the column.  Due to pressure drop.

 

In most vacuum columns these days there is a transfer line coming for the feed furnace which has some significant pressure drop with vapour travelling at allmost the speed of sound, it is not at equilibrium btw but thats another story.  The outlet of the furnace is therefore somewhat higher pressure than the column flash zone, due to pressure drop.  The furnace itself has significant pressure drop and the feed to the furnace is likely to be at a couple of bar above atmospheric pressure, having been delived by a pump from the bottom of the crude unit, typically.

[Art Montemayor]

Panagiotis:

 

I’m adding further comments because now that you have further explained your query, I suspect that you may need additional details in order to fully appreciate the basic and correct reply you received from Katmar and Silvershaded.  I wish I had the time to quickly sketch a flow diagram of a crude distillation tower under vacuum operation (VDU) because I suspect you need further detailed explanations.  For example, you state: “…the inlet feed to the column is at atmospheric conditions and then the vacuum is created inside the column with the aid of the steam ejectors?  Why do you mention that the condenser creates the vacuum inside the column?”

 

Katmar has correctly advised that the Overheads Condenser is the creator of the vacuum.  The supplied vacuum system only maintains the vacuum because of practical needs:

 

• Dissolved air and gases in the tower feed;
• Mechanical leaks in the vessels and system;
• Possible decomposition of feed components; and
• Variations in feed composition.

 

We engineers have to confront the every-day reality of real-life conditions in the field application of ideal ideas and concepts.  Katmar’s recommendation of studying concepts of vapor pressure, partial pressure, and phase equilibria cannot be emphasized enough in order to fully understand and implement this unit operation.

 

In a perfect world, we would only need to generate a vacuum in the VDU initially at startup and afterwards we would not need it anymore during a production run.  However, this is not a perfect world and vacuum maintenance must be incorporated and controlled.  The key component establishing the degree of vacuum is the Overheads Condenser that establishes a phase change that allows the generation of the needed tower liquid reflux and the ultimate liquid overheads products.  The phase change generated by the Overheads Condenser is the key and that is why Katmar’s experienced recommendations are so important.

[panoska]

Art thank you for your elaborate answer. Indeed it would be nice if you had time to sketch a VDU for me. However, I will search on the internet to find a sketch of it.

 

Your replies contain a bit of fundamental and experience. I will dedicate more time to understanding them.

 

What I can understand from your replies generally is:

 

a) The ejectors create the vacuum during start-up.

b )  The condenser maintains this vacuum inside the column.

 

Also, I have concluded that the condenser and the condensing temperature defines the pressure in the column. (To be honest, this is not crystal clear to me yet. Maybe I have to challenge myself with the fundamentals. And then I will come back to you for further discussion.)

Edited by panagiotis, 20 February 2022 - 03:38 PM.

[Art Montemayor]

Panagiotis:

 

While I haven't the time to generate the kind of flow diagram I would like to make, I found the following information which contains a very simplified diagram on how a vacuum tower is controlled - and what's more, an explanation of the controls and instruments used to do it.

 

You may not only find the article useful, but also very interesting in showing how the vacuum is controlled in a VDU.

 

I also make the following comments on your previous statements to emphasize what I meant:

 

a) The ejectors create the vacuum during start-up. (The steam ejectors used work to evacuate any non-condensable vapors or gases migrating into the system - just as Katmar states)

 

b )  The condenser maintains this vacuum inside the column.  (The condenser establishes the vacuum pressure by condensing the produced vapors at the appropriate vapor pressure they produce - which is in the selected vacuum range.  That is why the condenser is the key item.)

 

I hope I have been helpful.

Attached Files

•  Controlling Vacuum Distillation.docx   123.72KB   117 downloads

[katmar]

There is a very important aspect of the vacuum control system that is shown in some diagrams in Art's article. None of us had brought it up here and it is not actually discussed in the article.

The vacuum pump has to be sized for start-up and worst-case conditions and this means that it is usually over-sized for normal operation. If there is not enough incondensible gas in the system to satisfy the vacuum pump's curve there will be problems. At the very least it will lead to carry-over and probable loss of product from the condenser. And for some types of pumps (such as liquid ring) it can lead to cavitation and physical damage to the pump.

The solution to the problem is to bleed air or some inert gas into the line before the vacuum pump to satisfy the curve. I have seen the air bleed into the system located either between the column and the condenser or between the condenser and the vacuum pump. The air has a greater effect if it is located upstream of the condenser because the air flowing through the condenser decreases the Heat Transfer Coefficient of the condenser but this can lead to product losses and can only be used in some instances.

[breizh]

Hi,

To add to Harvey's comments let you consider the docs attached .

Note : You will find a second document about air leakage estimation for vessels under vacuum.

Good luck

Breizh 

[SilverShaded]

Specifically for refinery vacuum columns which i think is the topic we are discussing, there is no overhead condenser (other than to condense ejector steam and slop oil) and no top reflux besides that which comes from the top pumparound.  The vapour travelling overhead to the ejectors is mostly steam from bottoms stripping and furnace coil steam and some cracked hydrocarbons.  Air leakage is included in design estimates but is relatively insignifcant compared to the steam load.  Cracked hydrocarbon estimates are essentially rule-of-thumb or based on known values from operating units.

 

A good HAZOP question to ask yourself is what happens to column pressure if either you lose the ejector steam or a pumparound pump fails.

Edited by SilverShaded, 21 February 2022 - 02:52 AM.

[katmar]

I agree with SilverShaded's comments and there are certainly vacuum distillation units without overhead condensers. However, the heat exchangers in the side stream pumparounds then become the primary vacuum generators. If these heat exchangers do not condense the hydrocarbon vapors rising up the column these vapors would be carried over to the ejectors and overload them. The temperature control on the pumparounds will be critical to getting the correct split of products in each side draw and the correct pressure/temperature profile in the column.

[gegio1960]

the level of vacuum in a VDU refinery tower is defined by an economic optimization.

it could be in the range 10 (dry tower) to 75 mmHg (wet tower).

while dry towers have been discussed in detail, wet towers merit some additional words.

the operating conditions are obtained by considering the addition of steam, for decreasing the partial pressure of HCs, at the bottom of the tower.

steam is also injected in other two points:

- the heater tubes, to increase the velocity in the final path of the tubes (velocity steam, that also avoids eccessive fouling)

- the bottom of the side-strippers, to improve the product qualities (stripping steam)

the steam presents in the tower shall be condensed (top condenser) before the first ejector to decrease the size of the vacuum generation package (vacuum ejectors/pumps, condensers).

good luck :-)

[breizh]

Hi,

Why do you bring the subject to refinery ?  To me it's related to Vacuum distillation unit in general .

my 2 cents

Breizh 

[panoska]

Hello everyone for your time that you dedicate in order to explain me. The documents are many and I will go through them. 

Art thank you for including in your document some information about control. 

 

My next challenge is to understand  the vacuum of a deodorizer column, which is actually a stripping column.

 

From all the different information you gave me, I think it is essential to keep in mind that the condensers even if they are on the top of the column or side heat exchangers are the primary vacuum generators.

[gegio1960]

panagiotis,

the comment by breizh about the right forum choice is simple but very important to obtain more specific info and to avoid misunderstandings.

anyway, i like very much this post because of the high level contributions of some of the best engineers in cheres (with particular happiness about the return of art montemayor after a long time!).

specifically, a deodorizer column is very different from a VDU vacuum tower (that is the "big queen" of the vacuum towers).

i designed about 20 VDU vacuum towers (3 in greece, from where panagiotis appears to come...) but only 1 deodorizing system (for a battery of bitumen tanks).

while the diameter of the VDU towers is often above 10 m coupled with heights that can go up to 50 m and more, the deodorizer systems have very little tower, if any, and don't involve high temperatures. often, a vacuum pump, without condenser is enough for deodorizing.

good luck :-)

[SilverShaded]

Hello everyone for your time that you dedicate in order to explain me. The documents are many and I will go through them. 

Art thank you for including in your document some information about control. 

 

My next challenge is to understand  the vacuum of a deodorizer column, which is actually a stripping column.

 

From all the different information you gave me, I think it is essential to keep in mind that the condensers even if they are on the top of the column or side heat exchangers are the primary vacuum generators.

The condensers are generally not the vaccum generators, on the vast majority of vaccum columns the vacuum is generated either by a vacuum pump or an ejector set.  Barometric condensers can generate a slight vaccum but never for a deep vaccum as found on a refinery for example (except on very very old designs with exceptionally poor performance).

It's more correct to say that without for example a pumparound on a refinery vac column, the ejectors would be overloaded and unable to pull a vacuum leading to overpressure of the column and an emergency shutdown.

 

In other industries, where the vacuum may not need to be so deep then barometric condensers maybe are still used, thats outside of my area.

[katmar]

SilverShaded, whether the condenser or the vacuum pump is the main generator of the vacuum depends on the relative quantities of the process vapor and the incondensible gas.  In my experience with distillation and evaporation there has always been vastly more process vapor than incondensible gas and that is why I have said that the condenser is the major generator of the vacuum.  But of course this does not always have to be the case.

 

If there were no incondensible gases present then there would be no need for the vacuum pump at all.  All you would need is sufficiently cold coolant for the condenser.  But the next leak-tight column I see will be the first one!

[SilverShaded]

Different industries have different conditions i'm sure, however in oil refineries the condenser is operating above atmospheric pressure and contributes nothing at all to the vacuum.  The exhaust steam leaving the final ejector stage, or from the vacuum pump, is allready above atmospheric pressure.

Barometric condensers, ie those pulling a vacuum, really don't exist on most oil refineries, except for very old/small/backward designs.

[Bobby Strain]

Silver,

   I would heed katmar's discussion. You seem to lack experience. Learn by listening.

 

Bobby

[latexman]

^^^^^

Ditto

 

My understanding aligns with katmar's.  The vacuum pump removes inerts, which allows operation within the VLE P&T envelope.  The condenser makes 500-1000 volume units of sat'd vapor collapse into 1 volume unit of liquid.  With no inerts there to fill the void, pressure plummets.

[SilverShaded]

On a refinery vacuum column there is a range of components and actual inerts being the least present.  The majority of vapour pulled at the first stage is steam, typically 20-30 te/hr.  The primary purpose of the first stage condenser is to remove water vapour allowing the second stage ejector to be sized smaller.  The same principle applies to the third and if present, fourth ejector stages.  At each stage hydrocarbon vapours will also be condensed and recovered as slop oil.

The pressure ratios of the ejectors determine the operating pressures of the intermediate stage condensers.  If cooling capacity drops then pressure will increase largely as less vapours are condensed and the next stage ejector duty is increased, therefore suction pressure has to increase.  As i mentioned previously the final stage condenser operates above atmospheric pressure.

This is how refinery vacuum columns operate, you need to dstinguish between a multi stage ejector set on a refinery (which is not all inerts) and what your describing which is essentially a barometric condenser with a vaccum pump and only pumping inerts.

 

you may want to read this;

 

https://www.graham-m...article-PTQ.pdf

Edited by SilverShaded, 23 February 2022 - 07:49 AM.