4 replies to this topic

[musrey]

This is a rather open ended question, but I am in the early design phase and need some observations and thoughts.

We plan on compressing the vapor overhead of one of our distillation columns. The column seperates an alcohol and water. The overhead is at 1 psig and must be compressed to 40 psig for use downstream of the column. Increasing the column pressure is not an option.

I am uncertain about the dynamics involved in starting the compressor and the impact on the operating column. I am also concerned about starting up the compressor with a stream at its dew point. What capabilities should be included in the design to mitigate these types of issues.

Any experiences from people who operated this type of process would be helpful.

[Art Montemayor]

This is an interesting application.

You write as if your organization has already thought out and planned the use of a compressor on the column overheads as opposed to condensing the vapors and transporting the product as a conventional liquid with subsequent vaporization downstream – if that is needed there.

First and foremost, you have now inherited a new column pressure control system. Distillation columns are conventionally controlled as to their operating pressure by overhead condensers that establish the vapor pressure (which is the column pressure at the top) inside the column’s top vapor space. Since you now do not want to condense the total overhead vapor (you must still use a reflux partial condenser) you must ensure that the vapor draw off rate is controlled. This is one of the tricky deals that you are proposing. Why don’t you tell us what you have planned and how you intend to implement it?

Next comes the problem of selecting the appropriate vapor compressor to do the job. Even if your vapor flow rate is huge, I doubt you will employ a centrifugal type – albeit, its inherent characteristics are appealing in this particular case. A centrifugal can tolerate a saturated vapor better than a reciprocating type, for example. A centrifugal is smaller and simpler to operate and maintain than a positive displacement type. It can also probably tolerate the high temperatures that must be confronted on the suction side – as well as on the discharge side. However, the centrifugal has some drawbacks as well. It is suited to large volumes and a small one is hard to find. It is suited to stable, steady flow; it can’t tolerate a marked turn-down or turn-up in capacity without going through a severe surge analysis and potential problem potential control problem(s). Any column excursion or process upset will affect a centrifugal more seriously than it would a positive displacement type of compressor. If your alcohol is toxic, you will have a formidable problem in finding and selecting a proper set of seals for the centrifugal unit. A centrifugal is only as effective as its seals – and these can often turn out to be the Achilles’ Heel of the operation. If you have an upset or excursion, what will you do with excess process vapors? If you are using a centrifugal, we are talking about a large flow rate. If you can’t condense (remember: that process decision was made when you decided to go with compression instead of pumps), then you are facing a rather large amount of vapors to dispose of or flare. This issue alone would throw this proposal away in many industrial plant situations. The environmental downside potential is a large one if you have to dispose of large alcohol vapor streams into the atmosphere.

So it would seem that you are probably planning on a positive displacement (PD) type of compressor. Is that correct? Confiding in us with your Basic Data and Scope of Work would be a lot easier than making us go through all this guessing and conjecturing. A PD machine will give better control, but will be more sensitive to the alcohols (& other goodies in your stream). Corrosion won’t probably be a major problem, but the fact that you can’t tolerate any oil or other lubricants getting into your process stream means that you must use a non-lube machine. This really starts to narrow down your choices of machine(s). A screw type of compressor is probably the best mechanical match for what you want to do, but it also depends on seals – and since you are going to run it dry (not oil-flooded), you will suffer a lot of slippage through the rotors. This is inefficient, but it also starts to raise the discharge temperatures – making it tougher for seals, bearings, and clearances. Here is where the experience and capability of the fabricator plays an important and vital role.

Your proposed compression ratio of 3.7 might not seem like a lot when dealing with conventional gases, but the discharge temperature is dependent on the Cp/Cv ratio and we don’t know that – like we don’t know a lot of the information you have on hand. Issues such as starting up the machine with saturated vapor can be resolved by experienced designers and engineers. I would be more concerned (from what I know now) with maintaining a steady and constant distillation column - assuming that is the existing situation.

Depending on the compressor you have selected, the amount of money you are planning to throw at it, the experience and capability of the designers and fabricators, and the amount and type of instrumentation you can afford, you might be able to do it. I am pre-supposing that you have done your economic analysis and that vapor handling is the best way to go from an operability, economic, and safe point of view.

[pawan]

Could not understand your / your mgmt decision bcoz it the compression of OH vapors can have following adv/disadv.

1. It is much cheaper to pump liquid than vapors.

2. It is not energy efficient unless U plant to recover heat from compressed vapors (Heat Pump) OR need to re-vaporize it. Economics depends on final pressure ratio for compression energy vis-a-vis latent heat.

3. Disturbances in Distillation process, Maintenance, and associated problems will contribute significantly to the LCC of the system apart from energy.

4. I think LCC will make it very clear for easy judgement if U consider production / process losses due to problems associated with the direct compression. To avoid this U may need very large surge tank volume for reduced impact.

5. In any case U need to provide some reflux which still requires condensation.

Anyway Unless we know specific use, we can't suggest a good option. So if it is possible to share, we can suggest more effective way to achieve your goal.

[abhi_agrawa]

musrey,

I was once involved in a feasibility study, where we studied something similar to to what you have in mind. However, there we used the hot and compressed vapor from the compressor a reboiling medium for the column. There are few such operating columns in petrochemical applications (the one we studied, was a propylene fractionator in a cracker). There the objective was to get all the required reboiling by vapor compression. the control; scheme for the same was quite complicated.

abhishek

[musrey]

Thanks for all the replies. A few comments in the way of explanation:

(1) The purpose of the project is energy savings. All the distillate from this column is currently revaporized before it is used. However, the users are at either 12 or 32 psig. The column overhead pressure is currently 2 psig. It is heat integrated to another distillation column and the pressure can not be increased.

(2) The distilate is 100 % methanol.

(3) The pressure control on the tower is via varying the liquid level (hence heat transfer area) in the condenser.

(4) The users of the methanol vapor are flow controlled. We plan on slowly ramping the flows to minimize any disturbance in the column.