6 replies to this topic

[Ichigo]

I am student in Nottingham University and working in designing the upper distillation column, which is a low pressure Packed column. The suggested packings for this column are Mellapak structure packing produced by Sulzer.

I have designed the column in HYSYS and now I am trying to convert and validate this design into a real working column. I have to discuss all the mechanical design of this column.

So, Could I get some help in identifying some ideas of what should I design or even look at in term of the mechanical design?

The column is as shown:
Number of theoritical stages: 83
The feeds are to be connected at these stages:
•Liquid crude oxygen feed : stage no. 22.
•Vapour crude oxygen feed : stage no. 30.
•Liquid returned from argon column: stage no. 41

Side-stream take-offs:
•Argon side-stream : stage no. 40
•Waste nitrogen side-stream: stage no. 4

And the the flows as follows:
The table below show the amounts of each streams (feeds and side streams):

Stream type, The stream, The amount/ (kg/h)

Feeds: Reflux N2 (56400)
Liquid crude O2 (44710)
Vapour crude O2 (29800)
Liquid returned from argon column (36200)
Bottom feed (80000)

Draws: Bottom (O2) product (93400)
Argon side-stream (5610)
Waste N2 (37800)
Top product (N2) (11000)


And due to all this division for feed and sidestreams, the packing bed will be divided into 5 divisions.

So, what typical values for the spacing between these beds (divisions)?

And, can I use different packings for one bed and another one in the other beds? This is because the different flow of vapour at different places in the column, causing to have different HETPs for each division.

Also, is there any special features considering the cold box and the low temperatures involved??

Regards.

[Art Montemayor]

Ichigo:

Welcome to our Forums. This is a nostalgic thread for me since my first professional assignment in my first engineering job, straight out of university in 1960, was to go to Kingston, Jamaica and assume production manager responsibilities for various industrial gas plants – among them were 3 Air Products air separation plants. The air separation plants were old ones and frequently needed maintenance and I had to dismantle one and modify it. Therefore, I quickly acquired a first-hand experience on just what is inside the cold box and how it works. I later went on and operated several more air separation units in other countries – and modified those as well. So what I’m going to comment to you is based on personal, first-hand experience involving the mechanical design and operation of air separation units in the field.

First and foremost, always bear in the mind the basic and most important basic data about your design: the fluids you are dealing with are not only all in the cryogenic zone, but they are all PURE and basic ELEMENTS. By this, I mean that you are designing a process that is devoid of impurities and can be easily designed on the basis of pure fluid properties that are well defined and databased. That is a great advantage and certainly very good news for a design engineer.

Now for the BAD NEWS! All the rest of the design is not easy - nor is it conventional - and will cause you and others a lot of brain cell consumption in working out the unique and difficult problems confronting a cryogenic system. As it always goes in engineering, there are no free rides and no free lunches; you always have a trade off to confront in order to solve a problem. It comes with the territory.

1. Your distillation columns – both the HP and the LP ones as well as the intermediate Nitrogen condenser – will essentially have to be 100% special stainless steel and integrally welded throughout. What this means is that there are no conventional “manways” or manual “access” entries into the system. Therefore, once you make the last seal weld on the columns, you are at the mercy of your operations maintaining a perfectly 100% pure air feed into the cryogenic section because you will not have access to the internals in the future after you startup the unit.
2. Of course, the real life experience will be that there will always be a slippage of contaminants as time goes on. Primarily water and CO₂ will slip through – even in parts per million they will accumulate within the column and after a certain time, you will be forced to shut down the unit and go into a “defrost” operation where you force in hot air throughout the system to purge it of accumulated frozen impurities. Depending on the mechanical design, this can be done every year or two – on a planned turn-around. What this means for a mechanical design is that your columns are going to have to withstand -350 ^oF or less and then be subjected to +250 ^oF or more. These are terrific and abnormal stresses to impose on any process equipment. Therefore, your equipment – especially the piping - has to be designed to flex and expand/contract as the temperatures require it.
3. You must design for the ability to defrost (“de-rime”) the cold box on a given notice. Therefore, you must build in expansion/contraction flexibility. You cannot rely or employ conventional “corrugated bellows” type of expansion joints. Conventional expansion joints cannot be trusted within the cold box. Part of the basic Scope of Work for the cold box should be that you cannot tolerate a fluid leak from the cryogenic equipment. To have this happen is tantamount to a potential catastrophic disaster within a separation plant. Therefore, your mechanical design must ensure 100% dependable for integrity and safety. This part of the design is not negotiable nor can anything less be acceptable.
4. Together with equipment integrity, you must ensure the same integrity with the shut off valves employed throughout the system – even those outside the cold box. The valves handling cryogenic fluids must have the ability to positively shut off all fluid and undergo rigid and tough performance tests before being found acceptable. Needless to say, the cryogenic valve design is a special one – especially in that the operator must have access to the manual valve handle and yet not be subjected to having water ice formation stop the turning of the handles. This is usually done by using special extended valve stems that are insulated by stagnant gas blanket.
5. The need for allowance of expansion/contraction forces to dissipate safely has been mentioned for the piping and valves. The column itself had the same requirement and – just as important – also has the basic Scope of Work basis that it must be maintained 100% plumb while it is moving in a vertical direction. This is extremely important in order to ensure that product purity levels are maintained. The product purity, of course, depends on the successful distribution of liquid and vapor throughout the column. Any liquid channeling within packed beds will result in off-spec product. Therefore, strategically placed expansion/contraction guide rails must be installed on four locations around the column at several heights in order to ensure that expansion/contraction of the column takes place in a perfect plumb direction all the time.
6. Because of the required cryogenic operating temperatures, the column itself and all other related cryogenic cold box equipment has to be insulated from transferring outside conduction heat through any structural stainless steel supports (such as legs, stiffeners, etc.). The insulation material selected should have very low thermal conductivity values but also have strong and tough stress properties as well. Do not forget that no oil, grease, hydrocarbon, or any flammable substance can be allowed in the system or inside the cold box. You must assume that sooner or later, it could come in contact with pure liquid or gaseous Oxygen and the result would be instantaneous combustion - otherwise known as an explosion! This is just not acceptable. If you have done your homework in thoroughly researching air separation plant operations and their history, you will have discovered that cold box explosions have taken place before because of what I have mentioned. This has caused needless loss of human life and equipment and as engineers we must learn by our past mistakes.
7. Always remember that the operators will only see the external surface of the cold box and, if you succeed in producing a good design, will never get to see the internal vessels, piping, and structural pieces. However, the real truth is that eventually you will have to access the cold box internals (if at least for inspection) so that you must allow for gasketed paneling of the cold box external walls. You also have to furnish outside platforms, ladders, and lighting that has to be fixed to the external panel walls. Therefore, you have to have a sound and strong box structure for the cold box containing all the cryogenic equipment. You also have to devise a method for efficiently removing all the internal loose insulation (usually "Perlite") that fills the inside of the cold box and keeps the cryogenic equipment insulated.
8. Typical values for the spacing between the packed beds depends on the size of the column and the type of liquid re-distributor as well as the allowable superficial vapor velocity. In the case of a 100% welded and sealed column, you don’t have to figure on human access in between packed beds. Usually, if you have to access the column, you have to remove all the packed beds – starting from the top and working down. That’s why it is so important to do an excellent column mechanical design from the very onset. The need for cutting a column open in order to gain internal access triggers an economic downturn for the operation and is a major negative experience. You do not want that to happen and you don't want to go there. Make damn sure all your internal mechanical design has strength and process integrity. Once installed, there is no turning back.
9. There is no reason why you can’t employ different sizes of packing materials – or even different types – in each packed bed. As long as the design is done correctly and the vapor/liquid distribution is kept efficient, there should be nothing wrong. Don’t forget: with pure fluids and stainless construction, you have almost “zero” corrosion allowance. Therefore, you should have no concern about your stainless steel beds and construction requiring frequent inspection for corrosion or erosion.

That’s all I can come up for now; I’ll probably recall other factors and items to comment on, but I’ll pass on this information for now. Perhaps you will have some questions or comments yourself.

Good Luck.

[Ichigo]

First of all, thank you alot for the discussion above. It is really helpful, giving the fact I have no real life experience in real working plants, despite some visits.

As for the material to be used, what kind of special stainless steel is to be used? I am guessing alloyed with some material. Could I have some insight and specifics on that?

Also, what about the feed inlets and nozzels? and how the their size can be determined? And is there any special cases in dealing with these?

My warm regards, the above ideas were much of a help.

[Art Montemayor]

Ichigo:

The Stainless Steels used in cryogenic services are generally of the Austenitic grades - such as the "300" series. However, in the case of designing an air separation column what is always done is that the steel company(ies) are consulted as to the best alloy to use under the stress and temperature considerations. You have nothing to worry about with respect to corrosion; but the allowable stresses at the temperature conditions have to be identified with accuracy. Steel manufacturers are continuously improving the quality of the stainless product and that decision is best left up to them.

I have discussed nozzle sizing in another thread and I recommend you read that. The size of the nozzle should correspond to the size of the pipe it is connected to and the to the pressure drop allowed through it. It is that simple. There is no "magic" equation to calculate nozzle sizes.

I don't understand your question about "special cases" dealing with nozzles. Nozzle are simply a means to introduce a fluid into a vessel or to remove that same fluid from the vessel. Their size is predicated on fitting the corresponding pipe and not causing fluid flow problems - such as large pressure drops.

[Ichigo]

Thank you.

The material to use is clear now, although I still have to find the typical stress design at such low temperatures. Any recomended data book to use? (for deciding the thickness, preminilary design).

As for asking about any special cases, I meant any special case dealing with the fact that there is low temperature in the system.

Regards.

[Art Montemayor]

Ichigo:

Regarding special situations involving cryogenic fluids, there is always the need to have a drain line and valve connected to each low spot of your columns and condenser. This is the part where you have to use your common sense and design for the ability to drain the cryogenic fluids out of the system before shutting the unit down for a defrost (or de-rime). Remember what I said at the very beginning: "the design is not easy - nor is it conventional - and will cause you and others a lot of brain cell consumption in working out the unique and difficult problems confronting a cryogenic system."

If you have to make that type of drain connection you must ensure that the line and the drain valve are free of cryogenic temperatures in order to avoid any contamination or plugging with condensed and frozen atmospheric water & CO₂ deposits. The way this is done is that a "vapor trap" is created. This vapor trap is nothing more than a creation of a gas atmosphere at the end of the drain line, right up against the drain valve. This vapor trap is nothing more than a line that goes up vertically well beyond the level of the liquid inside the vessels (usually as high as you can go) and then vertically down and out through the cold box panels and up to the drain valve. This type of piping configuration creates a gas trap and, since gas has a horrible coefficient of thermal conductivity, an insulated barrier between the cryogenic fluid and the drain valve. You will find that this maintains the external drain valve at atmospheric temperature and causes a minimum of heat leak through the piping and valve. This is just another ingenious way to overcome a serious and complex problem with a simple technique.

You will surely come up with more challenges in your design - like: How do you measure and monitor the liquid levels in the sumps of the HP column and LP column - as well as in the Nitrogen condenser? I think you can figure that one out on your own. Just consider what I described above.

[hosein]

Dear friend
I am student at ch.sharif.edu
working on the same matter, But I am at the first steps.
thanks for the data. have you find your answer anywhere? was it useful? can you share some data with me.
I am working in ASPEN and I have only modeled bottom column and argon columns I am now focusing on low pressure oxygen column as you and my goal is to understand this column.
you will make me happy if there is more data on this column.

I will share new data with you as soon as i got some.