2 replies to this topic

[rbagster]

Hello all, please be kind. I have been thrown into the process engineering department at my company with no air separation experience (I am an ME). I have been asked to determine the unit powers associated with an oxygen plant - separation power, product compression power, and liquefaction power (not from a separate liquefier); all associated with the ASU. I have been trying to find any writeups about this calculation online or in a book to no avail.

The system has a base load air compressor with booster, a lower column turbine booster, upper and lower column, Argon column and the primary heat exchanger. All pretty standard ASU equipment. I have the Hysys model for the plant listing pressures and temperatures and I have the design data for the air compressor and plant design power. The ASU puts out gaseous and liguid Ox, N, and Ar.

Question, what is the equation for calculating unit separation power? P=(FairxTxln(Pd/Pi))/efficiency? Then to get unit power do I divide by the product flow; GOx + LOx since it is an oxygen plant?

And lastly, how do I calculate the liquefaction unit power? I know that all 3 unit powers have to add up to the total power and if I can calculate 2 of the 3 I've got the 3rd. Or I can just combine product compression and liquefaction into one number and then I only have to calculate separation unit power.

Any help would be appreciated. I've been banging my head on my table for weeks and the flat spot groweth and I can get 3 different answers which isn't good. It takes me 2 weeks to get in to talk to someone for 15min and that isn't long enough to help but plenty long enough to make things worse.

Thanks for anything
rbagster

[Art Montemayor]

Rbagster:

You state you have been asked to determine the unit powers associated with an oxygen plant:

• separation power;
• product compression power; and,
• liquefaction power

You also state you are operating a “standard” Air Separation Unit. But you don’t tell us if you are to determine the theoretical (calculated) values or the actual, empirical values. In other words, does your supervisor want you to calculate the power expected to be consumed? - or is the actual value of the metered power the answer desired? Once the ASU is in operation, the calculated values only serve as an indication of the operating efficiency of the unit. The real, important power consumption figures for an operating ASU are the metered quantities that the unit is supposed to pay for.

I wouldn’t worry a minute about being a mechanical engineer in charge of an ASU. When I graduated as a chemical engineer and I was assigned to operate 3 ASU units, I was very lucky in getting an experienced and knowledgeable engineering mentor – and he was a mechanical engineer. Most of my early hands-on experience in air separation units I owe to him. And the first thing he taught me was to approach the separation of air from a common-sense level. And he was right, there is nothing simpler and easier than separating oxygen, nitrogen, and argon from an air mixture. Allow me to tell you why:

• The basic process of air separation is based on a simple Unit Operation – which deals only with physical changes – as opposed to a Process Operation which deals with chemical changes. Hell, I’d rather deal with Unit Operations any day of the week. Things such as pumping, heat transfer, distillation, filtering, compression, etc. only deal with simple physical operations. There are no complex reactions or side results to understand or handle. Therefore, air separation is right up there as one of the simplest operations.
• The process is still the same, basic operation as Karl von Linde developed it in 1895. It has been improved by incorporating Georges Claude’s invention of the expansion engine (1902) and a method for removing water and other impurities by adsorption (usually molecular sieves). So we are talking about a process that is over 100 years old! This is not nuclear science; it is much easier.
• The original, basic principle of liquefying air was to introduce enough compression energy that allowed you to auto-refrigerate the air by free expansion, subsequently liquefying it. At first, only a Joule-Thomson expansion was used; later, Claude introduced the expansion engine and the economics really took a positive turn. The total energy requirements for the production of pure Oxygen and/or Nitrogen was the compression of the air – that was all!
• Now, in today’s economy, we require more flexibility and methods to reduce costs. Therefore we have introduced more efficient turboexpanders, centrifugal compressors (which are less efficient that reciprocating units, but handle bulk quantities and introduce no oil into the air), pre-cooling cycles, and switching exchangers to remove bulk water and other impurities. Also, the emphasis is on large, bulk quantity plants producing liquids in order to reduce the unit production and distribution costs. But the basis of the process remains the same: you consume energy (usually electrical power) to produce the pure products – either as liquids or gases. There are other minor power costs involved – such as for pumping liquids and regenerating some adsorbent. But this is a small quantity. The basic reality is that all you need to produce pure Oxygen, Nitrogen, and Argon are power and air – period!
• There is no logical separation of the power requirement as you have expressed it. There is no such, detectable or calculable quantity of “separation” energy. The entire distillation process is self-contained and no external, identifiable energy is required to carry out the distillation. The heat required for distillation is the incoming, relatively hot air.
• There may be some product compression requirements. If only gaseous oxygen and nitrogen are produced, then these are compressed for either storage or distribution driving force through pipeline. You don’t give us a description of what (and how) you are producing your products, so we can’t digress any further.
• The energy needed for liquefaction is in the initial compression of the air. However, there are some liquefaction cycles that are optimized and based on pre-chilling and using turboexpanders. How you are running your unit is not revealed by you, so we are left in the dark. There are various ways you can design your ASU – it all depends on what you want and how you want to operate it.
• The bottom line is that all you should have to calculate (if that is what you have been assigned) is the power requirement to compress the air feed. Then you should add to this requirement the power used by any additional chiller or refrigeration units. To this energy input you may have to discount the power generated by any turboexpanders that you may be running – if you are putting the generated power back into the power line and not into your compression trains. Any need to pump or consume energy to store or move the product (such as compressors) should also logically be added or accounted for.

I hope I have helped you lessen your headaches and convinced you that the task before you is a rather simple one and not as dark and complex as some may think.

[jeffreyfrog]

Air separation plants have been considered which are meant for simultaneous production of gaseous and liquid oxygen, nitrogen and argon, and which are designed according to the schematic diagrams of low pressure, medium pressure and two pressures with a nitrogen cycle. On the basis of the procedure developed for technical and economical optimization, the optimum values of oxygen and argon recovery from air have been defined; the relationships between power consumption and reduced costs as well as between other parameters and the fraction of the products in a liquid state have been described. Presented are the ranges of the liquid product fraction variations for which the use of a particular schematic diagram of the plant is essential.