Main Air Compressor for an Air Separation Plant

[Panipat]

I am trying to calculate the required Air Flow for an Main Air Compressor.
Example: The Air Separation Plant Size is 3600tons (metric) of Oxygen per day.
What will be the air flow for the Main Air Compressor?

Thanks for your help!

 

[gtsim]

1 kmole of Air contains 0.2095 kmoles of O2

To get 1 kmole of O2 we need 1/0.2095 kmoles of Air

3600 x 1000 kg/day of O2 = 3600 x 1000/31.999 = 112503.52 kmoles of O2/day

Therefore to get 112503.52 kmoles/day of O2 we need 112503.52/0.2095 kmoles/day of air, which is 5370009.62 kmoles/day of air.

5370009.62 kmoles/day of air = 5370009.62 x 28.95/1000 = 15546.43 Tonnes of Air

31.998 is the Mole Wt of O2 and 28.95 is the Mol Wt of Air

 

[zdas04]

I'm in the process of designing a project to take the other part of air (nitrogen obviously) and it is REALLY important to define your process first. If you are doing membrane separation you need to compress about 3 times as much air as you expect to recover of your target product. For a well-designed cryo plant the number is closer to 1.4 times.

I would multiply the appropriate factor times gtsim's value (and then add 20% so that I'm not operating the compressor at its surge line).

You also need to be pretty careful of your compression technology--recips and oil-flooded screws will always put too much oil in the stream for downstream processes. Most folks I've been talking to use axial or centrifigul compressors for air compression.

David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

The harder I work, the luckier I seem

 

[cryotechnic]

First of all,
In our air separation plants where are talking in units like Nm3/h. (I think it's common in the whole air separation industry to talk in Nm3/h)

So I converted the figures above in the reply of "gtsim" into m3/h.
Density Air = 1.29kg/m3
Density Oxygen = 1.43kg/m3

(15546.43*1000/1.29)/24 = 502.131 m3/h air
(3600*1000/1.43)/24 = 104.895 m3/h oxygen
In this calculation I considered that you want Oxygen with a purity of 99,5%. If you want lower purity, you need less air.

Above calculation shows that you will need a very large air compressor. I have to say, that we have a few large ASU's in our company, but, this large we don't have them..... :-(
(the largest we have is 300.000 m3/h air and produces 65000 m3/h oxygen, and this is allready a large plant)

A compressor this big, that's need to be a axial, or centrifugal one (or a combination of both for that matter).
The whole proces would be cryogenic. For production on this scale, there is no membrane technology I think.

By the way, in our small onsite unit's (2500m3/h air) we are using oil-lubricated screw compressors. It's not a problem, as long as you use oil/water separators, and oil filters before the air enters the mol-sieve.

Good luck
Cryotechnic


"Math is the ruler of your potential succes...."

 

[Panipat]

Ladies and Gentlemen, thanks a lot for that information.

the calculation would then be:
3600 tpd x 1000 / 24 / 31,99 / 0,2095 x 28,95 / 1,29299 = 500984 Nm³/h

There are 2 issues which I still need to solve:

1.) Is there a common understanding that a 3600tpd Oxygen Plant would produce 3600tpd O2 under Norm conditions (0°C, 1013mbar, 60% humidity)? Or do I have to consider site conditions? Is that a rated or a nominal figure?

2.) according to zdas, I need 1.4 times more air than expected (It's a cryo plant). That would mean 1.4 x 552429 Nm³/h = 701378 Nm³/h

By the way, the compressors are axial type, API machines.

 

[cryotechnic]

I'm not an ASU designer, but from my experience, when a plant is designed, the design is based on what the plant needs to produce. That would be 3600t Oxygen p/d, purity of (?) in your case.

I would discuss the factor 1.4.
Personally I think that's too much. A factor between 1.16, 1.20 would be more the case.
The recovery of the ASU's these day's is getting better. And the engineers will design more and more "on the limits", because the ASU's need to be more and more cost competitve.

What kind of plant are you designing? I mean, would it only be GOX or also LOX? Are you using the nitrogen? Do you have Argon production?
Do you have expirience what operation a cryo plant?

Good luck!
Cryotechnic

"Math is the ruler of your potential succes...."

 

[zdas04]

cryotechnic,
Things may be getting better, but I got quotes from 4 different companies in the last two months and in every case the air-compressor capacity was really close to 40% over the nitrogen-volume requirements.

Must be a coincidence.

David

 

[cryotechnic]

zdas04,
I thought about what you said. And the last day's I did some calculations.
I found what's going on. Partly your right, BUT, you are talking about the nitrogen part. I was talking about the oxygen part.

When I did the calculations on the nitrogen part there will be factor of 1.4 (just like you said)

You will not recover all the nitrogen, there will be allways waste-gas/regenerationgas which is mostly nitrogen. (depending on the plant, it has about 2% oxygen and some argon in it)

In our large ASU's the production is completly based on the oxygen, ofcourse we use the nitrogen, but it's actually a wastegas of the oxygen production.

What I was wandering, in your first reply you said: "and then add 20% so that I'm not operating the compressor at its surge line"
I mean, is there a "rule of thumb" for that?
I'm not a compressor engineer, I allways assumed that the compressor was designed according to the desired discharge pressure and the total flow. From that there would be an expected surge-line, and during the commissioning the real surge line is determined, and finally the vent-line/action-line is set a little under the surge-line.

Greetings,
Cryo

"Math is the ruler of your potential succes...."

 

[zdas04]

Motors, engines, and compressors all come in discrete sizes and capacities. No compressor-engineer worth his salt will recommend equipment that doesn't fit these discrete sizes. Consequently, if the calculations say you need a 375 hp engine you'll look for an engine that has at least 375 hp plus a factor for auxilliary loads plus some size of fudge factor. I find that auxilliary loads and the fudge factor often come in at about 20% over the simple gas-compression calculations.

For compressors, folks tend to throw in a bit of fudge for valve problems, seal leakage, oil problems (depending on the compressor technology). I don't know that 20% is any kind of "rule of thumb", but you have to make sure that you have enough machine to deal with reasonable deviations from ideal conditions.

David

 

[Panipat]

Hi zdas04,
in smaller sizes you are right: compressors come in discrete sizes and capacities. But for large ASU the rotating equipment is fit for purpose, customised for the specified duty.
As you mentioned, an important aspect is the turndown capacity of the compressor as you have to start up the compressor train.
We are also looking just for the Oxygen, and I assumed that this above mentioned factor is not 1.4 but 1.07.

 

[cryotechnic]

I agree with Panipat,
for these large compressors it's all fit for purpose, it's not an off the self product. Maybe the compressor houses are in some cases the same, but the rotors and the compressorwheels are definitly designed for the project.

Ofcourse during the design they have to deal with pressure losses in the line from the compressor to the ASU etc. But that's where the design of the rotor is based on.

Like I said before, in such projects, they want to design on the limits, they want to keep that factor as small as possible, because every Nm3/h overdimension extra cost a lot of money. Not only in the design, but later also in the totalcosts because you will use much extra energy.
And since energy is the main part of the costs of cryogenic production they're are allways challenging to cut the energy consumption.

Cryotechnic

"Math is the ruler of your potential succes...."