CO2 compression 4

[npf]

All:

I am working on a case where I have to compresss CO2 gas(97% CO2, 3% H2O) from 2 bar to 275 bar to be injected into wellhead. Right now I developed a simulation model w/ 5 stages, I have used 80% efficiency for the compressors.Also I have used intersatge coolers. I have limited my discharge temp after each stage to around 300F.
The CO2 is in supercritical phase after last stage.
I have couple of questions:
How do I determine the need for dehydration system?
Also on the interstage coolers(water), I have used a temperature approach of 10F.Any comments?
On Compressor efficiency, I know we can determine from the manufacturers curves but at this phase I do not have those?
Is 80% good for all the stages?

Your comments would be appreciated

Regards,
npf

 

[dcasto]

The Peng Robison Equation of state is one of the best for this similation.
Here is what the EOS will tell you.

CO2 be come hydroscopic and there will be no frewater after the 4th stage. NACE 0175 2003 recommends all carbon stell construction for all stages.

A 10 Degree F maybe agressive for air cooling, but don't worry about approach as much heat rate, most people forget about the heat of condensation of the water in the intercoolers.

Reciprocating compressor effiency is more a function of cylinder and gas passage area and because of this, only the manufacturer can give you that. I've matched manufacturers data to thermo data and I've seen as low as 70% and as high as 85%. Valve loss is the biggest problem for effiency and molecular size and density maybe come a problem. CO2 will have this and the way to overcome the issue is to slow the unit down and run it under its maximum RPM.

300 F is very conservative, 350 F would be fine, manufactures set this limit mainly for watching the derating of the steel yield strenght and problems with expansion due to temperature differences.

You will find that your third and fourth stage suction temperatures may have to be raise to keep the CO2 under saturated and away from the criticle point. One design hade the intercooler set at 140 F instaed of 120 F per the original design.

If the volume is large, you may want to look at a 4 stage unit going to 1250 psig and the installing a pump to go to the 4000 psig (sorry, I'm an english units type of guy).

NOW the good news, go to

http://www.arielcorp.com

and you can download their compressor design similation program. The program will all info you need. Make sure you use the VGM database for physical properties.

 

[zdas04]

To add to dcasto's points. While each cylinder may very well be 70-85% efficient, you'll never get those kind of numbers if you compare the energy input to the increased enthalpy across the total compressor. Aux loads, friction in piping and coolers eat this up. The number that I usually come up with in a rigorous analysis is 68% for a single stage unit and drop that 6% for every added stage. Your 5 stage skid should be expected to be around 38% efficient (based on energy input vs. enthalpy increase, ). Anything above that is just wishful thinking. Oddly, your 80% raised to the 5th power is 32%

dcasto's point about watching stage-inlet temp closely is really important. If you try to stay at 200F inlet to stage 5, then your maximum allowable discharge temp will limit you to about 1.6 ratios and you'd probably be better off with a pump.

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

 

[npf]

dcasto and zdas04:

Thanks for your valuable comments. I think I need to stay away from the CO2 triple point also .One more question on HYSYS. Can we generate phase envelope for CO2 in HYSYS. If so how can we do it?

Thanks & Regards,

npf

 

[BigInch]

No stars yet?

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[Montemayor]

npf:

If you are simulating a 5-stage reciprocating compressor, your staged pressures should be something in the vicinity of:

1st Stage suction = 29 psia
2nd Stage suction = 78 psia
3rd Stage suction = 208 psia
4th Stage suction = 557 psia
5th Stage suction = 1,490 psia
5th Stage discharge = 3,989 psia

You’ve failed to state whether your pressures are absolute or gauge; I’ve opted for absolute; otherwise the compression ratios can’t be calculated since they are in absolute pressure.

Contrary to what you state, the CO₂ starts to go “critical” after 4th stage, not the last stage. This is assuming the above discharge pressures, which are based on the normal and conventional method used to determine the number of stages: that the work is equal in each stage. That means you’ve got a special situation and, as dcasto says, the best (&most efficient) bet is to pump liquid CO₂ instead of compressing supercritical fluid.

Regarding dehydration, we don’t know what your other requirements are, but you’ve a need to separate liquid water prior to each compression stage. After the 4th stage, the remaining water of saturation will probably go into solution. At this stage you are in the supercritical and a lot of funny and strange things that aren’t documented, start to happen. You may get by with minimal carbonic acid corrosion; I would research this before being certain. If you can’t tolerate any subsequent corrosion, then you’ll have to dehydrate – probably with an adsorption unit using activated alumina. This can drop your water content down to less than -90 ^oF dew point, if you want it that low. I would dry the gas at the 3rd stage discharge, after intercooling. You have to state your constraints or accept the basis.

A 10 ^oF approach should not be a problem to efficient water intercoolers.

Compression efficiency is the ratio of the theoretical to the actual gas horsepower and , as used by the industry, does not include mechanical friction losses. These are added later either through the use of a mechanical efficiency or by adding actual mechanical losses previously determined. The mechanical efficiencies of positive displacement compressors range from 88 to 95%, depending upon the size and type of unit. Dynamic compressors commonly add actual mechanical bearing and seal friction horsepower. Dynamic units also have certain relatively small hydraulic losses that are often disregarded for estimating purposes. Positive displacement machines (such as reciprocating compressors) are compared to the adiabatic cycle while dynamic units (centrifugal machines) generally use the polytropic. Here, I am quoting Charlie Gibbs who wrote the classic “Compressed Air and Gas Data” for Ingersoll-Rand Company way back when. Since you are only doing simulation (or wishful thinking) 80% is a safe conservative estimate at that stage of design.

I believe you are on the rational trail in keeping your compression discharge temperartures below 300 ^oF. In fact, I always design for 250 ^oF because I know that any reciprocating compressor valve is going to fail seating sooner or later. When it does, the discharge valve temperature will start to clim and any cylinder lubricating oil on the market (except for the synthetics) will start to decompose and gum-up at 300-350 ^oF. any discharge valve failure will very quickly attain 400-500 ^oF. This, I’ve proven out in the field.

I don’t understand your concern for getting to the triple point. You have no reason to go there – and nor do you want to even wander there. The triple point for CO₂ is approximately -70 ^oF and what reason would you want to go there for your application? I think that point is best forgotten.

Why would you or any one else want to generate a phase envelope for CO₂ in HYSYS when you have so many, proven and established T-S diagrams for this common gas? I guess you can always ask your Hysys rep.

I hope this experience in CO₂ compressors helps.

 

[dcasto]

The GPSA data book has the PH diagram. The Ariel compression program will map one too along with the plot of your compressor pressures and temperatures.

 

[npf]

All:

Thanks for your valuable comments.I hsve 2 cases- 4 stg + pumping supercritical CO2 and 5 stages w/ high intercooler temperaure.
The dewpoint specs are -20F.Is there a way to determine water dewpoint for known composition in HYSYS.Or do I have to use y/x (equilibrium ratios). My outlet CO2 is 99.5% and H20 is 0.5%

Any comments would be appreciated

Regards,
npf

 

[npf]

All:

My previous e-mail is unclear and numbers werent accurate.
The dewpoint specs are 23F in an operation range between 750 and 3900 psi. The outlet CO2 is 99.8% and H20 is 0.2%. Is there any way to determine dewpoint in HYSYS. If I use y/x equilibrium ratio method, where can i find charts for H20 and C02

Any help wouild be appreciated

Regards,
npf

 

[dcasto]

Peng robinson EOS will give good results, as for Hysys, I don't know the call function, in Winsim its a check box on the fluid stream od a "Phase envelope" process.