Pumping Supercritical CO2

[GaTechTheron]

I came across an application were I need to pump supercritical CO2 in a multistage centrifugal pump at ambient temps. I have a few questions:

1) What considerations beyond that of a standard pump application do I need to consider?
2) How do I determine HP draw with a varying specific gravity?
3) How do I consider compressibility?

I have looked through some threads on the webpage, but none have been much help on the applications side, and how the fluid is affected as it flows through the pump. If I can somehow estimate the compressibility across each stage, then I can develop a customer set of curves, and determine power draw (I think).

Thoughts?

Thanks in advance.

 

[BigInch]

http://www.criticalprocesses.com/Calculation of density, enthalpy and entropy of carbon dioxide.htm

See if you get the same answers by comparing against this paper.

Independent events are seldomly independent.

 

[GaTechTheron]

Biginch,

How have you approached this in the past?

Thanks,
Theron

 

[BigInch]

The use of Stoner Pipeline Simulator helps a great deal.

Independent events are seldomly independent.

 

[BigInch]

isothermal properties of CO2
(NIST website)

Independent events are seldomly independent.

 

[GaTechTheron]

BigInch,

Thanks for your help. Here is the plan for calculating the pump performance:

1) Do a simple pump curve selection assuming this is an incompressible liquid. Select something that is less than BEP because of the pump will have to account for compressibility.
2) Determine incremental pressure gradient between stages inside of the pump.
3) Assume mass flow rate INTO pump equals mass flow OUT of pump. Use constant mass & varying desnity at each stage to calculate volumetric flow through each stage.
4) Use volumetric flow at each stage to find a point on the curve that each stage will operate.
5) Combine TDH, and HP performance of each stage to get a combined performance.
6) Compare new datapoint with origional pump selection, and post results:

Flow is used as a reference for both scenario, so this is help constant. Otherwise:

Flow: 0% (Used as reference point)
Head: 9% Low
Eff: 10% Low

Based on the data, the stages operating at the lowest pressure (1400psi range) are what are at far less effeciency, less head, etc etc.

2) Now do I need to evaluate rotor support and/or lateral analysis?
3) How about head soak from the atmosphere?
4) How about heat from compression?
5) Any additional seal considerations?

Thanks again,
Theron

 

[BigInch]

Your logic (1)->(6) looks good.
I don't do detailed pump design, so I'm no help to your additional questions.
Evaluating compressive heat is always a good idea. You won't likely get reasonable densities going into multiple stages if you don't.

Independent events are seldomly independent.

 

[25362]

May be contacting a reputable pump manufacturer would be of help, for example:

http://www.sulzer.com/en/-/media/Do...ures/MSDAxiallySplitMultistagePump_E00577.pdf

This company, among others, deals with supercritical CO₂ multistage pumping.