supersonic compressor

[rotw]

http://www.dresser-rand.com/news-in...pression-represents-game-changing-technology/

I am quite convinced that supersonic compression represents a serious market potential for CO2 applications and equivalent.

- But do you think this is a technology that can be deployed, say in the short/mid term, for lower molecular weight applications; e.g. natural gas (say in the 17-22 MW range) or even much lighter gases (such as hydrogen)?

- What would you expect as limitations / challenges for extending the range of applications to lower MW?

I would like to quote the following article:

http://www.ramgen.com/tech_shock_rampressor.html

and the following table:

Gas Molecular Weight Mach 1 (ft/sec)
Hydrogen 2.0 4290
Methane 16.0 1440
Ammonia 17.0 1410
Water (water vapor or steam) 18.0 1400
Air 28.9 1130
CO2 44.0 880

As a matter of general discussion and being quite ignorant on this new technology, I suspect that when going toward lower MW - the following problems will become even more acute:

- Increase of peripheral speed.
- Increase of volume flow / capacity thereby increasing diameter / frame size.

For example - as far as tip speeds are concerned, here are some of the achieved performance:
Tip speed for CO2 (MW 44) : 1600 ft/s
Tip speed for Air (MW 28.9): 2100 ft/s

I am curious about which material would be capable to withstand 2100 ft/s and above while lighter MW are envisaged?

I will be glad if you can share your thoughts.

 

[georgeverghese]

From the first DR link, there's a clue to what the compression eff for this machine is, given that it results in a discharge temp of some 288degC for a compression ratio of 10:1 for CO2. Assuming that suction temp was 25degC in this test, I get a polytropic eff of 80% for a Cp/Cv of 1.28 applicable to CO2 in this P/T range, going from 100kpa abs to 1000kpa abs.

This is no different from what one would get in a typical centrifugal compressor running at subsonic impeller tip speeds these days. In any case, for natural gas / other refinery gases that may contain small amounts of components that could thermally crack at high temps, going beyond 150degC discharge temp is not a good idea no matter what technology you use.

The narrative is also somewhat misleading in the later paras on the topic of waste heat recovery - one can derive the same waste heat recovery benefits with regular centrigugal compressors also.

 

[zdas04]

You can get 80% eff per stage in a centrifugal, but to get 10 compression ratios would require 3 stages, and since skid efficiency is the product of the stage efficiency, you get something closer to 50% doing it with a centrifugal.

Also the waste heat with 10 ratios per stage is concentrated enough to be worth bothering with. The waste heat in a typical centrifugal stage is too dilute to be worth bothering with.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[rotw]

For an overall compression from 1 to 200 bar or higher, a typical centrifugal compressor train for CO2 would be either integrally geared type or of between bearing design type.

In the first case, only very few manufacturers are capable to manufacture integrally geared compressors (often with up to 8 wheels) with well established references and proven technology. Its also actually quite complex equipment requiring significant maintenance and their reliability is definitely lower than between bearing design type compressors.

The cost is also relatively comparable to between bearing type centrifugal compressors for same duty - probably slightly cheaper but not that much. The possible real advantage is reduction of foot print and efficiency advantage -
although not huge - due to the fact that the flow coefficient can be increased by spinning the wheels faster, e.g. use of open impeller - titanium made, and put IGV in between each impeller.

As for between bearing design, its quite long skid involved, common configuration would be two casings back to back type each.

You would need to inter-cool at about 4:1 compression ratio starting from say 40C inlet temperature. If you push to extreme discharge temperature, probably 5 or 6:1 can be achieved per stage before you need to inter-cool. The polytropic efficiency could be well above 80 % per stage (means before to cool down) but when going toward high pressures (100 bar and up) efficiency would drastically drop (e.g. 60%) because actual volume flow would reduce drastically unless you put in huge mass flow but that would make the overall train huge, possibly not what is commonly done.

What I find definitely interesting with supersonic CO2 compression is that equipment seems to be very compact and with lesser complexity. Efficiency seems not to be limited by actual volume flow. But these are just my assumptions. I could be wrong. And I still wonder how this would translate for lower MW gas applications...

 

[rotw]

Zdas,

I have one question relatively to this, just to clarify my knowledge:

You can get 80% eff per stage in a recip, but to get 10 compression ratios would require 3 stages, and since skid efficiency is the product of the stage efficiency, you get something closer to 50% doing it with a centrifugal.


I guess your point is 0.8^3 per three stages equals roughly 50%.

I thought that for polytropic efficiency, I would consider the average of each stage efficiency for a thermodynamic section (which has NO inter-cooling) to arrive at the overall efficiency. So if each stage is 80% then overall it would be about 80% as average.

However in your statement quoted, have you multiplied stage efficiency to get the an overall efficiency (which you rightly called skid efficiency), because inter-cooling was considered between stages? Could you clarify ? I really might be faulty at my basic understandings of thermodynamics.

Thanks for your help

 

[zdas04]

rotw,
First, the word "recip" in my post and your quotation was a typo, I've edited it to be "centrifugal" in my post.

If we define "efficiency" of any compressor as the change in fluid enthalpy divided by the input energy (other methods and definitions were developed because fluid enthalpy is not the most accessible of parameters), then a 3-stage unit's stage efficiency would be that stage's change in enthalpy divided by 1/3 of total input power. When I've done that calculation carefully on multistage recips (I do mean to say "recip" this time), I get that the skid efficiency was the product of the efficiency of the three stages. Which closely matches the number I get if I divide the change in enthalpy from the skid inlet flange to the skid outlet flange divided by total power input.

I've never done the calculation without interstage cooling and do not know from my personal experience how that impacts's the calculation. The way it seems to work is that the second and third stages must add the energy transferred to the coolant medium plus the energy of raising the pressure. Without the cooler the fixed power input will raise the pressure more per stage than a unit with an interstage cooler, but the change in enthalpy would be about the same in either case since we don't do interstage cooling to improve efficiency but we do it to keep the fluid temperature within manageable limits.

The same concept holds true for skid compression ratios vs. stage compression ratios--a 3 stage compressor with 2.5 ratios per stage would increase skid pressure 2.5³=15.6 times instead of the sum of the average (i.e. 2.5*3=7.5).

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[georgeverghese]

Have never heard of any compressor, recip or centrifugal, operating at beyond 180-190degC discharge for clean gas, thermally stable chemical component(s) application, and I presume this is due to some upper temperature limits for the materials of construction used for the shaft seals. Dont know how DR set this machine up to allow continuous operating temperatures going beyond 250degC. Oil free air compressors run up to discharge temperatures of approx 200degC, but these dont last for long (less than 2years cont. run time or so?) , from what I've heard from plant operations staff.

 

[rotw]

georgeverghese,
As for centrifugals, shaft seals are not systematically limiting the temperature to the levels you have stated, even though it is true that what you have indicated is generally considered as an operating temperature limit. There are several applications where the use of labyrinth type shaft seals allows to go beyond 200 and even 220 degC discharge temperature. Example: CO2, Air or N2 applications. In addition back to back typologies and alike do not immediately subject the shaft seals to high discharge temperature since the discharge side would often be located toward the "middle" of the casing span ; such typologies would mitigate the problem to a certain extent and relief the shaft end seals from such excessive thermal constraints.