PUMP OR COMPRESSOR CO2 AT SUPERCRITICAL CONDITIONS 1

[ratash]

How do you determine whether to pump or compress 97% co2 into a pipeline, when it is in supercritical region?

 

[dcasto]

it doesn't matter, as long as you don't go into the liquid zone.

For example, we have 3-3000 horsepower electric drive IR HHE 2 throw (8" by 14" stroke") compressors on C2=, suction 850 psig at 60F and discharge 2400 psig. There were time we needed a lower rate so we installed a Bingham MSD split case pump with 8 impellars. suction 850 psig 60F (on the same header as the reciprocating compressors), discharge 1650 psig.

The pump was far cheaper than a reciprocating unit, but its eff is lower. NOW, the impellars are not a stainless steel as in most pump operations, they are aluminum alloy to dissapate heat better.


We also has booster pumps on the ethylene, suction from 850 to 1250 psig, discharge 1300 to 2000 psig.

 

[ratash]

Thanks.

How does one determine to pump or use a compressor?

From your example, it seems to me capacity is one factor. Correct?

 

[dcasto]

once you know the design parameters of the system, like flow rates, temperature, pressures, then you can size the equipment. Then get quotes. If the head required gets to large, then pump may not work.

But the real test will come when looking at the conditions. If the pipeline pressure goes from 1600 to 2100 psig, you will waste a lot of horsepower using the pump. The pumps tend to have lower maintenance isssues.

So you should do a little life cycle cost for the first 15 years (after 15 years, the NPV values don't show much influance). My gut says use a pump along the pipeline, but if this is at the same location where you are gathering and compressing in the first few stages, then just continue with the compressor...

 

[ratash]

Thanks.

What is the issue with density in determing whether to use a pump or a compressor?

I have to compress it until I get to the critical pressure of 950 psig in my case, and then pump it. Is that a good assumption? My flowrate is only 1 MMscfd and my final pressure is 1800 psig.

 

[maddocks]

I would stick with the compressor all the way. Pumping it is certainly possible, but there are some additional considerations:

- compositional changes can change your critical point and play havoc with your design.

- having a pump and compressor in series slaves them to each other with maintenance and makes each of them a critical piece of equip.

- If you have a compressor already, I'd limit myself to a single piece of rotating equipment to limit leaks and seals.

just my opinion...

 

[ratash]

Thanks.

 

[dcasto]

you'll need to watch the temperature on the suction to the last stage. Sometimes the system will go into the two phase envelope so the solution is to stop the cooling on that stage's suction.

a 1 MMSCFD unit is a cake walk....

 

[maddocks]

Good call dcasto, but being nearly pure CO2, the phase envelope looks pretty good as long as you don't drop below 110 Deg F. However, my simplistic phase line is no substitute for a detailed look at phase behaviour and how the system will perform.

Be aware of compressor bypass arrangements, obviously liquid or subcooled CO2 going through a compressor bypass valve will hit cryogenic temp's. We've had good success with a dual-stage bypass to minimize the potential for hydrates and freezing temps.

 

[ratash]

Thanks.

Could you please explain your dual-stage bypass to minimize the potential for hydrates and freezing temps? Thanks.

 

[maddocks]

The primary bypass is cool 2nd stage discharge gas bypassed on pressure control to maintain the suction pressure on stage 1. Because this robs gas from 3rd stage suction, we then bypass hot 5th stage discharge gas to the inlet of the 2nd stage aftercooler. It sounds complex but works quite well, although it does need a good compressor control panel.

Setpoints and loop ramping speeds are important, especially during startup, but overall, we're really happy with the performance. Note that the compressor blowdown MUST also be on the hot 5th stage discharge gas.

 

[ratash]

Thanks, you are very smart!

 

[JimCasey]

You said that the CO2 is supercritical, which means it is a gas at greater than 31C.
The distinction between a pump and a compressor is largely semantic, but pumping a gas IS compressing it.

Maddocks demonstrated impressive in-depth knowledge of a multistage centrifugal compressor.

I once had interaction with a company called Compressor Controls Corp. Their Compressor management technology and surge control technology are both state-of-the-art. Expensive, but not as expensive as a ruined compressor.

 

[ratash]

Thanks for taking the time and commenting.

 

[ratash]

I have it designed with 4 stages recipe compressors (2 MMscfd), the gas is dehydrated after 3rd stage and the 4th stage discharge goes to a pipeline. I have added ESD's at inlet to 1st stage and outlet of 4th stage, and within the TEG dehydrator system. I will take into consideration temperature issue of co2 during exapnsion. Do I you need recycles on my recipe compressors? Also, why did you choose your recyles off the 2nd and last stages? Is there anything else I need to worry about when it comes to control? I am on the gas side of the phase envelope all the way to 2000 psig. Regards.

 

[maddocks]

Capacity control of your acid gas machine is pretty critical.

Is this system variable speed drive? You need a way to vary your system capacity, either through auto-bypassing or through speed control or preferably both for maximum control. Do you have a way to bypass the dehydration unit? You will also need some startup bypass assembly to get the system up and running. Ideally, its nice to have a hot discharge to flare so you can run the system and pretend you're injecting - it allows for a really nice startup and you can play with your loops and tune effectively.

 

[ratash]

Thanks. Currently the stream is vented to atmosphere, so I have it going to atnosphere in case of emergency shutdown. As far as flow control, I have not added anything, that is why I was asking how do I set up the control on the compression train. The dehydrator system is on pressure and level control. I will add bypass around the dehydration, but do not know how to do it around the compressors efficiently. That is why I was asking the reason behind your choice of recycle points. By the way, I have sour water coming off my scrubbers. I am thinking of collecting them into a sump tank and trucking it off. What kind of material does my sour water tank need to be? Regards

 

[maddocks]

This forum can be a bit difficult to discuss everything you need to know for an acid gas machine. However, a few key tips to keep in mind:

1.) Cascading water dumps can allow for some excellent degassing of water so that when it finally dumps to the tank, it has been degassed to a great degree. if you dump from stage 4 suction direct to a tank, there some considerable risks including ripping the tank apart when the dump valve sticks open. As well, if you overcool 4th stage suction, you can send condensed CO2 to the tank where it expands and...well, you get the picture.

2.) I described the bypass pretty well on one of the above postings.

3.) I've written this up in a couple of recent design papers - let me know if you can't track them down with google.

 

[ratash]

Thanks, I iwll look for them.

I will add one by pass from last to 1st, the rest will be designed by the compressor vendor.

Thanks for the dump valve issue, good point.

 

[dcasto]

the liquid lines need to be 304L, some people believe there may be Cl attack and use 316L.

Here is another tip. Put a sweet gas purge system in place. If the unit ESD'd, the suction and discharge lines are closed and a sweet natural gas source is sent through the compressor to a flare. You could even use N2.