Compressor Cavitation-Corrosion

You could read up on cavitation and see if your design could have cavitation damage in that location. Google can provide good pictures showing where the damage happens.

Just in case you don't know how it happens...cavitation damage is caused when the fluid being pumped doesn't have sufficient back pressure. The low pressure areas often on the trailing side of an impeller will develop vapor bubbles which collapse causing tiny but powerful shockwaves. When they occur close enough to the surface the damage caused can look like the metal was sandblasted with heavy medium.

Drew K

1) Where are you getting the CO₂? Is it clean? Dirty?

2) What pressure are you up to at the compressor outlet?

3) How much water is in with the CO₂?

4) What are the cylinders made of?

1) CO2 is taken from the reservoir. The scrubbers are supposed to extract all of the liquid in the CO2. However, sometimes it can have petroleum and water droplets in it.

2) Usual pressure and temperature values at the exit of each stage:

1st stage exit: 60-80 psi, 80-90 C

2nd stage exit: 150-200 psi, 90-100 C

3rd stage exit: 500-600 psi, 100-115 C

4th stage exit: 1150-1300 psi, 120-140 C

3) That question is almost impossible to answer because it totally depends on t he reservoir. There is also water injection to the reservoir. So the amount of water in the CO2 varies. However, the mixture of water-oil-CO2 goes through seperators and many scrubbers. So, at the end, before entering the compressor, it should not have a dangerous amount of water. It has been working fine in the last more than 3 years.

4) Carbon steel.

The third stage is the stage you are having issues with correct?

Is it a coincidence that the third stage is the stage that any water droplets in this system will reach above boiling point, and possibly turn into gasses? Sometimes in the right conditions, small amounts of trapped water, turning to steam can be quite explosive, and may cause cavitation-like issues within the system.

I would make sure you don't have ANY water in there.

I gave a GA and was going to state great observation, but then I looked at the 3rd stage pressure. Water vaporizes at 100C at atmospheric pressure, at the listed pressures in the 3rd stage, I do not think water is close to vapor pressure.

But I do reserve the right to be completely off-base with this comment also!

That sounds reasonable. what do u think about the carbondioxide? I think it might be about a local phase change of carbondioxide, any ideas?

Hi aggie,

Thanks you for answering my questions. Sorry I haven't gotten back until now.

I was looking at a CO2 phase diagram and I think you may be right about a local phase change. That doesn't explain however, why you haven't seen this before. In the fourth cylinder, your CO2 is supercritical.

Without doing some complex equation of state calculations, I can't confirm what phase (or how close you are to phase change) is in cylinder 3.

Aside from the shaft breakage, has anything changed in the operation/maintenance of the compressor?

I thought of something else. The temp and pressure data you gave for each stage; is this historical data? Have these been measured recently?

Everyday they are recorded each hour. Those data I gave to you is the average data that I observed in the last few weeks.

Could the pressure in the pump, or near the intake portion of the pump possibly lower the pressure in a localized manner causing some type of out-gassing? I'm way out of my normal comfort zone on this one... I know in hydraulics, you can see very strange things happen sometimes on the inlet of a pump, and sometimes, depending on temps/fluids, see some interesting things happen inside the pump itself.

Like I said, this isn't my field... But it is interesting

I echo your comment that this has been an interesting thread! I've enjoyed the problem and the educational exercise of watching various talents work out a solution.

Thanks to the OP who kept us informed and offered relevant feedback!

for water to change into steam, the temperature should be 450-500 F at the third stage pressure, which is around 550 psia. So, I do not think it is about water evaporation.
Do u agree?

Yes, there is an inter cooler. At the exit of each stage, CO2 goes through an inter cooler before entering the next stage. However, there is not a condensate drain system in the inter cooler. therefore, what you think makes perfect sense so we can go with it.

Lets say that, the CO2(including water steam) exits the 2nd stage and goes through the inter cooler. The temperature of the inter cooler should not as low as to let the water condensate. However, we also had a problem with the inter cooler months ago. Assume that the inter cooler cools the CO2 more than it is supposed to do and thus the water condensates. Now, we have CO2 gas along with some water droplets and they are moving to the 3rd stage. At the 3rd stage, the pressure and temperature suddenly increases. Next, under very high pressure, that water droplets some kind of explodes, right ? Does that make sense to you?

Thanks.

We would need to know entering/exiting temp and Press at the intercoolers. I doubt that water phase change is the issue, but the temperature data would serve to confirm or reject this possibility.

People smarter than me have pointed to possible CO2 phase change and CO2/water reaction to form carbolic acid. My gut feel is those are the two leads worth following up on.

I presume you meant carbonic acid, not carbolic acid. Carbonic acid is much more likely to be formed, and is significantly more acidic than carbolic acid.

The problem is the entering-exiting temperature&pressure conditions of inter cooler have not been recorded. that's why I am producing lots of guesses.

Please see the pictures below. Thermodynamic properties of CO2 and water for exit pressure of each stage.

CO2

WATER

My initial guess was CO2 phase change but I do not go with it anymore. Considering the stage tem.&pres. values and the above graphs, it does not make good sense that either CO2 or water has gone through a phase change in the cylinder while being compressed. Thus, I think the best I have now is the phase change of water through the inter cooler.

About chemical things between water-CO2, honestly speaking I have no idea, but I will be more than happy to hear yours and regarding possible solutions.

One more thing;

Today I went and checked the inter cooler. It does not have an auto drain system on it.

Then I checked the exit line of 2nd stage (which is the inlet line of 3rd stage). I opened the plug on the line. The empty space at the top of the plug was full of water-lets say approximately 6-7 cm3 of water at atmospheric conditions. In fact, the amount of water travelling to the 3rd stage should be much more because some portion is entraiend by high pressure CO2. Only a little amount is entrapped at the top of the slug due to its geometrical shape.

Thus, I can conclude that definitely some water travels through the system. So, considering this information and my previous answer to 'pritam' , what do you think might be the reason? if it is entrapped water from the inter cooler, why it explodes in the cylinder?

GA to pritam and you. I think the best thing would be to provide condensate drains as suggested by pritam.

I do too!

Lots of great answers and I was thinking it may be possible that water and co2 are combining to form a strong acid causing some of the issues. It might be interesting to run some tests on PH on condensate.

I AM GRATEFUL TO HEAR LOTS OF GREAT IDEAS FROM YOU GREAT GUYS ALL.

THANKS TO EVERYBODY WHO SPEND TIME THINKING ON THIS PROBLEM.