Liquid CO2 Pumping

[thegasgiant]

G'day all,

I have a problem with a liquid CO2 pump failing every 4 to 9 months. The CO2 is at its boiling point ~-22°C at a pressure of ~1750kPag. The pump has 2 metres of liquid height above the suction. The pump suction is 1.25 inchs with approx 2.5 metres of the same size pipework between it and the vessel. Flow rate required is 3,000 kg/hr. Operation is 24-7-350.

Basically what sort of pump best suits this application?

Many thanks,

Paul.

 

[25362]

BTW, you didn't specify the type of failure, nor the type of pump.
The CO₂ is, as you say, just at equilibrium (boiling) conditions in the reservoir, thus friction drop in the suction line plus heat gains from the surroundings may well provoque vaporization of the liquid. Is there a possibility that the liquid vaporizes in the suction line or in the pump ? Why isn't liquid elevation measured at the bottom of the reservoir; is there a way to keep the liquid level surely constant ? Are there any chances for drastic weather changes every so many months to become sources of heat gain ? Since the pump works around-the-clock for 8400 hours, can the flow of about 48 L/min change upwards so as to increase the friction in the suction line, including entering and leaving losses, or, otherwise, downwards to create recirculation in the pump itself? Do you have flow controls of any kind? What is the ID of your suction pipe ? Can it be increased ?

These questions are meant for your own consideration and analysis on the probable reasons for the pump's sporadic failure. Good luck.

 

[Montemayor]

thegasgiant:

The pumping of pure, saturated LCO2 (Liquified CO2) is an ordinary, common occurance today. I've done it many times and I've left installations that may still be in service. Your major Industrial Gas Producers, like Praxair, deliver thousands of tons of LCO2 on a daily basis using transfer pumps mounted on tanker trucks that transport the LCO2 to bottlers (like Coca-Cola, Pepsi) on a daily basis. This pumping of saturated LCO2 takes place under saturated conditions, like your own, and the trucker pumps out the contents of his tanker down to essentially zero level - far less than your good 2 meters of head.

In the early days, 40 years ago, we tried a lot of pumps out at Liquid Carbonic Corp (now Praxair) where the book was written on CO2. We settled on a Smith MC2 model that seemed to work very well. It was a conventional gear pump with cast iron construction. However, I must alert you to the fact that what 25362 is asking has the basis for success. You must TRANSPORT your heat pick up that the pump generates. In other words, you can't throttle, constrain, or allow the pump to heat up without having a heat sink - and that is essentially what you have in a LCO2 pump, no external heat sink. In other words, you don't have a colder fluid (like liquid N2) cooling down the pump and removing the heat generated by pumping. You are totally reliant on the flow of the LCO2 to take the heat built up away with it. You've got to keep it moving forward.

I used centrifugal pumps (both conventional and regenerative turbine) as well as gear pumps and piston pumps to transport LCO2. Their construction material has been cast iron, stainless, and bronze. There are suppliers out there with plenty of experience now and they'll be of great help to you. I realize it's difficult to describe your system without a sketch ability on this forum, but you could at least tell us basic data like 25362 mentions.

You should not have any problems, but you haven't told us how or why the pump(s) fail after the months of service. Is it a materials' failure or a process failue?

Art Montemayor
Spring, TX

 

[thegasgiant]

Thanks for the answers folks.

I deliberately did not indicate which pump we are already using as I wanted an impartial response. The pump is a cast iron vane pump that is normally used on tanker operations and is quite reliable on tankers. But due to the infrequent operation on tankers this tends to skew the 'reliability' perception, where 2 to 3+ yrs without a problem is the norm.

Failures tend to be worn vanes, which if the pump is left to run (which is normally the case) leads to bearing and then coupling failure as well. I believe that it all can be traced back to cavitation.

25362, I understand all that. I have suggested to the plant folks that the suction line be increased in size to 2 inches from 1.25 inches to reduce the pressure drop in the line, but as Montemayor suggests we have to keep the liquid moving to avoid boiling due to heat ingress, so it is a trade off. The suction piping has 2 inchs of insulation around it and the good news is there is no ice build up on the outside of the insulation, so the insulation is working well. The pump was originally installed with insulation on the pump itself, but it was removed and not re-installed hence a large amount of water ice build up on it is the norm. I will suggest the insulation be reinstated.

The pump is on a VSD to provide level control on the vessel, our flowrate into the vessel is quite constant. There are also speed limits in the VSD to avoid excessive cavitation at higher pump speeds and also a low speed cut out for very low incoming flows to avoid 'stalling' the pump (i.e running at very low speeds with only gas in it). A suggestion has been made to increase the level set point to help increase the NPSH available.

I will consult the local suppliers again for their suggestions and see if they come up with a good suggestion.

Thanks,

Paul.

 

[25362]

You may not have ice on the surface of the insulation but moisture penetration may intensify thermal conductance, if the vapour barrier is not effective enough.

Take care.

 

[chemtiger]

I think you should enlarge the inlet piping and make a recycling piping from the outlet of pump to tank. This can solve your problem without changing your pump. If you have anyquestion, do not hesitate to contact to me.

 

[vsquared]

thegasgiant,
you have recieved quite a bit of valuable information. What everyone is saying is correct. The secret to pumping liquid CO2 is to keep it as a liquid. Is it posible that the liquid source is flashing prior to going into the pump? How far away is your source tank? Also keeping the pump cool is another key factor. We have used piston pumps in our gas plants and have found them very reliable.