Tank design for liquid CO2

[Crodark]

Does anybody have experience identifying and preventing stratification and rollover in a liquid CO2 storage tank?
Let's assume the tank is a vertical cylinder, volume 2000 -3000 m3, with liquid CO2 at 18 barg pressure.

 

[LittleInch]

What's the design pressure and temperature of your tank / Pressure vessel.

Does it matter?

A rollover would simply raise the pressure, but you can't have a fixed temperature surely?

18 bar is minus 22 C no?

A PV with that size is going to be rather thick or you're in a rock cavern.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[hacksaw]

"Let's assume the tank is a vertical cylinder, volume 2000 -3000 m3, with liquid CO2 at 18 barg pressure."

Quite an assumption. Perhaps it would be more beneficial to describe what you are designing, process wise.

 

[georgeverghese]

With single component liquids, rollover can still occur with thermal stratification. With a colder-denser feed displacing a warmer-lighter layer which was previously the heel in the tank. And with tank pressure/boiloff gas compressor suction pressure commensurately slightly higher to keep the colder feed at bubble point temp.

 

[LittleInch]

Yes, but are rollover issues relevant for a PV?

They are for low pressure tanks where you can't allow the pressure to rise and hence need to deal with a lot of BOG, but a PV where you assume the MAWP is higher than the 18 bar quoted?

Hence my questions. What is the MAWP? Does it matter if the pressure went up to 20 bar, or 25 bar? This is at minus 22 C remember so something needs to keep it at that temperature and allow some warming surely?

It all seems very speculative to me at the moment....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

Obvious remedies to prevent thermal stratification and minimise the risk of rollover would be
(a)recirculating from tank bottoms to tank top during fill operations or whenever you see thermal stratification.
(b) Fill from the tank top and not to tank bottoms.

Colder denser feed can also be due to subcooled feed.

Vapor generation during rollover would be well beyond safety valve capacity, so these remedies should be part of critical safety operations

 

[Crodark]

Thanx, gent on your comments; here are more details.
There are two 2500 m3 tanks i.e. pressure vessels, to be designed for liquid CO2 at 18 barg and -22 C. The Current "working" scenario is that one will be in the discharge mode with an outflow of 16 m3/h of liquid CO2, while the second one will be in the "filling" mode with max inflow of 100 m3/h 12 hours per day. Liquid CO2 will be supplied at a pressure of 16 barg and then have to be pumped to the tank/PV.
Liquid feed to the tank will be at the top of the tank.
Tanks will be provided with a condensing system for vapor, as well as an evaporator to maintain the pressure of the tank in the "discharge mode."
MAWP not decided but as low as possible to optimize cost. Especially coz there is a requirement from the client to go for low-temperature (-70 C) steel to cover the scenario of temperature drop due to a tank leak/rupture.
For the early detection of stratification, we were thinking of a "couple" thermoelements across tank height. What do you think?
There will also be a recirculation of the tank inventory possible, don't want to run the pump continuously but only if stratification is observed. What is the required recirculation rate?

If rollover happens, how much vapor generation will be, i.e., how to size PSV and condensing system?
Our current main drive for the condensing system is the filling scenario since tanks will be located in a cold climate area with "good" PUR insulation.

 

[Compositepro]

I am not understanding this thread about how rollover can cause a problem in a CO2 vessel. Warm liquid at the top has a higher vapor pressure and determines the vessel pressure. Any mixing with colder/denser liquid will just lower the vessel pressure. Rollover can occur in a water tank containing dissolved CO2, or a hot oil tank with some water at the bottom.

 

[LittleInch]

As said previously, you need to do some process simulations to see if this is in fact an issue or not and choosing your design pressure for the tank, I suppose similar to LPG bullets will determine this.

I would have thought your issue is thermal heat load heating up the liquid CO2 when no outlet resulting in higher pressure and hence determining your vapour handling to keep the tank below a certain temperature / pressure.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

In most cases, liquid feed will be subcooled with respect to tank pressure. Since subcooled feed is to tank top, I dont see a possibility for rollover. This is because colder, denser liquid arriving at the liquid top will sink down into the warmer residual liquid in the tank and get mixed. It may be a good idea to do tank recirculation if you notice thermal stratification in any case - this in my view is optional with fresh feed to tank top.

Check if fresh feed is always subcooled, as this would affect the duty of the vapor recondensing system.

 

[Compositepro]

Let's be clear. The important point is that there is no possibility of rollover under any circumstances for the given facts. So why check for sub-cooled feed?

 

[georgeverghese]

@compositepro, feed should be subcooled, else the vapor fraction in an otherwise 2 phase feed stream will contribute to vapor load at the overheads condensing system, and could add more relief load also at the tank top PSV.

Temperature profiler instruments are available in a single assembly which should help you keep tabs on the tank temp profile - optional in this case with tank top feed.

https://www.emerson.com/documents/a...temperature-water-level-sensors​-en-81216.pdf

 

[georgeverghese]

On further thought, a rollover like event is still possible even with liquid feed at the top. And this can happen in stagnant liquids in tall tanks that soak in heat from the ambient. And in this case, you may not see thermal stratification at all, so a tank temp profiler isnt going to be of much use. Continuous recirculation of a small flow of liquid would keep the tank contents well mixed. Use a collector grid spread out to cover the tank bottom leading to the recirc pump. Its your guess what flow to use for this recirculation leading to the tank top feed.

 

[georgeverghese]

@crodark,
My final thought on this is that if you have a single component liquid feed at tank top, it is NOT possible to have rollover. Even in a stagnant tank, liquid "trapped" at the lower layer will get warmer than the cooler upper layer, resulting in a thermal convection current.
Rollover can happen with tank top feed when you have a multicomponent feed, where 2 liquids with slightly different compositions at the same P/T have different densities.

 

[pierreick]

Hi,
A few documents to support your study with a link to describe the phenomena, applicable to LNG and a Handbook with CO2 properties.

https://www.sigtto.org/media/1398/guidance-for-the-prevention-of-rollover-on-lng_ships.pd

Pierre

 

[Crodark]

Thanks gents, for your interesting views and feedback.
Bear in mind that the tank will be fed 12 hours per day only, and on Sunday not.
There might be some other occasion when the tank will be idle (logistic problems, problems with the downstream plant, turnaround...).

Dimension of one tank: 6 meters dia and 50 meters in length (tangent to tangent line), volume around 1500 m3

So, is stratification and rollover possible or not?