4 replies to this topic

[Roland]

Dear All,

I am currently involved in the design of a supercritical CO2 spray-drying system and one question that has come up repeatedly is how to safely vent the CO2 when everything that can go wrong has. We have relief valves for all the separate parts of the system that each don't hold much CO2 at any time - except for the storage vessel that is a bit bigger. Data: Volume: 1m3, maximum fill level 75% liquid CO2 at 50bar and 14degC. The vessel's maximum operating pressure is 80 bar. The supplier (Niezgodka) suggested a relief valve with a max flow of 11m3/h (water) / 1500m3/h (steam) /1370m3/h (CO2, gaseous, I suppose but not stated by the supplier) if triggered at 95 bar.

Our concerns are: When the pressure has reached 95bar the CO2 is supercritical and at this flow rate would vacate the vessel rather quickly (which is fine) but how do we design the piping system after the relief valve? How far would you expand it from the outlet of the relief valve (3/4")? Can a system like that have any 'bends' in it ?
Would a relief valve with a lower max. flow be more suited for this?

Comments would be much appreciated!

Cheers,
roland

[Andrei]

Roland,

As a general rule, your PSV backpressure cannot exceed 10% of the set pressure. Please see Ch. 3.3.3.1.3 from API 520. So, if the discharge is in the atmosphere than the pressure drop on the discharge line should be max. 10% of the set pressure.
You have to consider phase changes during relief and you have to compensate for any thermal effects during relief.
The best thing to start is to have a look at the CO2 phase diagram and identify what is the process line you will follow during relief. I think you will discover a significant thermal effect, most probably significant cooling.
Regarding how far you have to expand your discharge that is a safety concern. PSV vent shouldn't interfere with any other surrounding processes, has to be outside from the building and at a safe location. Careful, CO2 is heavier than air, tends to accumulate in confined spaces, and is lethal.

Regards

[Art Montemayor]

roland:

You fail to identify whether your stated pressures are absolute or gauge. I have to assume the former.

You say your concern is your storage vessel which is 1.0 m3 in total volume capacity – of which 75% is liquid CO2 and the remainder being vapor CO2 at 50 barA and 14 oC. My data states saturated CO2 liquid at 50 barA has 14.284 oC – which is as close as practicality can take us. Therefore, my guess at your pressures being absolute is correct.

I would worry more about the relative height and location of your PSV on top of your storage container than I would about the 75% liquid contained in the vessel. The reason I state this is because it is very important when dealing with liquid CO2 to ensure that NO LIQUID gets entrained into the PSV when this is activated. Should liquid CO2 be introduced into the PSV what will result is Dry Ice snow being produced in the PSV orifice and outlet nozzle, causing these to be subject to clogging and pluggage with the produced solid. Your PSV could easily become ineffectual due to the plug-up. Therefore, ALWAYS locate your PSV as far away (as high) as you can from the surface of the liquid CO2 contents in the vessel.

You state that your "concerns are that when the pressure has reached 95 bar the CO2 is supercritical and at this flow rate would vacate the vessel rather quickly". I totally disagree with this statement. CO2 fluid at 95 barA and 14 oC IS NOT SUPERCRITICAL. IT IS SATURATED LIQUID. However, CO2 fluid at 95 barA and 31 oC IS SUPERCRITICAL. Now, which one do you mean?

Additionally, how do you surmise that the CO2 would exit the vessel "rather quickly"? It will exit the vessel in accordance with the proper sizing of the PSV orifice.

You ask: "how do we design the piping system after the relief valve?" GOOD QUESTION!

If you analyze the thermodynamic state and phase that the CO2 finds itself in under both conditions of relief and exit pressures, you will discover that you are expanding a supercritical fluid into the region where only saturated vapor and Dry Ice can exist. This is because you are expanding a supercritical fluid directly to atmospheric pressure. You are going to have a LOT OF TROUBLE with your exhaust piping and system. As I stated previously, you will be producing Dry Ice snow and it will clog up and plug your exhaust piping to the point where it will not be functioning. This is a dangerous and hazardous situation that you could create for yourself. I highly recommend that you not do what you propose.

The reason your system will produce Dry Ice upon PSV relief is due to the fact that when you reach 95 barA and 31 oC, the CO2 is totally in the supercritical phase. This is so because there is ONLY ONE PHASE IN THE VESSEL – THE SUPERCRITICAL PHASE. Therefore, what your PSV relieves is not a vapor but a supercritical phase that is linked isenthalpically to two (2) additional phases: a gas phase and a solid phase (Dry Ice). You can't avoid this.

A relief valve cannot control or fix the relief rate. The relief rate is fixed by the process it works under and the relief valve is sized accordingly to the required relief rate.

If you require additional engineering recommendations or answers on this application, please furnish ALL THE RELEVANT BASIC DATA AND PROCESS INFORMATION. This application is very serious and subject to a dangerous hazard if not applied and designed correctly. If we are expected to help or aid you, we must have what I have specified before proceeding any further.

Await your reply.

[Qalander (Chem)]

Dear roland Hello/Good Morning,

May I add another dimension to this venting at high velocity; It was practically observed that static charge is generated and accumulated on the body and if sufficient grounding/bonding continuity is not available this apparently hazard-free outflow may have the potential to discharge in air at nearby available grounded structure and may prove very risky.

As such sufficient height and fool proof and proper grounding is imperative I believe!
Best Regards
Qalander

[Roland]

Thank you for your help! I've been traveling for a while so I didn't reply earlier.

Andrei:
We will use a venting pipework that we have in place for the air coming from molecular sieve vessels during regeneration, it will be quite hot, large in diameter (starting at 3" and expanding to a 5" outlet) and only a few meters long. We do not have any rooms underground (the whole plant is pretty much open-air) and will have gas sensors installed. For the few closed, air-conditioned rooms that we have we are currently designing a system to have automatic (i.e. spring loaded) doors that open with power failure, CO2 pressure drop in the system or low O2 levels.

Art:
You're right, CO2 is not supercritical at 95bar and 14 degree C, I neglected to write that I assumed ambient temperature (if the cooling fails) which is close to or above 30C here most of the time. Thanks for the comments on the location of the PSV valve, we are thinking now to place it higher than planned - and set the trigger pressure below the critical pressure (73bar) to keep the CO2 in liquid and gaseous phases and avoid the inconvenience of having to relieve supercritical CO2. Would you expect this to work?

Qualander:
Thanks! I hadn't even thought about a static charge build-up during relief - we'll ensure proper grounding of all the related pipework.

Best regards,
roland