Leak from low pressure to high pressure??? 1

[seasar]

I have a process in which liquified CO2 with less than 1ppm of entrained O2 stored at 240 psig on one side of a ball valve with air at a few psig on the other appears to not only leak CO2 through the valve (small amounts) but the CO2 appears to absorb O2 through it as well (very small amounts but measureable). The equipment with air in it doesn't normally have it (it is a condenser for the CO2). This only happens when air is introduced to that unit and goes away when the unit is purged. Is it possible that Henry's law is having some affect? I have noticed this affect in a number of processes where O2 has been removed from a liquid (notably water with O2 less than 10 ppb in a pressurized pipe with a leaking flange gasket will pick up O2 and when the flange is fixed the problem goes away).

I would greatly appreciate any help someone could provide in articulating this, I can explain henry's law but have difficulty explaining how the leak appears to occur from the low pressure to the high pressure side.

Thanks!

 

[moltenmetal]

This isn't leakage- it's diffusion, if it's happening at all. It could also be cross-contamination due to valve operation (the contents of a valve cavity slowly dissolving into the parent fluid after the valve has changed position etc.), or bad measurements, or any number of other things.

Diffusion can happen down a concentration gradient, even when there's a pressure gradient in the opposite direction. Witness the loss of helium through the intact glass boretube of a helium-neon laser- in this case, the helium is diffusing from a high concentration of helium, under vacuum, into a lower concentration fo helium in the surrounding atmosphere. That the atmospheric pressure is higher than the vacuum inside the tube is immaterial- the diffusional concentration gradient remains.

If indeed it's occurring, diffusion in your case is happening against an advective gradient (the bulk leakage of fluid across the pressure differential). If the advective flow is high enough (the leak is big enough), diffusion will be overwhelmed- think fish swimming up river. If the river is flowing fast enough, the fish won't be able to move relative to the shore.

 

[zdas04]

MM,
I like the fish analogy. It even goes a step farther--the mass flow rate of the fish is tiny in comparison to the mass flow rate of water in the river. If the fish was a lot bigger then the drag force would make it a lot harder to go against the flow. That kind of explains why volume of something like CO2 that can difuse is much smaller than the volume of He that can difuse.

David

 

[Chance17]

This is pure speculation, but sometimes ball valves have a small hole drilled in the ball.
This is especially common for liquified gas service.
The hole vents pressure if the ball happens to trap liquid. Otherwise trapped liquid would depressure through the stem packing.
In your case, the ball could be venting trapped air. Think about it.

 

[unclesyd]

We have a polymer process were O2 in any quantity is very detrimental to the quality of the polymer so we purge the system with N2 with non-detectable O2 with Teledyne O2 meters. We want startup if the O2 in system measures > 1 ppm O2. If we get a reading of 10 ppm O2 we immediately go around an tighten all the connections on the N2 system while running a continuous O2 analysis. i would say 80% of the time it is the valve packing on non-bellows sealed valves where the O2 is diffusing into the N2.
We had the same problem with a high pressure hydrogenation process, O2 would diffuse into the high pressure H2 and make some very undesirable byproducts

You find this hard to believe that one of the first papers on the diffusion of O2 into a high pressure system, in this case Helium was published many years ago. The helium system was operating at 22,000 psig. The paper was published by the old Gulf Oil Research Corporation.

 

[rmw]

Sesar,

Do an "advanced search" on this site on the word "diffusion" (without the quotation marks. I think you will find some interesting reading as difussion has been discussed quite a few times before.

rmw