Gas Displacement in an enclosed space

[cmcbain]

Hi everybody.

I am working with a large (7000 gal) horizontal tank that has a single vent in the top. I am trying to solve the following problem:

Before opening the tank, conditions are STP with a mixture of hydrogen gas and air in the tank (say 50/50). I will then open the vent, which is a 24" diameter flange, and allow the the hydrogen in the tank to be displaced by air through natural circulation. I'm trying to figure out some sort of function that will allow me to estimate the hydrogen concentration in the tank with respect to time.

Your suggestions appreciated.

 

[GregLocock]

It's a buoyancy problem.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[zdas04]

I don't think it is a buoyancy problem. Mechanical gas-on-gas separation is really difficult. If you take a tall, skinny vessel filled with a mixture of (say) CO2 and H2. A gas analysis taken at the top will be the same (within the uncertainty of the instruments) as one taken at the bottom. In an unconstrained system the lighter (and more energetic) component will tend to bounce out of the control volume, but within a tank I'd think that the vent process would have to replace about 3 times the void volume with uncontaminated air.

From what was described in the OP I think that when you open the vent, very little will happen. What is the driving force for the circulation? The vent is big, but it sounds like there is only one so you can't get a syphon going. I've watched single-vent tanks several times and they tend to breathe in, then breathe out, but neither the intake or exhaust flow is very fast or very consistent. The best air exchange I've ever seen is a tank with two 3-inch vents, one in shadow and one in sun. There was always flow out the sunny vent and in the shaded vent--void-space samples showed that the void space was filled with air (the stuff we wanted out was mostly gone).

David

 

[IRstuff]

The gas that's right at the opening will be depleted of hydrogen, because its bouyancy will allow the hydrogen to diffuse quickly into the outside air.

The gas inside will see a diffusion problem alone, and the concentration gradient will drive the diffusion. However, it'll be an awfully slow process.

TTFN

FAQ731-376

 

[GregLocock]

The buoyant pressure on the column directly under the vent is roughly (1/2*rho air)*g*h where h is the height of the tank, assuming that hydrogen has zero density.

Quickest way of getting hydrogen out would be to split the vent into two, and fit a magic one way valve to each.

Now work out the flow through an orifice with of that size with the pressure differential above.

Then allow for an increase in the tank gas density as mixing takes place.

That'll overestimate the rate of dilution, in reality the system has a series of burps rather than using magic one way valves.







Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[cmcbain]

I think I'd like to go at this from a chemical concentration gradient point of view.

Greg, I started out on this problem before I made this post with your "magic valve" method, basically modeling the system as a U tube with the air on one side and Hydrogen on the other. I just don't think that it's a reasonable approximation.

I have a feeling I'm going to do all this work to find out that daily temperature cycling is still faster than diffusion...

 

[Compositepro]

The density difference between air and 50% hydrogen is so great that a closer analogy would be a 2-liter soda bottle full of water turned up-side down. The It will go glug-glug (now that's scientific).

 

[cpretty]

The problem with a single vent hole is that it is very difficult to have things go in and out of a hole at once. Very similar to trying to pull an upside down glass out of water.

The 'magic one way valve' could be accomplished by placing a tube of approximately half the cross-sectional area through the vent hole to the bottom of the tank. You could even run this vent to the other end to the tank for improvend venting.

If the tube stuck up into the air, then this could allow fresh air into the bottom of the tank, with the hydrogen rising and exiting via the remaining open area.

I feel that you will get combinations of diffusion, bouyancy, temperature effects etc that are at such a slow rate that any assumptions will have orders of magnitude errors compared to the real answer.

Possible easier would be to optain measuring equipment to measure hydrogen concentration in the tank. Measuring this a number of times and charting may be able to be extrapolated to give an indication of when you will reach the conditions you require.

I assume that forced ventilation is not an option. A blower into the furtherest end of the tank could possibly acheive what you require in a fraction of the time.

Craig Pretty
Tru-Design Plastics

 

[cmcbain]

cpretty & compositepro,
The purpose of my post was to seek ideas on a strategy to evaluate the problem. I am performing this evaluation to determine the relative benefits of adding a purge system to the tank.

One of the problems is that hydrogen remains combustible in air down to about 4% concentration. At 50% the gasses might be relatively well separated, but at 4% there isn't going to be a nice "cream" layer of hydrogen above the air, or at least I don't believe there will be. I believe the air and hydrogen will be well mixed...

 

[Compositepro]

If you want to analyze it you better read a lot books on diffusion and vortex shedding and still have no precise answer. If you want to vent the tank quickly drop a 12" flex duct to the bottom of the tank so that the air can enter the tank and go to the bottom without mixing to displace the hydrogen. You will get a pretty high velocity flow - like a rising helium balloon.

 

[eadwine]

One practical concern above and beyond the time needed to evacuate the H2 in the tank, hopefully, it's being vented to a safe dispersal location and not an operating area.

As for purging, use an inert gas like N2. Less chance of explosion and fire. And above all don't use O2 as the purging medium.

 

[cmcbain]

Fick's Law... That seems to get me the best guess if anyone's interested in my progress. I wanted to be as conservative as I could, and Fick's just gives a pure diffusion model. Like the heat equation for diffusion, pretty nifty...

eadwine, I was probably just going to purge it with a pipe bomb...

 

[eadwine]

cmcbain, Just checking. Sometimes you never know... Murphy's always on site. I haven't been irradiated, contaminated, blown up, or burnt yet and don't intend to start now.

 

[MiketheEngineer]

All seem plausible - just light a match - problem over.