Design of an Oxygen Transfer Vessel

[Ajax001]

Hi Folks,

We are working on trying to come up with a reasonable design of a pipe system to efficiently transfer pure oxygen into solution. This vessel will be designed as follows:

- Three pipes in series, configured vertically. Each pipe will be approximately 30" long. First pipe is 4" dia, second or middle pipe is 6" dia and last pipe is 8" dia.
- Water will enter at the top through a 3" short section of pipe before entering the 4" pipe.
- Pure oxygen will be fed into the 3" pipe where it is expected to be carried down into the reaction vessel.
- Water will be pumped through the vessel at a continuous flow rate of 130 gpm (seawater) at a constant pressure of 2 atm.

Our goal is to transfer as much oxygen to solution as possible inside this chamber - ideally like to achieve 90% or better oxygen transfer eff.

My main question is - what should my pipe sizes be to ensure that my bubbles stay entrapped inside my pressure vessel as long as possible?

As you see, I have chosen pipes sizes of 4", 6" and 8" with each section being 30" in length. These pipe diameters are based on past correspondence on these boards which suggested that Froude Number be used. A Froude Number <0.31 will result in bubbles going upward and a Froude Number greater then 1.0 would result in bubbles being carried out of the vessel. (original article here:

http://www.eng-tips.com/viewthread.cfm?qid=218923).

The pipe sizes I have chosen above result in Froude Number of 2.13 in the 3" section leading into the pressure vessel,0.79 in the 4" pipe, 0.33 in the 6" pipe and 0.16 in the 8" pipe.

Does this seem reasonable?

One question that seems significant is - what size are the bubbles to begin with? The Froude calculation does not seem to account for this.

On the note of bubble size - how would I calculate this and is there any information regarding an oxygen bubbles size versus its upward velocity in water?

And finally, I have read that inlet velocity of the gas is important with flows less then 0.3 m / sec achieving better transfer rates if a sparger is used. However, our inlet velocities will be around 0.8 - 1.4 m / sec via a 3/4" hose. Does this mean, we will get adequate bubble production and dispersion with need to shear up the air flow?

Thanks...

 

[unclesyd]

Personally I don't think you will get enough contact with your proposal. We ran a system for years similar to your proposal at a very low efficiency until we started injecting very small bubbles, actually through diffusers until we got the Komax mixer. I don;t have numbers but do remember that we picked up quite a few points in efficiency in respect to O2 usage.

We used a product from Komax to mix O2 with a process stream. I would give them a call.

http://www.komax.com/det-pnpsteam.html

 

[Ajax001]

This is our second design. We achieved ~85% transfer efficiency with our first attempt using a similar approach. I am hoping to do better then that in this second round. Point being that this design does work but any feedback that might help us do better is always appreciated.

 

[Gator]

I'm assuming that you have already contacted your oxygen supplier's parent company's (Air Products, Air Liquide, Praxair, to name a few biggies) engineering departments. They will have application engineers that may have some insight/solutions.