Cooling Reactor Contents Using an Air Sparge 2

[saffer]

Does anyone have experience of cooling reactor contents using an air sparge?
We have just taken over a process in which a 30m3 glass fiber reactor (vented to atmosphere) contains 25m3 of an aqueous solution which requires cooling from 110 deg c to 70 deg C.

We do not want to use a graphite type heat exchanger (cost) and I would like to know if anyone has experience of using an air sparge as a means of cooling.
Thanks
Saffer

 

[moltenmetal]

Wow, you really need to provide a LOT more detail if you expect an intelligent response to that question!

 

[saffer]

Thanks, Initially I just wanted to know if anyone had tried this approach to cooling a reactors contents before and was it successful?
Happy to supply more details if you think the concept is feasible.
Regards Saffer

 

[StoneCold]

Air sparging will either take a lot of air or a long time.
You could calculate it out simply just based on heat capacity of the two fluids assuming perfect heat transfer.
25m3 aqueous solution is a lot to cool down. If your tank is vacuum rated you would have better luck with autorefrigeration than the air sparge. I would either pump it around to a heat exchanger or have submerged coils in the mix tank.

Regards
Stonecold

 

[moltenmetal]

If you're starting with a BOILING solution at 110 C (assuming your "glass fiber" (presume you mean FRP/GRP) reactor is not under much higher than atmospheric pressure) and you're taking the temperature down to 70 C by sparging air, what you're actually going to be doing is evaporating water from your aqueous solution, humidifying the air. It's a simultaneous heat and mass transfer problem, but it's way more favourable than treating it as if it were a pure sensible heating/cooling problem the way Stonecold recommended.

Evaporation will remove quite a bit of heat and fairly quickly- especially at the beginning. How much heat and how quickly you'll get down to 70 C depends on how fast and how effectively you add and remove the air (i.e. the kla), and what else is in your "aqueous solution" (i.e. the partial pressure of water above that solution versus temperature)- and of course, the pressure.

What's the source of your air? If you're feeding it from a blower or compressor directly, it won't be cold, and that will matter a great deal at the end of your cooling cycle.

In future, more information helps us give you better advice.

 

[moltenmetal]

The trouble of course is that there's obviously something ELSE in that "aqueous solution" which is why you'd have to splash out on a carbon block exchanger- big $$$- rather than buying a cheap plate exchanger in stainless steel or the like. It would be foolish to assume that the hot vapour you'd be producing by means of an air sparge wouldn't contain that nasty substance, at very least in droplets leaving with the humidified air. Perhaps this is all going to an emissions scrubber which is sized to handle all this vapour and all is well, but if not, a carbon block condenser is going to cost you even more than a carbon block recirc stream cooler.

 

[saffer]

Many thanks for the response the following background info will put the situation in perspective.
We have just taken over the plant which manufactures ferric sulphate solution (very unsuccessfully) from iron oxide water and 98% sulphuric acid using the said GRP reactor under atmospheric pressure.
There is no cooling capability on the reactor (dont ask why I dont know either?) so after a 24 hour period when the reaction is complete the reactor contents are at about 110 deg C. The GRP reactor then takes 3 days to cool to 70 deg C the temperature at which it can be filtered (filter limitations).
You are right there is a scrubber (not effective) with a 3.5 kw blower drawing off vapours from the reactor to atmosphere.The nasties you refer to would be ferric sulphate droplets.
Because of the plant history I am required to to spend "too much" on rectifing the problems hence my idea of using an air sparge to cool powered by a blower or compressor fitted with an air cooler.
I would appreciate some guide lines in sizing the compressor/blower to cool the reactor contents in say 12 hours.
The reactor vent is 4 inch as is the reactor outlet where I propose to inject the air.
Regards and Thanks
Saffer

 

[moltenmetal]

Sorry, I won't be doing your fishing for you- teaching you to fish for yourself is all I can offer.

If you were making ferrous sulphate, you'd be out of luck obviously!

Assuming you don't have a simulator handy, I'd suggest you set up a very crude model- assume that the bubbling tank is, say, 80% of a theoretical stage (i.e. you get the exiting air up to about 80% saturation of water vapour at the temperature of the liquor at any point in time). That's a total guess- you can get more accurate guesses from a simulator or more detailed calcs- or you can just set the % approach to a theoretical stage as another variable.

Set up a finite difference table in Excel, with a time period of about a minute or so. You know the mass and heat capacity of the material in the tank. Set an air mass flow, and see how much heat you remove with a minute's worth of air, how much the temperature drops, how much water you lose, and what the new vapour pressure of water is at that temperature. If you think the sensible heat capacity of the air itself will help, you can add that in too, but it's probably negligible. By difference, set up the conditions for the next minute, and the next...

Once you come up with an air mass flow that's necessary to do what you want, you can dig through Perry's or McCabe, Smith & Harriot or Treybal or the like to find out if that's a reasonable amount of air to bubble through that much water- what gas hold up you should expect, whether or not that will overflow your tank, and whether or not the cost of the blower will bankrupt you or the air will overwhelm the scrubber etc. Of course, if your solution foams, you're screwed.

Best of luck- others here might want to suggest better models, or you might come up with one yourself.

 

[Compositepro]

To air cool requires a large mass flow rate of air. You can get much larger flow rates by blowing air over the surface of your liquid than by blow air through your liquid. Cooling your coffee by blowing over it is more effective than using a straw and blowing through the coffee. And it is far easier and less messy.

Your are basically air cooling your reactor now, but very inefficiently. Just blowing air through the head space will greatly improve your efficiency.

 

[moltenmetal]

compositepro is of course right- unless your tank is unmixed, which I've assumed that it is. Perhaps that was a dumb assumption, as it's used for dissolving iron oxide in sulphuric acid, which implies mixing.

Your cup of coffee can cool effectively by blowing across it only because it's not very deep. Your cup of coffee is 25,000 litres and probably more than 30 tonnes. But if you have a mixer, the mixer will ensure that the surface is rapidly replaced with hot liquid, so bubbling through the liquid isn't required. A fan could work quite well- it will achieve a much lower approach to a theoretical stage, but would allow you to move a MUCH larger mass of air per minute. That said, your 4" inlet and outlet are a problem. You'll still need quite a blower to get enough air through there to do the trick.

 

[Compositepro]

Actually, convection currents are very effective at circulating tank contents as long as the liquid viscosity is not too high. Cooling from the top and heating from the bottom generally does not require agitation to get pretty uniform temperature in a tank.

To elaborate on my straw analogy, compare using a straw to blow air at the surface of the coffee cup to blowing through the straw to the bottom of the coffee cup. You have to throttle way back on airflow because air velocity cannot be higher than the velocity of rising bubbles or you blow the coffee out of the cup.

A blower on a 4" nozzle will give at least an order of magnitude increase in cooling versus a covered tank. You do want the air to impinge onto the surface and not merely sweep the head space for maximum heat transfer.

 

[moltenmetal]

Sounds like the OP is already sucking quite a bit of air through the headspace using that 5 hp blower on the scrubber. Perhaps just opening the 4" inlet to the atmosphere would be enough to get the flow of air up to the max the scrubber can handle, or is the OP already doing that?

 

[saffer]

Moltenmetal,Compositepro and StoneCold thanks for your help.Will remember "site etiquette" re fishing in future.
Both the air sparge and air blow over the surface (the tank is stirred)worked and will provide a cost effective solution to our problem.
Will post details when the GRP reactor is assembled after repairs and in operation.
Many Thanks and Regards
Saffer