Minimum flow over Carbonbed adsorber

[willem79]

Weve got an adsorber for VOC's in tank ofgas, nothing special. I want to raise the caloric value for a downstream RTO by (partially) bypassing the adsorber. During normal operation, the flow of air and vapour through the carbon helps to remove heat by convection. I was warned that very low air flow rates can result in a buildup of flammable vapour in the bed, which may increase the likelihood of hot spots. How do I know what flow is safe?

 

[Latexman]

When the temperature reading of the carbon bed is below the alarm point. The low flow or too rich scenarios probably justifies a high temperature alarm and shutdown.

Good luck,
Latexman

 

[willem79]

Thats true and already done. I'll try to further explain my prediciment. At the point where the VOC laiden air enters the bed, the VOC conc (not offered amount) and available facing bed surface dictates the amount of VOC's absorbed creating a more or less fixed amount of heat while the volume flow of air is variable. Now: If the flow of air (to disperse heat) is lowered, the bed heats up, et voila"we have a hot spot. There must be soome way to predict this point. Is there information on this phenomena for different vapour components and kinds of Carbon?

 

[laliyo]

your heat generation in he bed will depend upon type of contaminants in air.Each specie will have a different heat of adsorption.
I believe, if your air flowrate and VOC concn vary constantly,it would be difficult to come at a perticular number of flow...

Best you can do is keep a certain minimum flow based on your temperature reading of the bed.

You can also mitigate the risk by installing water spray around your bed , though it may be a litle costly if your bed is hugh in size.

 

[Latexman]

willem79,

I'd suggest a different approach. First, at what temperature(s) do bad things start to happen in your system? One example is the auto-ignition temperature of your VOC? There may be other temperatures of concern. You will want the system to stay below the lowest temperature of concern. Calculate the minimum air flow that guarantees that.

Good luck,
Latexman

 

[willem79]

We've got all sorts of goodies such as benzene, mek, phenol, toluene and many more. But i think thf and cresol are the worst for flamability. On the other hand Benzene has the highest mass flow and heat of adsorption.

 

[cheute79]

You have a significant likelihood of a fire at some time. MEK is your biggest problem. Benzene and toluene are pretty easy to deal with. oxygenated HCs, particularly ketones, are known to oxidize once adsorbed (the carbon acts as a catalyst), and generate heat in the oxidation. Your adsorber is essentially a gas chromatograph, so the solvents are preferentially adsorbed in the bed based upon boiling point, if the air flow is not uniform (it never is) and gets worse with the flows below 200mm/sec superficial velocity. The flow distribution in adsorber beds is terrible - you get dead spots under the inlet nozzles, and around the outlet in most designs.

I have worked on several adsorber systems, and any that had ketones had fires. MEK was the best, but still had fires. detection of fires is a combination of art, experience, and luck. The issue is that a fire starts in a small spot, with low air flow, so that the heat doesn't move readily to the outlet stream until its significant, and the carbon is a fantastic insulator, so you can't detect a hot spot 20 mm away. You may find that the best indicator of a fire before it gets big is a sudden and unusual spike (of just a few percent) in the VOCs in the outlet stream or stack. this is caused by thermal desorption of the carbon near the fire. Humidity control is crucial - make sure you have no superheat in the steam, and that the bed gets thoroughly resteamed every few hours that it is "down" or in reduced flow, and GOOD LUCK!

 

[willem79]

Oke,
So let me het this streight. An average carbon bed will not show hotspots (under normal operation and good handling)if we keep average vapour velocity in the bed above 200mm/sec?
Or is there a margin to this?

Thanks anyway (to al you guys for giving me your thougts) becouse I'me starting to see were getting ourselves into a quagmire of problems, just to save some on the gasbill.

Maybe someone has an idea to accieve what i set out to do without the risks involved with lowering gas velocity in the beds

Thanks an greets

 

[cheute79]

Willem:

A "rule of thumb" for sizing a carbon adsorber is a minimum 200mm/sec superficial velocity. Very high heats of adsorption require additional velocity. Higher heats of reaction (such as with oxidation of ketones) require even higher velocities. The real problem is that uniform flow distribution is assumed, and the geometry of most carbon beds is not favorable to uniform flow (there is added cost to the equipment and piping).

If you want to send more fuel to the RTO, you could reduce the efficiency of the adsorbers by adding outside air to the flow to the adsorber. You should do an analysis of the total operating cost for such an option. Include safety, carbon change frequency changes, steaming cycle changes (more water per mol of VOC to the RTO), and other things. These systems are rather complex.