How to calculate Carbon Usage Rate

[esass13]

I am familiar with GAC filters, but how do I realistically calculate the carbon usage rate? I've used all the references for Freundlich Isotherms and have come up with a GAC filter life cycle, but my numbers seem way too high when compared to any real experience. Is this sort of thing really more of an estimate and then adjust in the field? Is there a way to be more accurate?

Are there any really good references on GAC filter design out there? All the references I've seen basically give all the isotherm equations and then state what the field study results are. They never seem to match.

Any help is appreciated. Thanks

 

[ccor]

You've done all you can do, and you're right the calculations are basically worthless. The Fruendlich isotherms are determined with DI water and a single compound. Bed life is usually controlled by the NOM in the system, but since all NOM is different there is no way to do the calcs based on your NOM concentration. This is where the rapid small-scall column tests (RSSCTs) and pilot testing come in. The RSSCT's are provide great reswults as long as you have someone who knows what they are doing perform them (only a few do). Pilot testing is also reliable and scale-up is pretty easy, but it will take months longer to come to the same conclusion as the RSSCT.

 

[ivanhoe374]

ccor,

Do you have any additional information on the rapid small-scall column tests ?

I imagine that it involves running a sample of water to be treated through a very small amount of carbon and then testing for break through of contaminants?

 

[ccor]

Yes the method uses laws of similitude to scale down the governing transport equations.

Water Treatment: Principles and Design by MWH (2005) has a decent section on RSSCT design written by John Crittenden one of the original developers of the RSSCT method. Its a good reference if you do a lot of treatment work anyway.

Though I haven't found it on the web, the EPA's ICR Manual for Bench- and Pilot-Scale Treatment Studies (EPA814-B-96-003) has a chapter specifically on this. It was written by Scott Summers, one of the other primary developers of the method that still performs these studies for consultants on a regular basis.

The crux of the method is determining whether a CD (constant diffusivity) or PD (proportional diffusivity) set-up is going to result in the best prediction of full-scale performance. You'll see alot about this issue as you learn more about RSSCT's but beware, recent results have been contradicting long used rules of thumb over CD vs PD.