Jeff:
Rich2001 identifies the main stumbling block that most engineers trip over when they naively believe that "filters" will remove oil vapors from a compressed air or gas stream. As any experienced engineer in the compressed gas industry will confirm: No filter will remove oil vapors. A filter will remove solids readily - and maybe some liquid droplets by coalescing them. But they will not - I repeat: NOT - remove any oil vapor. Do not fall for the bull that filter guys like Parker, etc., try to feed you. There is no filtering process available to do that. Over 10 years in Air Fractionation for Oxygen, Nitrogen and Argon production taught me this. And you absolutely cannot tolerate ppms of oil in air fractionation without a firm last will & testament and a knowledge that it will be the last time you live to do that.
There is only one industrial-proven unit operation for removing oil (I'm talking total oil content, not only liquid) contamination down to 1 ppm (or less) and that is ADSORPTION. This can be done with activated carbon, but I prefer to dry my air simultaneously and therefore opt for Activated Alumina or Mol Sieves. This is done under pressure, with at least 100 psig.
I don't know whose screw compressor you're using, but beware of the oil contaminated product! If you can't tolerate oil downstream, I would never employ a screw on this service. The screw started out as a "blue sky" design in the late 1950's that was supposed to save the world by replacing the reciprocating compressor with a non-lubricated feature. After hitting a brick wall, application-wise during the 1960 -1970 era, fabricators and users finally conceded that a non-lubricated screw was only a pipe dream and the only way that the machine was going to be used industrially was to flood it with oil in order to "caulk" the inherent mechanical clearances. I've used them in ammonia refrigeration where the oil is not a terrible factor and in CO2 production and liquefaction where the oil is a strong liability issue. As in air liquefaction plants, the CO2 application also is subjected to subsequent fixed bed adsorption for mutual drying and to capture the inherent oil vapors that always are present in this type of machine. Of course, the adsorbent that captures the oil is forever poisoned and re-generation is not cost effective. So what I have always done is installed adsorbent "guard beds" in front of my adsorbent dryer unit. The adsorbent that I use for the guard bed(s) is that main dryer adsorbent that I replace on an annual basis due to aging or attrition characteristics. So I make use of what is an accepted annual waste product.
In the old days (1930-1950) Potassium Permanganate was used as an oxidant to remove the oil or hydrocarbon contaminants in the gas stream. This was a wet mess. The working solution has to be replaced from time-to-time and it is, like I say, a Holy Mess. I wouldn't recommend this type of unit operation today with the convenience of dry and available adsorbents. This is a much more effective, controllable, safer, cheaper method that effecively removes the contaminant down to less than a ppm if designed and operated correctly. For cleaning up a sample stream, much as you describe, this should be a piece of cake. I would use Activate Alumina (F-1 or H-151 grade) rather than Carbon because you can visually "see" the degree of the oil captured in the adsorbent - something you can't do with carbon. You're going to scrap the used or depleted adsorbent anyway - I doubt if you can justify regeneration. If you have other gaseous contaminants you want to remove, you can resort to more expensive Mol Sieves - which are custom-tailored for the application and the contaminants.
As a reinforcing bit of process history, bear in mind that a gas chromatograph is, in reality, nothing more than an adsorption unit. This hint tells you what the right process is.
Art Montemayor
