We are planning to buy an RGA(Refinery Gas analyzer) for our refinery plant.Is the response for hydrocarbons in FID linear? i.e. can we calibrate the FID with only one standard mixture at a particular concentration and can we use this standard mixture for all concentrations of hydrocarbons from 0.01 % to 100% concentration?
What is the general experience with RGA's from Agilent ,Perkinelmer,Varian,etc ?

Thanks in advance

Dear A

I wonder why you are not purchasing a unit from the major players in the analyzer field:

AAI AAB Yokogawa Rosemount etc

Buying a unit based on a lab GC is not a bad decision but the main four have an established durability and service record you may want to compare. There are others in the market and I do not wish to imply that the four above are the only players, or that companies associated with Agilent, Perkin Elmer or Varian do not make excellent equipment.

There are a lot of service and calibration issues in a purchase of an analyzer. I suggest you seek out information before you buy.

Concerning your questions: Yes, FIDs are 'linear' for hydrocarbons over several orders of magnitude BUT.... definitions are everything, and you must understand that your definition of 'linear' may not be the same as another person in the industry. Sampling and measurements can both be non-linear.

Generally, any standard is accurate within an order of magnitude so several are used for BEST RESULTS:
one std for 1 to 1000ppm -- centered on 500 ppm
one std for 0.1 to 1% -- centered on 5000 ppm
one std for 1 to 10% -- centered on 50000ppm(5%)
one std for 10 to 90% -- centered on 50%
one std for 90 to 100% centered on 95%

Do you need BEST RESULTS? Only you can answer that question. And do you want to pay for it?

Different stream concentrations usually require different calibration stds.

Remember: no calibration is linear, all show some curvature. You have to decide what you need and what you want to pay for.

Good luck.

Rod

We have a PE/Arnel RGA. I've found this unit to be very robust and reliable. I don't particularly like PE GC's. Ovens are small, unit is large.
As for calibration standards, we have 3 levels of "fuel gas" containing H2,CO,CO2,N2,C1-C3, and a 2% C3-C5 in N2. Most of our samples on the FID are not high in concentration, therefore I don't see the need for more than 1 cal level above ethane/ethylene. We don't run LPG samples on our RGA. All LPG is run on plot columns with Liq Sampling Valves using theoretical response factors. We have a validation liquid cylinder to adjust response factors. I've only analysed the cylinder a handful of times without any RF adjustments.