Product Analysis: Light Gases plus C1-C7 Hydrocarbons

[rakib]

Hi,

We want to detect the components and find the composition of reactor gas product, which is supposed to contain C1 to C7 hydrocarbon components, CO2 , CO, H2 , H2O, O2 and N2. The hydrocarbon components will most likely include CH4, C2H6, C2H4, C2H2, C3H8, C3H6, n-C4H10, i-C4H10, C4H8 (1-Butene), n-C5H12, n-C6H14, and n-C7H16. The reactor product (gases only), is supposed to be coming out at a temperature of 500 to 700 C, depending on experiments.

We have 2 GC-s which we can use: GC-8A (with TCD), and GC-14A (with TCD and FID).

If the product gases are passed through a condenser, H2O, n-Hexane, and n-Heptane will condense out. So, instead of condensing out these, I was thinking of maintaining the temperature of these to around 200 C for one stream, and much lower temp for another stream and feeding to the GC-s.

But since I am a very new user of GC, I am not confident if such an arrangement is going to work or not. Also, how do I co-ordinate between the results from the 2 GC-s?

I would be waiting for your experts suggestions. With many thanks,

Rakib

[chromatographer1]

You have essentially a plant gas analysis which is a common application.

You will require gas sampling valves and multidimensional switching valves.

Your model/brand of GC may have these as options.

You will need a sample system to cool and filter your gas sample and remove as much water as possible from your sample stream, unless you wish to monitor the composition as it leaves the reactor.

Your two GCs can perform the analysis you desire. One will will have two trains to analyze H2, O2, N2, C1, and CO on one column and CO2, C2, C2=, acetylene, and water on another column while backflushing the residual stream composition.

The second GC will measure will measure the C3, C3=, iC4, nC4, iC5, nC5, C6 and C7 hydrocarbons while backflushing the water on one or two trains. This may be done by TCD or FID.

By the use of calibration standards running on the GCs you will be able to determine the mole % of each component and then determine the mass% of your stream composition.

See the various vendors of process control analyzers for the exact applications which will serve you best. They usually include the software to perform the calculations for calibration and stream composition.

best wishes,

Rod

[rakib]

Thanks for the suggestions, Rod.

The GC-8A model, which is equipped with TCD only, is actually a common use GC, and the temperature settings is not supposed to be disturbed. Can you suggest if this only can be used for H2, N2, C1, and CO, CO2, C2, C2=, acetylene, and water ? O2 is supposed to be negligible, and since most likely it won't appear in product, this might not need detection. No temperature programming in this GC.

The other one, GC-14A, which I can pretty control according to my needs (unfortunately this one only has both TCD and FID) can then be used to find mole% of C3, C3=, iC4, nC4, iC5, nC5, C6 and C7 hydrocarbons. The lighter components might leave as a single peak (???. Please suggest if I am wrong. I have no experience in this). Temperature programming possible with this GC.

Since I would like to co-ordinate between results from these 2 GC-s, and find out the molar flow rate of each component in the reactor gas, I was thinking of having at least one common component to be detected by both GC-s. Can you suggest which one might be more feasible? I was guessing C3.

Please note that I am interested in quantifying H2O as well, if possible with these 2 GC-s.

With thanks again, Rakib

[chromatographer1]

The first GC with one TCD can be outfitted with a sampling and backflush valve and a second valve to divide the chromatography train into two parts, one of which will be trapped and released. This is called a backflush/trap configuration.

It will require a mole sieve column to separate the H2, N2, C1, and CO peaks (3 meters should be ok) and a N or A type porous polymer column, Waters or Hayes brand, 2-3 meters in length. Tubing should be SS with fused silica coating if you will be measuring water and CO at low levels. Temperature should be 80-90°C isothermal.

YOU MUST HAVE A CALIBRATION STANDARD FOR MEASUREMENT OF GASES. This standard may be used for both GCs. You should not try to accurately measure and calculate from a common peak.

The second GC can have a baclflush to vent configuration to measure the hydrocarbons propane and higher. A SP-1700 coated chromosorb P column is possibility and there are others. If you wanted to measure propane and propylene together a simple BTU column set from Supelco, my employer, would work nicely. Call Technical Service and they will be able to assist you. www.sial.com or email techservice@sial.com

If higher hydrocarbons are very low you may wish to use the FID and a capillary column to measure them.

best wishes,

Rodney George
Senior Research and Development Scientist
Gas Separations Research
Supelco
595 North Harrison Road
Bellefonte, PA 16823

[rakib]

Thanks a lot for the useful suggestions, Rodney.

Rakib